I had read (in some old post) that: "...with nvidia gpu''s the quickest is using VAR in conjunction
with fence...". But I haven''t these extensions.
I tried Compilex Vertex Arrays but it make my engine a bit slower.
There are some equivalent extensions for ATI Radeon based video cards?
If ATI Radeon have totally different way to handle T&L in OpenGL... then, I need to write two
different procedure for rendering? IMHO it''s really incredible that nowadays still there isn''t a
STANDARD way (extension) to exploit T&L in OpenGL. In this sense, I think that DirectX is
better than OpenGL.
Thanks.
- AGPX
OpenGL vs T&L...
Author
Hi guys,
I''m really unable to exploit hardware transformation and lighting of my ATI Radeon 7200 when I
use OpenGL
I had read (in some old post) that: "...with nvidia gpu''s the quickest is using VAR in conjunction
with fence...". But I haven''t these extensions.
I tried Compilex Vertex Arrays but it make my engine a bit slower.
There are some equivalent extensions for ATI Radeon based video cards?
If ATI Radeon have totally different way to handle T&L in OpenGL... then, I need to write two
different procedure for rendering? IMHO it''s really incredible that nowadays still there isn''t a
STANDARD way (extension) to exploit T&L in OpenGL. In this sense, I think that DirectX is
better than OpenGL.
Thanks.
- AGPX
I had read (in some old post) that: "...with nvidia gpu''s the quickest is using VAR in conjunction
with fence...". But I haven''t these extensions.
I tried Compilex Vertex Arrays but it make my engine a bit slower.
There are some equivalent extensions for ATI Radeon based video cards?
If ATI Radeon have totally different way to handle T&L in OpenGL... then, I need to write two
different procedure for rendering? IMHO it''s really incredible that nowadays still there isn''t a
STANDARD way (extension) to exploit T&L in OpenGL. In this sense, I think that DirectX is
better than OpenGL.
Thanks.
- AGPX
Actually, I''d like to know more about the T&L topic too concerning OpenGL. I have absolutely no clue how to use hardware T&L
As far as I know, you can get extensive documentation on the appropriate extensions in the respective SDKs (Radeon SDK and NVSDK) - download them off the nVidia and ATI developer sites.
And obviously, that still won''t make it compatible to all T&L capable cards, just the ATI and nVidia ones, so basically you''re limited to whatever you can test on.
Same thing applies to shaders by the way, which again, are standardised in DirectX but nonstandard in OpenGL.
- JQ
Full Speed Games. Are back.
And obviously, that still won''t make it compatible to all T&L capable cards, just the ATI and nVidia ones, so basically you''re limited to whatever you can test on.
Same thing applies to shaders by the way, which again, are standardised in DirectX but nonstandard in OpenGL.
- JQ
Full Speed Games. Are back.
T&L is automatic in OpenGL. If you use the standard OpenGL pipeline, it will always use T&L, given that the 3D card supports it. The use of T&L is totally vendor independent, it is an basic feature of OpenGL.
VAR has nothing to do with T&L. VAR is a transfer method to efficiently stream vertex data to the 3D card. The ATI version of VAR is called VAO (vertex_array_object), both will be standarized in a very near future (ARB is working on it). YOur card will use T&L regardless of you using VAR/VAO or not. Another point lots of people mix up is the vertex cache: it's an onboard cache used to store recently transformed coordinates. It can only be used with vertex arrays (or VAR, or VAO), but not in immediate mode. The vertex cache again has nothing to do with T&L, it's a fully separate entity.
Now, vertex shaders are an extension to T&L. You can see them as a user programmable T&L version. They are standarized in the ARB_vertex_program extension.
[edited by - Yann L on December 2, 2002 12:45:47 PM]
VAR has nothing to do with T&L. VAR is a transfer method to efficiently stream vertex data to the 3D card. The ATI version of VAR is called VAO (vertex_array_object), both will be standarized in a very near future (ARB is working on it). YOur card will use T&L regardless of you using VAR/VAO or not. Another point lots of people mix up is the vertex cache: it's an onboard cache used to store recently transformed coordinates. It can only be used with vertex arrays (or VAR, or VAO), but not in immediate mode. The vertex cache again has nothing to do with T&L, it's a fully separate entity.
Now, vertex shaders are an extension to T&L. You can see them as a user programmable T&L version. They are standarized in the ARB_vertex_program extension.
[edited by - Yann L on December 2, 2002 12:45:47 PM]
damn i am annoyed by crosposters. i have answered this question i think yesterday (or early in the morning today).. its so annyoing.
"take a look around" - limp bizkit
www.google.com
"take a look around" - limp bizkit
www.google.com
This topic is closed to new replies.
Advertisement
Popular Topics
Advertisement