Re: Redshift3D Render
Redshift requires an Nvidia chip based card. A gaming card works fine, I have a mid-level GTX 670(3GB RAM) to go with it and am already happy every time I render a region. You can use the redshift beta with both Maya and Softimage. Might help ease the pain of transition... Cheers, tim On 21.03.2014 02:08, Ed Manning wrote: We should all put our money into 3-person startups. Better ROI than subscription.
Re: Redshift3D Render
I have GTX 680 installed on iMac, and use RS with SI now. It shows me great render speed. It doesn't matter hair, displacement map or any expensive options, RS shows always stable and good speed :) We have 7 Arnold license but it's replaced to RS already :) --- Daniel Kim Animation Director Professional 3D Generalist http://www.danielkim3d.com ---
Re: Redshift3D Render
They just recent added the RS Hair Shader. Rob did an amazing job. And again. They listen, care about the user and have an amazing support. Even that they are working hard to get the first release, they are always going through the bugs and wish list. Congratulations Redshift Team! A true game changer for Softimage. Here is a link to the docs. http://docs.redshift3d.com/Default.html --- Emilio Hernández VFX 3D animation. 2014-03-21 1:33 GMT-06:00 Daniel Kim danielki...@gmail.com: I have GTX 680 installed on iMac, and use RS with SI now. It shows me great render speed. It doesn't matter hair, displacement map or any expensive options, RS shows always stable and good speed :) We have 7 Arnold license but it's replaced to RS already :) --- Daniel Kim Animation Director Professional 3D Generalist http://www.danielkim3d.com ---
Re: Redshift3D Render
Agreed. The Redshift team listens and addresses things very quickly. They are currently supporting Softimage Maya with Max up next. After that I've heard they're looking at everything from Houdini Modo to doing something with Fabric. -Paul On Fri, Mar 21, 2014 at 4:11 AM, Emilio Hernandez emi...@e-roja.com wrote: They just recent added the RS Hair Shader. Rob did an amazing job. And again. They listen, care about the user and have an amazing support. Even that they are working hard to get the first release, they are always going through the bugs and wish list. Congratulations Redshift Team! A true game changer for Softimage. Here is a link to the docs. http://docs.redshift3d.com/Default.html --- Emilio Hernández VFX 3D animation. 2014-03-21 1:33 GMT-06:00 Daniel Kim danielki...@gmail.com: I have GTX 680 installed on iMac, and use RS with SI now. It shows me great render speed. It doesn't matter hair, displacement map or any expensive options, RS shows always stable and good speed :) We have 7 Arnold license but it's replaced to RS already :) --- Daniel Kim Animation Director Professional 3D Generalist http://www.danielkim3d.com ---
Re: Redshift3D Render
Guys, today i bought one license, half an hour later was already delivering a shot, must say i'm impressed! very powerful engine, and just for $100 ! F. On Thursday, February 27, 2014, Eric Lampi ericla...@gmail.com wrote: I just scratched the surface with RS early in the beta test last summer. My wife was doing pro-bono design work for the NYC Human Rights Campaign fundraising gala, and one afternoon I whipped up a neon sign graphic for her. Rendering was a breeze and of course very very fast compared to Mental Ray. Just go spend the $100 and play with it. It's well worth it! Eric On Feb 27, 2014 9:34 AM, olivier jeannel olivier.jean...@noos.frjavascript:_e(%7B%7D,'cvml','olivier.jean...@noos.fr'); wrote: Bumping that thread, to share enthousiasm. I've just switched from RS Alpha 0.2.1 to the Beta 0.3.46. Spent a huge 100$ bill Today is my testing day, doodeling, trying things that were not implemented. You know, just re-descovering. Well, the speed is there. I'm doing an interior (ok semi interior, walls are opened), in rather dark color and it's noise free. But what amaze me is the integration. I'm mixing several bumps, some are repeating some are not, with several different set of UVs, and it's doing exactly what it is supposed to do. ... And dof is activated on preview, because it's free. Le 18/02/2014 16:17, Ed Manning a écrit : Yes, I AM ignoring the RAM requirements of Elysium-style scenes. So none of those in my scenario. On Tue, Feb 18, 2014 at 10:15 AM, Ed Manning etmth...@gmail.comjavascript:_e(%7B%7D,'cvml','etmth...@gmail.com'); wrote: On Mon, Feb 17, 2014 at 6:04 PM, Steven Caron car...@gmail.comjavascript:_e(%7B%7D,'cvml','car...@gmail.com'); wrote: doesn't work like that... i have to convince someone to buy it for the studio, then the graphics cards you guys talk about... 3 titans!? we don't have those types of investments. we have an existing farm with cpus and lots of ram. if i want to render a sequence with redshift... i have to render it on workstations only. also, i am not going to convert elysium to work for redshift on my free time ;) You might be able to write a script to convert the materials, since the parameters are pretty close to Arnold's (they're VERY similar to MR's so going from there would be relatively easy). One possible selling point to management -- since your workstations are probably pretty well-equipped in GPU, and those GPUs are idle all night, you'd be leveraging capacity that's already paid-for. You wouldn't even need to take the workstations off the CPU farm, just earmark a couple of cores on each for scene loading and conversion for Redshift. Network and server might get stressed a bit, but that's kind of normal... Also see my other post on the costs to transition to GPU from CPU. Speaking as a small business owner, I gotta say the GPU path looks MORE attractive financially.
Re: Redshift3D Render
temporarily $100 to be fair (2 months then more than likely it is another 400 to keep using it), but yes, it really is awesome, and a real pleasure to use. From: Francisco Criado Sent: Thursday, March 20, 2014 8:16 PM To: softimage@listproc.autodesk.com Subject: Re: Redshift3D Render Guys, today i bought one license, half an hour later was already delivering a shot, must say i'm impressed! very powerful engine, and just for $100 ! F. On Thursday, February 27, 2014, Eric Lampi ericla...@gmail.com wrote: I just scratched the surface with RS early in the beta test last summer. My wife was doing pro-bono design work for the NYC Human Rights Campaign fundraising gala, and one afternoon I whipped up a neon sign graphic for her. Rendering was a breeze and of course very very fast compared to Mental Ray. Just go spend the $100 and play with it. It's well worth it! Eric On Feb 27, 2014 9:34 AM, olivier jeannel javascript:_e(%7B%7D,'cvml','olivier.jean...@noos.fr'); wrote: Bumping that thread, to share enthousiasm. I've just switched from RS Alpha 0.2.1 to the Beta 0.3.46. Spent a huge 100$ bill Today is my testing day, doodeling, trying things that were not implemented. You know, just re-descovering. Well, the speed is there. I'm doing an interior (ok semi interior, walls are opened), in rather dark color and it's noise free. But what amaze me is the integration. I'm mixing several bumps, some are repeating some are not, with several different set of UVs, and it's doing exactly what it is supposed to do. ... And dof is activated on preview, because it's free. Le 18/02/2014 16:17, Ed Manning a écrit : Yes, I AM ignoring the RAM requirements of Elysium-style scenes. So none of those in my scenario. On Tue, Feb 18, 2014 at 10:15 AM, Ed Manning javascript:_e(%7B%7D,'cvml','etmth...@gmail.com'); wrote: On Mon, Feb 17, 2014 at 6:04 PM, Steven Caron javascript:_e(%7B%7D,'cvml','car...@gmail.com'); wrote: doesn't work like that... i have to convince someone to buy it for the studio, then the graphics cards you guys talk about... 3 titans!? we don't have those types of investments. we have an existing farm with cpus and lots of ram. if i want to render a sequence with redshift... i have to render it on workstations only. also, i am not going to convert elysium to work for redshift on my free time ;) You might be able to write a script to convert the materials, since the parameters are pretty close to Arnold's (they're VERY similar to MR's so going from there would be relatively easy). One possible selling point to management -- since your workstations are probably pretty well-equipped in GPU, and those GPUs are idle all night, you'd be leveraging capacity that's already paid-for. You wouldn't even need to take the workstations off the CPU farm, just earmark a couple of cores on each for scene loading and conversion for Redshift. Network and server might get stressed a bit, but that's kind of normal... Also see my other post on the costs to transition to GPU from CPU. Speaking as a small business owner, I gotta say the GPU path looks MORE attractive financially.
Re: Redshift3D Render
Well, until it gets released end of march, then the price will be $500 for the full version. Still a bargain in my eyes, and a fantastic renderer. On 21 March 2014 00:16, Francisco Criado malcriad...@gmail.com wrote: Guys, today i bought one license, half an hour later was already delivering a shot, must say i'm impressed! very powerful engine, and just for $100 ! F. On Thursday, February 27, 2014, Eric Lampi ericla...@gmail.com wrote: I just scratched the surface with RS early in the beta test last summer. My wife was doing pro-bono design work for the NYC Human Rights Campaign fundraising gala, and one afternoon I whipped up a neon sign graphic for her. Rendering was a breeze and of course very very fast compared to Mental Ray. Just go spend the $100 and play with it. It's well worth it! Eric On Feb 27, 2014 9:34 AM, olivier jeannel olivier.jean...@noos.fr wrote: Bumping that thread, to share enthousiasm. I've just switched from RS Alpha 0.2.1 to the Beta 0.3.46. Spent a huge 100$ bill Today is my testing day, doodeling, trying things that were not implemented. You know, just re-descovering. Well, the speed is there. I'm doing an interior (ok semi interior, walls are opened), in rather dark color and it's noise free. But what amaze me is the integration. I'm mixing several bumps, some are repeating some are not, with several different set of UVs, and it's doing exactly what it is supposed to do. ... And dof is activated on preview, because it's free. Le 18/02/2014 16:17, Ed Manning a écrit : Yes, I AM ignoring the RAM requirements of Elysium-style scenes. So none of those in my scenario. On Tue, Feb 18, 2014 at 10:15 AM, Ed Manning etmth...@gmail.com wrote: On Mon, Feb 17, 2014 at 6:04 PM, Steven Caron car...@gmail.comwrote: doesn't work like that... i have to convince someone to buy it for the studio, then the graphics cards you guys talk about... 3 titans!? we don't have those types of investments. we have an existing farm with cpus and lots of ram. if i want to render a sequence with redshift... i have to render it on workstations only. also, i am not going to convert elysium to work for redshift on my free time ;) You might be able to write a script to convert the materials, since the parameters are pretty close to Arnold's (they're VERY similar to MR's so going from there would be relatively easy). One possible selling point to management -- since your workstations are probably pretty well-equipped in GPU, and those GPUs are idle all night, you'd be leveraging capacity that's already paid-for. You wouldn't even need to take the workstations off the CPU farm, just earmark a couple of cores on each for scene loading and conversion for Redshift. Network and server might get stressed a bit, but that's kind of normal... Also see my other post on the costs to transition to GPU from CPU. Speaking as a small business owner, I gotta say the GPU path looks MORE attractive financially. -- www.matinai.com
Re: Redshift3D Render
Redshift is definitivley a game changer. I have switched the entire Softimage pipeline to it. It has returned the joy of lighting and shading. --- Emilio Hernández VFX 3D animation. 2014-03-20 18:16 GMT-06:00 Francisco Criado malcriad...@gmail.com: Guys, today i bought one license, half an hour later was already delivering a shot, must say i'm impressed! very powerful engine, and just for $100 ! F. On Thursday, February 27, 2014, Eric Lampi ericla...@gmail.com wrote: I just scratched the surface with RS early in the beta test last summer. My wife was doing pro-bono design work for the NYC Human Rights Campaign fundraising gala, and one afternoon I whipped up a neon sign graphic for her. Rendering was a breeze and of course very very fast compared to Mental Ray. Just go spend the $100 and play with it. It's well worth it! Eric On Feb 27, 2014 9:34 AM, olivier jeannel olivier.jean...@noos.fr wrote: Bumping that thread, to share enthousiasm. I've just switched from RS Alpha 0.2.1 to the Beta 0.3.46. Spent a huge 100$ bill Today is my testing day, doodeling, trying things that were not implemented. You know, just re-descovering. Well, the speed is there. I'm doing an interior (ok semi interior, walls are opened), in rather dark color and it's noise free. But what amaze me is the integration. I'm mixing several bumps, some are repeating some are not, with several different set of UVs, and it's doing exactly what it is supposed to do. ... And dof is activated on preview, because it's free. Le 18/02/2014 16:17, Ed Manning a écrit : Yes, I AM ignoring the RAM requirements of Elysium-style scenes. So none of those in my scenario. On Tue, Feb 18, 2014 at 10:15 AM, Ed Manning etmth...@gmail.com wrote: On Mon, Feb 17, 2014 at 6:04 PM, Steven Caron car...@gmail.comwrote: doesn't work like that... i have to convince someone to buy it for the studio, then the graphics cards you guys talk about... 3 titans!? we don't have those types of investments. we have an existing farm with cpus and lots of ram. if i want to render a sequence with redshift... i have to render it on workstations only. also, i am not going to convert elysium to work for redshift on my free time ;) You might be able to write a script to convert the materials, since the parameters are pretty close to Arnold's (they're VERY similar to MR's so going from there would be relatively easy). One possible selling point to management -- since your workstations are probably pretty well-equipped in GPU, and those GPUs are idle all night, you'd be leveraging capacity that's already paid-for. You wouldn't even need to take the workstations off the CPU farm, just earmark a couple of cores on each for scene loading and conversion for Redshift. Network and server might get stressed a bit, but that's kind of normal... Also see my other post on the costs to transition to GPU from CPU. Speaking as a small business owner, I gotta say the GPU path looks MORE attractive financially.
Re: Redshift3D Render
If you have bought the beta you have a $100 credit on the release. --- Emilio Hernández VFX 3D animation. 2014-03-20 18:24 GMT-06:00 Matt Morris matt...@gmail.com: Well, until it gets released end of march, then the price will be $500 for the full version. Still a bargain in my eyes, and a fantastic renderer. On 21 March 2014 00:16, Francisco Criado malcriad...@gmail.com wrote: Guys, today i bought one license, half an hour later was already delivering a shot, must say i'm impressed! very powerful engine, and just for $100 ! F. On Thursday, February 27, 2014, Eric Lampi ericla...@gmail.com wrote: I just scratched the surface with RS early in the beta test last summer. My wife was doing pro-bono design work for the NYC Human Rights Campaign fundraising gala, and one afternoon I whipped up a neon sign graphic for her. Rendering was a breeze and of course very very fast compared to Mental Ray. Just go spend the $100 and play with it. It's well worth it! Eric On Feb 27, 2014 9:34 AM, olivier jeannel olivier.jean...@noos.fr wrote: Bumping that thread, to share enthousiasm. I've just switched from RS Alpha 0.2.1 to the Beta 0.3.46. Spent a huge 100$ bill Today is my testing day, doodeling, trying things that were not implemented. You know, just re-descovering. Well, the speed is there. I'm doing an interior (ok semi interior, walls are opened), in rather dark color and it's noise free. But what amaze me is the integration. I'm mixing several bumps, some are repeating some are not, with several different set of UVs, and it's doing exactly what it is supposed to do. ... And dof is activated on preview, because it's free. Le 18/02/2014 16:17, Ed Manning a écrit : Yes, I AM ignoring the RAM requirements of Elysium-style scenes. So none of those in my scenario. On Tue, Feb 18, 2014 at 10:15 AM, Ed Manning etmth...@gmail.comwrote: On Mon, Feb 17, 2014 at 6:04 PM, Steven Caron car...@gmail.comwrote: doesn't work like that... i have to convince someone to buy it for the studio, then the graphics cards you guys talk about... 3 titans!? we don't have those types of investments. we have an existing farm with cpus and lots of ram. if i want to render a sequence with redshift... i have to render it on workstations only. also, i am not going to convert elysium to work for redshift on my free time ;) You might be able to write a script to convert the materials, since the parameters are pretty close to Arnold's (they're VERY similar to MR's so going from there would be relatively easy). One possible selling point to management -- since your workstations are probably pretty well-equipped in GPU, and those GPUs are idle all night, you'd be leveraging capacity that's already paid-for. You wouldn't even need to take the workstations off the CPU farm, just earmark a couple of cores on each for scene loading and conversion for Redshift. Network and server might get stressed a bit, but that's kind of normal... Also see my other post on the costs to transition to GPU from CPU. Speaking as a small business owner, I gotta say the GPU path looks MORE attractive financially. -- www.matinai.com
Re: Redshift3D Render
We should all put our money into 3-person startups. Better ROI than subscription.
Re: Redshift3D Render
The speed of the hair rendering has just blown me away, can finally render hair and strands using GI, motion blur, DoF, and ice color attributes. No more MR cheats to get stuff rendered in time. Really is a game changer for us, and I don't say that lightly. On 27 February 2014 14:34, olivier jeannel olivier.jean...@noos.fr wrote: Bumping that thread, to share enthousiasm. I've just switched from RS Alpha 0.2.1 to the Beta 0.3.46. Spent a huge 100$ bill Today is my testing day, doodeling, trying things that were not implemented. You know, just re-descovering. Well, the speed is there. I'm doing an interior (ok semi interior, walls are opened), in rather dark color and it's noise free. But what amaze me is the integration. I'm mixing several bumps, some are repeating some are not, with several different set of UVs, and it's doing exactly what it is supposed to do. ... And dof is activated on preview, because it's free. Le 18/02/2014 16:17, Ed Manning a écrit : Yes, I AM ignoring the RAM requirements of Elysium-style scenes. So none of those in my scenario. On Tue, Feb 18, 2014 at 10:15 AM, Ed Manning etmth...@gmail.com wrote: On Mon, Feb 17, 2014 at 6:04 PM, Steven Caron car...@gmail.com wrote: doesn't work like that... i have to convince someone to buy it for the studio, then the graphics cards you guys talk about... 3 titans!? we don't have those types of investments. we have an existing farm with cpus and lots of ram. if i want to render a sequence with redshift... i have to render it on workstations only. also, i am not going to convert elysium to work for redshift on my free time ;) You might be able to write a script to convert the materials, since the parameters are pretty close to Arnold's (they're VERY similar to MR's so going from there would be relatively easy). One possible selling point to management -- since your workstations are probably pretty well-equipped in GPU, and those GPUs are idle all night, you'd be leveraging capacity that's already paid-for. You wouldn't even need to take the workstations off the CPU farm, just earmark a couple of cores on each for scene loading and conversion for Redshift. Network and server might get stressed a bit, but that's kind of normal... Also see my other post on the costs to transition to GPU from CPU. Speaking as a small business owner, I gotta say the GPU path looks MORE attractive financially. -- www.matinai.com
Re: Redshift3D Render
I just scratched the surface with RS early in the beta test last summer. My wife was doing pro-bono design work for the NYC Human Rights Campaign fundraising gala, and one afternoon I whipped up a neon sign graphic for her. Rendering was a breeze and of course very very fast compared to Mental Ray. Just go spend the $100 and play with it. It's well worth it! Eric On Feb 27, 2014 9:34 AM, olivier jeannel olivier.jean...@noos.fr wrote: Bumping that thread, to share enthousiasm. I've just switched from RS Alpha 0.2.1 to the Beta 0.3.46. Spent a huge 100$ bill Today is my testing day, doodeling, trying things that were not implemented. You know, just re-descovering. Well, the speed is there. I'm doing an interior (ok semi interior, walls are opened), in rather dark color and it's noise free. But what amaze me is the integration. I'm mixing several bumps, some are repeating some are not, with several different set of UVs, and it's doing exactly what it is supposed to do. ... And dof is activated on preview, because it's free. Le 18/02/2014 16:17, Ed Manning a écrit : Yes, I AM ignoring the RAM requirements of Elysium-style scenes. So none of those in my scenario. On Tue, Feb 18, 2014 at 10:15 AM, Ed Manning etmth...@gmail.com wrote: On Mon, Feb 17, 2014 at 6:04 PM, Steven Caron car...@gmail.com wrote: doesn't work like that... i have to convince someone to buy it for the studio, then the graphics cards you guys talk about... 3 titans!? we don't have those types of investments. we have an existing farm with cpus and lots of ram. if i want to render a sequence with redshift... i have to render it on workstations only. also, i am not going to convert elysium to work for redshift on my free time ;) You might be able to write a script to convert the materials, since the parameters are pretty close to Arnold's (they're VERY similar to MR's so going from there would be relatively easy). One possible selling point to management -- since your workstations are probably pretty well-equipped in GPU, and those GPUs are idle all night, you'd be leveraging capacity that's already paid-for. You wouldn't even need to take the workstations off the CPU farm, just earmark a couple of cores on each for scene loading and conversion for Redshift. Network and server might get stressed a bit, but that's kind of normal... Also see my other post on the costs to transition to GPU from CPU. Speaking as a small business owner, I gotta say the GPU path looks MORE attractive financially.
Re: Redshift3D Render
FWIW, I've done some pretty heavy archvis-type scenes with Redshift. Several million triangles (usually straight out of CAD, so, yuck), a few hundred ray-traced area lights, an HDRI dome, about 50 MB of textures (so not very much there), TONS of procedural noise textures, bump maps on most surfaces, up to a couple million instanced small objects with emissive materials, with depth of field and motion blur, for HD and print resolution, etc. RS has chomped through all of it in very reasonable times and without ever challenging the memory limit on the Titan. Can't share any scenes or images though. :-( One note on the expense of a Titan -- I have a pair of them in an early-2008 4-core Mac Pro Bootcamped into Windows 7. For a proper lighting/lookdev station running VRay, Arnold or Mental Ray, I would have to spend well north of $8000 on CPU RAM alone and still wouldn't have more than a fraction of the performance. The Mac Pro could be had for under $2K (although it was free to me as I bought it new and it's been fully depreciated) and the pair of Titans for about the same. By upgrading only GPU instead of CPU chassis, companies will be able to double the lifetime of their CPU purchases, perhaps more. And the upgrade cycle costs will be significantly lower per-seat. The Titans certainly paid for themselves on the first project they rendered. I'd feel differently about the changeover costs If I already had a sizable render farm, since most of the blades and 1Us that people have invested in can't even carry 1 GPU card. But if you are building out a new place or growing the farm, you can pick up 1U servers built for Tesla-style GPUC clusters that have power and slots for 3 beefy cards. Since you won't care too much about CPU specs, you can buy ones that are being retired on the secondary market pretty reasonably. Once it can handle ICE attributes, I'll have no reason to look back, EVER... Also, someone was wondering about SLI -- RS doesn't use it; it handles multiple CUDA cards just fine without it, even if they're very different cards.
