When running version 4 of the program w/o reflections on a machine that is approximately 500MHz, running NT, with a generic graphics card, I got 60 FPS _UNTIL_ it started drawing the FPS counter (the first time it drew the counter, it showed 60 FPS, then after that it showed <20 every update) so I would guess glDrawText is a ''little'' slow. IT would depend on the OS, card, and drivers, but it may be that its extremely slow on your problem PC.
"I believe; therefore, it is." -True Perception
"The Requested Information Is Unknown Or Classified" -Anonymous
Reflection Help & Slow texture mapping...
well
one thing..
you enable texgen for the spheremapping
you should disable it after drawing the floor
that way you don''t have those fancy "selfreflectmapping" cards
spheremaps are NOT for reflections, they are for environmental effects.. like looking like metall for example
they CAN be used for some reflection-effects but surely NOT for real reflections..
result:
DONT use envmaps for the planar reflection
"take a look around" - limp bizkit
www.google.com
one thing..
you enable texgen for the spheremapping
you should disable it after drawing the floor
that way you don''t have those fancy "selfreflectmapping" cards
spheremaps are NOT for reflections, they are for environmental effects.. like looking like metall for example
they CAN be used for some reflection-effects but surely NOT for real reflections..
result:
DONT use envmaps for the planar reflection
"take a look around" - limp bizkit
www.google.com
quote:Original post by zin
Your version 5 file is unavailable, version 4 runs 450 fps on my 1 gig athlon w/ geforce2.
zin
zintel.com - 3d graphics & more or less
This has been fixed now, sorry about that.. Version 5 does reflections using Stencil buffers.. I''m running this on a AMD Athlon 800mhz w/GeForce 2 and 1.5 Gigs of RAM and only getting about 2 frames per second... =(
quote:Original post by davepermen
well
one thing..
you enable texgen for the spheremapping
you should disable it after drawing the floor
that way you don''t have those fancy "selfreflectmapping" cards
spheremaps are NOT for reflections, they are for environmental effects.. like looking like metall for example
they CAN be used for some reflection-effects but surely NOT for real reflections..
result:
DONT use envmaps for the planar reflection
I''ve completely changed the way this works to now use stencil buffers, the new code is available at:
http://www.oeginc.com/downloads/RotateCard5.zip
or you can click here.
But it runs incredibly slow on my machine (see above).
just gave the above code a test
got a sweeet rock soild 170fps from it
desktop running @ 1280*1024*32
vsync off
CPU : AMD Athlon T-bird 1.4Ghz
RAM : 768SD @ 140Mhz
GFX : Geforce 3 Ti200 64Meg DDR, o/c 220Mhz core, 503Mhz Mem
OS : Win2K
the fps held at 170fps and I did run it for a while
all was nice and smooth and it does look sweet
got a sweeet rock soild 170fps from it
desktop running @ 1280*1024*32
vsync off
CPU : AMD Athlon T-bird 1.4Ghz
RAM : 768SD @ 140Mhz
GFX : Geforce 3 Ti200 64Meg DDR, o/c 220Mhz core, 503Mhz Mem
OS : Win2K
the fps held at 170fps and I did run it for a while
all was nice and smooth and it does look sweet
Well, I tried it out myself and I have bad news for you...
Less than 1 frame a second. Yes, you heard me right. 0.656 (average) frames a second. Presumably, I have the same problem that you do (except much worse, most likely due to my aging hardware).
Pentium III 650MhZ
256 MB RAM
TNT2 32MB
Less than 1 frame a second. Yes, you heard me right. 0.656 (average) frames a second. Presumably, I have the same problem that you do (except much worse, most likely due to my aging hardware).
Pentium III 650MhZ
256 MB RAM
TNT2 32MB
pfd.dwFlags = PFD_DRAW_TO_WINDOW | PFD_SUPPORT_OPENGL | PFD_DOUBLEBUFFER;
pfd.dwLayerMask = PFD_MAIN_PLANE; // We want the standard mask (this is ignored anyway)
pfd.iPixelType = PFD_TYPE_RGBA; // We want RGB and Alpha pixel type
pfd.cColorBits = SCREEN_DEPTH; // Here we use our #define for the color bits
pfd.cDepthBits = SCREEN_DEPTH; // Depthbits is ignored for RGBA, but we do it anyway
pfd.cAccumBits = 0; // No special bitplanes needed
pfd.cStencilBits = 0; // We desire no stencil bits
you don''t ask for any stencil bits.. how do you except that there will be a stencil buffer awailable? set this to 1, or 8, depends on hardware..
