Hey Chris Thanks for extensive cover. That is first concrete post with specific points that makes sense. Still stays question of killing Softimage in it's prime (even if it stagnating and Maya is on developers rise) before covering at least thsoe basic workflow things. All underlying tech is great stuff but have little meaning if it is disconnected from user on top. Tech makes fast viewport and faster then ever performance possible and workflow then cut that all in half or more. IF in two years is possible to improve what Maya is lacking in terms of workflow then it does makes sense. But if it takes another 5-6 years then Softimage is killed prematurely.
On Sat, Mar 15, 2014 at 12:58 PM, Chris Vienneau < [email protected]> wrote: > Hi Mirko, > > > > We have already shown the maya roadmap to about 60% of the softimage user > base under NDA. I have had calls with several people on this forum last > week and I apologize to those that I have not got to already. We will > launch our new software line up this week and GDC is here with Nvidia GTC > and NAB all back to back so things are crazy. Anyone that needs more > information for their transition that has asked for a road map will get one > in the next couple of months. There are users in a variety of different > situations that have different things to weigh. > > > > The largest community of softimage users is in Japan building games and > most use the modeling and animation tools but very little ICE. This > community is focusing on the transition to the PS4 and Xbox One console > cycle and want features like DX11 APIs and .net programming as well as back > end API for FBX and Middleware. The main work we did with Microsoft and > Sony for supporting the next gen consoles was done in Maya although the > transition for a games developer is pretty big when they have to swap out > engines or back end pipelines. These users upgrade usually once every five > years around the time a console changes. With the work you will see in Maya > 2015 this week we are addressing most of the big modeling feature concerns > in terms of legacy things like crappy Booleans and bevel and we have a > whole new retopology tool set. We also have done tons of work on supporting > UV tiling and workflows and ported over the Unfold 3D work into Maya. > > > > As for the workflows we have an internal project called Project H (or > Humanize Maya) where we are working with all sorts of users from students > to pros to studios to come up with proposals to the problems that have come > up here and in the Maya user base. We invite anyone here and many of you > have taken up our offer to contribute and it is up to us to show that we > are delivering over the next two years during the transition period. If we > don't at the end than you will all have choices and plenty of time to > evaluate your options. > > > > As for ICE and Bifrost you will see this week the liquid workflows for > generalists. We have three user personas we work with: generalists, tds, > and programmers. If you want a great view of our thoughts on Bifrost than > read this article from fx guide: > https://www.fxguide.com/featured/bifrost-the-return-of-the-naiad-team-with-a-bridge-to-ice/. > The underlying procedural core that powers was written by a core group of > developers that worked on ICE except this time they built it separate from > Maya so that we could potentially re-use it in other areas. So we will show > the benefits of a modern procedural core running on all CPUs but we will > not expose the node graphing network underneath in the first version. As > you will read in the article and see in the videos the graph which is very > much like ICE is there but we want to focus on enabling a key workflow like > liquids first for the big user base and then expose it to tds through a > node graph interface and then programmers through an API. We have been > working on this core for a long time so we are confident that it is ready > for prime time with the hundred of thousands of Maya users out there today. > Again those of you that want to take part in how we expose this procedural > core through a node interface are welcome to join the beta. > > > > As for modernizing Maya's core there are several components that make up > the core of an application. This is by no means exhaustive but for this > conversation let's imagine the primary components are UI, Viewport, Data > model, and Evaluation. Higher level things like API and scripting are also > in the equation. For Maya we ported the entire UI to QT over the last > couple of years and you will see us start to take advantage of that in the > 2015 release. The viewport 2.0 project will be completed in the 2015 > release so we will make it the default viewport. This supports all the > modern DX11 techniques like gpu tessellation, opensubdiv and will natively > display UV tiling, Ptex, Alembic streaming, and most elements of the > hypershade and good chunks of mental ray. With GPU caching and proper use > of the GPU we are loading huge scenes into Maya and showing them with a > much higher level of fidelity than ever before. With the new bifrost tiled > architecture we are showing 200-300 m particles in some cases and they are > interactive if you have a beefy enough gfx card. > > > > For data we have been working with a lot of the larger studios on a > project called open data. The alembic work we put into the open source came > out of this project and we are actively working on projects like USD ( > http://graphics.pixar.com/usd/) and the future of how materials are > stored openly (http://code.google.com/p/alembic/wiki/AbcMaterial) . The > key is that the way data is stored is not owned by any one application but > an open source framework