Re: Re[2]: Redshift3D Render
On Mon, Feb 17, 2014 at 6:04 PM, Steven Caron car...@gmail.com wrote: doesn't work like that... i have to convince someone to buy it for the studio, then the graphics cards you guys talk about... 3 titans!? we don't have those types of investments. we have an existing farm with cpus and lots of ram. if i want to render a sequence with redshift... i have to render it on workstations only. also, i am not going to convert elysium to work for redshift on my free time ;) You might be able to write a script to convert the materials, since the parameters are pretty close to Arnold's (they're VERY similar to MR's so going from there would be relatively easy). One possible selling point to management -- since your workstations are probably pretty well-equipped in GPU, and those GPUs are idle all night, you'd be leveraging capacity that's already paid-for. You wouldn't even need to take the workstations off the CPU farm, just earmark a couple of cores on each for scene loading and conversion for Redshift. Network and server might get stressed a bit, but that's kind of normal... Also see my other post on the costs to transition to GPU from CPU. Speaking as a small business owner, I gotta say the GPU path looks MORE attractive financially.
Re: Re[2]: Redshift3D Render
Yes, I AM ignoring the RAM requirements of Elysium-style scenes. So none of those in my scenario. On Tue, Feb 18, 2014 at 10:15 AM, Ed Manning etmth...@gmail.com wrote: On Mon, Feb 17, 2014 at 6:04 PM, Steven Caron car...@gmail.com wrote: doesn't work like that... i have to convince someone to buy it for the studio, then the graphics cards you guys talk about... 3 titans!? we don't have those types of investments. we have an existing farm with cpus and lots of ram. if i want to render a sequence with redshift... i have to render it on workstations only. also, i am not going to convert elysium to work for redshift on my free time ;) You might be able to write a script to convert the materials, since the parameters are pretty close to Arnold's (they're VERY similar to MR's so going from there would be relatively easy). One possible selling point to management -- since your workstations are probably pretty well-equipped in GPU, and those GPUs are idle all night, you'd be leveraging capacity that's already paid-for. You wouldn't even need to take the workstations off the CPU farm, just earmark a couple of cores on each for scene loading and conversion for Redshift. Network and server might get stressed a bit, but that's kind of normal... Also see my other post on the costs to transition to GPU from CPU. Speaking as a small business owner, I gotta say the GPU path looks MORE attractive financially.
Re: Redshift3D Render
I've just started to look into Redshift and I can only say wow! It's really quite amazing. DAN On Mon, Feb 17, 2014 at 8:22 AM, Angus Davidson angus.david...@wits.ac.zawrote: I believe the expression is *Vrooom* That is incredibly fast -- *From:* Emilio Hernandez [emi...@e-roja.com] *Sent:* 17 February 2014 02:42 AM *To:* softimage@listproc.autodesk.com *Subject:* Re: Redshift3D Render Just changed an old scene from Arnold to Redshift. I am impressed with the performance. They are still working on the hair shader, this one is with the Redshift arch shader. Judge by yourselfves. https://vimeo.com/86856132 2014-02-16 18:33 GMT-06:00 Emilio Hernandez emi...@e-roja.com: This is just XSI hair, but now it supports also strands. 2014-02-16 18:28 GMT-06:00 Steven Caron car...@gmail.com: ice strands or just xsi hair? On Sun, Feb 16, 2014 at 1:40 PM, Emilio Hernandez emi...@e-roja.comwrote: Latest release of redshift with hair. https://dl.dropboxusercontent.com/u/49626349/hair_01.jpg This communication is intended for the addressee only. It is confidential. If you have received this communication in error, please notify us immediately and destroy the original message. You may not copy or disseminate this communication without the permission of the University. Only authorised signatories are competent to enter into agreements on behalf of the University and recipients are thus advised that the content of this message may not be legally binding on the University and may contain the personal views and opinions of the author, which are not necessarily the views and opinions of The University of the Witwatersrand, Johannesburg. All agreements between the University and outsiders are subject to South African Law unless the University agrees in writing to the contrary.
Re: Redshift3D Render
Lofted strands aren't quite there yet, but are around the corner. All in all, Redshift really is rocking our little corner of the business. For us, it's been a game-changer. -Tim On 2/16/2014 6:33 PM, Emilio Hernandez wrote: This is just XSI hair, but now it supports also strands. 2014-02-16 18:28 GMT-06:00 Steven Caron car...@gmail.com mailto:car...@gmail.com: ice strands or just xsi hair? On Sun, Feb 16, 2014 at 1:40 PM, Emilio Hernandez emi...@e-roja.com mailto:emi...@e-roja.com wrote: Latest release of redshift with hair. https://dl.dropboxusercontent.com/u/49626349/hair_01.jpg -- Signature
Re: Redshift3D Render
out of curiosity, has anyone tried a heavy production scene in redshift? somewhere around 100-500m triangles with a bunch of 2k/4k textures? I'm seeing lots of single objects and simple scenes done with it, rendering really fast, but what happens when you throw something heavier at it? -Andreas On Tue, Feb 18, 2014 at 4:00 AM, Tim Crowson tim.crow...@magneticdreams.com wrote: Lofted strands aren't quite there yet, but are around the corner. All in all, Redshift really is rocking our little corner of the business. For us, it's been a game-changer. -Tim On 2/16/2014 6:33 PM, Emilio Hernandez wrote: This is just XSI hair, but now it supports also strands. 2014-02-16 18:28 GMT-06:00 Steven Caron car...@gmail.com: ice strands or just xsi hair? On Sun, Feb 16, 2014 at 1:40 PM, Emilio Hernandez emi...@e-roja.comwrote: Latest release of redshift with hair. https://dl.dropboxusercontent.com/u/49626349/hair_01.jpg -- -- Andreas Byström Weta Digital
Re: Redshift3D Render
I would gladly test to push it as much as possible on 4 titan system but don't really have any complex scene to test on :) On Mon, Feb 17, 2014 at 8:48 PM, Andreas Bystrom andreas.byst...@gmail.comwrote: out of curiosity, has anyone tried a heavy production scene in redshift? somewhere around 100-500m triangles with a bunch of 2k/4k textures? I'm seeing lots of single objects and simple scenes done with it, rendering really fast, but what happens when you throw something heavier at it? -Andreas On Tue, Feb 18, 2014 at 4:00 AM, Tim Crowson tim.crow...@magneticdreams.com wrote: Lofted strands aren't quite there yet, but are around the corner. All in all, Redshift really is rocking our little corner of the business. For us, it's been a game-changer. -Tim On 2/16/2014 6:33 PM, Emilio Hernandez wrote: This is just XSI hair, but now it supports also strands. 2014-02-16 18:28 GMT-06:00 Steven Caron car...@gmail.com: ice strands or just xsi hair? On Sun, Feb 16, 2014 at 1:40 PM, Emilio Hernandez emi...@e-roja.comwrote: Latest release of redshift with hair. https://dl.dropboxusercontent.com/u/49626349/hair_01.jpg -- -- Andreas Byström Weta Digital
Re: Redshift3D Render
Unfortunatley I don't have such a scene to test. I will gladly push it as far with such a kind of scene. 2014-02-17 13:53 GMT-06:00 Mirko Jankovic mirkoj.anima...@gmail.com: I would gladly test to push it as much as possible on 4 titan system but don't really have any complex scene to test on :) On Mon, Feb 17, 2014 at 8:48 PM, Andreas Bystrom andreas.byst...@gmail.com wrote: out of curiosity, has anyone tried a heavy production scene in redshift? somewhere around 100-500m triangles with a bunch of 2k/4k textures? I'm seeing lots of single objects and simple scenes done with it, rendering really fast, but what happens when you throw something heavier at it? -Andreas On Tue, Feb 18, 2014 at 4:00 AM, Tim Crowson tim.crow...@magneticdreams.com wrote: Lofted strands aren't quite there yet, but are around the corner. All in all, Redshift really is rocking our little corner of the business. For us, it's been a game-changer. -Tim On 2/16/2014 6:33 PM, Emilio Hernandez wrote: This is just XSI hair, but now it supports also strands. 2014-02-16 18:28 GMT-06:00 Steven Caron car...@gmail.com: ice strands or just xsi hair? On Sun, Feb 16, 2014 at 1:40 PM, Emilio Hernandez emi...@e-roja.comwrote: Latest release of redshift with hair. https://dl.dropboxusercontent.com/u/49626349/hair_01.jpg -- -- Andreas Byström Weta Digital
Re: Redshift3D Render
Hey Andreas, https://www.dropbox.com/s/1rhnxa9cvge0ddc/redshift_forest_dof_unifiedsampling.jpg I only have this as an example, was done during the alpha, and has a couple million polys in it (can't remember the exact number as i don't have the scene in front of me), sss for the leaves, brute force gi with ibl (dome) with dof and vignetting from the render. All instanced ice geo. Took about 7-8 mins if i remember correctly on a gtx470 with 1 gb of vram (ancient stuff) and seemed to handle the thing pretty well. Did not run into any memory issues with that amount of geo but large textures might be a problem for the vram though. -Octavian On Mon, Feb 17, 2014 at 9:48 PM, Andreas Bystrom andreas.byst...@gmail.comwrote: out of curiosity, has anyone tried a heavy production scene in redshift? somewhere around 100-500m triangles with a bunch of 2k/4k textures? I'm seeing lots of single objects and simple scenes done with it, rendering really fast, but what happens when you throw something heavier at it? -Andreas On Tue, Feb 18, 2014 at 4:00 AM, Tim Crowson tim.crow...@magneticdreams.com wrote: Lofted strands aren't quite there yet, but are around the corner. All in all, Redshift really is rocking our little corner of the business. For us, it's been a game-changer. -Tim On 2/16/2014 6:33 PM, Emilio Hernandez wrote: This is just XSI hair, but now it supports also strands. 2014-02-16 18:28 GMT-06:00 Steven Caron car...@gmail.com: ice strands or just xsi hair? On Sun, Feb 16, 2014 at 1:40 PM, Emilio Hernandez emi...@e-roja.comwrote: Latest release of redshift with hair. https://dl.dropboxusercontent.com/u/49626349/hair_01.jpg -- -- Andreas Byström Weta Digital -- Octavian Ureche +40 732 774 313 (GMT+2) Animation Visual Effects www.okto.ro
Re: Redshift3D Render
We did some tests, on a scene with about 20,000 instanced objects on particles with refraction and reflection to match a Vray shader we were using. The rendertimes were something like 4 minutes compared to 30+ minutes with Vray. Freelance 3D and VFX animator http://vimeopro.com/mybudoinc/animation On Mon, Feb 17, 2014 at 2:48 PM, Andreas Bystrom andreas.byst...@gmail.comwrote: out of curiosity, has anyone tried a heavy production scene in redshift? somewhere around 100-500m triangles with a bunch of 2k/4k textures? I'm seeing lots of single objects and simple scenes done with it, rendering really fast, but what happens when you throw something heavier at it? -Andreas On Tue, Feb 18, 2014 at 4:00 AM, Tim Crowson tim.crow...@magneticdreams.com wrote: Lofted strands aren't quite there yet, but are around the corner. All in all, Redshift really is rocking our little corner of the business. For us, it's been a game-changer. -Tim On 2/16/2014 6:33 PM, Emilio Hernandez wrote: This is just XSI hair, but now it supports also strands. 2014-02-16 18:28 GMT-06:00 Steven Caron car...@gmail.com: ice strands or just xsi hair? On Sun, Feb 16, 2014 at 1:40 PM, Emilio Hernandez emi...@e-roja.comwrote: Latest release of redshift with hair. https://dl.dropboxusercontent.com/u/49626349/hair_01.jpg -- -- Andreas Byström Weta Digital
Re[2]: Redshift3D Render
Don't have such heavy scenes, but did some architectural interiors, 1-2 mio polys, 240 light sources with soft shadows... according to Panos Zobolas, whom I sent the scene for testing, there was an extremely high shadow ray count, which brought it to a crawl (6h for 1920x1920px, GeForce 760). That was 3 months ago - there might be improvements now in that regard. And it did render... -- Originalnachricht -- Von: Andreas Bystrom andreas.byst...@gmail.com An: softimage@listproc.autodesk.com Gesendet: 17.02.2014 20:48:35 Betreff: Re: Redshift3D Render out of curiosity, has anyone tried a heavy production scene in redshift? somewhere around 100-500m triangles with a bunch of 2k/4k textures? I'm seeing lots of single objects and simple scenes done with it, rendering really fast, but what happens when you throw something heavier at it? -Andreas On Tue, Feb 18, 2014 at 4:00 AM, Tim Crowson tim.crow...@magneticdreams.com wrote: Lofted strands aren't quite there yet, but are around the corner. All in all, Redshift really is rocking our little corner of the business. For us, it's been a game-changer. -Tim On 2/16/2014 6:33 PM, Emilio Hernandez wrote: This is just XSI hair, but now it supports also strands. 2014-02-16 18:28 GMT-06:00 Steven Caron car...@gmail.com: ice strands or just xsi hair? On Sun, Feb 16, 2014 at 1:40 PM, Emilio Hernandez emi...@e-roja.com wrote: Latest release of redshift with hair. https://dl.dropboxusercontent.com/u/49626349/hair_01.jpg -- -- Andreas Byström Weta Digital --- Diese E-Mail ist frei von Viren und Malware, denn der avast! Antivirus Schutz ist aktiv. http://www.avast.com
Re: Redshift3D Render
it's a nice example Octavian, but would be fun to try on something really heavy, if it's only a couple of million polys it's still a pretty light scene. if I have time I might prepare something, it wouldn't have to be nice looking, just dump a bunch of high-poly object in the scene and render, you can just map a bunch of random high-rez textures on there as well.. my experience with gpu renderers is that they can be really fast on simple scenes, but once you go over a certain complexity they grind to a halt, redshift is supposed to deal with complexity better than some other gpu renderers, I just haven't seen any examples yet. On Tue, Feb 18, 2014 at 9:07 AM, Eric Lampi ericla...@gmail.com wrote: We did some tests, on a scene with about 20,000 instanced objects on particles with refraction and reflection to match a Vray shader we were using. The rendertimes were something like 4 minutes compared to 30+ minutes with Vray. Freelance 3D and VFX animator http://vimeopro.com/mybudoinc/animation On Mon, Feb 17, 2014 at 2:48 PM, Andreas Bystrom andreas.byst...@gmail.com wrote: out of curiosity, has anyone tried a heavy production scene in redshift? somewhere around 100-500m triangles with a bunch of 2k/4k textures? I'm seeing lots of single objects and simple scenes done with it, rendering really fast, but what happens when you throw something heavier at it? -Andreas On Tue, Feb 18, 2014 at 4:00 AM, Tim Crowson tim.crow...@magneticdreams.com wrote: Lofted strands aren't quite there yet, but are around the corner. All in all, Redshift really is rocking our little corner of the business. For us, it's been a game-changer. -Tim On 2/16/2014 6:33 PM, Emilio Hernandez wrote: This is just XSI hair, but now it supports also strands. 2014-02-16 18:28 GMT-06:00 Steven Caron car...@gmail.com: ice strands or just xsi hair? On Sun, Feb 16, 2014 at 1:40 PM, Emilio Hernandez emi...@e-roja.comwrote: Latest release of redshift with hair. https://dl.dropboxusercontent.com/u/49626349/hair_01.jpg -- -- Andreas Byström Weta Digital -- Andreas Byström Weta Digital
Re: Redshift3D Render
Andreas, I'm working on a project now (can't post any renders though...) that involves a pretty large outdoor environment (not as big as the WT's Athens demo... maybe 1/4th of that). I'm instancing grass clumps /everywhere/, and I have 11-12 Tree prototypes being scattered as proxies, each between 500,000 and 1.5M polys. The enviro is basically surrounded by a forest. Also have some extra scattered instances of undergrowth, bushes, shrubs, flowers. Some static ivy meshes which are very dense as well... the master scene from which I break out and publish assets contains about 15M raw polygons, excluding the proxies... When I render, XSI hangs up for a minute or two, not exactly sure what it's doing... then Redshift kicks in and takes another 2-3 min to export the scene, process shaders/textures, then renders a 1920x1080 frame (using full Monte Carlo only in my case) in about 20-30min, depending on what's in the frame. We have a triple-Titan box for testing and it does it in about 2.4x that. Redshift is excellent at caching anything it can, so it's fairly easy to iterate over local changes. To be clear, when Redshift goes out-of-core, it goes /into /your system ram. So in the end, you still need plenty of system ram for large scenes. But I've only got 20GB in my workstation and I'm getting very complex renders out of it without coming even close to that. More in the 6-8GB range so far. Proxies are awesome, and Redshift wisely had those from day one of the alpha. Now several months ago I rendered this http://timcrowson.smugmug.com/photos/i-qGFB7nv/1/O/i-qGFB7nv.jpg, but it's more along the lines of what Octavian posted. It's just one tree instanced a bunch of times. This rendered in about 30min at that resolution, on a GTX470. -Tim On 2/17/2014 2:18 PM, Andreas Bystrom wrote: it's a nice example Octavian, but would be fun to try on something really heavy, if it's only a couple of million polys it's still a pretty light scene. if I have time I might prepare something, it wouldn't have to be nice looking, just dump a bunch of high-poly object in the scene and render, you can just map a bunch of random high-rez textures on there as well.. my experience with gpu renderers is that they can be really fast on simple scenes, but once you go over a certain complexity they grind to a halt, redshift is supposed to deal with complexity better than some other gpu renderers, I just haven't seen any examples yet. On Tue, Feb 18, 2014 at 9:07 AM, Eric Lampi ericla...@gmail.com mailto:ericla...@gmail.com wrote: We did some tests, on a scene with about 20,000 instanced objects on particles with refraction and reflection to match a Vray shader we were using. The rendertimes were something like 4 minutes compared to 30+ minutes with Vray. Freelance 3D and VFX animator http://vimeopro.com/mybudoinc/animation On Mon, Feb 17, 2014 at 2:48 PM, Andreas Bystrom andreas.byst...@gmail.com mailto:andreas.byst...@gmail.com wrote: out of curiosity, has anyone tried a heavy production scene in redshift? somewhere around 100-500m triangles with a bunch of 2k/4k textures? I'm seeing lots of single objects and simple scenes done with it, rendering really fast, but what happens when you throw something heavier at it? -Andreas On Tue, Feb 18, 2014 at 4:00 AM, Tim Crowson tim.crow...@magneticdreams.com mailto:tim.crow...@magneticdreams.com wrote: Lofted strands aren't quite there yet, but are around the corner. All in all, Redshift really is rocking our little corner of the business. For us, it's been a game-changer. -Tim On 2/16/2014 6:33 PM, Emilio Hernandez wrote: This is just XSI hair, but now it supports also strands. 2014-02-16 18:28 GMT-06:00 Steven Caron car...@gmail.com mailto:car...@gmail.com: ice strands or just xsi hair? On Sun, Feb 16, 2014 at 1:40 PM, Emilio Hernandez emi...@e-roja.com mailto:emi...@e-roja.com wrote: Latest release of redshift with hair. https://dl.dropboxusercontent.com/u/49626349/hair_01.jpg -- -- Andreas Byström Weta Digital -- Andreas Byström Weta Digital -- Signature
Re[2]: Redshift3D Render
To be clear, when Redshift goes out-of-core, it goes into your system ram. So in the end, you still need plenty of system ram for large scenes. But I've only got 20GB in my workstation and I'm getting very complex renders out of it without coming even close to that. More in the 6-8GB range so far. Proxies are awesome, and Redshift wisely had those from day one of the alpha. Now several months ago I rendered this, but it's more along the lines of what Octavian posted. It's just one tree instanced a bunch of times. This rendered in about 30min at that resolution, on a GTX470. -Tim Very nice!! How did you light it? Sun + dome? --- Diese E-Mail ist frei von Viren und Malware, denn der avast! Antivirus Schutz ist aktiv. http://www.avast.com
Re: Redshift3D Render
just so you know... athens was a fraction of what we rendered for elysium. so... i think you guys should be aiming high... i mean REALLY high. if i don't see a billion somewhere (instancing is fine) then i am still not convinced the gpu renderer has overcome the memory limitations. but it sounds like the scene you describe is a beast... can't wait to see it! On Mon, Feb 17, 2014 at 1:48 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: Andreas, I'm working on a project now (can't post any renders though...) that involves a pretty large outdoor environment (not as big as the WT's Athens demo... maybe 1/4th of that). I'm instancing grass clumps *everywhere*, and I have 11-12 Tree prototypes being scattered as proxies, each between 500,000 and 1.5M polys. The enviro is basically surrounded by a forest. Also have some extra scattered instances of undergrowth, bushes, shrubs, flowers. Some static ivy meshes which are very dense as well... the master scene from which I break out and publish assets contains about 15M raw polygons, excluding the proxies... When I render, XSI hangs up for a minute or two, not exactly sure what it's doing... then Redshift kicks in and takes another 2-3 min to export the scene, process shaders/textures, then renders a 1920x1080 frame (using full Monte Carlo only in my case) in about 20-30min, depending on what's in the frame. We have a triple-Titan box for testing and it does it in about 2.4x that. Redshift is excellent at caching anything it can, so it's fairly easy to iterate over local changes. To be clear, when Redshift goes out-of-core, it goes *into *your system ram. So in the end, you still need plenty of system ram for large scenes. But I've only got 20GB in my workstation and I'm getting very complex renders out of it without coming even close to that. More in the 6-8GB range so far. Proxies are awesome, and Redshift wisely had those from day one of the alpha. Now several months ago I rendered thishttp://timcrowson.smugmug.com/photos/i-qGFB7nv/1/O/i-qGFB7nv.jpg, but it's more along the lines of what Octavian posted. It's just one tree instanced a bunch of times. This rendered in about 30min at that resolution, on a GTX470. -Tim
Re: Re[2]: Redshift3D Render
yep, looks nice. I'm also wondering if redshift can use the combined vram when using several cards in sli mode, if you have 3 titans, can you use 3x6gb or just 6gb for your scene? a tree scene like that, at that rez would probably take a few hours in arnold or vray, so even at 30m in redshift it's still very fast.. still, at 30m a frame you won't exactly be able to render full shots without a farm, and once you work with even heavier scenes I imagine you are looking at rendertimes of several hours per frame, and at that point I don't think the gpu will speed anything up, quite the opposite. On Tue, Feb 18, 2014 at 11:12 AM, Eugen Sares sof...@mail.sprit.org wrote: To be clear, when Redshift goes out-of-core, it goes *into *your system ram. So in the end, you still need plenty of system ram for large scenes. But I've only got 20GB in my workstation and I'm getting very complex renders out of it without coming even close to that. More in the 6-8GB range so far. Proxies are awesome, and Redshift wisely had those from day one of the alpha. Now several months ago I rendered thishttp://timcrowson.smugmug.com/photos/i-qGFB7nv/1/O/i-qGFB7nv.jpg, but it's more along the lines of what Octavian posted. It's just one tree instanced a bunch of times. This rendered in about 30min at that resolution, on a GTX470. -Tim Very nice!! How did you light it? Sun + dome? -- http://www.avast.com/ Diese E-Mail ist frei von Viren und Malware, denn der avast! Antivirushttp://www.avast.com/Schutz ist aktiv. -- Andreas Byström Weta Digital
Re: Re[2]: Redshift3D Render
Well there are some guys that are using Redshift with Royal Render without any issues. Just started playing with Melena and Redshift. 60,000 strands with 150 subdivisions in 99 seconds from sending the scene to redshift to final. After sending the scene the frame went down to 57 seconds. The scene extraction at the beginning took 22 seconds. https://dl.dropboxusercontent.com/u/49626349/Melena_prev_04.jpg 2014-02-17 16:19 GMT-06:00 Andreas Bystrom andreas.byst...@gmail.com: yep, looks nice. I'm also wondering if redshift can use the combined vram when using several cards in sli mode, if you have 3 titans, can you use 3x6gb or just 6gb for your scene? a tree scene like that, at that rez would probably take a few hours in arnold or vray, so even at 30m in redshift it's still very fast.. still, at 30m a frame you won't exactly be able to render full shots without a farm, and once you work with even heavier scenes I imagine you are looking at rendertimes of several hours per frame, and at that point I don't think the gpu will speed anything up, quite the opposite. On Tue, Feb 18, 2014 at 11:12 AM, Eugen Sares sof...@mail.sprit.orgwrote: To be clear, when Redshift goes out-of-core, it goes *into *your system ram. So in the end, you still need plenty of system ram for large scenes. But I've only got 20GB in my workstation and I'm getting very complex renders out of it without coming even close to that. More in the 6-8GB range so far. Proxies are awesome, and Redshift wisely had those from day one of the alpha. Now several months ago I rendered thishttp://timcrowson.smugmug.com/photos/i-qGFB7nv/1/O/i-qGFB7nv.jpg, but it's more along the lines of what Octavian posted. It's just one tree instanced a bunch of times. This rendered in about 30min at that resolution, on a GTX470. -Tim Very nice!! How did you light it? Sun + dome? -- http://www.avast.com/ Diese E-Mail ist frei von Viren und Malware, denn der avast! Antivirushttp://www.avast.com/Schutz ist aktiv. -- Andreas Byström Weta Digital
Re: Re[2]: Redshift3D Render
