Re: [Flightgear-devel] Final (?) 3D clouds patch
On lundi 08 décembre 2008, Stuart Buchanan wrote: Curt wrote: I wonder if there is some sort of floating point resolution / rounding problem with the sort? I see a lot of flickering myself. Also if I look some particular direction and the clouds get sorted ok, then look away for even a second, and then look back (by changing the view direction) the clouds seem to have totally lost their previous correct sort and need to be sorted again ... but that doesn't happen until the clouds come back in view. I'm not sure what the sort criteria is, but it seems strange that the sort order would get messed up in a brief second of not having a particular set of clouds in view. One of the performance improvements introduced with the last patch is to re-use clouds rather than generating unique clouds for each position. This makes a significant performance difference on my system (about 5fps IIRC), but means that occasionally you end up between two instances of the same cloud, and the sprites need resorting as you change the view. You can avoid this by increasing the number of cloud types in the Rendering Options dialog, then toggling Enable 3D clouds to cause re-generation. At the moment the slider goes to 20, which minimizes the probability of hitting this issue. Obviously, we could increase the limit still further if required. -Stuart Hello Stuart, First thanks to you and to everybody who work on it. We have got now a huge improvement. However i am , now, a bit disappointed with these flickering ( too often, may be it is my graphics cards NVIDIA 7800 GS 512 mo) which decrease dramatically the result (mainly with Metar) . And which make me to come back to the 2D clouds, with the GUI preference. I can understand that, with that last patch you have tried to answer to the low cpu performance, when we are using old computers. It was the case with the old 3D clouds version and nobody apologized about it since we had ever the choice 2D clouds or 3D clouds. I did like better one of your previous version. Yes we had that ugly blue edge ( sometime only when we had a 2d Clouds layer behind), but now, i have blue edge randomly coming on and flickering, which is worth. Wont it be possible to have a version which will come to a better eye candy, even if it is more cpu eater. ? Greeting -- Gérard http://pagesperso-orange.fr/GRTux/ J'ai décidé d'être heureux parce que c'est bon pour la santé. Voltaire -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Final (?) 3D clouds patch
gerard robin wrote: However i am , now, a bit disappointed with these flickering ( too often, may be it is my graphics cards NVIDIA 7800 GS 512 mo) which decrease dramatically the result (mainly with Metar) . And which make me to come back to the 2D clouds, with the GUI preference. I can understand that, with that last patch you have tried to answer to the low cpu performance, when we are using old computers. It was the case with the old 3D clouds version and nobody apologized about it since we had ever the choice 2D clouds or 3D clouds. I did like better one of your previous version. Yes we had that ugly blue edge ( sometime only when we had a 2d Clouds layer behind), but now, i have blue edge randomly coming on and flickering, which is worth. Wont it be possible to have a version which will come to a better eye candy, even if it is more cpu eater. ? Hi Gerard, Thanks for the feedback - much appreciated. I'll put in an extra option this evening so you can toggle whether the number of cloud types is limited or not. -Stuart -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Final (?) 3D clouds patch
dave perry wrote: You were correct. I had not set the weather scenario to METAR. I ran fgfs once with 3D clouds and once w/o 3D clouds, both with real-weather-fetch and scenario METAR. I only got 1 fps with the 3D clouds. Earlier with 3D clouds, I got about 21 fps. I assume you mean Earlier with 2D clouds, I got about 21fps ? That's very low. I'd expect a drop of about 10fps. What graphics card are you using? Also for both 2D and 3D clouds, the field elevation is not accounted for in applying the cloud base MSL height. The METAR for these 2 runs showed broken at 011 (translates to 1,100 ft AGL) but leaving KDSM field elevation of 957 ft MSL, I was in the clouds by 1100 ft MSL or only about 150 ft AGL. Are we not applying the metar field elevation + metar AGL to get the cloud level? This sounds like a bug, though I thought I saw something adding the field elevation. I'll check. Thanks for pointing it out. -Stuart -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Final (?) 3D clouds patch
