Radeon X1300 Material and Shader Problems
by Thomas Phillips · in Torque Game Engine Advanced · 11/12/2006 (3:15 pm) · 10 replies
I just upgraded to a Radeon X1300 (with latest drivers) and am now having problems with the materials system. One thing that stands out is that nothing is emissive any more.
The GG demos seem to work fine (the MS4 dojo and the feature walkthru). I'm guessing this has something to do with the procedural shaders, but I'm not sure. Has anyone else had any problems with a Radeon X1300?
Tom
The GG demos seem to work fine (the MS4 dojo and the feature walkthru). I'm guessing this has something to do with the procedural shaders, but I'm not sure. Has anyone else had any problems with a Radeon X1300?
Tom
#2
Tom
11/22/2006 (5:32 pm)
Well, I can't seem to figure out the cause, other than there a couple assets that seem to not work after I installed the X1300. Everything else, including all the GG demos, works fine, so for now I'll assume it's a problem peculiar to certain material settings. I'll follow up here if I ever figure out conclusively what the issue is.Tom
#3
11/27/2006 (2:31 pm)
Wait materials? I had a problem but what It turned out to be for me was, refining the difference between custom Materials, and just materials, cause there is a difference between your custom shaders and there calls and the ones already there, and only certain functions works for each in a different way. the answer is in TDN when reading about materials, ill post back with the link if it helps.
#4
Tom
11/30/2006 (9:21 pm)
It sounds like you're talking about fallbacks for different shader models and support by different cards. I'm not having any problems with custom shaders -- this is just the standard out-of-the box shaders that are giving the problem.Tom
#5
I now know what causes the problem -- enabling both transparency and glow in the same material. This seems to be a bug across all cards, but affects different cards in different ways. On my card it blacks out the texture. On other cards it seems to reduce the tranparency of the texture.
Has anyone else had a similar problem with both transparency and glow in the same material? My first thought is that it's a bug in the glow code, but then who's to say the glow algorithm really is meant to play with other effects? ;-)
Tom
12/21/2006 (5:05 pm)
Update!I now know what causes the problem -- enabling both transparency and glow in the same material. This seems to be a bug across all cards, but affects different cards in different ways. On my card it blacks out the texture. On other cards it seems to reduce the tranparency of the texture.
Has anyone else had a similar problem with both transparency and glow in the same material? My first thought is that it's a bug in the glow code, but then who's to say the glow algorithm really is meant to play with other effects? ;-)
Tom
#6
I think the only glow problems Nvidia had where the 6600 gpu and maybe 6200's.
TGEA's trans and glow are by the book so it's a driver / gpu issue.
After you updated the 1300 drivers did you reinstall DirectX?
Then do thru all the DirectX tests to see if they all pass.
12/30/2006 (7:03 am)
Other cards? You mean ATI cards or Nvidia?I think the only glow problems Nvidia had where the 6600 gpu and maybe 6200's.
TGEA's trans and glow are by the book so it's a driver / gpu issue.
After you updated the 1300 drivers did you reinstall DirectX?
Then do thru all the DirectX tests to see if they all pass.
#7
I found out that the issue is not glow and transparect on the same material (I was mistaken in my update), but a glowing material behind a transparent material. The polygons are separated, so it's not an explicit z issue. Tests for spacing ran from 1 to 31 units apart.
Tom
12/30/2006 (8:38 am)
All drivers (ATI and DirectX) were the latest as of 12 November, the date of the first post. All DirectX tests passed. The material problem is affecting both ATI and nVidia.I found out that the issue is not glow and transparect on the same material (I was mistaken in my update), but a glowing material behind a transparent material. The polygons are separated, so it's not an explicit z issue. Tests for spacing ran from 1 to 31 units apart.
Tom
#8
I have begun to scrub the materials. Looks like we have 687 materials in 31 files in our repository. This does not include any stray materials the art team may be using, but have not checked in. 486 of these materials are unique, meaning we have (at least) 201 duplicate materials. The 201 duplicates may or may not define the materials in the same way. And the materials stockpile continues to grow...
I suppose a complete overhaul of the materials is in order. :-(
Tom
01/02/2007 (2:29 pm)
And another update...I have begun to scrub the materials. Looks like we have 687 materials in 31 files in our repository. This does not include any stray materials the art team may be using, but have not checked in. 486 of these materials are unique, meaning we have (at least) 201 duplicate materials. The 201 duplicates may or may not define the materials in the same way. And the materials stockpile continues to grow...
I suppose a complete overhaul of the materials is in order. :-(
Tom
#9
As example, you can scan your shape and interior folders for textures and automatically create materials, and maybe use a naming convention to conveniently link diffuse and normal map textures.
01/06/2007 (11:11 am)
I recommend automating the material creation as much as you can to avoid such problems. You can organize the materials a lot just by using torquescript. As example, you can scan your shape and interior folders for textures and automatically create materials, and maybe use a naming convention to conveniently link diffuse and normal map textures.
#10
Tom
01/06/2007 (2:12 pm)
Our material system is automated. However, we have many people on the art team and it leads to further complications.Tom
Torque Owner Bobby Leighton
Imagn' Games