Well Steven, for $100 bucks for the Beta maybe you can try it. Will be awsome to have some feedback on such scene as Elysium. 2014-02-17 16:28 GMT-06:00 Emilio Hernandez emi...@e-roja.com: Well there are some guys that are using Redshift with Royal Render without any issues. Just started playing with Melena and Redshift. 60,000 strands with 150 subdivisions in 99 seconds from sending the scene to redshift to final. After sending the scene the frame went down to 57 seconds. The scene extraction at the beginning took 22 seconds. https://dl.dropboxusercontent.com/u/49626349/Melena_prev_04.jpg 2014-02-17 16:19 GMT-06:00 Andreas Bystrom andreas.byst...@gmail.com: yep, looks nice. I'm also wondering if redshift can use the combined vram when using several cards in sli mode, if you have 3 titans, can you use 3x6gb or just 6gb for your scene? a tree scene like that, at that rez would probably take a few hours in arnold or vray, so even at 30m in redshift it's still very fast.. still, at 30m a frame you won't exactly be able to render full shots without a farm, and once you work with even heavier scenes I imagine you are looking at rendertimes of several hours per frame, and at that point I don't think the gpu will speed anything up, quite the opposite. On Tue, Feb 18, 2014 at 11:12 AM, Eugen Sares sof...@mail.sprit.orgwrote: To be clear, when Redshift goes out-of-core, it goes *into *your system ram. So in the end, you still need plenty of system ram for large scenes. But I've only got 20GB in my workstation and I'm getting very complex renders out of it without coming even close to that. More in the 6-8GB range so far. Proxies are awesome, and Redshift wisely had those from day one of the alpha. Now several months ago I rendered thishttp://timcrowson.smugmug.com/photos/i-qGFB7nv/1/O/i-qGFB7nv.jpg, but it's more along the lines of what Octavian posted. It's just one tree instanced a bunch of times. This rendered in about 30min at that resolution, on a GTX470. -Tim Very nice!! How did you light it? Sun + dome? -- http://www.avast.com/ Diese E-Mail ist frei von Viren und Malware, denn der avast! Antivirushttp://www.avast.com/Schutz ist aktiv. -- Andreas Byström Weta Digital
Re[2]: Redshift3D Render
just so you know... athens was a fraction of what we rendered for elysium. so... i think you guys should be aiming high... i mean REALLY high. if i don't see a billion somewhere (instancing is fine) then i am still not convinced the gpu renderer has overcome the memory limitations. but it sounds like the scene you describe is a beast... can't wait to see it! If they really managed to render a scene like this... that would be something I'd say... A 3 guys company, in such a short time, and they blow everybody out of the lake? What's the Oscar for software development? --- Diese E-Mail ist frei von Viren und Malware, denn der avast! Antivirus Schutz ist aktiv. http://www.avast.com
Re: Redshift3D Render
@ Eugen... I think I lit that with a dome light... honestly I can't remember though @ Steve... yep I hear I ya. Rest assured, we're not doing anything close to elysium, and as I said not even as big as Athens. And it's cartoony, so although it has a lot of detail, the shading is exaggerated. @ Andreas... no, Redshift cannot combine the ram from the cards. If you have 3x6GB, you have 6GB. As for mass-rendering, we've already begun to outfit our farm with more GPU boxes. We're a small operation here, so our node numbers are low, but we definitely have to be able to mass-render frames. We have a history of CPU rendering with Mental Ray, and we're used to a certain volume. For us, the question isn't really the render time per frame, but per shot. And so far, even with the heavy scenes I descibed in my previous email, I think 35min may be the longest time my single-Titan has taken at 1920x1080 full MC. With multi-GPU machines (which is not an easy thing to figure out, either), we could get through these shots in very short order. I don't know whether it's the fact that RS is on the GPU, or whether it's RS' actual techniques (probably both), but we're getting great renders in very fast times. And we're using Redshift with Royal Render perfectly well. The Redshift guys worked with us to implement some environment variables so we can pull from a central location. It works well on the farm, and is stupid easy to update. -Tim On 2/17/2014 4:19 PM, Andreas Bystrom wrote: yep, looks nice. I'm also wondering if redshift can use the combined vram when using several cards in sli mode, if you have 3 titans, can you use 3x6gb or just 6gb for your scene? a tree scene like that, at that rez would probably take a few hours in arnold or vray, so even at 30m in redshift it's still very fast.. still, at 30m a frame you won't exactly be able to render full shots without a farm, and once you work with even heavier scenes I imagine you are looking at rendertimes of several hours per frame, and at that point I don't think the gpu will speed anything up, quite the opposite. On Tue, Feb 18, 2014 at 11:12 AM, Eugen Sares sof...@mail.sprit.org mailto:sof...@mail.sprit.org wrote: To be clear, when Redshift goes out-of-core, it goes /into /your system ram. So in the end, you still need plenty of system ram for large scenes. But I've only got 20GB in my workstation and I'm getting very complex renders out of it without coming even close to that. More in the 6-8GB range so far. Proxies are awesome, and Redshift wisely had those from day one of the alpha. Now several months ago I rendered this http://timcrowson.smugmug.com/photos/i-qGFB7nv/1/O/i-qGFB7nv.jpg, but it's more along the lines of what Octavian posted. It's just one tree instanced a bunch of times. This rendered in about 30min at that resolution, on a GTX470. -Tim Very nice!! How did you light it? Sun + dome? http://www.avast.com/ Diese E-Mail ist frei von Viren und Malware, denn der avast! Antivirus http://www.avast.com/ Schutz ist aktiv. -- Andreas Byström Weta Digital -- Signature *Tim Crowson */Lead CG Artist/ *Magnetic Dreams, Inc. *2525 Lebanon Pike, Bldg C, Suite 101, Nashville, TN 37214 *Ph* 615.885.6801 | *Fax* 615.889.4768 | www.magneticdreams.com tim.crow...@magneticdreams.com /Confidentiality Notice: This email, including attachments, is confidential and should not be used by anyone who is not the original intended recipient(s). If you have received this e-mail in error please inform the sender and delete it from your mailbox or any other storage mechanism. Magnetic Dreams, Inc cannot accept liability for any statements made which are clearly the sender's own and not expressly made on behalf of Magnetic Dreams, Inc or one of its agents./
Re: Re[2]: Redshift3D Render
solid angle isn't some massive company either... On Mon, Feb 17, 2014 at 2:38 PM, Eugen Sares sof...@mail.sprit.org wrote: If they really managed to render a scene like this... that would be something I'd say... A 3 guys company, in such a short time, and they blow everybody out of the lake? What's the Oscar for software development?
Re: Re[2]: Redshift3D Render
doesn't work like that... i have to convince someone to buy it for the studio, then the graphics cards you guys talk about... 3 titans!? we don't have those types of investments. we have an existing farm with cpus and lots of ram. if i want to render a sequence with redshift... i have to render it on workstations only. also, i am not going to convert elysium to work for redshift on my free time ;) i am sold on the value of redshift, i can see the value clearly! i just wanting to see how well this 'out of core memory' feature really works. On Mon, Feb 17, 2014 at 2:30 PM, Emilio Hernandez emi...@e-roja.com wrote: Well Steven, for $100 bucks for the Beta maybe you can try it. Will be awsome to have some feedback on such scene as Elysium.
Re: Redshift3D Render
It works well, but as I said earlier, going out-of-core means you go into system ram. So you still need system ram, but you needed that anyway You don't /need /Titans. The reason we mention those is because we've decided to invest in a few heftier cards. We also have some 770s and they're kicking ass too. Not as fast as the Titans or the new 780s, but they do extremely well. Even on my old 470 RS was impressive, and it only had like 1.2GB of vram on it. There's a free trial available. Full bells and whistles plus a watermark. It's not yet ready for /every /production, but you gotta try it. -Tim On 2/17/2014 5:04 PM, Steven Caron wrote: doesn't work like that... i have to convince someone to buy it for the studio, then the graphics cards you guys talk about... 3 titans!? we don't have those types of investments. we have an existing farm with cpus and lots of ram. if i want to render a sequence with redshift... i have to render it on workstations only. also, i am not going to convert elysium to work for redshift on my free time ;) i am sold on the value of redshift, i can see the value clearly! i just wanting to see how well this 'out of core memory' feature really works. On Mon, Feb 17, 2014 at 2:30 PM, Emilio Hernandez emi...@e-roja.com mailto:emi...@e-roja.com wrote: Well Steven, for $100 bucks for the Beta maybe you can try it. Will be awsome to have some feedback on such scene as Elysium. -- Signature
Re: Re[2]: Redshift3D Render
Yeah it would involve quite an adjustment for a trad farm - rackmounts would be a little useless. I also don't think RS is quite gunning for the arnold market, its aiming for the smaller shops who don't have massive render farms. I rendered 10 million hairs on a character today, ended up using 20Gb of system ram, as the gpu was only a quadro k4000, but it rendered pretty fast still. The extraction was the longest part. Also, just as a quick test, I've instanced 5000 trees using ice scatter, each tree has 1.2 million polys, so that makes it 6 Billion(?), used 6 Gb system ram. render stats below: # INFO : [Redshift] Rendering frame 0 # INFO : [Redshift] Scene extraction time: 4.623 s # INFO : [Redshift] Rendering time: 176.518 s (2 GPU(s) used) just the usual deformed grid test, nothing exciting: https://app.box.com/s/v3z5se5fl0a8zt1sw4ps The ground texture was 20k, that took a little while to convert before the render started, but after that, I was suprised by the system ram, particularly as I couldn't even begin to view the instances in the viewport, bounding box only... On 17 February 2014 23:04, Steven Caron car...@gmail.com wrote: doesn't work like that... i have to convince someone to buy it for the studio, then the graphics cards you guys talk about... 3 titans!? we don't have those types of investments. we have an existing farm with cpus and lots of ram. if i want to render a sequence with redshift... i have to render it on workstations only. also, i am not going to convert elysium to work for redshift on my free time ;) i am sold on the value of redshift, i can see the value clearly! i just wanting to see how well this 'out of core memory' feature really works. On Mon, Feb 17, 2014 at 2:30 PM, Emilio Hernandez emi...@e-roja.comwrote: Well Steven, for $100 bucks for the Beta maybe you can try it. Will be awsome to have some feedback on such scene as Elysium. -- www.matinai.com
Re: Re[2]: Redshift3D Render
i don't think they care what market, as long as people buy it ;) On Mon, Feb 17, 2014 at 3:13 PM, Matt Morris matt...@gmail.com wrote: Yeah it would involve quite an adjustment for a trad farm - rackmounts would be a little useless. I also don't think RS is quite gunning for the arnold market, its aiming for the smaller shops who don't have massive render farms.
Re: Re[2]: Redshift3D Render
Hey Matt cool scene! 2014-02-17 17:16 GMT-06:00 Steven Caron car...@gmail.com: i don't think they care what market, as long as people buy it ;) On Mon, Feb 17, 2014 at 3:13 PM, Matt Morris matt...@gmail.com wrote: Yeah it would involve quite an adjustment for a trad farm - rackmounts would be a little useless. I also don't think RS is quite gunning for the arnold market, its aiming for the smaller shops who don't have massive render farms.
Re: Redshift3D Render
To echo Matt, it definitely involes some adjustments to the farm! That may even be a stumbling block for some, or at least a weird surprise. There's a temptation to think that all you have to do is put a GPU in your render node and you're golden. Well... that may or may not be true... depends on all kinds of things. It could be that easy, but it could also require new hardware entirely. It really depends on what kind of GPU machines you're wanting to implement, what kind of scale you're going for. There is no standard way of dealing with this, not yet at any rate. It's just too early in the life of GPU rendering. We've found that for us, it's been worth it to devote some time and money to outfit a portion of our farm for Redshift. Our stuff just looks that much better in less time. When we saw what it could do for us, it was a no-brainer to investigate our hardware a little more seriously. Again, all relative to what you're doing... We still have racks though nothing wrong with a GPU in a rack -Tim On 2/17/2014 5:13 PM, Matt Morris wrote: Yeah it would involve quite an adjustment for a trad farm - rackmounts would be a little useless. I also don't think RS is quite gunning for the arnold market, its aiming for the smaller shops who don't have massive render farms. I rendered 10 million hairs on a character today, ended up using 20Gb of system ram, as the gpu was only a quadro k4000, but it rendered pretty fast still. The extraction was the longest part. Also, just as a quick test, I've instanced 5000 trees using ice scatter, each tree has 1.2 million polys, so that makes it 6 Billion(?), used 6 Gb system ram. render stats below: # INFO : [Redshift] Rendering frame 0 # INFO : [Redshift] Scene extraction time: 4.623 s # INFO : [Redshift] Rendering time: 176.518 s (2 GPU(s) used) just the usual deformed grid test, nothing exciting: https://app.box.com/s/v3z5se5fl0a8zt1sw4ps The ground texture was 20k, that took a little while to convert before the render started, but after that, I was suprised by the system ram, particularly as I couldn't even begin to view the instances in the viewport, bounding box only... On 17 February 2014 23:04, Steven Caron car...@gmail.com mailto:car...@gmail.com wrote: doesn't work like that... i have to convince someone to buy it for the studio, then the graphics cards you guys talk about... 3 titans!? we don't have those types of investments. we have an existing farm with cpus and lots of ram. if i want to render a sequence with redshift... i have to render it on workstations only. also, i am not going to convert elysium to work for redshift on my free time ;) i am sold on the value of redshift, i can see the value clearly! i just wanting to see how well this 'out of core memory' feature really works. On Mon, Feb 17, 2014 at 2:30 PM, Emilio Hernandez emi...@e-roja.com mailto:emi...@e-roja.com wrote: Well Steven, for $100 bucks for the Beta maybe you can try it. Will be awsome to have some feedback on such scene as Elysium. -- www.matinai.com http://www.matinai.com -- Signature
Re: Re[2]: Redshift3D Render
True dat. On 17 February 2014 23:16, Steven Caron car...@gmail.com wrote: i don't think they care what market, as long as people buy it ;) On Mon, Feb 17, 2014 at 3:13 PM, Matt Morris matt...@gmail.com wrote: Yeah it would involve quite an adjustment for a trad farm - rackmounts would be a little useless. I also don't think RS is quite gunning for the arnold market, its aiming for the smaller shops who don't have massive render farms. -- www.matinai.com
Re: Redshift3D Render
Does Redshift have something like .map or .tx for textures? and what about standins/procedurals? I think those are essentials for big productions, not even thinking on massive productions, just bigger than simple scenes... If it has (or plan to implement) something like that, I would say it's a serious contender, but if not, I think it's aimming to small studios or freelancers only, which of course is not a bad thing at all, just saying it shouldn't be compared to other renderers just because, well, they are renderers... Orlando. On Mon, Feb 17, 2014 at 4:39 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: @ Eugen... I think I lit that with a dome light... honestly I can't remember though @ Steve... yep I hear I ya. Rest assured, we're not doing anything close to elysium, and as I said not even as big as Athens. And it's cartoony, so although it has a lot of detail, the shading is exaggerated. @ Andreas... no, Redshift cannot combine the ram from the cards. If you have 3x6GB, you have 6GB. As for mass-rendering, we've already begun to outfit our farm with more GPU boxes. We're a small operation here, so our node numbers are low, but we definitely have to be able to mass-render frames. We have a history of CPU rendering with Mental Ray, and we're used to a certain volume. For us, the question isn't really the render time per frame, but per shot. And so far, even with the heavy scenes I descibed in my previous email, I think 35min may be the longest time my single-Titan has taken at 1920x1080 full MC. With multi-GPU machines (which is not an easy thing to figure out, either), we could get through these shots in very short order. I don't know whether it's the fact that RS is on the GPU, or whether it's RS' actual techniques (probably both), but we're getting great renders in very fast times. And we're using Redshift with Royal Render perfectly well. The Redshift guys worked with us to implement some environment variables so we can pull from a central location. It works well on the farm, and is stupid easy to update. -Tim On 2/17/2014 4:19 PM, Andreas Bystrom wrote: yep, looks nice. I'm also wondering if redshift can use the combined vram when using several cards in sli mode, if you have 3 titans, can you use 3x6gb or just 6gb for your scene? a tree scene like that, at that rez would probably take a few hours in arnold or vray, so even at 30m in redshift it's still very fast.. still, at 30m a frame you won't exactly be able to render full shots without a farm, and once you work with even heavier scenes I imagine you are looking at rendertimes of several hours per frame, and at that point I don't think the gpu will speed anything up, quite the opposite. On Tue, Feb 18, 2014 at 11:12 AM, Eugen Sares sof...@mail.sprit.orgwrote: To be clear, when Redshift goes out-of-core, it goes *into *your system ram. So in the end, you still need plenty of system ram for large scenes. But I've only got 20GB in my workstation and I'm getting very complex renders out of it without coming even close to that. More in the 6-8GB range so far. Proxies are awesome, and Redshift wisely had those from day one of the alpha. Now several months ago I rendered thishttp://timcrowson.smugmug.com/photos/i-qGFB7nv/1/O/i-qGFB7nv.jpg, but it's more along the lines of what Octavian posted. It's just one tree instanced a bunch of times. This rendered in about 30min at that resolution, on a GTX470. -Tim Very nice!! How did you light it? Sun + dome? -- http://www.avast.com/ Diese E-Mail ist frei von Viren und Malware, denn der avast! Antivirushttp://www.avast.com/Schutz ist aktiv. -- Andreas Byström Weta Digital -- *Tim Crowson **Lead CG Artist* *Magnetic Dreams, Inc. * 2525 Lebanon Pike, Bldg C, Suite 101, Nashville, TN 37214 *Ph* 615.885.6801 | *Fax* 615.889.4768 | www.magneticdreams.com tim.crow...@magneticdreams.com *Confidentiality Notice: This email, including attachments, is confidential and should not be used by anyone who is not the original intended recipient(s). If you have received this e-mail in error please inform the sender and delete it from your mailbox or any other storage mechanism. Magnetic Dreams, Inc cannot accept liability for any statements made which are clearly the sender's own and not expressly made on behalf of Magnetic Dreams, Inc or one of its agents.* -- -- IMPRESSUM: PiXABLE STUDIOS GmbH Co.KG, Sitz: Dresden, Amtsgericht: Dresden, HRA 6857, Komplementärin: Lenhard Barth Verwaltungsgesellschaft mbH, Sitz: Dresden, Amtsgericht: Dresden, HRB 26501, Geschäftsführer: Frank Lenhard, Tino Barth IMPRINT: PiXABLE STUDIOS GmbH Co.KG, Domicile: Dresden, Court of Registery: Dresden, Company Registration Number: HRA 6857, General Partner: Lenhard Barth Verwaltungsgesellschaft mbH, Domicile: Dresden, Court of Registery: Dresden, Company Registration Number: HRB
Re: Redshift3D Render
I believer Redshift converts textures at render time to a mip-mapped format, and quite rapidly. I don't know off the top of my head what it does with .map or .tx files. Proxies were a main feature from the start, and although their options are a bit limited at the moment, the core functionality has been solid so far. -Tim On 2/17/2014 5:56 PM, Orlando Esponda wrote: Does Redshift have something like .map or .tx for textures? and what about standins/procedurals? I think those are essentials for big productions, not even thinking on massive productions, just bigger than simple scenes... If it has (or plan to implement) something like that, I would say it's a serious contender, but if not, I think it's aimming to small studios or freelancers only, which of course is not a bad thing at all, just saying it shouldn't be compared to other renderers just because, well, they are renderers... Orlando. On Mon, Feb 17, 2014 at 4:39 PM, Tim Crowson tim.crow...@magneticdreams.com mailto:tim.crow...@magneticdreams.com wrote: @ Eugen... I think I lit that with a dome light... honestly I can't remember though @ Steve... yep I hear I ya. Rest assured, we're not doing anything close to elysium, and as I said not even as big as Athens. And it's cartoony, so although it has a lot of detail, the shading is exaggerated. @ Andreas... no, Redshift cannot combine the ram from the cards. If you have 3x6GB, you have 6GB. As for mass-rendering, we've already begun to outfit our farm with more GPU boxes. We're a small operation here, so our node numbers are low, but we definitely have to be able to mass-render frames. We have a history of CPU rendering with Mental Ray, and we're used to a certain volume. For us, the question isn't really the render time per frame, but per shot. And so far, even with the heavy scenes I descibed in my previous email, I think 35min may be the longest time my single-Titan has taken at 1920x1080 full MC. With multi-GPU machines (which is not an easy thing to figure out, either), we could get through these shots in very short order. I don't know whether it's the fact that RS is on the GPU, or whether it's RS' actual techniques (probably both), but we're getting great renders in very fast times. And we're using Redshift with Royal Render perfectly well. The Redshift guys worked with us to implement some environment variables so we can pull from a central location. It works well on the farm, and is stupid easy to update. -Tim On 2/17/2014 4:19 PM, Andreas Bystrom wrote: yep, looks nice. I'm also wondering if redshift can use the combined vram when using several cards in sli mode, if you have 3 titans, can you use 3x6gb or just 6gb for your scene? a tree scene like that, at that rez would probably take a few hours in arnold or vray, so even at 30m in redshift it's still very fast.. still, at 30m a frame you won't exactly be able to render full shots without a farm, and once you work with even heavier scenes I imagine you are looking at rendertimes of several hours per frame, and at that point I don't think the gpu will speed anything up, quite the opposite. On Tue, Feb 18, 2014 at 11:12 AM, Eugen Sares sof...@mail.sprit.org mailto:sof...@mail.sprit.org wrote: To be clear, when Redshift goes out-of-core, it goes /into /your system ram. So in the end, you still need plenty of system ram for large scenes. But I've only got 20GB in my workstation and I'm getting very complex renders out of it without coming even close to that. More in the 6-8GB range so far. Proxies are awesome, and Redshift wisely had those from day one of the alpha. Now several months ago I rendered this http://timcrowson.smugmug.com/photos/i-qGFB7nv/1/O/i-qGFB7nv.jpg, but it's more along the lines of what Octavian posted. It's just one tree instanced a bunch of times. This rendered in about 30min at that resolution, on a GTX470. -Tim Very nice!! How did you light it? Sun + dome? http://www.avast.com/ Diese E-Mail ist frei von Viren und Malware, denn der avast! Antivirus http://www.avast.com/ Schutz ist aktiv. -- Andreas Byström Weta Digital -- *Tim Crowson */Lead CG Artist/ *Magnetic Dreams, Inc. * 2525 Lebanon Pike, Bldg C, Suite 101, Nashville, TN 37214 *Ph* 615.885.6801 | *Fax* 615.889.4768 | www.magneticdreams.com http://www.magneticdreams.com tim.crow...@magneticdreams.com mailto:tim.crow...@magneticdreams.com /Confidentiality Notice: This email, including attachments, is confidential and should not be used by anyone who is not the original
Re: Redshift3D Render