"take a look around" - limp bizkit
www.google.com
pfd.dwLayerMask = PFD_MAIN_PLANE; // We want the standard mask (this is ignored anyway)
pfd.iPixelType = PFD_TYPE_RGBA; // We want RGB and Alpha pixel type
pfd.cColorBits = SCREEN_DEPTH; // Here we use our #define for the color bits
pfd.cDepthBits = SCREEN_DEPTH; // Depthbits is ignored for RGBA, but we do it anyway
pfd.cAccumBits = 0; // No special bitplanes needed
pfd.cStencilBits = 0; // We desire no stencil bits
you don''t ask for any stencil bits.. how do you except that there will be a stencil buffer awailable? set this to 1, or 8, depends on hardware..
"take a look around" - limp bizkit
www.google.com
Run your app in 32bit colour mode. If youre running in 16bit colour and trying to use stencil on your gf2/tnt2 cards, youre using software rendering which is why its so damn slow. Put it on 32bit and hardware stencilling is enabled and youll have hardware rendering.
Geez Im amazed nobody else pointed this out.
-----------------------
"When I have a problem on an Nvidia, I assume that it is my fault. With anyone else''s drivers, I assume it is their fault" - John Carmack
Geez Im amazed nobody else pointed this out.
-----------------------
"When I have a problem on an Nvidia, I assume that it is my fault. With anyone else''s drivers, I assume it is their fault" - John Carmack
quote:Original post by davepermen
pfd.dwFlags = PFD_DRAW_TO_WINDOW | PFD_SUPPORT_OPENGL | PFD_DOUBLEBUFFER;
pfd.dwLayerMask = PFD_MAIN_PLANE; // We want the standard mask (this is ignored anyway)
pfd.iPixelType = PFD_TYPE_RGBA; // We want RGB and Alpha pixel type
pfd.cColorBits = SCREEN_DEPTH; // Here we use our #define for the color bits
pfd.cDepthBits = SCREEN_DEPTH; // Depthbits is ignored for RGBA, but we do it anyway
pfd.cAccumBits = 0; // No special bitplanes needed
pfd.cStencilBits = 0; // We desire no stencil bits
you don't ask for any stencil bits.. how do you except that there will be a stencil buffer awailable? set this to 1, or 8, depends on hardware..
"take a look around" - limp bizkit
www.google.com
That made all the difference in the world, on one of my machines... still only get 4fps on the other... (I can't believe I missed that, doh!)
--- EDITED ---
quote:Original post by Maximus
Run your app in 32bit colour mode. If youre running in 16bit colour and trying to use stencil on your gf2/tnt2 cards, youre using software rendering which is why its so damn slow. Put it on 32bit and hardware stencilling is enabled and youll have hardware rendering.
Ding Ding Ding! We have a winner.. =) It now runs at 75fps! Thanks! I was unaware of this limitation... Now to throw some code in there to inform others if they aren't running in 32bit mode it may be slow...
[edited by - oeginc on May 8, 2002 9:24:01 AM]
ah yes, I've just looked at my tweeked OGL setup from NVMax :
Color depth 32bits
Fast Mipmaping one
Force S3TC v3 compression on
64tap texture filtering
Quincunx FSAA is on
the first one and 3rd one might well explain why I saw very good performance when I tested it.
Edit :
maybe not however got : 460fps with same settings as before, only changes are as follows :
FSAA off
16bit color depth
no texture compression
Bi or Trilinear filtering only
so I turned the Qunicunx FSAA back on and the fps dropped back to 190, this was still better than my last reading, so I'm starting to think that, on a gf3 at least, the color depth dont matter so much as it did on older hardware.
[edited by - _the_phantom_ on May 8, 2002 11:14:22 AM]
Color depth 32bits
Fast Mipmaping one
Force S3TC v3 compression on
64tap texture filtering
Quincunx FSAA is on
the first one and 3rd one might well explain why I saw very good performance when I tested it.
Edit :
maybe not however got : 460fps with same settings as before, only changes are as follows :
FSAA off
16bit color depth
no texture compression
Bi or Trilinear filtering only
so I turned the Qunicunx FSAA back on and the fps dropped back to 190, this was still better than my last reading, so I'm starting to think that, on a gf3 at least, the color depth dont matter so much as it did on older hardware.
[edited by - _the_phantom_ on May 8, 2002 11:14:22 AM]
This topic is closed to new replies.
Advertisement
Popular Topics
Advertisement