Check new water out.
by Phil Carlisle · in Torque Game Engine · 05/08/2002 (4:34 pm) · 8 replies
Guys (Melv) check out developer.nvidia.com for the new 3rd party demo's.
The water effect that guys done is incredible..
Of course, its crawling, so it wouldnt prolly work in torque, but damn if it dont look nice! :)
Phil.
The water effect that guys done is incredible..
Of course, its crawling, so it wouldnt prolly work in torque, but damn if it dont look nice! :)
Phil.
About the author
Recent Threads
#2
It's the problem I'm finding with my prototypes, cpu load and sheer graphics power needed. I've found some algorithms that work on a geForce2 with reasonable cpu load. They also have the ability to tessellate the surface which can reduce the computation load.
I'm trying to build on these to produce something that most people could use. It's damn hard work as I'm having to brush up on various math/rendering skills, hence my recent purchase of books totaling around
05/09/2002 (11:04 am)
It's certainly looks nice but I've seen **MUCH** faster implementations.It's the problem I'm finding with my prototypes, cpu load and sheer graphics power needed. I've found some algorithms that work on a geForce2 with reasonable cpu load. They also have the ability to tessellate the surface which can reduce the computation load.
I'm trying to build on these to produce something that most people could use. It's damn hard work as I'm having to brush up on various math/rendering skills, hence my recent purchase of books totaling around
#3
Others are catching up tho.
I must admit, that shader was slow as hell at work on my GF3, but it was pretty slick on the GF4 at home.
Looked good tho.. probably look a bit worse when used as a water plane.
Phil.
05/09/2002 (12:36 pm)
Yeah, thats true. Thats one of the issues Ive got right now really. I'm the only one i know with GF4 and Radeon 8500 level rendering available.Others are catching up tho.
I must admit, that shader was slow as hell at work on my GF3, but it was pretty slick on the GF4 at home.
Looked good tho.. probably look a bit worse when used as a water plane.
Phil.
#4
I'm in a lucky position that we need serious graphics power at work for most of the stuff we work on.
I just got my hands on a pair of nVidia Quadro4 900 XGLs and they seriously kick butt with my prototypes but at
05/09/2002 (1:27 pm)
Phil,I'm in a lucky position that we need serious graphics power at work for most of the stuff we work on.
I just got my hands on a pair of nVidia Quadro4 900 XGLs and they seriously kick butt with my prototypes but at
#5
1) Hardware first OGL
2) Hardware DX
3) Software (Quake3).
Or 1,3,2
or some other mix :)
I really think the Quake3 shader language will help for guys without hardware shaders. I guess that should almost be a priority. Once people are used to shaders its not a huge step to alter them to support hardware (other than the hardware shaders arent really anything like the software ones).
Phil.
05/10/2002 (12:56 am)
Yeah, Ive got the same feeling. I'm not sure in which order to implement the shaders.1) Hardware first OGL
2) Hardware DX
3) Software (Quake3).
Or 1,3,2
or some other mix :)
I really think the Quake3 shader language will help for guys without hardware shaders. I guess that should almost be a priority. Once people are used to shaders its not a huge step to alter them to support hardware (other than the hardware shaders arent really anything like the software ones).
Phil.
#6
I think you're dead on the money with your point of getting people used to shaders. I can just see the division created from people who can use hardware shaders and those than cannot. To be able to target almost everyone in a single attack would be preferable.
The development of the shader language implementation could progress whilsts the hardware support on D3D/GL could follow (in the background). I'd be glad to work on the GL hardware support for nVidia/ATI etc.
I'm afraid I'd have to follow your lead on this one so what kind of workload do you expect from the implementation of a shader language from modelling to rendering? Weeks, months?
- Melv.
05/10/2002 (2:25 am)
Perhaps 3,1,2 then?I think you're dead on the money with your point of getting people used to shaders. I can just see the division created from people who can use hardware shaders and those than cannot. To be able to target almost everyone in a single attack would be preferable.
The development of the shader language implementation could progress whilsts the hardware support on D3D/GL could follow (in the background). I'd be glad to work on the GL hardware support for nVidia/ATI etc.
I'm afraid I'd have to follow your lead on this one so what kind of workload do you expect from the implementation of a shader language from modelling to rendering? Weeks, months?
- Melv.
#7
Specially for all of us who still cant get access to a geforce3/4 or radeon card.
Then? I dont think DX support is a rush since most people are using OGL for their work (at least from what i saw in the forum posts), so that wont be a hurry. So first OpenGL support is first and then a shader software language.
Btw, isn't there any 'wrapper' dll for emulating hardware shaders calls to normal opengl calls? Thus making normal cards able to render shader code (of course using the software dll language and at a much slower speed).
05/10/2002 (8:39 am)
Well guys, i think software shaders is a must :)Specially for all of us who still cant get access to a geforce3/4 or radeon card.
Then? I dont think DX support is a rush since most people are using OGL for their work (at least from what i saw in the forum posts), so that wont be a hurry. So first OpenGL support is first and then a shader software language.
Btw, isn't there any 'wrapper' dll for emulating hardware shaders calls to normal opengl calls? Thus making normal cards able to render shader code (of course using the software dll language and at a much slower speed).
#8
Some links:
OGL 2.0 whitepapers
[url=http://www.microsoft.com/mscorp/corpevents/ meltdown2001/ppt/DXG9.ppt]MS Powerpoint presentation on DX9[/url] (not too useful, but does cover the major changes. No sample of the new high level shader lang though)
It looks like OGL2.0 will have a c-like shader language, with functions and whatnot. Has anyone seen the DX9 high level language?
I guess the way to go as far as shader support would be to make it generic enough so that people can provide either shader asm or a high level language. You also need to let people provide a bunch of different versions, so that Torque can use the best shader language available.
It seems like they way to go as far as supporting shaders is to roll that into the eventual transition to a XML based model format (or whatever we eventually decide should replace DTS). Is anyone working on this yet? It seems like something of a waste to work on adding shader support to the DTS file format, if we are just going to throw that out sometime in the future.
Josh
05/10/2002 (11:24 am)
My only worry with working on GL/D3D shader support is that it looks like GL 2.0 and/or DX9 may drastically change the way the shaders work. GL 2.0 doesn't sound too far off, though I suspect driver support will take quite a while. I haven't played with D3D9 yet, but I heard that shader support works differently than it does in D3D8.Some links:
OGL 2.0 whitepapers
[url=http://www.microsoft.com/mscorp/corpevents/ meltdown2001/ppt/DXG9.ppt]MS Powerpoint presentation on DX9[/url] (not too useful, but does cover the major changes. No sample of the new high level shader lang though)
It looks like OGL2.0 will have a c-like shader language, with functions and whatnot. Has anyone seen the DX9 high level language?
I guess the way to go as far as shader support would be to make it generic enough so that people can provide either shader asm or a high level language. You also need to let people provide a bunch of different versions, so that Torque can use the best shader language available.
It seems like they way to go as far as supporting shaders is to roll that into the eventual transition to a XML based model format (or whatever we eventually decide should replace DTS). Is anyone working on this yet? It seems like something of a waste to work on adding shader support to the DTS file format, if we are just going to throw that out sometime in the future.
Josh
Associate Melv May
I've downloaded the source but I'm at work so I'm not going to be able to try it until later. His site seems to be down unfortunately.
- Melv.