Just recently started a small team here, only 8 of us here, 8 workstations. Nothing fancy for now just i7 machines with 770 cards, and one big 4 titan system for now. For setup like this Redshift is great option. Adding 2nd card on all those workstations will practically double the power of our rendering. Also with projects mostly done that are more cartoon it is perfect. It eats through everything sent to it. Also last time I heard from them guys from Deadline render manager are making option to use concurrent rendering with multy GPU systems as well. There are a lot of simple projects where frames are rendered ni like 5-10 sec no matter how many cards used, really simple scenes and then having 4 cards rendering 4 frames at once is effectively making things almost 4 times faster as well. So yes depending on your setup and projects working on Redshift is jackpot, where Arnold would be a bit of a overkill. Also adding more cards to existing workstations is still more cost effective then building new render farm. On the other hand if you already have medium to big render farm in racks then it is a bit different story... On Tue, Feb 18, 2014 at 2:40 AM, Tim Crowson tim.crow...@magneticdreams.com wrote: I believer Redshift converts textures at render time to a mip-mapped format, and quite rapidly. I don't know off the top of my head what it does with .map or .tx files. Proxies were a main feature from the start, and although their options are a bit limited at the moment, the core functionality has been solid so far. -Tim On 2/17/2014 5:56 PM, Orlando Esponda wrote: Does Redshift have something like .map or .tx for textures? and what about standins/procedurals? I think those are essentials for big productions, not even thinking on massive productions, just bigger than simple scenes... If it has (or plan to implement) something like that, I would say it's a serious contender, but if not, I think it's aimming to small studios or freelancers only, which of course is not a bad thing at all, just saying it shouldn't be compared to other renderers just because, well, they are renderers... Orlando. On Mon, Feb 17, 2014 at 4:39 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: @ Eugen... I think I lit that with a dome light... honestly I can't remember though @ Steve... yep I hear I ya. Rest assured, we're not doing anything close to elysium, and as I said not even as big as Athens. And it's cartoony, so although it has a lot of detail, the shading is exaggerated. @ Andreas... no, Redshift cannot combine the ram from the cards. If you have 3x6GB, you have 6GB. As for mass-rendering, we've already begun to outfit our farm with more GPU boxes. We're a small operation here, so our node numbers are low, but we definitely have to be able to mass-render frames. We have a history of CPU rendering with Mental Ray, and we're used to a certain volume. For us, the question isn't really the render time per frame, but per shot. And so far, even with the heavy scenes I descibed in my previous email, I think 35min may be the longest time my single-Titan has taken at 1920x1080 full MC. With multi-GPU machines (which is not an easy thing to figure out, either), we could get through these shots in very short order. I don't know whether it's the fact that RS is on the GPU, or whether it's RS' actual techniques (probably both), but we're getting great renders in very fast times. And we're using Redshift with Royal Render perfectly well. The Redshift guys worked with us to implement some environment variables so we can pull from a central location. It works well on the farm, and is stupid easy to update. -Tim On 2/17/2014 4:19 PM, Andreas Bystrom wrote: yep, looks nice. I'm also wondering if redshift can use the combined vram when using several cards in sli mode, if you have 3 titans, can you use 3x6gb or just 6gb for your scene? a tree scene like that, at that rez would probably take a few hours in arnold or vray, so even at 30m in redshift it's still very fast.. still, at 30m a frame you won't exactly be able to render full shots without a farm, and once you work with even heavier scenes I imagine you are looking at rendertimes of several hours per frame, and at that point I don't think the gpu will speed anything up, quite the opposite. On Tue, Feb 18, 2014 at 11:12 AM, Eugen Sares sof...@mail.sprit.orgwrote: To be clear, when Redshift goes out-of-core, it goes *into *your system ram. So in the end, you still need plenty of system ram for large scenes. But I've only got 20GB in my workstation and I'm getting very complex renders out of it without coming even close to that. More in the 6-8GB range so far. Proxies are awesome, and Redshift wisely had those from day one of the alpha. Now several months ago I rendered thishttp://timcrowson.smugmug.com/photos/i-qGFB7nv/1/O/i-qGFB7nv.jpg, but it's more along the lines of what Octavian
Re: Redshift3D Render
Latest release of redshift with hair. https://dl.dropboxusercontent.com/u/49626349/hair_01.jpg 2014-02-16 15:24 GMT-06:00 Emilio Hernandez emi...@e-roja.com: Redshift has hair now! 2014-01-10 1:10 GMT-06:00 Mirko Jankovic mirkoj.anima...@gmail.com: just played a bit with overclocking... now I pushed PCI GEN in bios back to auto instead of GEN 3 so it is now PCI 2.0 x16 instead of 3.. That alone gave me slightly higher score on cinebench.. and after overclocking CPU to 4.5 now new result in cinebench for titan is 100, and script in softimage gives me ~200 fps. Oh well... On Thu, Jan 9, 2014 at 8:24 PM, Mirko Jankovic mirkoj.anima...@gmail.com wrote: yes fps is score. hmm maybe if I got some time I could get one of titans and test it out on maximus MBO to see if that is the diference... On Thu, Jan 9, 2014 at 8:09 PM, Emilio Hernandez emi...@e-roja.comwrote: Dumb question here. Where do I see the score? I only get a fps result for OpenGl and a 759cb in the CPU results in Cinebench. and I get 75.78fps at the OpenGl test. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com cinebench you mean? so what is score? :) On Thu, Jan 9, 2014 at 7:47 PM, Emilio Hernandez emi...@e-roja.comwrote: Now I am starting to scratch my head to figure out the maze of Nvidia... At least my benchmark said the Titatn beated up a Quadro 4000K ha ha 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com getting around 95 score in cinebench with titan, and again 780 in home comp getting 140 score in cinebench... wondering how can 780 crash titan so much in opengl. titan system beside 4 titans have i7 3930K, asus p9x79-e ws MBO and home comp i7 4770k on asus maximus VI hero Will test some redshift rendering later to compare single GPU 780 vs titan On Thu, Jan 9, 2014 at 7:11 PM, Emilio Hernandez emi...@e-roja.comwrote: Yes specailly if you are getting more than the double speed. I ran the test and I got 75.78 fps Thx 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com yea I guess that woul dbe the case but just tried.. strange really no idea.. any chance to run cinebench 15 opengl test then? if that gives too low result as well then something is not good On Thu, Jan 9, 2014 at 6:58 PM, Emilio Hernandez emi...@e-roja.com wrote: It is set to ALL. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com hmm another gues.. is it set to RT or ALL in play? On Thu, Jan 9, 2014 at 6:53 PM, Emilio Hernandez emi...@e-roja.com wrote: I already did that and still getting the 65 fps limit with the Titan. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com just got back , yes 60 is vsync on, turn of vsync in nvida control panel On Thu, Jan 9, 2014 at 5:52 PM, Stefan Kubicek s...@tidbit-images.com wrote: I guess 60fps is the refresh rate of your display, right? Have you disabled VSync in the driver settings? I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.com wrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000,
Re: Redshift3D Render
Yeah Emilio ! definitely starting to look interesting On 16 February 2014 22:40, Emilio Hernandez emi...@e-roja.com wrote: Latest release of redshift with hair. https://dl.dropboxusercontent.com/u/49626349/hair_01.jpg 2014-02-16 15:24 GMT-06:00 Emilio Hernandez emi...@e-roja.com: Redshift has hair now! 2014-01-10 1:10 GMT-06:00 Mirko Jankovic mirkoj.anima...@gmail.com: just played a bit with overclocking... now I pushed PCI GEN in bios back to auto instead of GEN 3 so it is now PCI 2.0 x16 instead of 3.. That alone gave me slightly higher score on cinebench.. and after overclocking CPU to 4.5 now new result in cinebench for titan is 100, and script in softimage gives me ~200 fps. Oh well... On Thu, Jan 9, 2014 at 8:24 PM, Mirko Jankovic mirkoj.anima...@gmail.com wrote: yes fps is score. hmm maybe if I got some time I could get one of titans and test it out on maximus MBO to see if that is the diference... On Thu, Jan 9, 2014 at 8:09 PM, Emilio Hernandez emi...@e-roja.comwrote: Dumb question here. Where do I see the score? I only get a fps result for OpenGl and a 759cb in the CPU results in Cinebench. and I get 75.78fps at the OpenGl test. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com cinebench you mean? so what is score? :) On Thu, Jan 9, 2014 at 7:47 PM, Emilio Hernandez emi...@e-roja.comwrote: Now I am starting to scratch my head to figure out the maze of Nvidia... At least my benchmark said the Titatn beated up a Quadro 4000K ha ha 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com getting around 95 score in cinebench with titan, and again 780 in home comp getting 140 score in cinebench... wondering how can 780 crash titan so much in opengl. titan system beside 4 titans have i7 3930K, asus p9x79-e ws MBO and home comp i7 4770k on asus maximus VI hero Will test some redshift rendering later to compare single GPU 780 vs titan On Thu, Jan 9, 2014 at 7:11 PM, Emilio Hernandez emi...@e-roja.com wrote: Yes specailly if you are getting more than the double speed. I ran the test and I got 75.78 fps Thx 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com yea I guess that woul dbe the case but just tried.. strange really no idea.. any chance to run cinebench 15 opengl test then? if that gives too low result as well then something is not good On Thu, Jan 9, 2014 at 6:58 PM, Emilio Hernandez emi...@e-roja.com wrote: It is set to ALL. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com hmm another gues.. is it set to RT or ALL in play? On Thu, Jan 9, 2014 at 6:53 PM, Emilio Hernandez emi...@e-roja.com wrote: I already did that and still getting the 65 fps limit with the Titan. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com just got back , yes 60 is vsync on, turn of vsync in nvida control panel On Thu, Jan 9, 2014 at 5:52 PM, Stefan Kubicek s...@tidbit-images.com wrote: I guess 60fps is the refresh rate of your display, right? Have you disabled VSync in the driver settings? I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.com wrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , ,
Re: Redshift3D Render
ice strands or just xsi hair? On Sun, Feb 16, 2014 at 1:40 PM, Emilio Hernandez emi...@e-roja.com wrote: Latest release of redshift with hair. https://dl.dropboxusercontent.com/u/49626349/hair_01.jpg
Re: Redshift3D Render
This is just XSI hair, but now it supports also strands. 2014-02-16 18:28 GMT-06:00 Steven Caron car...@gmail.com: ice strands or just xsi hair? On Sun, Feb 16, 2014 at 1:40 PM, Emilio Hernandez emi...@e-roja.comwrote: Latest release of redshift with hair. https://dl.dropboxusercontent.com/u/49626349/hair_01.jpg
Re: Redshift3D Render
Just changed an old scene from Arnold to Redshift. I am impressed with the performance. They are still working on the hair shader, this one is with the Redshift arch shader. Judge by yourselfves. https://vimeo.com/86856132 2014-02-16 18:33 GMT-06:00 Emilio Hernandez emi...@e-roja.com: This is just XSI hair, but now it supports also strands. 2014-02-16 18:28 GMT-06:00 Steven Caron car...@gmail.com: ice strands or just xsi hair? On Sun, Feb 16, 2014 at 1:40 PM, Emilio Hernandez emi...@e-roja.comwrote: Latest release of redshift with hair. https://dl.dropboxusercontent.com/u/49626349/hair_01.jpg
RE: Redshift3D Render
I believe the expression is *Vrooom* That is incredibly fast From: Emilio Hernandez [emi...@e-roja.com] Sent: 17 February 2014 02:42 AM To: softimage@listproc.autodesk.com Subject: Re: Redshift3D Render Just changed an old scene from Arnold to Redshift. I am impressed with the performance. They are still working on the hair shader, this one is with the Redshift arch shader. Judge by yourselfves. https://vimeo.com/86856132 [http://img694.imageshack.us/img694/8965/erojamailpleca.jpg] 2014-02-16 18:33 GMT-06:00 Emilio Hernandez emi...@e-roja.commailto:emi...@e-roja.com: This is just XSI hair, but now it supports also strands. [http://img694.imageshack.us/img694/8965/erojamailpleca.jpg] 2014-02-16 18:28 GMT-06:00 Steven Caron car...@gmail.commailto:car...@gmail.com: ice strands or just xsi hair? On Sun, Feb 16, 2014 at 1:40 PM, Emilio Hernandez emi...@e-roja.commailto:emi...@e-roja.com wrote: Latest release of redshift with hair. https://dl.dropboxusercontent.com/u/49626349/hair_01.jpg table width=100% border=0 cellspacing=0 cellpadding=0 style=width:100%; tr td align=left style=text-align:justify;font face=arial,sans-serif size=1 color=#99span style=font-size:11px;This communication is intended for the addressee only. It is confidential. If you have received this communication in error, please notify us immediately and destroy the original message. You may not copy or disseminate this communication without the permission of the University. Only authorised signatories are competent to enter into agreements on behalf of the University and recipients are thus advised that the content of this message may not be legally binding on the University and may contain the personal views and opinions of the author, which are not necessarily the views and opinions of The University of the Witwatersrand, Johannesburg. All agreements between the University and outsiders are subject to South African Law unless the University agrees in writing to the contrary. /span/font/td /tr /table
Re: Redshift3D Render
Well I finally got nice reason to get 4 Titans.. mmm 4way SLI and 3 27 mon stereo gaming. Soo nice to have same comp good for both job and gaming. On Thu, Jan 9, 2014 at 3:49 AM, Sebastien Sterling sebastien.sterl...@gmail.com wrote: Call me immature, but i kind of love the idea of my gaming pc also being my render farm :P Sod of Nvidia and your flaby Quadro cards there expensive as fuck and you can't play Crysis on them :) On 8 January 2014 23:13, Tim Leydecker bauero...@gmx.de wrote: For a current list of features available as well as a roadmap, I would like to suggest to just go and give it a free try: https://www.redshift3d.com/get-redshift Yes, actually you don´t even have to commit to spending $100 directly, the Free Beta Trial gives you 30 days of full access to Redshift. A special benefit of this free trial option is that you could actually try out how a bunch of machines would run using redshift in a farm or knot. Reading the docs doesn´t require a login: http://docs.redshift3d.com/Default.html Redshift is a really well balanced renderer and I wholehartedly trust in it´s success. With the above opportunity available it is a good time to test it in your production scenario and wheight it against VRay and Arnold, which are also both very nice plattforms with enough momentum to also be around for quite a while. I am sure Redshift is a valuable addition to that arsenal. Cheers, tim On 08.01.2014 22:44, Emilio Hernandez wrote: Backing up Tim, in the forums there is actually a hair test in Softimage with a simple phong shader. And IMHO it looks nice. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com mailto: tim.crow...@magneticdreams.com They've stated pretty clearly that Hair and Strand support is the next big thing to come... shouldn't be too long now... -Tim On 1/8/2014 3:36 PM, Daniel Kim wrote: I found some weird result of displacement map with Redshift. Elevation is okay, but sometimes I could see weird connection of UVM. All UVM boders wasn't smooth and I had no idea how to fix it. Arnold has that option though : / But more option I need is... hair @__@ and ICE strand --- Daniel Kim Animation Director Professional 3D Generalist http://www.danielkim3d.com --- On Thu, Jan 9, 2014 at 10:31 AM, Emilio Hernandez emi...@e-roja.com mailto:emi...@e-roja.com wrote: Displacement and bump map are there and they work beautiful. They event implement a scalar change range into the displacement node. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com mailto: tim.crow...@magneticdreams.com I'll repost what I said in the other thread We started using Redshift back in March and pretty much use it exclusively now. Of course it all depends on the needs of the project (and there are still some real limitations). The RS dev team is top notch though. I'm really excited to see how things will be at the end of this calendar year. Redshift development is progressing at a fantastic rate, and the pricing is very competitive. For facilities, even small ones, it does require that you spend some time considering your hardware and infrastructure, especially if you want to start converting CPU farms for GPU rendering, or augmenting them. Fortunately, Redshift isn't licensed per GPU, but per machine, and that should provide some breathing room. To be honest (and I realize we have many Arnold folks here), here at Magnetic we evaluated our rendering options (MR, vRay, 3Delight, Arnold). I even started working on a Soft-to-Modo pipeline. Among these Arnold was the clear winner. That said, we felt that to be useful for us in production, Arnold was too costly a solution for us, both financially and in render time, /considering the kinds of projects we do/ . Then Redshift came along and despite its infancy, really turned our heads. We cautiously began using it on productions, and it has since proven itself for us. Again, it all depends on what kind of project you're working on! You need to evaluate it for yourself of course, but for smaller houses like us, it allows us to produce better looking content faster, while staying in Softimage. And in this economy, we can't argue with that. -Tim On 1/8/2014 2:53 PM, Paul Griswold wrote: I've been using it along side Arnold for quite a while now. I just finished a project for CES entirely in Redshift. I think Redshift falls more into the category of a VRay competitor rather than Arnold. Redshift isn't open the way Arnold is I don't think they intend it to
Re: Redshift3D Render
quick one, can your rendermap with redshift? On Thu, Jan 9, 2014 at 7:39 PM, Mirko Jankovic mirkoj.anima...@gmail.comwrote: Well I finally got nice reason to get 4 Titans.. mmm 4way SLI and 3 27 mon stereo gaming. Soo nice to have same comp good for both job and gaming. On Thu, Jan 9, 2014 at 3:49 AM, Sebastien Sterling sebastien.sterl...@gmail.com wrote: Call me immature, but i kind of love the idea of my gaming pc also being my render farm :P Sod of Nvidia and your flaby Quadro cards there expensive as fuck and you can't play Crysis on them :) On 8 January 2014 23:13, Tim Leydecker bauero...@gmx.de wrote: For a current list of features available as well as a roadmap, I would like to suggest to just go and give it a free try: https://www.redshift3d.com/get-redshift Yes, actually you don´t even have to commit to spending $100 directly, the Free Beta Trial gives you 30 days of full access to Redshift. A special benefit of this free trial option is that you could actually try out how a bunch of machines would run using redshift in a farm or knot. Reading the docs doesn´t require a login: http://docs.redshift3d.com/Default.html Redshift is a really well balanced renderer and I wholehartedly trust in it´s success. With the above opportunity available it is a good time to test it in your production scenario and wheight it against VRay and Arnold, which are also both very nice plattforms with enough momentum to also be around for quite a while. I am sure Redshift is a valuable addition to that arsenal. Cheers, tim On 08.01.2014 22:44, Emilio Hernandez wrote: Backing up Tim, in the forums there is actually a hair test in Softimage with a simple phong shader. And IMHO it looks nice. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com mailto: tim.crow...@magneticdreams.com They've stated pretty clearly that Hair and Strand support is the next big thing to come... shouldn't be too long now... -Tim On 1/8/2014 3:36 PM, Daniel Kim wrote: I found some weird result of displacement map with Redshift. Elevation is okay, but sometimes I could see weird connection of UVM. All UVM boders wasn't smooth and I had no idea how to fix it. Arnold has that option though : / But more option I need is... hair @__@ and ICE strand --- Daniel Kim Animation Director Professional 3D Generalist http://www.danielkim3d.com --- On Thu, Jan 9, 2014 at 10:31 AM, Emilio Hernandez emi...@e-roja.com mailto:emi...@e-roja.com wrote: Displacement and bump map are there and they work beautiful. They event implement a scalar change range into the displacement node. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com mailto: tim.crow...@magneticdreams.com I'll repost what I said in the other thread We started using Redshift back in March and pretty much use it exclusively now. Of course it all depends on the needs of the project (and there are still some real limitations). The RS dev team is top notch though. I'm really excited to see how things will be at the end of this calendar year. Redshift development is progressing at a fantastic rate, and the pricing is very competitive. For facilities, even small ones, it does require that you spend some time considering your hardware and infrastructure, especially if you want to start converting CPU farms for GPU rendering, or augmenting them. Fortunately, Redshift isn't licensed per GPU, but per machine, and that should provide some breathing room. To be honest (and I realize we have many Arnold folks here), here at Magnetic we evaluated our rendering options (MR, vRay, 3Delight, Arnold). I even started working on a Soft-to-Modo pipeline. Among these Arnold was the clear winner. That said, we felt that to be useful for us in production, Arnold was too costly a solution for us, both financially and in render time, /considering the kinds of projects we do/ . Then Redshift came along and despite its infancy, really turned our heads. We cautiously began using it on productions, and it has since proven itself for us. Again, it all depends on what kind of project you're working on! You need to evaluate it for yourself of course, but for smaller houses like us, it allows us to produce better looking content faster, while staying in Softimage. And in this economy, we can't argue with that. -Tim On 1/8/2014 2:53 PM, Paul Griswold wrote: I've been using it along side Arnold for quite a while now. I just finished a project for CES entirely in Redshift. I think Redshift falls
Re: Redshift3D Render
Afaik it's in development just like hair fur.quick one, can your rendermap with redshift?On Thu, Jan 9, 2014 at 7:39 PM, Mirko Jankovic mirkoj.anima...@gmail.com wrote: Well I finally got nice reason to get 4 Titans.. mmm 4way SLI and 3 27" mon stereo gaming.Soo nice to have same comp good for both job and gaming. On Thu, Jan 9, 2014 at 3:49 AM, Sebastien Sterling sebastien.sterl...@gmail.com wrote: Call me immature, but i kind of love the idea of my gaming pc also being my render farm :PSod of Nvidia and your flaby Quadro cards there expensive as fuck and you can't play Crysis on them :) On 8 January 2014 23:13, Tim Leydecker bauero...@gmx.de wrote: For a current list of features available as well as a roadmap, I would like to suggest to just go and give it a free try: https://www.redshift3d.com/get-redshift Yes, actually you don´t even have to commit to spending $100 directly, the Free Beta Trial gives you 30 days of full access to Redshift. A special benefit of this free trial option is that you could actually try out how a bunch of machines would run using redshift in a farm or knot. Reading the docs doesn´t require a login: http://docs.redshift3d.com/Default.html Redshift is a really well balanced renderer and I wholehartedly trust in it´s success. With the above opportunity available it is a good time to test it in your production scenario and wheight it against VRay and Arnold, which are also both very nice plattforms with enough momentum to also be around for quite a while. I am sure Redshift is a valuable addition to that arsenal. Cheers, tim On 08.01.2014 22:44, Emilio Hernandez wrote: Backing up Tim, in the forums there is actually a hair test in Softimage with a simple phong shader. And IMHO it looks nice. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com mailto:tim.crow...@magneticdreams.com They've stated pretty clearly that Hair and Strand support is the next big thing to come... shouldn't be too long now... -Tim On 1/8/2014 3:36 PM, Daniel Kim wrote: I found some weird result of displacement map with Redshift. Elevation is okay, but sometimes I could see weird connection of UVM. All UVM boders wasn't smooth and I had no idea how to fix it. Arnold has that option though : / But more option I need is... hair @__@ and ICE strand --- Daniel Kim Animation Director Professional 3D Generalist http://www.danielkim3d.com --- On Thu, Jan 9, 2014 at 10:31 AM, Emilio Hernandez emi...@e-roja.com mailto:emi...@e-roja.com wrote: Displacement and bump map are there and they work beautiful. They event implement a scalar change range into the displacement node. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com mailto:tim.crow...@magneticdreams.com I'll repost what I said in the other thread We started using Redshift back in March and pretty much use it exclusively now. Of course it all depends on the needs of the project (and there are still some real limitations). The RS dev team is top notch though. I'm really excited to see how things will be at the end of this calendar year. Redshift development is progressing at a fantastic rate, and the pricing is very competitive. For facilities, even small ones, it does require that you spend some time considering your hardware and infrastructure, especially if you want to start converting CPU farms for GPU rendering, or augmenting them. Fortunately, Redshift isn't licensed per GPU, but per machine, and that should provide some breathing room. To be honest (and I realize we have many Arnold folks here), here at Magnetic we evaluated our rendering options (MR, vRay, 3Delight, Arnold). I even started working on a Soft-to-Modo pipeline. Among these Arnold was the clear winner. That said, we felt that to be useful for us in production, Arnold was too costly a solution for us, both financially and in render time, /considering the kinds of projects we do/ . Then Redshift came along and despite its infancy, really turned our heads. We cautiously began using it on productions, and it has since proven itself for us. Again, it all depends on what kind of project you're working on! You need to evaluate it for yourself of course, but for smaller houses like us, it allows us to produce better looking content faster, while staying in Softimage. And in this economy, we can't argue with that. -Tim On 1/8/2014 2:53 PM, Paul Griswold wrote: I've been using it along side Arnold for quite a while now. I just finished a project for CES entirely in Redshift. I think Redshift falls more into the category of a VRay competitor rather than Arnold. Redshift isn't open the way Arnold is I don't think they intend it to be. I've found it to be extremely fast and
Re: Redshift3D Render