Maybe add an option to do full length bubble sort each update?Or quicksort, even. It should really be that much of a cpu eater, after the clouds are sorted. It would result in a bit of frame drop on pan around, but flickering would disappear. For users with modern system it would improve visuals. I'll test that later. greetings, yon On Wed, Dec 10, 2008 at 10:26 AM, gerard robin [EMAIL PROTECTED] wrote: On lundi 08 décembre 2008, Stuart Buchanan wrote: Curt wrote: I wonder if there is some sort of floating point resolution / rounding problem with the sort? I see a lot of flickering myself. Also if I look some particular direction and the clouds get sorted ok, then look away for even a second, and then look back (by changing the view direction) the clouds seem to have totally lost their previous correct sort and need to be sorted again ... but that doesn't happen until the clouds come back in view. I'm not sure what the sort criteria is, but it seems strange that the sort order would get messed up in a brief second of not having a particular set of clouds in view. One of the performance improvements introduced with the last patch is to re-use clouds rather than generating unique clouds for each position. This makes a significant performance difference on my system (about 5fps IIRC), but means that occasionally you end up between two instances of the same cloud, and the sprites need resorting as you change the view. You can avoid this by increasing the number of cloud types in the Rendering Options dialog, then toggling Enable 3D clouds to cause re-generation. At the moment the slider goes to 20, which minimizes the probability of hitting this issue. Obviously, we could increase the limit still further if required. -Stuart Hello Stuart, First thanks to you and to everybody who work on it. We have got now a huge improvement. However i am , now, a bit disappointed with these flickering ( too often, may be it is my graphics cards NVIDIA 7800 GS 512 mo) which decrease dramatically the result (mainly with Metar) . And which make me to come back to the 2D clouds, with the GUI preference. I can understand that, with that last patch you have tried to answer to the low cpu performance, when we are using old computers. It was the case with the old 3D clouds version and nobody apologized about it since we had ever the choice 2D clouds or 3D clouds. I did like better one of your previous version. Yes we had that ugly blue edge ( sometime only when we had a 2d Clouds layer behind), but now, i have blue edge randomly coming on and flickering, which is worth. Wont it be possible to have a version which will come to a better eye candy, even if it is more cpu eater. ? Greeting -- Gérard http://pagesperso-orange.fr/GRTux/ J'ai décidé d'être heureux parce que c'est bon pour la santé. Voltaire -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Final (?) 3D clouds patch
Stuart Buchanan wrote: dave perry wrote: You were correct. I had not set the weather scenario to METAR. I ran fgfs once with 3D clouds and once w/o 3D clouds, both with real-weather-fetch and scenario METAR. I only got 1 fps with the 3D clouds. Earlier with 3D clouds, I got about 21 fps. I assume you mean Earlier with 2D clouds, I got about 21fps ? No, that was with 3D clouds. But not via real weather fetch. That's very low. I'd expect a drop of about 10fps. What graphics card are you using? My graphics card: BFG GeForce 7800 GS OC which is an AGP 8x 256MB GDDR3 My system is an AMD Athlon XP 3200+ with 2G of ram. I typically get 70 to 80 fps with 2 D clouds and I had the frame rate throttled to 30 fps when I got the 1 fps with the latest 3D clouds. But I just tried running with 3D clouds and real weather fetch at KLMO and got 16 to 21 fps. Also for both 2D and 3D clouds, the field elevation is not accounted for in applying the cloud base MSL height. The METAR for these 2 runs showed broken at 011 (translates to 1,100 ft AGL) but leaving KDSM field elevation of 957 ft MSL, I was in the clouds by 1100 ft MSL or only about 150 ft AGL. Are we not applying the metar field elevation + metar AGL to get the cloud level? This sounds like a bug, though I thought I saw something adding the field elevation. I'll check. Thanks for pointing it out. -Stuart -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Final (?) 3D clouds patch