so you can share data between applications, > renderers, and frameworks. If you want to see how this comes together and > benefits the user you can watch this video ( > http://www.youtube.com/watch?v=Qsn6i1YVZo8). This is also the future of > how people will interact on the cloud as you will be have scenes that have > objects with multiple representations including streamable assets. You can > author in a lightweight environment and then pass around changes that are > stored in open formats like json or xml and use whatever you want (Maya, > Houdini, Fabric, C++) to bring it all together and back end tools like > Shotgun can work with the data without having to load Maya. I don't think > anyone on this list including Paul from Fabric that the future for storing > data is in open standards that are supported by all tools and pipelines. > The other key elements of the open data framework is working with remote > data like drop box and autodesk 360. As more and more workflows go cloud > having scenes that can reference remote elements and having compute on > demand will be key for any facility that wants to keep costs in line with > the production. > > > > Evaluation is what you do with the data to get the resulting geometry, > pixels, manipulators, etc... . Generalists make up a vast majority of the > 3D user base. They like the 3dsmax, Maya, XSI model of being able to do > modeling, animation, fx, and lighting in one package and the key constraint > here is that you have to fit all of that into your memory. Each application > has its own way of exposing that to user from a fully procedural > environment like Houdini to mixed models like XSI, Maya, and Softimage to > more direct always move forward models like Modo. We are currently > evaluating how we can expose the history of Maya in a more operator like > stack UI like Softimage but the underlying feature set is there. ICE is a > mixed evaluation model with a modern procedural core plugged into a more > interactive traditional model. I won't talk about the future on an open > forum but the article written on fxguide shows a lot of the intent of what > we are trying to do and we have been very clear that we want to capture the > best elements of the ICE workflow with this new approach. For API/Scripting > Maya is the industry standard for extensibility. We increased our python > support and added .net (games industry standard) over the last few years. > Maya is a great tool to plug into a pipeline. We are very confident that as > we expose our backend procedural core to programmers people will have a > choice between Fabric, Houdini Engine, and Maya and will probably mix them > all together. The other work done on the API was to properly expose the > rendering API to make Maya much more friendly to third party renderers. Now > you can choose between Mental Ray, Vray, vue, Arnold, Renderman, Maxwell, > Caustic, Iray, Furry Ball, Red Shift, Gorilla, and Otoy Octane. There are a > few more hiding in the weeds that will come out this year that are great. > > > > So for modernizing the application you have to start from the ground up > and work your way towards the artist facing work. So in Maya 2015 as a > games artist you will see exactly what you will see in the console or > mobile device and load levels that match what can be seen on the xbox one > or PS4. For a film/vfx artist you can work with high resolution subdivided > meshes (open subdiv rocks) that have uv tiles painted in Mari and see a > much closer view of the final look and you can choose between tons of great > renderers that are not on Houdini or Modo or Soft. > > > > Where we need to work on is user workflow and hiding the complexity of > Maya. If you think of all the work that went into the work you see above > the team is making this a top priority now that we have a modern base layer > of UI, data, evaluation, and viewport on which to work. Maya sucks at > lighting and rendering compared to Soft and we need to fix that along with > a bunch of other things. If you want to hedge your bets than sign up for > the roadmap or continue to write in here. Maya's modeling community have > been working with us for two years to bring it up to speed with > applications like modo and we have monthly drops and a tight community that > keeps us honest. The videos in two days will be testament to that. > > > > As for the whole NDA thing I am not going to apologize that is how we have > to do business as that is how Adobe, Apple, and anyone else who is a public > company has to operate. If you want to see what happens when you don't work > that way look at what happened to Avid. ( > http://www.forbes.com/sites/bobbyowsinski/2014/02/27/major-changes-around-the-corner-for-music-as-avid-delists/) > They played fast and loose with promises and they couldn't be honest with > the people who invested in them. I am so tired of the whole public evil > corporation thing as investors are people like you and me who just want to > retire one day. Investment companies just keep these companies honest and > the laws put in place prevent disasters like what happened with Enron where > people were stealing money and getting away with it. We need to be > responsible. There are no investors making calls at Autodesk. Everyone in > this industry from the Foundry (google Caryle Group) to Fabric (Marc Petit > big investor) have people that fund them and expect a return for their > investment. Considering we have spoken with 60% of the user base already > and plan on hitting 80% under NDA by Siggraph I think we are getting the > best of both worlds. You as the user are informed about what we intend to > do and we are not breaking any laws. > > > > There are developers on soft right now and we will