I just added 2 780ti's to my machine use my 680 just to drive my displays. One of the negatives right now with Redshift (AFAIK) is that it's not entirely optimized for more than 2 cards. Once you get 3-4 cards in place you start seeing diminishing returns. But that might be an nVidia problem, not a Redshift problem. I think the main thing you have to consider is - do you already have a renderfarm that is CPU based and are you willing to migrate to GPU-based rendering? I have rack-mounted machines that just have Intel motheboard graphics no space for a GPU, so for now Redshift is limited to just 2 machines. For larger companies there's a space heat issue as well. Running just the 2 780ti's all night long kept my office nice and toasty warm. You need to have a very large power supply as well as a case motherboard that can fit the cards. It'd be very interesting to see a comparison of space, heat and watts consumed for CPU's vs GPU's for rendering... -Paul ᐧ On Thu, Jan 9, 2014 at 3:39 AM, Mirko Jankovic mirkoj.anima...@gmail.comwrote: Well I finally got nice reason to get 4 Titans.. mmm 4way SLI and 3 27 mon stereo gaming. Soo nice to have same comp good for both job and gaming. On Thu, Jan 9, 2014 at 3:49 AM, Sebastien Sterling sebastien.sterl...@gmail.com wrote: Call me immature, but i kind of love the idea of my gaming pc also being my render farm :P Sod of Nvidia and your flaby Quadro cards there expensive as fuck and you can't play Crysis on them :) On 8 January 2014 23:13, Tim Leydecker bauero...@gmx.de wrote: For a current list of features available as well as a roadmap, I would like to suggest to just go and give it a free try: https://www.redshift3d.com/get-redshift Yes, actually you don´t even have to commit to spending $100 directly, the Free Beta Trial gives you 30 days of full access to Redshift. A special benefit of this free trial option is that you could actually try out how a bunch of machines would run using redshift in a farm or knot. Reading the docs doesn´t require a login: http://docs.redshift3d.com/Default.html Redshift is a really well balanced renderer and I wholehartedly trust in it´s success. With the above opportunity available it is a good time to test it in your production scenario and wheight it against VRay and Arnold, which are also both very nice plattforms with enough momentum to also be around for quite a while. I am sure Redshift is a valuable addition to that arsenal. Cheers, tim On 08.01.2014 22:44, Emilio Hernandez wrote: Backing up Tim, in the forums there is actually a hair test in Softimage with a simple phong shader. And IMHO it looks nice. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com mailto: tim.crow...@magneticdreams.com They've stated pretty clearly that Hair and Strand support is the next big thing to come... shouldn't be too long now... -Tim On 1/8/2014 3:36 PM, Daniel Kim wrote: I found some weird result of displacement map with Redshift. Elevation is okay, but sometimes I could see weird connection of UVM. All UVM boders wasn't smooth and I had no idea how to fix it. Arnold has that option though : / But more option I need is... hair @__@ and ICE strand --- Daniel Kim Animation Director Professional 3D Generalist http://www.danielkim3d.com --- On Thu, Jan 9, 2014 at 10:31 AM, Emilio Hernandez emi...@e-roja.com mailto:emi...@e-roja.com wrote: Displacement and bump map are there and they work beautiful. They event implement a scalar change range into the displacement node. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com mailto: tim.crow...@magneticdreams.com I'll repost what I said in the other thread We started using Redshift back in March and pretty much use it exclusively now. Of course it all depends on the needs of the project (and there are still some real limitations). The RS dev team is top notch though. I'm really excited to see how things will be at the end of this calendar year. Redshift development is progressing at a fantastic rate, and the pricing is very competitive. For facilities, even small ones, it does require that you spend some time considering your hardware and infrastructure, especially if you want to start converting CPU farms for GPU rendering, or augmenting them. Fortunately, Redshift isn't licensed per GPU, but per machine, and that should provide some breathing room. To be honest (and I realize we have many Arnold folks here), here at Magnetic we evaluated our rendering options (MR, vRay, 3Delight, Arnold). I even started working on a Soft-to-Modo pipeline. Among these Arnold was
Re: Redshift3D Render
how diminishing are we talking? some enthusiast motherboards support upto 6 cards, also does pci-e slot speed make a difference? On Thu, Jan 9, 2014 at 10:40 PM, Paul Griswold pgrisw...@fusiondigitalproductions.com wrote: I just added 2 780ti's to my machine use my 680 just to drive my displays. One of the negatives right now with Redshift (AFAIK) is that it's not entirely optimized for more than 2 cards. Once you get 3-4 cards in place you start seeing diminishing returns. But that might be an nVidia problem, not a Redshift problem. I think the main thing you have to consider is - do you already have a renderfarm that is CPU based and are you willing to migrate to GPU-based rendering? I have rack-mounted machines that just have Intel motheboard graphics no space for a GPU, so for now Redshift is limited to just 2 machines. For larger companies there's a space heat issue as well. Running just the 2 780ti's all night long kept my office nice and toasty warm. You need to have a very large power supply as well as a case motherboard that can fit the cards. It'd be very interesting to see a comparison of space, heat and watts consumed for CPU's vs GPU's for rendering... -Paul ᐧ On Thu, Jan 9, 2014 at 3:39 AM, Mirko Jankovic mirkoj.anima...@gmail.comwrote: Well I finally got nice reason to get 4 Titans.. mmm 4way SLI and 3 27 mon stereo gaming. Soo nice to have same comp good for both job and gaming. On Thu, Jan 9, 2014 at 3:49 AM, Sebastien Sterling sebastien.sterl...@gmail.com wrote: Call me immature, but i kind of love the idea of my gaming pc also being my render farm :P Sod of Nvidia and your flaby Quadro cards there expensive as fuck and you can't play Crysis on them :) On 8 January 2014 23:13, Tim Leydecker bauero...@gmx.de wrote: For a current list of features available as well as a roadmap, I would like to suggest to just go and give it a free try: https://www.redshift3d.com/get-redshift Yes, actually you don´t even have to commit to spending $100 directly, the Free Beta Trial gives you 30 days of full access to Redshift. A special benefit of this free trial option is that you could actually try out how a bunch of machines would run using redshift in a farm or knot. Reading the docs doesn´t require a login: http://docs.redshift3d.com/Default.html Redshift is a really well balanced renderer and I wholehartedly trust in it´s success. With the above opportunity available it is a good time to test it in your production scenario and wheight it against VRay and Arnold, which are also both very nice plattforms with enough momentum to also be around for quite a while. I am sure Redshift is a valuable addition to that arsenal. Cheers, tim On 08.01.2014 22:44, Emilio Hernandez wrote: Backing up Tim, in the forums there is actually a hair test in Softimage with a simple phong shader. And IMHO it looks nice. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com mailto: tim.crow...@magneticdreams.com They've stated pretty clearly that Hair and Strand support is the next big thing to come... shouldn't be too long now... -Tim On 1/8/2014 3:36 PM, Daniel Kim wrote: I found some weird result of displacement map with Redshift. Elevation is okay, but sometimes I could see weird connection of UVM. All UVM boders wasn't smooth and I had no idea how to fix it. Arnold has that option though : / But more option I need is... hair @__@ and ICE strand --- Daniel Kim Animation Director Professional 3D Generalist http://www.danielkim3d.com --- On Thu, Jan 9, 2014 at 10:31 AM, Emilio Hernandez emi...@e-roja.com mailto:emi...@e-roja.com wrote: Displacement and bump map are there and they work beautiful. They event implement a scalar change range into the displacement node. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com mailto: tim.crow...@magneticdreams.com I'll repost what I said in the other thread We started using Redshift back in March and pretty much use it exclusively now. Of course it all depends on the needs of the project (and there are still some real limitations). The RS dev team is top notch though. I'm really excited to see how things will be at the end of this calendar year. Redshift development is progressing at a fantastic rate, and the pricing is very competitive. For facilities, even small ones, it does require that you spend some time considering your hardware and infrastructure, especially if you want to start converting CPU farms for GPU rendering, or augmenting them. Fortunately, Redshift isn't licensed per GPU, but per machine, and that should provide some breathing room.
Re: Redshift3D Render
There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not rendering. Redshift has a proxy system like Vray Arnold, but you have to manually create proxies per object my scene had hundreds and hundreds of objects, so I didn't have time to create them. Therefore, it was creating a renderable mesh per frame - so on a frame that took 28 seconds to render, 20 seconds was spent exporting the mesh and 8 seconds were spent on rendering. But again, it's a beta and they're continuing to improve things like the proxy system. Once I'm caught up I'm hoping to try rendering the classroom scene and see how it does. -Paul ᐧ
Re: Redshift3D Render
Here are some details from testing I did a while ago: On Thu, Jan 9, 2014 at 1:11 PM, Paul Griswold pgrisw...@fusiondigitalproductions.com wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not rendering. Redshift has a proxy system like Vray Arnold, but you have to manually create proxies per object my scene had hundreds and hundreds of objects, so I didn't have time to create them. Therefore, it was creating a renderable mesh per frame - so on a frame that took 28 seconds to render, 20 seconds was spent exporting the mesh and 8 seconds were spent on rendering. But again, it's a beta and they're continuing to improve things like the proxy system. Once I'm caught up I'm hoping to try rendering the classroom scene and see how it does. -Paul ᐧ
Re: Redshift3D Render
Daniel, You mentioned using it in conjunction with Arnoldwhat way would you use both together...as in render some elements in Arnold and Redshift? On Thu, Jan 9, 2014 at 8:26 AM, Mirko Jankovic mirkoj.anima...@gmail.comwrote: Here are some details from testing I did a while ago: On Thu, Jan 9, 2014 at 1:11 PM, Paul Griswold pgrisw...@fusiondigitalproductions.com wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not rendering. Redshift has a proxy system like Vray Arnold, but you have to manually create proxies per object my scene had hundreds and hundreds of objects, so I didn't have time to create them. Therefore, it was creating a renderable mesh per frame - so on a frame that took 28 seconds to render, 20 seconds was spent exporting the mesh and 8 seconds were spent on rendering. But again, it's a beta and they're continuing to improve things like the proxy system. Once I'm caught up I'm hoping to try rendering the classroom scene and see how it does. -Paul ᐧ
Re: Redshift3D Render
We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it at ~15fps in Shaded mode. That ain't shabby! -Tim On 1/9/2014 6:11 AM, Paul Griswold wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not rendering. Redshift has a proxy system like Vray Arnold, but you have to manually create proxies per object my scene had hundreds and hundreds of objects, so I didn't have time to create them. Therefore, it was creating a renderable mesh per frame - so on a frame that took 28 seconds to render, 20 seconds was spent exporting the mesh and 8 seconds were spent on rendering. But again, it's a beta and they're continuing to improve things like the proxy system. Once I'm caught up I'm hoping to try rendering the classroom scene and see how it does. -Paul ᐧ -- Signature
Re: Redshift3D Render
Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.comwrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it at ~15fps in Shaded mode. That ain't shabby! -Tim On 1/9/2014 6:11 AM, Paul Griswold wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not rendering. Redshift has a proxy system like Vray Arnold, but you have to manually create proxies per object my scene had hundreds and hundreds of objects, so I didn't have time to create them. Therefore, it was creating a renderable mesh per frame - so on a frame that took 28 seconds to render, 20 seconds was spent exporting the mesh and 8 seconds were spent on rendering. But again, it's a beta and they're continuing to improve things like the proxy system. Once I'm caught up I'm hoping to try rendering the classroom scene and see how it does. -Paul ᐧ --
Re: Redshift3D Render
Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it at ~15fps in Shaded mode. That ain't shabby! -Tim On 1/9/2014 6:11 AM, Paul Griswold wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not rendering. Redshift has a proxy system like Vray Arnold, but you have to manually create proxies per object my scene had hundreds and hundreds of objects, so I didn't have time to create them. Therefore, it was creating a renderable mesh per frame - so on a frame that took 28 seconds to render, 20 seconds was spent exporting the mesh and 8 seconds were spent on rendering. But again, it's a beta and they're continuing to improve things like the proxy system. Once I'm caught up I'm hoping to try rendering the classroom scene and see how it does. -Paul ᐧ --
Re: Redshift3D Render
I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.com mailto:emi...@e-roja.com wrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com mailto:mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure :) But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com mailto:tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it at ~15fps in Shaded mode. That ain't shabby! -Tim On 1/9/2014 6:11 AM, Paul Griswold wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not rendering. Redshift has a proxy system like Vray Arnold, but you have to manually create proxies per object my scene had hundreds and hundreds of objects, so I didn't have time to create them. Therefore, it was creating a renderable mesh per frame - so on a frame that took 28 seconds to render, 20 seconds was spent exporting the mesh and 8 seconds were
Re: Redshift3D Render
As far as I remember from watching the specs in order to high a better rate the magic number is the memory speed transfer. I was going to buy a 680 but then realized that my 470 has more speed in the memory transfer. Don't know why but Nvidia choked the speed in most of the 600 series except for the Titan. And open again the hoose in the 690. In my experience and from what we have discussing in the Redshift forums the two magic numbers to see which GPU will perform faster is not only the Fill triangles per sec number but also the memory transfer speed. Below are the specs of the Titan, the 680 and the 470 I have. *GTX TITAN GPU Engine Specs:* 2688CUDA Cores 837Base Clock (MHz) 876Boost Clock (MHz) 187.5Texture Fill Rate (billion/sec) *GTX TITAN Memory Specs:* 6.0 GbpsMemory Clock 6144 MBStandard Memory Config GDDR5Memory Interface 384-bit GDDR5Memory Interface Width 288.4Memory Bandwidth (GB/sec) *GTX 680 GPU Engine Specs:* CUDA Cores 1006Base Clock (MHz) 1058Boost Clock (MHz) 128.8Texture Fill Rate (billion/sec) *GTX 680 Memory Specs:* 6.0 GbpsMemory Speed 2048MBStandard Memory Config 256-bit GDDR5Memory Interface Width 192.2Memory Bandwidth (GB/sec) GTX 470 GPU Engine Specs CUDA Cores 448 Graphics Clock (MHz) 607 MHz Processor Clock (MHz) 1215 MHz Texture Fill Rate (billion/sec) 34.0 Memory Specs Memory Clock 1674 MHz (3348 data rate) Standard Memory Config 1280 MB Memory Interface GDDR5 Memory Interface Width 320-bit Memory Bandwidth (GB/sec) 133.9 GB/sec So as you can see my 470 MB (Memory Bandwith) is not that far from the 680 but the Memory Clock speed, and the Processor Clock speed, of the 470 is higher than the 680 Not a guru here but that can be an explanation of why the 470 has faster fps than the 680... For rendering you make sure you compare the Memory Bandwith of the GPU. Beside the CUDA. Double the CUDAs but half the Memory Bandwith, and I cannot assure but it will render almost the same. 2014/1/9 Leonard Koch leonardkoch...@gmail.com I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.comwrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it
Re: Redshift3D Render
Yes Mirko tell the secret. I don't want to break my mind thinking about memory clocks and bandwiths 2014/1/9 Tim Crowson tim.crow...@magneticdreams.com I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.comwrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it at ~15fps in Shaded mode. That ain't shabby! -Tim On 1/9/2014 6:11 AM, Paul Griswold wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not rendering. Redshift has a proxy system like Vray Arnold, but you have to manually create proxies per object my scene had hundreds and hundreds of objects, so I didn't have time to create them. Therefore, it was creating a renderable mesh per frame - so on a frame that took 28 seconds to render, 20 seconds was spent exporting the mesh and 8 seconds were spent on rendering. But again, it's a beta and they're continuing to improve things like the proxy system. Once I'm caught up I'm hoping to try rendering the classroom scene and see how it does. -Paul ᐧ -- --
Re: Redshift3D Render
Ah interesting. That begins to explain it. Thanks Emilio. On Thu, Jan 9, 2014 at 5:36 PM, Emilio Hernandez emi...@e-roja.com wrote: Yes Mirko tell the secret. I don't want to break my mind thinking about memory clocks and bandwiths 2014/1/9 Tim Crowson tim.crow...@magneticdreams.com I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.comwrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it at ~15fps in Shaded mode. That ain't shabby! -Tim On 1/9/2014 6:11 AM, Paul Griswold wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not rendering. Redshift has a proxy system like Vray Arnold, but you have to manually create proxies per object my scene had hundreds and hundreds of objects, so I didn't have time to create them. Therefore, it was creating a renderable mesh per frame - so on a frame that took 28 seconds to render, 20 seconds was spent exporting the mesh and 8 seconds were spent on rendering. But again, it's a beta and they're continuing to improve things like the proxy system. Once I'm caught up I'm hoping to try rendering the classroom scene and see how it does. -Paul ᐧ -- --
Re: Redshift3D Render
For GPU speeds, you always need to consult this list, it is pretty representative of what to expect from things like Redshift3D: http://www.videocardbenchmark.net/high_end_gpus.html -ben On Thu, Jan 9, 2014 at 11:36 AM, Emilio Hernandez emi...@e-roja.com wrote: Yes Mirko tell the secret. I don't want to break my mind thinking about memory clocks and bandwiths 2014/1/9 Tim Crowson tim.crow...@magneticdreams.com I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.comwrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it at ~15fps in Shaded mode. That ain't shabby! -Tim On 1/9/2014 6:11 AM, Paul Griswold wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not rendering. Redshift has a proxy system like Vray Arnold, but you have to manually create proxies per object my scene had hundreds and hundreds of objects, so I didn't have time to create them. Therefore, it was creating a renderable mesh per frame - so on a frame that took 28 seconds to render, 20 seconds was spent exporting the mesh and 8 seconds were spent on rendering. But again, it's a beta and they're continuing to improve things like the proxy system. Once I'm caught up I'm hoping to try rendering the classroom scene and see how it does. -Paul ᐧ -- -- -- Best regards, Ben Houston Voice: 613-762-4113 Skype: ben.exocortex Twitter: @exocortexcom
Re: Redshift3D Render
Just a funny fact that I found... After opening several tabs for looking for the Nvidia specs, I left them open. I am using FireFox. I hit play again and the fps dropped by half 24 fps. Start scratching my head. Closed the addtional tabs of Firefox, restarted Softimage and the speed was back to 50.7 So my guess is that this modern browser suck a lot from the GPU... pfff. 2014/1/9 Leonard Koch leonardkoch...@gmail.com Ah interesting. That begins to explain it. Thanks Emilio. On Thu, Jan 9, 2014 at 5:36 PM, Emilio Hernandez emi...@e-roja.comwrote: Yes Mirko tell the secret. I don't want to break my mind thinking about memory clocks and bandwiths 2014/1/9 Tim Crowson tim.crow...@magneticdreams.com I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.comwrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it at ~15fps in Shaded mode. That ain't shabby! -Tim On 1/9/2014 6:11 AM, Paul Griswold wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not rendering. Redshift has a proxy system like Vray Arnold, but you have to manually create proxies per object my scene had hundreds and hundreds of objects, so I didn't have time to create them. Therefore, it was creating a renderable mesh per frame - so on a frame that took 28 seconds to render, 20 seconds was spent exporting the mesh and 8 seconds were spent on rendering. But again,
Re: Redshift3D Render
Title: Signature I guess 60fps is the refresh rate of your display, right? Have you disabled VSync in the driver settings? I just get "60.0 fps +" How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.com wrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim("Cube", "MeshSurface", "", "") Application.SetValue("cube.polymsh.geom.subdivu", 831, "") Application.SetValue("cube.polymsh.geom.subdivv", 800, "") Application.SetValue("cube.polymsh.geom.subdivbase", 800, "") Application.SetValue("Camera.camvis.refreshrate", True, "") Application.SetDisplayMode("Camera", "shaded") Application.DeselectAll() Application.SetValue("PlayControl.Out", 5000, "") Application.DeselectAll() Application.GetPrim("Null", "", "", "") Application.SelectObj("Camera_Root", "", "") Application.CopyPaste("Camera_Root", "", "null", 1) Application.SelectObj("null", "", "") Application.SaveKey("null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz", 1, "", "", "", "", "") Application.SetValue("PlayControl.Key", 5000, "") Application.SetValue("PlayControl.Current", 5000, "") Application.Rotate("", 0, 8000, 0, "siAbsolute", "siPivot", "siObj", "siY", "", "", "", "", "", "", "", 0, "") Application.SaveKey("null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz", 5000, "", "", "", "", "") Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu,
Re: Redshift3D Render
So does Chrome btw. I also notice this wehn running my laptop without power supply. The GPU sucks the battery dry in an hour.With Chrome closed it's 4 hrs. GPU acceleration can be turned off in Chrome though, don't know about Firefox.Just a funny fact that I found...After opening several tabs for looking for the Nvidia specs, I left them open. I am using FireFox.I hit play again and the fps dropped by half 24 fps. Start scratching my head. Closed the addtional tabs of Firefox, restarted Softimage and the speed was back to 50.7So my guess is that this modern browser suck a lot from the GPU... pfff. 2014/1/9 Leonard Koch leonardkoch...@gmail.com Ah interesting. That begins to explain it.Thanks Emilio.On Thu, Jan 9, 2014 at 5:36 PM, Emilio Hernandez emi...@e-roja.com wrote: Yes Mirko tell the secret. I don't want to break my mind thinking about memory clocks and bandwiths 2014/1/9 Tim Crowson tim.crow...@magneticdreams.com I just get "60.0 fps +" How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.com wrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim("Cube", "MeshSurface", "", "") Application.SetValue("cube.polymsh.geom.subdivu", 831, "") Application.SetValue("cube.polymsh.geom.subdivv", 800, "") Application.SetValue("cube.polymsh.geom.subdivbase", 800, "") Application.SetValue("Camera.camvis.refreshrate", True, "") Application.SetDisplayMode("Camera", "shaded") Application.DeselectAll() Application.SetValue("PlayControl.Out", 5000, "") Application.DeselectAll() Application.GetPrim("Null", "", "", "") Application.SelectObj("Camera_Root", "", "") Application.CopyPaste("Camera_Root", "", "null", 1) Application.SelectObj("null", "", "") Application.SaveKey("null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz", 1, "", "", "", "", "") Application.SetValue("PlayControl.Key", 5000, "") Application.SetValue("PlayControl.Current", 5000, "")
Re: Redshift3D Render
Emilio, did you try telling the application (firefox, chrome) to use the cpu instead of the gpu? that happened to me too and found it usefu,l through the nvidia control panel. F. 2014/1/9 Emilio Hernandez emi...@e-roja.com Just a funny fact that I found... After opening several tabs for looking for the Nvidia specs, I left them open. I am using FireFox. I hit play again and the fps dropped by half 24 fps. Start scratching my head. Closed the addtional tabs of Firefox, restarted Softimage and the speed was back to 50.7 So my guess is that this modern browser suck a lot from the GPU... pfff. 2014/1/9 Leonard Koch leonardkoch...@gmail.com Ah interesting. That begins to explain it. Thanks Emilio. On Thu, Jan 9, 2014 at 5:36 PM, Emilio Hernandez emi...@e-roja.comwrote: Yes Mirko tell the secret. I don't want to break my mind thinking about memory clocks and bandwiths 2014/1/9 Tim Crowson tim.crow...@magneticdreams.com I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.comwrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it at ~15fps in Shaded mode. That ain't shabby! -Tim On 1/9/2014 6:11 AM, Paul Griswold wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not rendering. Redshift has a proxy system like Vray Arnold, but you have to manually create proxies per object my scene had hundreds and hundreds of
Re: Redshift3D Render
I am going to try your suggestions right now. That will explain a lot of things. 2014/1/9 Tim Crowson tim.crow...@magneticdreams.com I wish it was always that simple. For instance, Redshift will perform better with more vram, and the Titan comes standard with 6GB, which is not even an option on the 780 or 780Ti. If you can get by with less vram though, the 780s are pretty sweet. Can't wait to see what they announce next. -Tim On 1/9/2014 10:39 AM, Ben Houston wrote: For GPU speeds, you always need to consult this list, it is pretty representative of what to expect from things like Redshift3D: http://www.videocardbenchmark.net/high_end_gpus.html -ben On Thu, Jan 9, 2014 at 11:36 AM, Emilio Hernandez emi...@e-roja.comwrote: Yes Mirko tell the secret. I don't want to break my mind thinking about memory clocks and bandwiths 2014/1/9 Tim Crowson tim.crow...@magneticdreams.com I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.comwrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it at ~15fps in Shaded mode. That ain't shabby! -Tim On 1/9/2014 6:11 AM, Paul Griswold wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not rendering. Redshift has a proxy system like Vray Arnold, but you have to manually create proxies per object my scene had hundreds and hundreds of objects, so I didn't have time to create them.
