tnt2 m64 help

Originally posted by dhkim

1. Enable Buffer region extension
2. allow the dual planes extension to use local video memory
3. use fast linear - midpmap- linear filtering
4. enable anisotropic filtering
5. enable alternate depth bufferin technique
6. disable support for enhanced cpu instruction sets

Click to expand...

You know what 1 and 2 do...

3. will boost your performance a bit at the cost of a little visual quality. just makes textures in the distance look a little worse....

4. this is a method of filtering textures to make them look better closer up but uses _huge_ amounts of video memory and is slower than bilinear, or trilinear filtering, best not use this unlesss you're playing halflife

5. This is slightly faster way of using the depth buffer ( helps sort out objects that need to be drawn in a scene ), no noticable visual difference.

6. This just makes it slower for you, disables use of MMX, SSE, SSE2, and 3DNow! Just leave this on....

There you go, hope I helped you out a bit, if you do need a little more help PM me.

C Ya
Matt

Just tryed Pete's opengl plugin and made all kinds of adjustments to the plugin but it still doesn't run as well as the d3d. The game(GT2) is very jumpy along with the sound. I'm not sure what sure what the deal is but, the d3d is far smother. The overall grafix are a little bit better in opengl but not enough for me to compromise playabilty. Is there something I am doing wrong?
Direct 3d settings:
Enable fog table emulation= on
Adjust Z-buffer depth to rendering depth if unequal=on
Mipmap detail leval=best image quality
PCI Texture Memory Size=63MB(most it will let me)
OpenGL settings:
enable buffer region extension=on
allow the dual planes extension to use local mem.=on
use fast linear-mipmap-lin.-filtering=off
Default color depth for textures:Use desktop color
Buffer filping mode: auto select
Vertical sync: on by default
PCI mode system mem.=14MB(all it will let me even if I lower the Direct 3D)
Other than that Iam not sure what else to do maybe the MB's are what is making the difference?
Thanks for your help!!!!!!!!!!!!!!!!!!!!!!!!!!!!

Originally posted by dhkim
Hello to all !
can anyone tell me what are these options for ?


OpenGl Settings
Enable Buffer region extension
allow the dual planes extension to use local video memory
use fast linear - midpmap- linear filtering
enable anisotropic filtering
enable alternate depth bufferin technique
disable support for enhanced cpu instruction sets

I got buffer region extension and allow the dual planes extension to use local video memory enable. Will it be good if I enable the others too ? what are those use for ?

My system is 900 duron, 384 ram, and a tnt2 m64.

And I almost forgot, what is this one use for ? Enable alternate depth buffering technique ? Will it be good if I enable it too ?

Click to expand...

Well, let's see. I'll go over them in the order that they appear.

1.) Enable Buffer region extension - This enables the KTX buffer region extension. (it was originally made for use with 3D Studio.) I haven't looked much into this, but as far as I understand, it this is a fast buffered space in memory generally used for 3D transformations (back when hardware T&L didn't exist.)

2.) allow the dual planes extension to use local video memory - Allows the KTX extension to use the video card's memory for the buffered region. I think the buffer defaults to the system memory otherwise.

3.) use fast linear - midpmap- linear filtering - From the looks of it, I'd guess that this uses a lower form of linear blending with mip maps to speed up rendering.

4.) enable anisotropic filtering - This is an advanced filtering feature which programs (like Quake 3) can use to slightly improve visual quality of textures that are at a distance from the viewer. I really haven't noticed THAT much of a difference in visual quality, so I usually leave this turned off.

5.) enable alternate depth bufferin technique - This says it will improve visual quality in 16-bit modes. I'm sure what they mean is that applications that use 16-bit z-buffers for rendering will actually be using a higher depth of z-buffer than intended (although it will still look like to the application that a 16-bit z-buffer is in use.) So, for example, UT in 16-bit OGL mode might use a 24-bit z-buffer with this option turned on. I do have to say, that would prevent A LOT of polygons at a distance from clipping through each other.

6.) disable support for enhanced cpu instruction sets - It would be a good idea not to use this feature. It's mostly there as a "safe mode" just incase your new processor crashes when the drivers try to take advantage of advanced instruction sets like MMX or the SIMD extensions.

Hope that helps!