keep them there for two > years. The list of things we can do to help the transition is in their > hands and we will see what we can do. Just a reminder there is the > softimage webinar tomorrow online. > > > > thx. > > > > cv/ > > > > > > > > ________________________________ > From: [email protected] [ > [email protected]] on behalf of Mirko Jankovic [ > [email protected]] > Sent: Saturday, March 15, 2014 4:26 AM > To: [email protected] > Subject: Re: Idea- Just keep Mental Ray and FBX support - Softimage free > w/Maya or Max or any Suite. > > All that "we are fully developing Maya so it will be great just you see"is > pure PR crap. > >From couple sides is heard that bifrost which is one big argument on > their side is still too young to be usable. It is not even full grown fluid > simulation and to replace ICE it will take years.. even after, by AD's > plan, SI is long gone they will still be not near the level of ICE. Not to > mention all other aspects of Maya workflow which are SOOOO wrong that won/t > get to table at all for years to come. > > If they are so confident in Maya development they should let people see it > for them self and move by own choice. > But having whole road map so closed and under all NDAs is selling a cat in > the bag. You have no idea where is Maya going at all what is planned and > WHEN it will be available. > That is big pile of crap if you have to plan transition. > > > On Sat, Mar 15, 2014 at 8:56 AM, skuby <[email protected]<mailto: > [email protected]>> wrote: > The above has a severe typo: Should read: With that, I don't think > Autodesk will want to encourage new users to Softimage, so not selling new > licenses directly makes sense for them, when they want to be > selling/promoting Maya and presumably Max licenses. > > sorry, my error > > > On Sat, Mar 15, 2014 at 2:52 PM, skuby <[email protected]<mailto: > [email protected]>> wrote: > > """"Martin [email protected]<mailto:[email protected]> > > > > [https://mail.google.com/mail/images/cleardot.gif] > > [https://mail.google.com/mail/images/cleardot.gif] > [https://mail.google.com/mail/images/cleardot.gif] > > > With SI retired you can't buy licenses for SI anymore, it doesn't come > with Max or Maya, you can only use the licenses you have right now. You > should add "keep selling SI licenses" to your idea. > > Martin > Sent from my iPhone"""""""""""""" > > Thing is, Softimage IS now depreciated, like it or not, no petition will > likely change that, that should be clear, -it's not going to move forward > like it used to and it will eventually fully die. With that, I don't think > Autodesk will want to encourage new users to Softimage, so not selling new > licenses directly makes sense for them to continue selling Softimage > directly, when they want to be selling/promoting Maya and presumably Max > licenses, -(yes it makes sense to us but not for them and we aren't going > to win that one). But, to keep it updated with FBX/Crosswalk and Mental > Ray is a great transition boon for those interested in migrating to either > Maya or Max. > > This will allow people in the middle of a transition to easily return to > familiar territory and get work done fast and clean and go back and forth > with ease, at their own pace as they re-learn all of their skills as well > as totally new features in their new package of choice. > > Two years, now until 2016, may very well not be enough time for some > studios/freelancers if they are buried in much needed work and it may not > be so easy to get all of the functionality out of the interfaces they are > accustomed to integrated into Max/Maya in such a short time at a level that > would be acceptable to all users. This keeps Soft development in a > depreciated state, but gives a little bit more growing room than the 2016 > deadline which is rightfully ambitious. And none of this encourages new > Softimage users, which is I suspect a quality they are wisely after given > their total gameplan. > > > > On Sat, Mar 15, 2014 at 1:17 PM, skuby <[email protected]<mailto: > [email protected]>> wrote: > Ok you are very right on the SDK/Plug-in front, I really didn't think > about it like that. However, I think that dominated your response and your > accidentally throwing out the baby with the bathwater in effect. (without > question, no matter what, Softimage is depreciated but I'm trying to get > the best effect out of a transition here) > > Soft, as is, without plug-ins or new features, is STILL usable for games, > movies, commercials and more. Removing it as a stand-alone product, and > packing it SOLEY as a tie in bonus with Maya, Max or any Suite, ONLY > further promotes the packages AD wants to keep alive while appeasing the > small studios (Japan/Vancouver/etc.) and freelance users of Softimage quite > a bit better than just a 2016 cutoff. By keeping alive Crosswalk/FBX > functionality for some time longer, -modeling, animation, UV, rigging, etc. > can all still be done as is in Soft and if needed, easily sent over to > Maya, as that features work right now between those two packages quite well > (a little trickier with max as you don't get all the automated 2-way > crosswalk features last I checked but you still have FBX handy and a TD can > easily work with that). Keeping up to date Mental Ray, allows existing > pipe-lines that go in the other direction, or use Softimage entirely, able > to keep up and extend the transition period beyond 2016. Now, remember, > all the while anyone primarily using Softimage with these Crosswalk/Mental > Ray add-ons, are also going to have full up to date copies of at least one > of the 2 non-depreciated softwares that AD wants to continue and actively > promote at this time, which is only going to increase AD's revenue