Re: Redshift3D Render
I wish it was always that simple. For instance, Redshift will perform better with more vram, and the Titan comes standard with 6GB, which is not even an option on the 780 or 780Ti. If you can get by with less vram though, the 780s are pretty sweet. Can't wait to see what they announce next. -Tim On 1/9/2014 10:39 AM, Ben Houston wrote: For GPU speeds, you always need to consult this list, it is pretty representative of what to expect from things like Redshift3D: http://www.videocardbenchmark.net/high_end_gpus.html -ben On Thu, Jan 9, 2014 at 11:36 AM, Emilio Hernandez emi...@e-roja.com mailto:emi...@e-roja.com wrote: Yes Mirko tell the secret. I don't want to break my mind thinking about memory clocks and bandwiths 2014/1/9 Tim Crowson tim.crow...@magneticdreams.com mailto:tim.crow...@magneticdreams.com I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.com mailto:emi...@e-roja.com wrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com mailto:mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure :) But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com mailto:tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot,
Re: Redshift3D Render
Well switch displays to the Titan and I am getting with the Vsync option off the same as Tim 63.6fps Maybe this Rubikjancovik cube has a trick :) 2014/1/9 Emilio Hernandez emi...@e-roja.com I am going to try your suggestions right now. That will explain a lot of things. 2014/1/9 Tim Crowson tim.crow...@magneticdreams.com I wish it was always that simple. For instance, Redshift will perform better with more vram, and the Titan comes standard with 6GB, which is not even an option on the 780 or 780Ti. If you can get by with less vram though, the 780s are pretty sweet. Can't wait to see what they announce next. -Tim On 1/9/2014 10:39 AM, Ben Houston wrote: For GPU speeds, you always need to consult this list, it is pretty representative of what to expect from things like Redshift3D: http://www.videocardbenchmark.net/high_end_gpus.html -ben On Thu, Jan 9, 2014 at 11:36 AM, Emilio Hernandez emi...@e-roja.comwrote: Yes Mirko tell the secret. I don't want to break my mind thinking about memory clocks and bandwiths 2014/1/9 Tim Crowson tim.crow...@magneticdreams.com I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.comwrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it at ~15fps in Shaded mode. That ain't shabby! -Tim On 1/9/2014 6:11 AM, Paul Griswold wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not
Re: Redshift3D Render
just got back , yes 60 is vsync on, turn of vsync in nvida control panel On Thu, Jan 9, 2014 at 5:52 PM, Stefan Kubicek s...@tidbit-images.com wrote: I guess 60fps is the refresh rate of your display, right? Have you disabled VSync in the driver settings? I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.comwrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it at ~15fps in Shaded mode. That ain't shabby! -Tim On 1/9/2014 6:11 AM, Paul Griswold wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not rendering. Redshift has a proxy system like Vray Arnold, but you have to manually create proxies per object my scene had hundreds and hundreds of objects, so I didn't have time to create them. Therefore, it was creating a renderable mesh per frame - so on a frame that took 28 seconds to render, 20 seconds was spent exporting the mesh and 8 seconds were spent on rendering. But again, it's a beta and they're continuing to improve things like the proxy system. Once I'm caught up I'm hoping to try rendering the classroom scene and see how it does. -Paul ᐧ -- -- -- --- Stefan Kubicek --- keyvis digital imagery Alfred Feierfeilstraße 3 A-2380 Perchtoldsdorf bei Wien Phone: +43/699/12614231 www.keyvis.at ste...@keyvis.at -- This email and its
Re: Redshift3D Render
I already did that and still getting the 65 fps limit with the Titan. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com just got back , yes 60 is vsync on, turn of vsync in nvida control panel On Thu, Jan 9, 2014 at 5:52 PM, Stefan Kubicek s...@tidbit-images.comwrote: I guess 60fps is the refresh rate of your display, right? Have you disabled VSync in the driver settings? I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.comwrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it at ~15fps in Shaded mode. That ain't shabby! -Tim On 1/9/2014 6:11 AM, Paul Griswold wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not rendering. Redshift has a proxy system like Vray Arnold, but you have to manually create proxies per object my scene had hundreds and hundreds of objects, so I didn't have time to create them. Therefore, it was creating a renderable mesh per frame - so on a frame that took 28 seconds to render, 20 seconds was spent exporting the mesh and 8 seconds were spent on rendering. But again, it's a beta and they're continuing to improve things like the proxy system. Once I'm caught up I'm hoping to try rendering the classroom scene and see how it does. -Paul ᐧ -- -- -- --- Stefan Kubicek --- keyvis digital imagery Alfred
Re: Redshift3D Render
hmm another gues.. is it set to RT or ALL in play? On Thu, Jan 9, 2014 at 6:53 PM, Emilio Hernandez emi...@e-roja.com wrote: I already did that and still getting the 65 fps limit with the Titan. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com just got back , yes 60 is vsync on, turn of vsync in nvida control panel On Thu, Jan 9, 2014 at 5:52 PM, Stefan Kubicek s...@tidbit-images.comwrote: I guess 60fps is the refresh rate of your display, right? Have you disabled VSync in the driver settings? I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.comwrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it at ~15fps in Shaded mode. That ain't shabby! -Tim On 1/9/2014 6:11 AM, Paul Griswold wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not rendering. Redshift has a proxy system like Vray Arnold, but you have to manually create proxies per object my scene had hundreds and hundreds of objects, so I didn't have time to create them. Therefore, it was creating a renderable mesh per frame - so on a frame that took 28 seconds to render, 20 seconds was spent exporting the mesh and 8 seconds were spent on rendering. But again, it's a beta and they're continuing to improve things like the proxy system. Once I'm caught up I'm hoping to try rendering the classroom scene and see how it does. -Paul ᐧ -- -- --
Re: Redshift3D Render
It is set to ALL. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com hmm another gues.. is it set to RT or ALL in play? On Thu, Jan 9, 2014 at 6:53 PM, Emilio Hernandez emi...@e-roja.comwrote: I already did that and still getting the 65 fps limit with the Titan. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com just got back , yes 60 is vsync on, turn of vsync in nvida control panel On Thu, Jan 9, 2014 at 5:52 PM, Stefan Kubicek s...@tidbit-images.comwrote: I guess 60fps is the refresh rate of your display, right? Have you disabled VSync in the driver settings? I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.comwrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it at ~15fps in Shaded mode. That ain't shabby! -Tim On 1/9/2014 6:11 AM, Paul Griswold wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not rendering. Redshift has a proxy system like Vray Arnold, but you have to manually create proxies per object my scene had hundreds and hundreds of objects, so I didn't have time to create them. Therefore, it was creating a renderable mesh per frame - so on a frame that took 28 seconds to render, 20 seconds was spent exporting the mesh and 8 seconds were spent on rendering. But again, it's a beta and they're continuing to improve things like the proxy system. Once I'm caught up I'm hoping to try
Re: Redshift3D Render
yea I guess that woul dbe the case but just tried.. strange really no idea.. any chance to run cinebench 15 opengl test then? if that gives too low result as well then something is not good On Thu, Jan 9, 2014 at 6:58 PM, Emilio Hernandez emi...@e-roja.com wrote: It is set to ALL. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com hmm another gues.. is it set to RT or ALL in play? On Thu, Jan 9, 2014 at 6:53 PM, Emilio Hernandez emi...@e-roja.comwrote: I already did that and still getting the 65 fps limit with the Titan. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com just got back , yes 60 is vsync on, turn of vsync in nvida control panel On Thu, Jan 9, 2014 at 5:52 PM, Stefan Kubicek s...@tidbit-images.comwrote: I guess 60fps is the refresh rate of your display, right? Have you disabled VSync in the driver settings? I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.comwrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it at ~15fps in Shaded mode. That ain't shabby! -Tim On 1/9/2014 6:11 AM, Paul Griswold wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not rendering. Redshift has a proxy system like Vray Arnold, but you have to manually create proxies per object my scene had hundreds and hundreds of objects, so I didn't have time to create them. Therefore, it was creating a renderable mesh per
Re: Redshift3D Render
Yes specailly if you are getting more than the double speed. I ran the test and I got 75.78 fps Thx 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com yea I guess that woul dbe the case but just tried.. strange really no idea.. any chance to run cinebench 15 opengl test then? if that gives too low result as well then something is not good On Thu, Jan 9, 2014 at 6:58 PM, Emilio Hernandez emi...@e-roja.comwrote: It is set to ALL. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com hmm another gues.. is it set to RT or ALL in play? On Thu, Jan 9, 2014 at 6:53 PM, Emilio Hernandez emi...@e-roja.comwrote: I already did that and still getting the 65 fps limit with the Titan. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com just got back , yes 60 is vsync on, turn of vsync in nvida control panel On Thu, Jan 9, 2014 at 5:52 PM, Stefan Kubicek s...@tidbit-images.comwrote: I guess 60fps is the refresh rate of your display, right? Have you disabled VSync in the driver settings? I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.comwrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it at ~15fps in Shaded mode. That ain't shabby! -Tim On 1/9/2014 6:11 AM, Paul Griswold wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the speed of the card. Also, although it doesn't load the entire scene into your card's memory, the more memory your card has, the better it is. But overall, for the type of work I'm mainly doing these days, it's extremely fast. In fact, it's so fast that I was finding the bottleneck was the time taken to export the mesh to Redshift, not rendering. Redshift has a proxy system like Vray Arnold, but you have to manually
Re: Redshift3D Render
getting around 95 score in cinebench with titan, and again 780 in home comp getting 140 score in cinebench... wondering how can 780 crash titan so much in opengl. titan system beside 4 titans have i7 3930K, asus p9x79-e ws MBO and home comp i7 4770k on asus maximus VI hero Will test some redshift rendering later to compare single GPU 780 vs titan On Thu, Jan 9, 2014 at 7:11 PM, Emilio Hernandez emi...@e-roja.com wrote: Yes specailly if you are getting more than the double speed. I ran the test and I got 75.78 fps Thx 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com yea I guess that woul dbe the case but just tried.. strange really no idea.. any chance to run cinebench 15 opengl test then? if that gives too low result as well then something is not good On Thu, Jan 9, 2014 at 6:58 PM, Emilio Hernandez emi...@e-roja.comwrote: It is set to ALL. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com hmm another gues.. is it set to RT or ALL in play? On Thu, Jan 9, 2014 at 6:53 PM, Emilio Hernandez emi...@e-roja.comwrote: I already did that and still getting the 65 fps limit with the Titan. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com just got back , yes 60 is vsync on, turn of vsync in nvida control panel On Thu, Jan 9, 2014 at 5:52 PM, Stefan Kubicek s...@tidbit-images.comwrote: I guess 60fps is the refresh rate of your display, right? Have you disabled VSync in the driver settings? I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.comwrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it at ~15fps in Shaded mode. That ain't shabby! -Tim On 1/9/2014 6:11 AM, Paul Griswold wrote: There was a discussion on the RS forums about it. I don't recall the numbers, though. I don't think the speed of the PCIe slot made a huge difference. It's really all about the
Re: Redshift3D Render
Now I am starting to scratch my head to figure out the maze of Nvidia... At least my benchmark said the Titatn beated up a Quadro 4000K ha ha 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com getting around 95 score in cinebench with titan, and again 780 in home comp getting 140 score in cinebench... wondering how can 780 crash titan so much in opengl. titan system beside 4 titans have i7 3930K, asus p9x79-e ws MBO and home comp i7 4770k on asus maximus VI hero Will test some redshift rendering later to compare single GPU 780 vs titan On Thu, Jan 9, 2014 at 7:11 PM, Emilio Hernandez emi...@e-roja.comwrote: Yes specailly if you are getting more than the double speed. I ran the test and I got 75.78 fps Thx 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com yea I guess that woul dbe the case but just tried.. strange really no idea.. any chance to run cinebench 15 opengl test then? if that gives too low result as well then something is not good On Thu, Jan 9, 2014 at 6:58 PM, Emilio Hernandez emi...@e-roja.comwrote: It is set to ALL. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com hmm another gues.. is it set to RT or ALL in play? On Thu, Jan 9, 2014 at 6:53 PM, Emilio Hernandez emi...@e-roja.comwrote: I already did that and still getting the 65 fps limit with the Titan. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com just got back , yes 60 is vsync on, turn of vsync in nvida control panel On Thu, Jan 9, 2014 at 5:52 PM, Stefan Kubicek s...@tidbit-images.com wrote: I guess 60fps is the refresh rate of your display, right? Have you disabled VSync in the driver settings? I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.com wrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day, and this thing was letting me tumble around it at ~15fps in Shaded mode. That ain't shabby! -Tim On 1/9/2014 6:11
Re: Redshift3D Render
cinebench you mean? so what is score? :) On Thu, Jan 9, 2014 at 7:47 PM, Emilio Hernandez emi...@e-roja.com wrote: Now I am starting to scratch my head to figure out the maze of Nvidia... At least my benchmark said the Titatn beated up a Quadro 4000K ha ha 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com getting around 95 score in cinebench with titan, and again 780 in home comp getting 140 score in cinebench... wondering how can 780 crash titan so much in opengl. titan system beside 4 titans have i7 3930K, asus p9x79-e ws MBO and home comp i7 4770k on asus maximus VI hero Will test some redshift rendering later to compare single GPU 780 vs titan On Thu, Jan 9, 2014 at 7:11 PM, Emilio Hernandez emi...@e-roja.comwrote: Yes specailly if you are getting more than the double speed. I ran the test and I got 75.78 fps Thx 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com yea I guess that woul dbe the case but just tried.. strange really no idea.. any chance to run cinebench 15 opengl test then? if that gives too low result as well then something is not good On Thu, Jan 9, 2014 at 6:58 PM, Emilio Hernandez emi...@e-roja.comwrote: It is set to ALL. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com hmm another gues.. is it set to RT or ALL in play? On Thu, Jan 9, 2014 at 6:53 PM, Emilio Hernandez emi...@e-roja.comwrote: I already did that and still getting the 65 fps limit with the Titan. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com just got back , yes 60 is vsync on, turn of vsync in nvida control panel On Thu, Jan 9, 2014 at 5:52 PM, Stefan Kubicek s...@tidbit-images.com wrote: I guess 60fps is the refresh rate of your display, right? Have you disabled VSync in the driver settings? I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.com wrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame, hardware/software costs, rack space used, etc. But hey, Redshift aside, the Titan in my workstation is doing wonders for my viewport performance in Soft. I had a 58M, 2500-item model derived from a CAD file the other day,
Re: Redshift3D Render
Dumb question here. Where do I see the score? I only get a fps result for OpenGl and a 759cb in the CPU results in Cinebench. and I get 75.78fps at the OpenGl test. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com cinebench you mean? so what is score? :) On Thu, Jan 9, 2014 at 7:47 PM, Emilio Hernandez emi...@e-roja.comwrote: Now I am starting to scratch my head to figure out the maze of Nvidia... At least my benchmark said the Titatn beated up a Quadro 4000K ha ha 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com getting around 95 score in cinebench with titan, and again 780 in home comp getting 140 score in cinebench... wondering how can 780 crash titan so much in opengl. titan system beside 4 titans have i7 3930K, asus p9x79-e ws MBO and home comp i7 4770k on asus maximus VI hero Will test some redshift rendering later to compare single GPU 780 vs titan On Thu, Jan 9, 2014 at 7:11 PM, Emilio Hernandez emi...@e-roja.comwrote: Yes specailly if you are getting more than the double speed. I ran the test and I got 75.78 fps Thx 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com yea I guess that woul dbe the case but just tried.. strange really no idea.. any chance to run cinebench 15 opengl test then? if that gives too low result as well then something is not good On Thu, Jan 9, 2014 at 6:58 PM, Emilio Hernandez emi...@e-roja.comwrote: It is set to ALL. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com hmm another gues.. is it set to RT or ALL in play? On Thu, Jan 9, 2014 at 6:53 PM, Emilio Hernandez emi...@e-roja.comwrote: I already did that and still getting the 65 fps limit with the Titan. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com just got back , yes 60 is vsync on, turn of vsync in nvida control panel On Thu, Jan 9, 2014 at 5:52 PM, Stefan Kubicek s...@tidbit-images.com wrote: I guess 60fps is the refresh rate of your display, right? Have you disabled VSync in the driver settings? I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.com wrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they could really go before it started to cost more money than it was worth. In our case, I'm thinking 3 might be our max, based on some nerdy mathematics by one of our IT guys analyzing render times per shot, per frame,
Re: Redshift3D Render
yes fps is score. hmm maybe if I got some time I could get one of titans and test it out on maximus MBO to see if that is the diference... On Thu, Jan 9, 2014 at 8:09 PM, Emilio Hernandez emi...@e-roja.com wrote: Dumb question here. Where do I see the score? I only get a fps result for OpenGl and a 759cb in the CPU results in Cinebench. and I get 75.78fps at the OpenGl test. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com cinebench you mean? so what is score? :) On Thu, Jan 9, 2014 at 7:47 PM, Emilio Hernandez emi...@e-roja.comwrote: Now I am starting to scratch my head to figure out the maze of Nvidia... At least my benchmark said the Titatn beated up a Quadro 4000K ha ha 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com getting around 95 score in cinebench with titan, and again 780 in home comp getting 140 score in cinebench... wondering how can 780 crash titan so much in opengl. titan system beside 4 titans have i7 3930K, asus p9x79-e ws MBO and home comp i7 4770k on asus maximus VI hero Will test some redshift rendering later to compare single GPU 780 vs titan On Thu, Jan 9, 2014 at 7:11 PM, Emilio Hernandez emi...@e-roja.comwrote: Yes specailly if you are getting more than the double speed. I ran the test and I got 75.78 fps Thx 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com yea I guess that woul dbe the case but just tried.. strange really no idea.. any chance to run cinebench 15 opengl test then? if that gives too low result as well then something is not good On Thu, Jan 9, 2014 at 6:58 PM, Emilio Hernandez emi...@e-roja.comwrote: It is set to ALL. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com hmm another gues.. is it set to RT or ALL in play? On Thu, Jan 9, 2014 at 6:53 PM, Emilio Hernandez emi...@e-roja.com wrote: I already did that and still getting the 65 fps limit with the Titan. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com just got back , yes 60 is vsync on, turn of vsync in nvida control panel On Thu, Jan 9, 2014 at 5:52 PM, Stefan Kubicek s...@tidbit-images.com wrote: I guess 60fps is the refresh rate of your display, right? Have you disabled VSync in the driver settings? I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.com wrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim Crowson tim.crow...@magneticdreams.com wrote: We've been testing 1 Titan vs. 3 and so far, the speed increase of the triple-Titan box is holding at about 2.45x. In an email exchange (or maybe it was on the forums, can't recall) it was mentioned that on the topic parallelization, Pixar had determined that even for them, 4 units together (of whatever, not necessarily Titans) was the max they
Re: Redshift3D Render