Hi, My graphics card: BFG GeForce 7800 GS OC which is an AGP 8x 256MB GDDR3 My system is an AMD Athlon XP 3200+ with 2G of ram. I typically get 70 to 80 fps with 2 D clouds and I had the frame rate throttled to 30 fps when I got the 1 fps with the latest 3D clouds. But I just tried running with 3D clouds and real weather fetch at KLMO and got 16 to 21 fps. Let me guess: the real weather showed a complete overcast with something looking like nimbustratus? Yes, with that I got the same result. But with fair weather I have no problems. Maybe we have some cloud-modellers beside me here, who can edit the clouds with textures and by xml which may improve the perfomance. That's why it is customizable... -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Final (?) 3D clouds patch
Hi, Hi All, Attached is what I'm hoping will be the final 3D clouds patch. It does the following: - Replaces simple shader attributes with vectors (this was missed out of the last patch by mistake) - Includes Yon's Fog update code (Thanks!) - Fixes a bug since 1.0 where --enable-real-weather-fetch stopped the other weather scenarios from working. Let me know of any bugs. -Stuart Thanks for all your work, Yon and Stuart! The clouds looking great now- the order problem is 99% solved so much as I can see! I see only some few problems still: -against a second 3d-clouds layer, the problem with z-drawing appears again -setting thunderstorm: the clouds has this transparency problem again, perfomance is weak, no lightning and thunder back (o.k. missing feature) maybe (just an idea) we can create a special set of a thunderstorm which is loaded instead the usual set which seems to be changed for fitting. -ns and st-layers covers now full, but also decrease dramatically the fps. Even on my fast machine I only get around 8-12fps- not usuable! Maybe it helps a bit if we couple the cloud visibility range with the generell visibility. Or we should use of the old, 2d-layers instead! Cheers HHS -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Final (?) 3D clouds patch
Heiko wrote: The clouds looking great now- the order problem is 99% solved so much as I can see! Yes - I think we're pretty much done. I see only some few problems still: -against a second 3d-clouds layer, the problem with z-drawing appears again I don't know how to solve this at the moment. Sorry :( -setting thunderstorm: the clouds has this transparency problem again, perfomance is weak, no lightning and thunder back (o.k. missing feature) maybe (just an idea) we can create a special set of a thunderstorm which is loaded instead the usual set which seems to be changed for fitting. The Thuderstorm scenario has a very specific METAR. We could easily change this to something that looks better. One of the enhancements I'd like to make after the release is to allow the scenario METAR strings to be defined in a properties file, so a user can save METARs they want to fly in the future. -ns and st-layers covers now full, but also decrease dramatically the fps. Even on my fast machine I only get around 8-12fps- not usuable! Maybe it helps a bit if we couple the cloud visibility range with the generell visibility. Or we should use of the old, 2d-layers instead! I don't think we're likely to find massive performance improvements, so I think using 2-D layers for stratus might be the best solution for the moment. This is very easy to do : The 3D cloud code defaults to a 2D layer if it can't find an appropriate cloud definition in the cloudlayers.xml file. So, you can simple remove the st and ns sections from cloudlayers.xml. If you agree, I'll look at doing that tonight. This may also help with the Thunderstorm scenario. -Stuart -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Final (?) 3D clouds patch
Heiko Schulz wrote: Hi, Hi All, Attached is what I'm hoping will be the final 3D clouds patch. It does the following: - Replaces simple shader attributes with vectors (this was missed out of the last patch by mistake) - Includes Yon's Fog update code (Thanks!) - Fixes a bug since 1.0 where --enable-real-weather-fetch stopped the other weather scenarios from working. Let me know of any bugs. -Stuart Thanks for all your work, Yon and Stuart! The clouds looking great now- the order problem is 99% solved so much as I can see! I see only some few problems still: -against a second 3d-clouds layer, the problem with z-drawing appears again -setting thunderstorm: the clouds has this transparency problem again, perfomance is weak, no lightning and thunder back (o.k. missing feature) maybe (just an idea) we can create a special set of a thunderstorm which is loaded instead the usual set which seems to be changed