by > ensuring many Japanese/Vancouver studios and others stay in the Autodesk > family, and that alone will likely provide more than enough revenue that > would potentially be lost (even if only in small amounts) to the obvious > hoard of disgruntled orphans. > > I don't want to make this too long, but I had hoped by now, AD would have > already merged the 3 major userbases (I expect with a Maya base above all > else because big studios have a lot invested in existing pipelines with it > and that just makes financial sense) . I think the above will really help > with that transition, and they can promote in the future a refined Maya > with a well needed updated hypershade (Softimage's render tree kills the > current, dated, hypershade without question usability wise) and with a much > needed update to the interaction model/keyboard shortcut/user interface > that Softimage has over every other package out there, out of the box as a > default. When you tack on the upcoming ICE replacement, and the best is > done not to alienate hurt Softimage users, I think many if not most will > naturally migrate to a well refined Maya without much cajoling with the > above mentioned treats and perhaps some love given to the > animation/keying/rigging/weighting systems. > > Anyways, that's my peace, I put it out there with the best of intentions. > Mudbox is one of my favorites, I started with Max, moved onto Maya and > ended up with Softimage. And in the current forms I have to wait for the > future to move on from Softimage, because nothing, inside or outside of the > AD family is going to replace it today, and I want that extended for as > long as possible as Maya is refined into the obvious front-runner than > merges the 3 bases. > > > On Sat, Mar 15, 2014 at 11:19 AM, Matt Lind <[email protected] > <mailto:[email protected]>> wrote: > Short version: > > 1) Softimage supports more than just film/video. There are people working > in many industries which won't be served by mental ray and FBX updates. My > studio is a prime example (games). > > 2) The SDK is not open enough for plugin developers to do what you > suggest. Many of the 3rd party efforts to date usually required some > amount of help from the Softimage developers to make possible. Softimage > tended to implement new features, then expand the SDK to support those new > features. For 3rd party development, you need the SDK expanded ahead of the > features. In other words, while the SDK is mature, it's also not expanded > too far outside of what you can already do out of the box. New plugin > efforts would largely amount to using existing levers and switches instead > of making new levers and switches like you and everybody else would like. > > There's still plenty of tools that can be written, but probably not > anything earth shattering as the SDK is single threaded and some of the > components it leans on are getting long in the tooth too as Luc-Eric has > recently mentioned. Basically, Softimage is that car that still performs > really great, but has high mileage and about to enter that period where > major parts will need repairing/replacing making it a very expensive > upkeep. Since the driver hasn't seen that breakage firsthand yet, he's > under the impression there isn't a problem and the car can drive on > indefinitely into the sunset. > > Autodesk was at our site just as the press release was officially hitting > the wire. While no specifics were mentioned by Autodesk or our reseller in > our offices, my personal poking, prodding and reading between the lines > lead me to believe a lot of work (refactoring) has gone into Maya the past > few years to accommodate some of this burden in migration, but some (much) > work still has to come. I think we'll see parts of that work in Maya 2015 > and probably explains why the Softimage Montreal team has been busy for the > past 2 years on Maya instead of Softimage. I have no idea if there's any > truth to this, but it's what I've come to believe based on the sparse bits > and pieces I could put together, and witnessing how such migrations were > handled in the past. > > I'll share the rest of my thoughts from 21+ years of Softimage experience > in another thread at another time. Right now I'm neck deep in getting the > project out the door Ben Houston so kindly posted the other day. > > > > Matt > > > > > > > > > From: [email protected]<mailto: > [email protected]> [mailto: > [email protected]<mailto: > [email protected]>] On Behalf Of skuby > Sent: Friday, March 14, 2014 8:38 PM > To: [email protected]<mailto:[email protected] > > > Subject: Idea- Just keep Mental Ray and FBX support - Softimage free > w/Maya or Max or any Suite. > > Softimage is ready out of the box now with no new features added. Why > not, leave it as is, but continue to support new Mental Ray and FBX updates > so that it still continues to fit nicely into any pipeline and leave > continued development past 2016 at those simple two things. Leave the rest > up to plug-in developers and forever-more include Softimage as a free > throw-in with any Maya or Max or Suite purchase or rental, so there will be > many ways to own it or rent it while exclusively promoting your other > packages. The other great thing, since it can only now come with Max/Maya > or a Suite, is that Soft users are never a financial burden to their > employers since it will always be a free seat with any existing Max/Maya > licence. > > This could be a very easy, welcomed add-on to the existing Softimage > retirement announcement. Please please, consider this as an option > Autodesk and please, everyone, support and promote this idea to Autodesk, I > think it's our best bet and we might actually be able to get this. > > -regards > > > > >