just played a bit with overclocking... now I pushed PCI GEN in bios back to auto instead of GEN 3 so it is now PCI 2.0 x16 instead of 3.. That alone gave me slightly higher score on cinebench.. and after overclocking CPU to 4.5 now new result in cinebench for titan is 100, and script in softimage gives me ~200 fps. Oh well... On Thu, Jan 9, 2014 at 8:24 PM, Mirko Jankovic mirkoj.anima...@gmail.comwrote: yes fps is score. hmm maybe if I got some time I could get one of titans and test it out on maximus MBO to see if that is the diference... On Thu, Jan 9, 2014 at 8:09 PM, Emilio Hernandez emi...@e-roja.comwrote: Dumb question here. Where do I see the score? I only get a fps result for OpenGl and a 759cb in the CPU results in Cinebench. and I get 75.78fps at the OpenGl test. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com cinebench you mean? so what is score? :) On Thu, Jan 9, 2014 at 7:47 PM, Emilio Hernandez emi...@e-roja.comwrote: Now I am starting to scratch my head to figure out the maze of Nvidia... At least my benchmark said the Titatn beated up a Quadro 4000K ha ha 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com getting around 95 score in cinebench with titan, and again 780 in home comp getting 140 score in cinebench... wondering how can 780 crash titan so much in opengl. titan system beside 4 titans have i7 3930K, asus p9x79-e ws MBO and home comp i7 4770k on asus maximus VI hero Will test some redshift rendering later to compare single GPU 780 vs titan On Thu, Jan 9, 2014 at 7:11 PM, Emilio Hernandez emi...@e-roja.comwrote: Yes specailly if you are getting more than the double speed. I ran the test and I got 75.78 fps Thx 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com yea I guess that woul dbe the case but just tried.. strange really no idea.. any chance to run cinebench 15 opengl test then? if that gives too low result as well then something is not good On Thu, Jan 9, 2014 at 6:58 PM, Emilio Hernandez emi...@e-roja.comwrote: It is set to ALL. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com hmm another gues.. is it set to RT or ALL in play? On Thu, Jan 9, 2014 at 6:53 PM, Emilio Hernandez emi...@e-roja.com wrote: I already did that and still getting the 65 fps limit with the Titan. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com just got back , yes 60 is vsync on, turn of vsync in nvida control panel On Thu, Jan 9, 2014 at 5:52 PM, Stefan Kubicek s...@tidbit-images.com wrote: I guess 60fps is the refresh rate of your display, right? Have you disabled VSync in the driver settings? I just get 60.0 fps + How are you getting it display a value higher than 60? I'm pretty sure it the actual fps is higher, but the value in the viewport is capped at 60 -Tim On 1/9/2014 10:12 AM, Leonard Koch wrote: I get about 28-31 out of my 680. Does anyone have a common explanation for that? On Thu, Jan 9, 2014 at 5:10 PM, Emilio Hernandez emi...@e-roja.com wrote: Hey Mirko I ran your script and I got 50.7 fps... But then I remembered I have my displays plugged in to my 470.. hahaha. Don't ask why, but when using AE with the displays plugged into the Ti, AE does not like it and disables GPU for calculations... P. 2014/1/9 Mirko Jankovic mirkoj.anima...@gmail.com Hey Tim Would you be able to take 2 minutes of your tmie and run this ol python script for SI with your titan? I'm getting weird results with an 780 in my home system outperforming titan a lot... well here is copy paste from forum if you are able to check it out as well.. thanks!: itan: ~170 fps 780: ~245 fps Go figure [image: :)] But I'm suspecting something weird with my titan system for some time will have to test further but would be great if anyone with titan as well could run it too? This old python script: Application.CreatePrim(Cube, MeshSurface, , ) Application.SetValue(cube.polymsh.geom.subdivu, 831, ) Application.SetValue(cube.polymsh.geom.subdivv, 800, ) Application.SetValue(cube.polymsh.geom.subdivbase, 800, ) Application.SetValue(Camera.camvis.refreshrate, True, ) Application.SetDisplayMode(Camera, shaded) Application.DeselectAll() Application.SetValue(PlayControl.Out, 5000, ) Application.DeselectAll() Application.GetPrim(Null, , , ) Application.SelectObj(Camera_Root, , ) Application.CopyPaste(Camera_Root, , null, 1) Application.SelectObj(null, , ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 1, , , , , ) Application.SetValue(PlayControl.Key, 5000, ) Application.SetValue(PlayControl.Current, 5000, ) Application.Rotate(, 0, 8000, 0, siAbsolute, siPivot, siObj, siY, , , , , , , , 0, ) Application.SaveKey(null.kine.local.rotx,null.kine.local.roty,null.kine.local.rotz, 5000, , , , , ) Application.FirstFrame() Just paste in python script run and hit play. Thakns! On Thu, Jan 9, 2014 at 3:34 PM, Tim
Redshift3D Render
I just wanted to start this thread so we can kill the Sofimage is dead thread. :) So..Redshift3D enthusiasts...post here. I think we will benefit the group, as a whole, if we spread the word on this amazing GPU rendering company. They have made my workflow SO much faster. -- Best Regards, * Stephen P. Davidson* *(954) 552-7956*sdavid...@3danimationmagic.com *Any sufficiently advanced technology is indistinguishable from magic* - Arthur C. Clarke http://www.3danimationmagic.com
Re: Redshift3D Render
+1
Re: Redshift3D Render
I've been using it along side Arnold for quite a while now. I just finished a project for CES entirely in Redshift. I think Redshift falls more into the category of a VRay competitor rather than Arnold. Redshift isn't open the way Arnold is I don't think they intend it to be. I've found it to be extremely fast and stable in most cases, but being just out of alpha there are times when it's warts show. For $100, it's certainly worth a test drive. -Paul ᐧ On Wed, Jan 8, 2014 at 3:46 PM, Stephen Davidson magic...@bellsouth.netwrote: I just wanted to start this thread so we can kill the Sofimage is dead thread. :) So..Redshift3D enthusiasts...post here. I think we will benefit the group, as a whole, if we spread the word on this amazing GPU rendering company. They have made my workflow SO much faster. -- Best Regards, * Stephen P. Davidson* *(954) 552-7956 %28954%29%20552-7956 *sdavid...@3danimationmagic.com *Any sufficiently advanced technology is indistinguishable from magic* - Arthur C. Clarke http://www.3danimationmagic.com
Re: Redshift3D Render
I used Redshift when it was alpha (it's beta now) I really loved it, but Redshift is developing few features that I always need like hair, ICE strand rendering, and better displacement map result. But still Redshift is incredibly fast than any other renderers (Arnold, V-Ray, MRay) and it's really stable. Once it supports those features well, I am going to use it with Arnold. Daniel --- Daniel Kim Animation Director Professional 3D Generalist http://www.danielkim3d.com --- On Thu, Jan 9, 2014 at 9:53 AM, Paul Griswold pgrisw...@fusiondigitalproductions.com wrote: I've been using it along side Arnold for quite a while now. I just finished a project for CES entirely in Redshift. I think Redshift falls more into the category of a VRay competitor rather than Arnold. Redshift isn't open the way Arnold is I don't think they intend it to be. I've found it to be extremely fast and stable in most cases, but being just out of alpha there are times when it's warts show. For $100, it's certainly worth a test drive. -Paul ᐧ On Wed, Jan 8, 2014 at 3:46 PM, Stephen Davidson magic...@bellsouth.netwrote: I just wanted to start this thread so we can kill the Sofimage is dead thread. :) So..Redshift3D enthusiasts...post here. I think we will benefit the group, as a whole, if we spread the word on this amazing GPU rendering company. They have made my workflow SO much faster. -- Best Regards, * Stephen P. Davidson* *(954) 552-7956 %28954%29%20552-7956 * sdavid...@3danimationmagic.com *Any sufficiently advanced technology is indistinguishable from magic* - Arthur C. Clarke http://www.3danimationmagic.com
Re: Redshift3D Render
I'll repost what I said in the other thread We started using Redshift back in March and pretty much use it exclusively now. Of course it all depends on the needs of the project (and there are still some real limitations). The RS dev team is top notch though. I'm really excited to see how things will be at the end of this calendar year. Redshift development is progressing at a fantastic rate, and the pricing is very competitive. For facilities, even small ones, it does require that you spend some time considering your hardware and infrastructure, especially if you want to start converting CPU farms for GPU rendering, or augmenting them. Fortunately, Redshift isn't licensed per GPU, but per machine, and that should provide some breathing room. To be honest (and I realize we have many Arnold folks here), here at Magnetic we evaluated our rendering options (MR, vRay, 3Delight, Arnold). I even started working on a Soft-to-Modo pipeline. Among these Arnold was the clear winner. That said, we felt that to be useful for us in production, Arnold was too costly a solution for us, both financially and in render time, /considering the kinds of projects we do/ . Then Redshift came along and despite its infancy, really turned our heads. We cautiously began using it on productions, and it has since proven itself for us. Again, it all depends on what kind of project you're working on! You need to evaluate it for yourself of course, but for smaller houses like us, it allows us to produce better looking content faster, while staying in Softimage. And in this economy, we can't argue with that. -Tim On 1/8/2014 2:53 PM, Paul Griswold wrote: I've been using it along side Arnold for quite a while now. I just finished a project for CES entirely in Redshift. I think Redshift falls more into the category of a VRay competitor rather than Arnold. Redshift isn't open the way Arnold is I don't think they intend it to be. I've found it to be extremely fast and stable in most cases, but being just out of alpha there are times when it's warts show. For $100, it's certainly worth a test drive. -Paul ᐧ On Wed, Jan 8, 2014 at 3:46 PM, Stephen Davidson magic...@bellsouth.net mailto:magic...@bellsouth.net wrote: I just wanted to start this thread so we can kill the Sofimage is dead thread. :) So..Redshift3D enthusiasts...post here. I think we will benefit the group, as a whole, if we spread the word on this amazing GPU rendering company. They have made my workflow SO much faster. -- Best Regards, * Stephen P. Davidson** **(954) 552-7956 tel:%28954%29%20552-7956 * sdavid...@3danimationmagic.com /Any sufficiently advanced technology is indistinguishable from magic/ - Arthur C. Clarke http://www.3danimationmagic.com -- Signature *Tim Crowson */Lead CG Artist/ *Magnetic Dreams, Inc. *2525 Lebanon Pike, Bldg C, Suite 101, Nashville, TN 37214 *Ph* 615.885.6801 | *Fax* 615.889.4768 | www.magneticdreams.com tim.crow...@magneticdreams.com /Confidentiality Notice: This email, including attachments, is confidential and should not be used by anyone who is not the original intended recipient(s). If you have received this e-mail in error please inform the sender and delete it from your mailbox or any other storage mechanism. Magnetic Dreams, Inc cannot accept liability for any statements made which are clearly the sender's own and not expressly made on behalf of Magnetic Dreams, Inc or one of its agents./
Re: Redshift3D Render
Displacement and bump map are there and they work beautiful. They event implement a scalar change range into the displacement node. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com I'll repost what I said in the other thread We started using Redshift back in March and pretty much use it exclusively now. Of course it all depends on the needs of the project (and there are still some real limitations). The RS dev team is top notch though. I'm really excited to see how things will be at the end of this calendar year. Redshift development is progressing at a fantastic rate, and the pricing is very competitive. For facilities, even small ones, it does require that you spend some time considering your hardware and infrastructure, especially if you want to start converting CPU farms for GPU rendering, or augmenting them. Fortunately, Redshift isn't licensed per GPU, but per machine, and that should provide some breathing room. To be honest (and I realize we have many Arnold folks here), here at Magnetic we evaluated our rendering options (MR, vRay, 3Delight, Arnold). I even started working on a Soft-to-Modo pipeline. Among these Arnold was the clear winner. That said, we felt that to be useful for us in production, Arnold was too costly a solution for us, both financially and in render time, *considering the kinds of projects we do* . Then Redshift came along and despite its infancy, really turned our heads. We cautiously began using it on productions, and it has since proven itself for us. Again, it all depends on what kind of project you're working on! You need to evaluate it for yourself of course, but for smaller houses like us, it allows us to produce better looking content faster, while staying in Softimage. And in this economy, we can't argue with that. -Tim On 1/8/2014 2:53 PM, Paul Griswold wrote: I've been using it along side Arnold for quite a while now. I just finished a project for CES entirely in Redshift. I think Redshift falls more into the category of a VRay competitor rather than Arnold. Redshift isn't open the way Arnold is I don't think they intend it to be. I've found it to be extremely fast and stable in most cases, but being just out of alpha there are times when it's warts show. For $100, it's certainly worth a test drive. -Paul ᐧ On Wed, Jan 8, 2014 at 3:46 PM, Stephen Davidson magic...@bellsouth.netwrote: I just wanted to start this thread so we can kill the Sofimage is dead thread. :) So..Redshift3D enthusiasts...post here. I think we will benefit the group, as a whole, if we spread the word on this amazing GPU rendering company. They have made my workflow SO much faster. -- Best Regards, * Stephen P. Davidson* *(954) 552-7956 %28954%29%20552-7956 * sdavid...@3danimationmagic.com *Any sufficiently advanced technology is indistinguishable from magic* - Arthur C. Clarke http://www.3danimationmagic.com -- *Tim Crowson **Lead CG Artist* *Magnetic Dreams, Inc. *2525 Lebanon Pike, Bldg C, Suite 101, Nashville, TN 37214 *Ph* 615.885.6801 | *Fax* 615.889.4768 | www.magneticdreams.com tim.crow...@magneticdreams.com *Confidentiality Notice: This email, including attachments, is confidential and should not be used by anyone who is not the original intended recipient(s). If you have received this e-mail in error please inform the sender and delete it from your mailbox or any other storage mechanism. Magnetic Dreams, Inc cannot accept liability for any statements made which are clearly the sender's own and not expressly made on behalf of Magnetic Dreams, Inc or one of its agents.*
Re: Redshift3D Render
I found some weird result of displacement map with Redshift. Elevation is okay, but sometimes I could see weird connection of UVM. All UVM boders wasn't smooth and I had no idea how to fix it. Arnold has that option though : / But more option I need is... hair @__@ and ICE strand --- Daniel Kim Animation Director Professional 3D Generalist http://www.danielkim3d.com --- On Thu, Jan 9, 2014 at 10:31 AM, Emilio Hernandez emi...@e-roja.com wrote: Displacement and bump map are there and they work beautiful. They event implement a scalar change range into the displacement node. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com I'll repost what I said in the other thread We started using Redshift back in March and pretty much use it exclusively now. Of course it all depends on the needs of the project (and there are still some real limitations). The RS dev team is top notch though. I'm really excited to see how things will be at the end of this calendar year. Redshift development is progressing at a fantastic rate, and the pricing is very competitive. For facilities, even small ones, it does require that you spend some time considering your hardware and infrastructure, especially if you want to start converting CPU farms for GPU rendering, or augmenting them. Fortunately, Redshift isn't licensed per GPU, but per machine, and that should provide some breathing room. To be honest (and I realize we have many Arnold folks here), here at Magnetic we evaluated our rendering options (MR, vRay, 3Delight, Arnold). I even started working on a Soft-to-Modo pipeline. Among these Arnold was the clear winner. That said, we felt that to be useful for us in production, Arnold was too costly a solution for us, both financially and in render time, *considering the kinds of projects we do* . Then Redshift came along and despite its infancy, really turned our heads. We cautiously began using it on productions, and it has since proven itself for us. Again, it all depends on what kind of project you're working on! You need to evaluate it for yourself of course, but for smaller houses like us, it allows us to produce better looking content faster, while staying in Softimage. And in this economy, we can't argue with that. -Tim On 1/8/2014 2:53 PM, Paul Griswold wrote: I've been using it along side Arnold for quite a while now. I just finished a project for CES entirely in Redshift. I think Redshift falls more into the category of a VRay competitor rather than Arnold. Redshift isn't open the way Arnold is I don't think they intend it to be. I've found it to be extremely fast and stable in most cases, but being just out of alpha there are times when it's warts show. For $100, it's certainly worth a test drive. -Paul ᐧ On Wed, Jan 8, 2014 at 3:46 PM, Stephen Davidson magic...@bellsouth.netwrote: I just wanted to start this thread so we can kill the Sofimage is dead thread. :) So..Redshift3D enthusiasts...post here. I think we will benefit the group, as a whole, if we spread the word on this amazing GPU rendering company. They have made my workflow SO much faster. -- Best Regards, * Stephen P. Davidson* *(954) 552-7956 %28954%29%20552-7956 * sdavid...@3danimationmagic.com *Any sufficiently advanced technology is indistinguishable from magic* - Arthur C. Clarke http://www.3danimationmagic.com -- *Tim Crowson **Lead CG Artist* *Magnetic Dreams, Inc. *2525 Lebanon Pike, Bldg C, Suite 101, Nashville, TN 37214 *Ph* 615.885.6801 | *Fax* 615.889.4768 | www.magneticdreams.com tim.crow...@magneticdreams.com *Confidentiality Notice: This email, including attachments, is confidential and should not be used by anyone who is not the original intended recipient(s). If you have received this e-mail in error please inform the sender and delete it from your mailbox or any other storage mechanism. Magnetic Dreams, Inc cannot accept liability for any statements made which are clearly the sender's own and not expressly made on behalf of Magnetic Dreams, Inc or one of its agents.*
Re: Redshift3D Render
Sorry hit the send button... I was saying... Yes. They even implemented a scalar change range into the displacement node and they have an auto bump feature. It can render such fine detail, that really is amazing. The bump map by itself looks awsome. 2014/1/8 Emilio Hernandez emi...@e-roja.com Displacement and bump map are there and they work beautiful. They event implement a scalar change range into the displacement node. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com I'll repost what I said in the other thread We started using Redshift back in March and pretty much use it exclusively now. Of course it all depends on the needs of the project (and there are still some real limitations). The RS dev team is top notch though. I'm really excited to see how things will be at the end of this calendar year. Redshift development is progressing at a fantastic rate, and the pricing is very competitive. For facilities, even small ones, it does require that you spend some time considering your hardware and infrastructure, especially if you want to start converting CPU farms for GPU rendering, or augmenting them. Fortunately, Redshift isn't licensed per GPU, but per machine, and that should provide some breathing room. To be honest (and I realize we have many Arnold folks here), here at Magnetic we evaluated our rendering options (MR, vRay, 3Delight, Arnold). I even started working on a Soft-to-Modo pipeline. Among these Arnold was the clear winner. That said, we felt that to be useful for us in production, Arnold was too costly a solution for us, both financially and in render time, *considering the kinds of projects we do* . Then Redshift came along and despite its infancy, really turned our heads. We cautiously began using it on productions, and it has since proven itself for us. Again, it all depends on what kind of project you're working on! You need to evaluate it for yourself of course, but for smaller houses like us, it allows us to produce better looking content faster, while staying in Softimage. And in this economy, we can't argue with that. -Tim On 1/8/2014 2:53 PM, Paul Griswold wrote: I've been using it along side Arnold for quite a while now. I just finished a project for CES entirely in Redshift. I think Redshift falls more into the category of a VRay competitor rather than Arnold. Redshift isn't open the way Arnold is I don't think they intend it to be. I've found it to be extremely fast and stable in most cases, but being just out of alpha there are times when it's warts show. For $100, it's certainly worth a test drive. -Paul ᐧ On Wed, Jan 8, 2014 at 3:46 PM, Stephen Davidson magic...@bellsouth.netwrote: I just wanted to start this thread so we can kill the Sofimage is dead thread. :) So..Redshift3D enthusiasts...post here. I think we will benefit the group, as a whole, if we spread the word on this amazing GPU rendering company. They have made my workflow SO much faster. -- Best Regards, * Stephen P. Davidson* *(954) 552-7956 %28954%29%20552-7956 * sdavid...@3danimationmagic.com *Any sufficiently advanced technology is indistinguishable from magic* - Arthur C. Clarke http://www.3danimationmagic.com -- *Tim Crowson **Lead CG Artist* *Magnetic Dreams, Inc. *2525 Lebanon Pike, Bldg C, Suite 101, Nashville, TN 37214 *Ph* 615.885.6801 | *Fax* 615.889.4768 | www.magneticdreams.com tim.crow...@magneticdreams.com *Confidentiality Notice: This email, including attachments, is confidential and should not be used by anyone who is not the original intended recipient(s). If you have received this e-mail in error please inform the sender and delete it from your mailbox or any other storage mechanism. Magnetic Dreams, Inc cannot accept liability for any statements made which are clearly the sender's own and not expressly made on behalf of Magnetic Dreams, Inc or one of its agents.*
Re: Redshift3D Render
They've stated pretty clearly that Hair and Strand support is the next big thing to come... shouldn't be too long now... -Tim On 1/8/2014 3:36 PM, Daniel Kim wrote: I found some weird result of displacement map with Redshift. Elevation is okay, but sometimes I could see weird connection of UVM. All UVM boders wasn't smooth and I had no idea how to fix it. Arnold has that option though : / But more option I need is... hair @__@ and ICE strand --- Daniel Kim Animation Director Professional 3D Generalist http://www.danielkim3d.com --- On Thu, Jan 9, 2014 at 10:31 AM, Emilio Hernandez emi...@e-roja.com mailto:emi...@e-roja.com wrote: Displacement and bump map are there and they work beautiful. They event implement a scalar change range into the displacement node. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com mailto:tim.crow...@magneticdreams.com I'll repost what I said in the other thread We started using Redshift back in March and pretty much use it exclusively now. Of course it all depends on the needs of the project (and there are still some real limitations). The RS dev team is top notch though. I'm really excited to see how things will be at the end of this calendar year. Redshift development is progressing at a fantastic rate, and the pricing is very competitive. For facilities, even small ones, it does require that you spend some time considering your hardware and infrastructure, especially if you want to start converting CPU farms for GPU rendering, or augmenting them. Fortunately, Redshift isn't licensed per GPU, but per machine, and that should provide some breathing room. To be honest (and I realize we have many Arnold folks here), here at Magnetic we evaluated our rendering options (MR, vRay, 3Delight, Arnold). I even started working on a Soft-to-Modo pipeline. Among these Arnold was the clear winner. That said, we felt that to be useful for us in production, Arnold was too costly a solution for us, both financially and in render time, /considering the kinds of projects we do/ . Then Redshift came along and despite its infancy, really turned our heads. We cautiously began using it on productions, and it has since proven itself for us. Again, it all depends on what kind of project you're working on! You need to evaluate it for yourself of course, but for smaller houses like us, it allows us to produce better looking content faster, while staying in Softimage. And in this economy, we can't argue with that. -Tim On 1/8/2014 2:53 PM, Paul Griswold wrote: I've been using it along side Arnold for quite a while now. I just finished a project for CES entirely in Redshift. I think Redshift falls more into the category of a VRay competitor rather than Arnold. Redshift isn't open the way Arnold is I don't think they intend it to be. I've found it to be extremely fast and stable in most cases, but being just out of alpha there are times when it's warts show. For $100, it's certainly worth a test drive. -Paul ᐧ On Wed, Jan 8, 2014 at 3:46 PM, Stephen Davidson magic...@bellsouth.net mailto:magic...@bellsouth.net wrote: I just wanted to start this thread so we can kill the Sofimage is dead thread. :) So..Redshift3D enthusiasts...post here. I think we will benefit the group, as a whole, if we spread the word on this amazing GPU rendering company. They have made my workflow SO much faster. -- Best Regards, * Stephen P. Davidson** **(954) 552-7956 tel:%28954%29%20552-7956 * sdavid...@3danimationmagic.com mailto:sdavid...@3danimationmagic.com /Any sufficiently advanced technology is indistinguishable from magic/ - Arthur C. Clarke http://www.3danimationmagic.com -- *Tim Crowson */Lead CG Artist/ *Magnetic Dreams, Inc. *2525 Lebanon Pike, Bldg C, Suite 101, Nashville, TN 37214 *Ph* 615.885.6801 tel:615.885.6801 | *Fax* 615.889.4768 tel:615.889.4768 | www.magneticdreams.com http://www.magneticdreams.com tim.crow...@magneticdreams.com mailto:tim.crow...@magneticdreams.com /Confidentiality Notice: This email, including attachments, is confidential and should not be used by anyone who is not the original intended recipient(s). If you have received