for fitting. -ns and st-layers covers now full, but also decrease dramatically the fps. Even on my fast machine I only get around 8-12fps- not usuable! Maybe it helps a bit if we couple the cloud visibility range with the generell visibility. Or we should use of the old, 2d-layers instead! The frame rate is doing interesting things here. When I select 3d clouds, the frame rate drops from 28 to 15 at KSFO, but when I deselect the clouds the frame rate goes back only to 23, and when I reselect 3d clouds the frame rate goes to 7 and stays there, whatever I do. The 3d stratus looks horrid, sorry. I think the 2d solution is just fine for stratus, and for cirrus as well. Vivian -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Final (?) 3D clouds patch
On lundi 08 décembre 2008, Stuart Buchanan wrote: Heiko wrote: The clouds looking great now- the order problem is 99% solved so much as I can see! Yes - I think we're pretty much done. I see only some few problems still: -against a second 3d-clouds layer, the problem with z-drawing appears again I don't know how to solve this at the moment. Sorry :( -setting thunderstorm: the clouds has this transparency problem again, perfomance is weak, no lightning and thunder back (o.k. missing feature) maybe (just an idea) we can create a special set of a thunderstorm which is loaded instead the usual set which seems to be changed for fitting. The Thuderstorm scenario has a very specific METAR. We could easily change this to something that looks better. You answer to my previous question ( with the snapshot) on that topic . The blue edges are not on purpose :) One of the enhancements I'd like to make after the release is to allow the scenario METAR strings to be defined in a properties file, so a user can save METARs they want to fly in the future. SNIP -Stuart Cheers -- Gérard http://pagesperso-orange.fr/GRTux/ J'ai décidé d'être heureux parce que c'est bon pour la santé. Voltaire -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Final (?) 3D clouds patch
On lundi 08 décembre 2008, Stuart Buchanan wrote: Heiko wrote: The clouds looking great now- the order problem is 99% solved so much as I can see! Yes - I think we're pretty much done. I see only some few problems still: -against a second 3d-clouds layer, the problem with z-drawing appears again I don't know how to solve this at the moment. Sorry :( -setting thunderstorm: the clouds has this transparency problem again, perfomance is weak, no lightning and thunder back (o.k. missing feature) maybe (just an idea) we can create a special set of a thunderstorm which is loaded instead the usual set which seems to be changed for fitting. The Thuderstorm scenario has a very specific METAR. We could easily change this to something that looks better. One of the enhancements I'd like to make after the release is to allow the scenario METAR strings to be defined in a properties file, so a user can save METARs they want to fly in the future. -ns and st-layers covers now full, but also decrease dramatically the fps. Even on my fast machine I only get around 8-12fps- not usuable! Maybe it helps a bit if we couple the cloud visibility range with the generell visibility. Or we should use of the old, 2d-layers instead! I don't think we're likely to find massive performance improvements, so I think using 2-D layers for stratus might be the best solution for the moment. This is very easy to do : The 3D cloud code defaults to a 2D layer if it can't find an appropriate cloud definition in the cloudlayers.xml file. So, you can simple remove the st and ns sections from cloudlayers.xml. If you agree, I'll look at doing that tonight. This may also help with the Thunderstorm scenario. -Stuart An other snapshoot, feedback, with metar. Getting the same blue edge than with thunderstorm. May be it is my graphic card http://pagesperso-orange.fr/GRTux/3DClouds-img9.jpg -- Gérard http://pagesperso-orange.fr/GRTux/ J'ai décidé d'être heureux parce que c'est bon pour la santé. Voltaire -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Final (?) 3D clouds patch
gerard robin wrote: On lundi 08 décembre 2008, Stuart Buchanan wrote: Heiko wrote: The clouds looking great now- the order problem is 99% solved so much as I can see! Yes - I think we're pretty much done. I see only some few problems still: -against a second 3d-clouds layer, the problem with z-drawing appears again I don't know how to solve this at the moment. Sorry :( -setting thunderstorm: the clouds has this transparency problem again, perfomance is weak, no lightning and thunder back (o.k. missing feature) maybe (just an idea) we can create a special set of a thunderstorm which is loaded instead the usual set which seems to be changed for fitting. The Thuderstorm scenario has a very specific METAR. We could easily change this to something that looks better. You answer to my previous question ( with the snapshot) on that topic . The blue edges are not on purpose :) One of the enhancements I'd like to make after the release is to allow the scenario METAR strings to be defined in a properties file, so a user can save METARs they want to fly in the future. SNIP -Stuart Cheers The 3D cloud appearance is much improved. Thanks to all involved! Several questions and comments. 1. At night, the emmissive seems very very bright. 2. Are you intending that the 3D cloud base should match the lowest level in the current METAR? I just flew with a KDSM METAR using real weather fetch (current METAR copied from ADDS:* KDSM 081954Z 10007KT 10SM BKN130 OVC160 01/M03 A2964 RMK AO2 SLP047 T00111033. * ) This gives a broken layer at 13000 ft AGL but the 3D clouds started at 2000 AGL. 3. When I took off, the outside view showed the clouds flickering. -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Final (?) 3D clouds patch
On Mon, Dec 8, 2008 at 2:18 PM, dave perry [EMAIL PROTECTED] wrote: The 3D cloud appearance is much improved. Thanks to all involved! Several questions and comments. 1. At night, the emmissive seems very very bright. 2. Are you intending that the 3D cloud base should match the lowest level in the current METAR? I just flew with a KDSM METAR using real weather fetch (current METAR copied from ADDS:* KDSM 081954Z 10007KT 10SM BKN130 OVC160 01/M03 A2964 RMK AO2 SLP047 T00111033. * ) This gives a broken layer at 13000 ft AGL but the 3D clouds started at 2000 AGL. 3. When I took off, the outside view showed the clouds flickering. I wonder if there is some sort of floating point resolution / rounding problem with the sort? I see a lot of flickering myself. Also if I look some particular direction and the clouds get sorted ok, then look away for even a second, and then look back (by changing the view direction) the clouds seem to have totally lost their previous correct sort and need to be sorted again ... but that doesn't happen until the clouds come back in view. I'm not sure what the sort criteria is, but it seems strange that the sort order would get messed up in a brief second of not having a particular set of clouds in view. Regards, Curt. -- Curtis Olson: http://baron.flightgear.org/~curt/ -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Final (?) 3D clouds patch
Dave Perry wrote: The 3D cloud appearance is much improved. Thanks to all involved! Several questions and comments. 1. At night, the emmissive seems very very bright. 2. Are you intending that the 3D cloud base should match the lowest level in the current METAR? I just flew with a KDSM METAR using real weather fetch (current METAR copied from ADDS:* KDSM 081954Z 10007KT 10SM BKN130 OVC160 01/M03 A2964 RMK AO2 SLP047 T00111033. * ) This gives a broken layer at 13000 ft AGL but the 3D clouds started at 2000 AGL. Was the weather scenario set to METAR as well - one of the bugs I fixed with the latest patch was that previously --enable-real-weather-fetch over-wrote the various scenarios. Now, you will only get METAR if you have METAR as the scenario, as well as --enable-real-weather-fetch. -Stuart -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Final (?) 3D clouds patch
Curt wrote: I wonder if there is some sort of floating point resolution / rounding problem with the sort? I see a lot of flickering myself. Also if I look some particular direction and the clouds get sorted ok, then look away for even a second, and then look back (by changing the view direction) the clouds seem to have totally lost their previous correct sort and need to be sorted again ... but that doesn't happen until the clouds come back in view. I'm not sure what the sort criteria is, but it seems strange that the sort order would get messed up in a brief second of not having a particular set of clouds in view. One of the performance improvements introduced with the last patch is to re-use clouds rather than generating unique clouds for each position. This makes a significant performance difference on my system (about 5fps IIRC), but means that occasionally you end up between two instances of the same cloud, and the sprites need resorting as you change the view. You can avoid this by increasing the number of cloud types in the Rendering Options dialog, then toggling Enable 3D clouds to cause re-generation. At the moment the slider goes to 20, which minimizes the probability of hitting this issue. Obviously, we could increase the limit still further if required. -Stuart -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Final (?) 3D clouds patch