Re: Redshift3D Render
Yep. I see that's coming. I can't wait to use paid beta version --- Daniel Kim Animation Director Professional 3D Generalist http://www.danielkim3d.com --- On Thu, Jan 9, 2014 at 10:42 AM, Tim Crowson tim.crow...@magneticdreams.com wrote: They've stated pretty clearly that Hair and Strand support is the next big thing to come... shouldn't be too long now... -Tim On 1/8/2014 3:36 PM, Daniel Kim wrote: I found some weird result of displacement map with Redshift. Elevation is okay, but sometimes I could see weird connection of UVM. All UVM boders wasn't smooth and I had no idea how to fix it. Arnold has that option though : / But more option I need is... hair @__@ and ICE strand --- Daniel Kim Animation Director Professional 3D Generalist http://www.danielkim3d.com --- On Thu, Jan 9, 2014 at 10:31 AM, Emilio Hernandez emi...@e-roja.comwrote: Displacement and bump map are there and they work beautiful. They event implement a scalar change range into the displacement node. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com I'll repost what I said in the other thread We started using Redshift back in March and pretty much use it exclusively now. Of course it all depends on the needs of the project (and there are still some real limitations). The RS dev team is top notch though. I'm really excited to see how things will be at the end of this calendar year. Redshift development is progressing at a fantastic rate, and the pricing is very competitive. For facilities, even small ones, it does require that you spend some time considering your hardware and infrastructure, especially if you want to start converting CPU farms for GPU rendering, or augmenting them. Fortunately, Redshift isn't licensed per GPU, but per machine, and that should provide some breathing room. To be honest (and I realize we have many Arnold folks here), here at Magnetic we evaluated our rendering options (MR, vRay, 3Delight, Arnold). I even started working on a Soft-to-Modo pipeline. Among these Arnold was the clear winner. That said, we felt that to be useful for us in production, Arnold was too costly a solution for us, both financially and in render time, *considering the kinds of projects we do* . Then Redshift came along and despite its infancy, really turned our heads. We cautiously began using it on productions, and it has since proven itself for us. Again, it all depends on what kind of project you're working on! You need to evaluate it for yourself of course, but for smaller houses like us, it allows us to produce better looking content faster, while staying in Softimage. And in this economy, we can't argue with that. -Tim On 1/8/2014 2:53 PM, Paul Griswold wrote: I've been using it along side Arnold for quite a while now. I just finished a project for CES entirely in Redshift. I think Redshift falls more into the category of a VRay competitor rather than Arnold. Redshift isn't open the way Arnold is I don't think they intend it to be. I've found it to be extremely fast and stable in most cases, but being just out of alpha there are times when it's warts show. For $100, it's certainly worth a test drive. -Paul ᐧ On Wed, Jan 8, 2014 at 3:46 PM, Stephen Davidson magic...@bellsouth.net wrote: I just wanted to start this thread so we can kill the Sofimage is dead thread. :) So..Redshift3D enthusiasts...post here. I think we will benefit the group, as a whole, if we spread the word on this amazing GPU rendering company. They have made my workflow SO much faster. -- Best Regards, * Stephen P. Davidson* *(954) 552-7956 %28954%29%20552-7956 * sdavid...@3danimationmagic.com *Any sufficiently advanced technology is indistinguishable from magic* - Arthur C. Clarke http://www.3danimationmagic.com -- *Tim Crowson **Lead CG Artist* *Magnetic Dreams, Inc. *2525 Lebanon Pike, Bldg C, Suite 101, Nashville, TN 37214 *Ph* 615.885.6801 | *Fax* 615.889.4768 | www.magneticdreams.com tim.crow...@magneticdreams.com *Confidentiality Notice: This email, including attachments, is confidential and should not be used by anyone who is not the original intended recipient(s). If you have received this e-mail in error please inform the sender and delete it from your mailbox or any other storage mechanism. Magnetic Dreams, Inc cannot accept liability for any statements made which are clearly the sender's own and not expressly made on behalf of Magnetic Dreams, Inc or one of its agents.* --
Re: Redshift3D Render
Backing up Tim, in the forums there is actually a hair test in Softimage with a simple phong shader. And IMHO it looks nice. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com They've stated pretty clearly that Hair and Strand support is the next big thing to come... shouldn't be too long now... -Tim On 1/8/2014 3:36 PM, Daniel Kim wrote: I found some weird result of displacement map with Redshift. Elevation is okay, but sometimes I could see weird connection of UVM. All UVM boders wasn't smooth and I had no idea how to fix it. Arnold has that option though : / But more option I need is... hair @__@ and ICE strand --- Daniel Kim Animation Director Professional 3D Generalist http://www.danielkim3d.com --- On Thu, Jan 9, 2014 at 10:31 AM, Emilio Hernandez emi...@e-roja.comwrote: Displacement and bump map are there and they work beautiful. They event implement a scalar change range into the displacement node. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com I'll repost what I said in the other thread We started using Redshift back in March and pretty much use it exclusively now. Of course it all depends on the needs of the project (and there are still some real limitations). The RS dev team is top notch though. I'm really excited to see how things will be at the end of this calendar year. Redshift development is progressing at a fantastic rate, and the pricing is very competitive. For facilities, even small ones, it does require that you spend some time considering your hardware and infrastructure, especially if you want to start converting CPU farms for GPU rendering, or augmenting them. Fortunately, Redshift isn't licensed per GPU, but per machine, and that should provide some breathing room. To be honest (and I realize we have many Arnold folks here), here at Magnetic we evaluated our rendering options (MR, vRay, 3Delight, Arnold). I even started working on a Soft-to-Modo pipeline. Among these Arnold was the clear winner. That said, we felt that to be useful for us in production, Arnold was too costly a solution for us, both financially and in render time, *considering the kinds of projects we do* . Then Redshift came along and despite its infancy, really turned our heads. We cautiously began using it on productions, and it has since proven itself for us. Again, it all depends on what kind of project you're working on! You need to evaluate it for yourself of course, but for smaller houses like us, it allows us to produce better looking content faster, while staying in Softimage. And in this economy, we can't argue with that. -Tim On 1/8/2014 2:53 PM, Paul Griswold wrote: I've been using it along side Arnold for quite a while now. I just finished a project for CES entirely in Redshift. I think Redshift falls more into the category of a VRay competitor rather than Arnold. Redshift isn't open the way Arnold is I don't think they intend it to be. I've found it to be extremely fast and stable in most cases, but being just out of alpha there are times when it's warts show. For $100, it's certainly worth a test drive. -Paul ᐧ On Wed, Jan 8, 2014 at 3:46 PM, Stephen Davidson magic...@bellsouth.net wrote: I just wanted to start this thread so we can kill the Sofimage is dead thread. :) So..Redshift3D enthusiasts...post here. I think we will benefit the group, as a whole, if we spread the word on this amazing GPU rendering company. They have made my workflow SO much faster. -- Best Regards, * Stephen P. Davidson* *(954) 552-7956 %28954%29%20552-7956 * sdavid...@3danimationmagic.com *Any sufficiently advanced technology is indistinguishable from magic* - Arthur C. Clarke http://www.3danimationmagic.com -- *Tim Crowson **Lead CG Artist* *Magnetic Dreams, Inc. *2525 Lebanon Pike, Bldg C, Suite 101, Nashville, TN 37214 *Ph* 615.885.6801 | *Fax* 615.889.4768 | www.magneticdreams.com tim.crow...@magneticdreams.com *Confidentiality Notice: This email, including attachments, is confidential and should not be used by anyone who is not the original intended recipient(s). If you have received this e-mail in error please inform the sender and delete it from your mailbox or any other storage mechanism. Magnetic Dreams, Inc cannot accept liability for any statements made which are clearly the sender's own and not expressly made on behalf of Magnetic Dreams, Inc or one of its agents.* --
Re: Redshift3D Render
For a current list of features available as well as a roadmap, I would like to suggest to just go and give it a free try: https://www.redshift3d.com/get-redshift Yes, actually you don´t even have to commit to spending $100 directly, the Free Beta Trial gives you 30 days of full access to Redshift. A special benefit of this free trial option is that you could actually try out how a bunch of machines would run using redshift in a farm or knot. Reading the docs doesn´t require a login: http://docs.redshift3d.com/Default.html Redshift is a really well balanced renderer and I wholehartedly trust in it´s success. With the above opportunity available it is a good time to test it in your production scenario and wheight it against VRay and Arnold, which are also both very nice plattforms with enough momentum to also be around for quite a while. I am sure Redshift is a valuable addition to that arsenal. Cheers, tim On 08.01.2014 22:44, Emilio Hernandez wrote: Backing up Tim, in the forums there is actually a hair test in Softimage with a simple phong shader. And IMHO it looks nice. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com mailto:tim.crow...@magneticdreams.com They've stated pretty clearly that Hair and Strand support is the next big thing to come... shouldn't be too long now... -Tim On 1/8/2014 3:36 PM, Daniel Kim wrote: I found some weird result of displacement map with Redshift. Elevation is okay, but sometimes I could see weird connection of UVM. All UVM boders wasn't smooth and I had no idea how to fix it. Arnold has that option though : / But more option I need is... hair @__@ and ICE strand --- Daniel Kim Animation Director Professional 3D Generalist http://www.danielkim3d.com --- On Thu, Jan 9, 2014 at 10:31 AM, Emilio Hernandez emi...@e-roja.com mailto:emi...@e-roja.com wrote: Displacement and bump map are there and they work beautiful. They event implement a scalar change range into the displacement node. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com mailto:tim.crow...@magneticdreams.com I'll repost what I said in the other thread We started using Redshift back in March and pretty much use it exclusively now. Of course it all depends on the needs of the project (and there are still some real limitations). The RS dev team is top notch though. I'm really excited to see how things will be at the end of this calendar year. Redshift development is progressing at a fantastic rate, and the pricing is very competitive. For facilities, even small ones, it does require that you spend some time considering your hardware and infrastructure, especially if you want to start converting CPU farms for GPU rendering, or augmenting them. Fortunately, Redshift isn't licensed per GPU, but per machine, and that should provide some breathing room. To be honest (and I realize we have many Arnold folks here), here at Magnetic we evaluated our rendering options (MR, vRay, 3Delight, Arnold). I even started working on a Soft-to-Modo pipeline. Among these Arnold was the clear winner. That said, we felt that to be useful for us in production, Arnold was too costly a solution for us, both financially and in render time, /considering the kinds of projects we do/ . Then Redshift came along and despite its infancy, really turned our heads. We cautiously began using it on productions, and it has since proven itself for us. Again, it all depends on what kind of project you're working on! You need to evaluate it for yourself of course, but for smaller houses like us, it allows us to produce better looking content faster, while staying in Softimage. And in this economy, we can't argue with that. -Tim On 1/8/2014 2:53 PM, Paul Griswold wrote: I've been using it along side Arnold for quite a while now. I just finished a project for CES entirely in Redshift. I think Redshift falls more into the category of a VRay competitor rather than Arnold. Redshift isn't open the way Arnold is I don't think they intend it to be. I've found it to be extremely fast and stable in most cases, but being just out of alpha there are times when it's warts show. For $100, it's certainly worth a test drive. -Paul ᐧ On Wed, Jan 8, 2014 at 3:46 PM, Stephen Davidson magic...@bellsouth.net mailto:magic...@bellsouth.net wrote: I just wanted to start this thread so we can kill the Sofimage is dead thread. :) So..Redshift3D enthusiasts...post here. I think we will benefit
Re: Redshift3D Render
I'm pretty excited about GPU rendering as a concept and this is the first time I've tried it in a DCC app. I have to say it's really nice to be doing Render Region renders and never peg out the CPU. I typically have a lot of things open and switch around a lot and having the CPU free to run other things is a big plus. On Wed, Jan 8, 2014 at 5:13 PM, Tim Leydecker bauero...@gmx.de wrote: For a current list of features available as well as a roadmap, I would like to suggest to just go and give it a free try: https://www.redshift3d.com/get-redshift Yes, actually you don´t even have to commit to spending $100 directly, the Free Beta Trial gives you 30 days of full access to Redshift. A special benefit of this free trial option is that you could actually try out how a bunch of machines would run using redshift in a farm or knot. Reading the docs doesn´t require a login: http://docs.redshift3d.com/Default.html Redshift is a really well balanced renderer and I wholehartedly trust in it´s success. With the above opportunity available it is a good time to test it in your production scenario and wheight it against VRay and Arnold, which are also both very nice plattforms with enough momentum to also be around for quite a while. I am sure Redshift is a valuable addition to that arsenal. Cheers, tim On 08.01.2014 22:44, Emilio Hernandez wrote: Backing up Tim, in the forums there is actually a hair test in Softimage with a simple phong shader. And IMHO it looks nice. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com mailto:tim.crowson@ magneticdreams.com They've stated pretty clearly that Hair and Strand support is the next big thing to come... shouldn't be too long now... -Tim On 1/8/2014 3:36 PM, Daniel Kim wrote: I found some weird result of displacement map with Redshift. Elevation is okay, but sometimes I could see weird connection of UVM. All UVM boders wasn't smooth and I had no idea how to fix it. Arnold has that option though : / But more option I need is... hair @__@ and ICE strand --- Daniel Kim Animation Director Professional 3D Generalist http://www.danielkim3d.com --- On Thu, Jan 9, 2014 at 10:31 AM, Emilio Hernandez emi...@e-roja.commailto: emi...@e-roja.com wrote: Displacement and bump map are there and they work beautiful. They event implement a scalar change range into the displacement node. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com mailto: tim.crow...@magneticdreams.com I'll repost what I said in the other thread We started using Redshift back in March and pretty much use it exclusively now. Of course it all depends on the needs of the project (and there are still some real limitations). The RS dev team is top notch though. I'm really excited to see how things will be at the end of this calendar year. Redshift development is progressing at a fantastic rate, and the pricing is very competitive. For facilities, even small ones, it does require that you spend some time considering your hardware and infrastructure, especially if you want to start converting CPU farms for GPU rendering, or augmenting them. Fortunately, Redshift isn't licensed per GPU, but per machine, and that should provide some breathing room. To be honest (and I realize we have many Arnold folks here), here at Magnetic we evaluated our rendering options (MR, vRay, 3Delight, Arnold). I even started working on a Soft-to-Modo pipeline. Among these Arnold was the clear winner. That said, we felt that to be useful for us in production, Arnold was too costly a solution for us, both financially and in render time, /considering the kinds of projects we do/ . Then Redshift came along and despite its infancy, really turned our heads. We cautiously began using it on productions, and it has since proven itself for us. Again, it all depends on what kind of project you're working on! You need to evaluate it for yourself of course, but for smaller houses like us, it allows us to produce better looking content faster, while staying in Softimage. And in this economy, we can't argue with that. -Tim On 1/8/2014 2:53 PM, Paul Griswold wrote: I've been using it along side Arnold for quite a while now. I just finished a project for CES entirely in Redshift. I think Redshift falls more into the category of a VRay competitor rather than Arnold. Redshift isn't open the way Arnold is I don't think they intend it to be. I've found it to be extremely fast and stable in most cases, but being just out of alpha there
Re: Redshift3D Render
Call me immature, but i kind of love the idea of my gaming pc also being my render farm :P Sod of Nvidia and your flaby Quadro cards there expensive as fuck and you can't play Crysis on them :) On 8 January 2014 23:13, Tim Leydecker bauero...@gmx.de wrote: For a current list of features available as well as a roadmap, I would like to suggest to just go and give it a free try: https://www.redshift3d.com/get-redshift Yes, actually you don´t even have to commit to spending $100 directly, the Free Beta Trial gives you 30 days of full access to Redshift. A special benefit of this free trial option is that you could actually try out how a bunch of machines would run using redshift in a farm or knot. Reading the docs doesn´t require a login: http://docs.redshift3d.com/Default.html Redshift is a really well balanced renderer and I wholehartedly trust in it´s success. With the above opportunity available it is a good time to test it in your production scenario and wheight it against VRay and Arnold, which are also both very nice plattforms with enough momentum to also be around for quite a while. I am sure Redshift is a valuable addition to that arsenal. Cheers, tim On 08.01.2014 22:44, Emilio Hernandez wrote: Backing up Tim, in the forums there is actually a hair test in Softimage with a simple phong shader. And IMHO it looks nice. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com mailto:tim.crowson@ magneticdreams.com They've stated pretty clearly that Hair and Strand support is the next big thing to come... shouldn't be too long now... -Tim On 1/8/2014 3:36 PM, Daniel Kim wrote: I found some weird result of displacement map with Redshift. Elevation is okay, but sometimes I could see weird connection of UVM. All UVM boders wasn't smooth and I had no idea how to fix it. Arnold has that option though : / But more option I need is... hair @__@ and ICE strand --- Daniel Kim Animation Director Professional 3D Generalist http://www.danielkim3d.com --- On Thu, Jan 9, 2014 at 10:31 AM, Emilio Hernandez emi...@e-roja.commailto: emi...@e-roja.com wrote: Displacement and bump map are there and they work beautiful. They event implement a scalar change range into the displacement node. 2014/1/8 Tim Crowson tim.crow...@magneticdreams.com mailto: tim.crow...@magneticdreams.com I'll repost what I said in the other thread We started using Redshift back in March and pretty much use it exclusively now. Of course it all depends on the needs of the project (and there are still some real limitations). The RS dev team is top notch though. I'm really excited to see how things will be at the end of this calendar year. Redshift development is progressing at a fantastic rate, and the pricing is very competitive. For facilities, even small ones, it does require that you spend some time considering your hardware and infrastructure, especially if you want to start converting CPU farms for GPU rendering, or augmenting them. Fortunately, Redshift isn't licensed per GPU, but per machine, and that should provide some breathing room. To be honest (and I realize we have many Arnold folks here), here at Magnetic we evaluated our rendering options (MR, vRay, 3Delight, Arnold). I even started working on a Soft-to-Modo pipeline. Among these Arnold was the clear winner. That said, we felt that to be useful for us in production, Arnold was too costly a solution for us, both financially and in render time, /considering the kinds of projects we do/ . Then Redshift came along and despite its infancy, really turned our heads. We cautiously began using it on productions, and it has since proven itself for us. Again, it all depends on what kind of project you're working on! You need to evaluate it for yourself of course, but for smaller houses like us, it allows us to produce better looking content faster, while staying in Softimage. And in this economy, we can't argue with that. -Tim On 1/8/2014 2:53 PM, Paul Griswold wrote: I've been using it along side Arnold for quite a while now. I just finished a project for CES entirely in Redshift. I think Redshift falls more into the category of a VRay competitor rather than Arnold. Redshift isn't open the way Arnold is I don't think they intend it to be. I've found it to be extremely fast and stable in most cases, but being just out of alpha there are times when it's warts show. For $100, it's certainly worth a test drive. -Paul ᐧ