Stuart Buchanan a écrit : Was the weather scenario set to METAR as well - one of the bugs I fixed with the latest patch was that previously --enable-real-weather-fetch over-wrote the various scenarios. Now, you will only get METAR if you have METAR as the scenario, as well as --enable-real-weather-fetch. -Stuart Hi, just to say something about real weather fetch and METAR, it seems to me that metar information are only used once, after thet next metar update is not taken into account (in the different clouds layers or in /environment properties) , but i think you know this (i saw a TODO in fgclouds.cxx). an other thing is a concern about: /environment/temperature-sea-level-degc /environment/dewpoint-sea-level-degc and the temperatures properties in the clouds layers wich are not changed (always 15 and 5) . i tried this formula in FGClouds::update_env_config (): fgDefaultWeatherValue( temperature-degc,( fgGetDouble(/environment/metar/temperature-degc) + 0.0065 * 0.3048 * station_elevation_ft )) but my knowledge of c++ being close to 0, station elevation was not taken into account but temperature was updated. cheers jano -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Final (?) 3D clouds patch
Stuart Buchanan wrote: Dave Perry wrote: The 3D cloud appearance is much improved. Thanks to all involved! Several questions and comments. 1. At night, the emmissive seems very very bright. 2. Are you intending that the 3D cloud base should match the lowest level in the current METAR? I just flew with a KDSM METAR using real weather fetch (current METAR copied from ADDS:* KDSM 081954Z 10007KT 10SM BKN130 OVC160 01/M03 A2964 RMK AO2 SLP047 T00111033. * ) This gives a broken layer at 13000 ft AGL but the 3D clouds started at 2000 AGL. Was the weather scenario set to METAR as well - one of the bugs I fixed with the latest patch was that previously --enable-real-weather-fetch over-wrote the various scenarios. Now, you will only get METAR if you have METAR as the scenario, as well as --enable-real-weather-fetch. -Stuart You were correct. I had not set the weather scenario to METAR. I ran fgfs once with 3D clouds and once w/o 3D clouds, both with real-weather-fetch and scenario METAR. I only got 1 fps with the 3D clouds. Earlier with 3D clouds, I got about 21 fps. Also for both 2D and 3D clouds, the field elevation is not accounted for in applying the cloud base MSL height. The METAR for these 2 runs showed broken at 011 (translates to 1,100 ft AGL) but leaving KDSM field elevation of 957 ft MSL, I was in the clouds by 1100 ft MSL or only about 150 ft AGL. Are we not applying the metar field elevation + metar AGL to get the cloud level? Dave P. -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Final (?) 3D clouds patch
I wrote: Hi All, Attached is what I'm hoping will be the final 3D clouds patch. Nope, it wasn't attached, because I hit Send rather than Attach. This time it is attached. Sorry for the noise. It does the following: - Replaces simple shader attributes with vectors (this was missed out of the last patch by mistake) - Includes Yon's Fog update code (Thanks!) - Fixes a bug since 1.0 where --enable-real-weather-fetch stopped the other weather scenarios from working. Let me know of any bugs. -Stuart clouds.tar.gz Description: GNU Zip compressed data -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Final (?) 3D clouds patch
Hi, reading the patch, in the callback: fogC[3] = 0.0; I believe this sliped from my testing, not sure if it's needed. I was doing some alpha-channel tests. If it works with this set, all is fine :) greetings, yon On Sat, Dec 6, 2008 at 11:35 PM, Stuart Buchanan [EMAIL PROTECTED] wrote: I wrote: Hi All, Attached is what I'm hoping will be the final 3D clouds patch. Nope, it wasn't attached, because I hit Send rather than Attach. This time it is attached. Sorry for the noise. It does the following: - Replaces simple shader attributes with vectors (this was missed out of the last patch by mistake) - Includes Yon's Fog update code (Thanks!) - Fixes a bug since 1.0 where --enable-real-weather-fetch stopped the other weather scenarios from working. Let me know of any bugs. -Stuart -- SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel