256fps off. 65 on
On my Dell Inspiron 9100 Laptop (with ATI Mobility 9800 graphics)
Please test my first shader test app!
'lo,
I got this error message : Fulscreen Display Mode of 1024x768x32 is not supported.
1024x768x32 is my current resolution. Other OGL application work fine. (even mine ;p)
graphic card : Geforce3ti200
I got this error message : Fulscreen Display Mode of 1024x768x32 is not supported.
1024x768x32 is my current resolution. Other OGL application work fine. (even mine ;p)
graphic card : Geforce3ti200
Quote:Original post by Corrail
There is no difference between gloom and no-gloom mode here.
ForceWare 70.41 / GeForce 6800
Disable V-Sync !
256 without glow / 113 with glow.
Athlon XP 3200+ / 1.5G RAM / GeForce 6800 Ultra
i find it weird that it's only 256fps without glow when the other person with a 6800U got 340... (text shows up okay without the glow. when using the glow, it's not readable at all)
Athlon XP 3200+ / 1.5G RAM / GeForce 6800 Ultra
i find it weird that it's only 256fps without glow when the other person with a 6800U got 340... (text shows up okay without the glow. when using the glow, it's not readable at all)
Odd, but i think it has more to do with what else is running on the computer and how your driver is set up(i have 66.93).
I rebooted and ran it again, now it tops just under 400 without glow but only around 80 with.
(i got 340 and 100 last time)
So i started to fiddle with the settings, i turned on maximum FSAA(8xs) and maximum ansiotriopic(16) filtering and got 150 glow off, glow 55 on.
A more resonable setting of FSAA 2x and a ansiotropic filtering of 4 yealds 310 glow off and 70 glow on.
turning on preformance mode instead of quality mode(i even made shure you didn't use trilinear filtering or something like that) gives me the figures of 420 and 75.
I rebooted and ran it again, now it tops just under 400 without glow but only around 80 with.
(i got 340 and 100 last time)
So i started to fiddle with the settings, i turned on maximum FSAA(8xs) and maximum ansiotriopic(16) filtering and got 150 glow off, glow 55 on.
A more resonable setting of FSAA 2x and a ansiotropic filtering of 4 yealds 310 glow off and 70 glow on.
turning on preformance mode instead of quality mode(i even made shure you didn't use trilinear filtering or something like that) gives me the figures of 420 and 75.
Quote:Original post by Anonymous Poster
256 without glow / 113 with glow.
Athlon XP 3200+ / 1.5G RAM / GeForce 6800 Ultra
i find it weird that it's only 256fps without glow when the other person with a 6800U got 340... (text shows up okay without the glow. when using the glow, it's not readable at all)
Maybe it's the driver version or driver settings. Since the 6800 series is still young, every driver version is important actually.
Not to hijack this thread...
I was looking at how the effect was done. I was trying to think of a way to combine the blur step into 1 draw, instead of looping it many times with alot of glCopyTexImage2D's. The end result in speading this up a great amount. Instead I would image one pass where the frament shader has ALOT of lookups. Somthing where it's weighted on its surrounding pixel's as follows:
This would do 49 texture look ups and only one CopyTexImage2D, vs 64 lookups + 16 CopyTexImage2D.
Are there any flaws in doing it this way that I'm not thinking of?
I was looking at how the effect was done. I was trying to think of a way to combine the blur step into 1 draw, instead of looping it many times with alot of glCopyTexImage2D's. The end result in speading this up a great amount. Instead I would image one pass where the frament shader has ALOT of lookups. Somthing where it's weighted on its surrounding pixel's as follows:
0.000458271 0.002011732 0.004886957 0.006569435 0.004886957 0.002011732 0.0004582710.002011732 0.008831155 0.021452898 0.028838685 0.021452898 0.008831155 0.0020117320.004886957 0.021452898 0.052114004 0.070055774 0.052114004 0.021452898 0.0048869570.006569435 0.028838685 0.070055774 0.094174523 0.070055774 0.028838685 0.0065694350.004886957 0.021452898 0.052114004 0.070055774 0.052114004 0.021452898 0.0048869570.002011732 0.008831155 0.021452898 0.028838685 0.021452898 0.008831155 0.0020117320.000458271 0.002011732 0.004886957 0.006569435 0.004886957 0.002011732 0.000458271This would do 49 texture look ups and only one CopyTexImage2D, vs 64 lookups + 16 CopyTexImage2D.
Are there any flaws in doing it this way that I'm not thinking of?
beautiful, quite impressive,
radeon 9700 128 mb
p4 2.4 ghz
768 mb ram
45.something fps with bloom, and that was with windows media player and several other apps running in the background... bravo
-Dan
radeon 9700 128 mb
p4 2.4 ghz
768 mb ram
45.something fps with bloom, and that was with windows media player and several other apps running in the background... bravo
-Dan
Quote:Original post by skow
Not to hijack this thread...
I was looking at how the effect was done. I was trying to think of a way to combine the blur step into 1 draw, instead of looping it many times with alot of glCopyTexImage2D's. The end result in speading this up a great amount. Instead I would image one pass where the frament shader has ALOT of lookups. Somthing where it's weighted on its surrounding pixel's as follows:
********* stuff removed *************
This would do 49 texture look ups and only one CopyTexImage2D, vs 64 lookups + 16 CopyTexImage2D.
Are there any flaws in doing it this way that I'm not thinking of?
that sounds like _alot_ of overkill, i use a 2pass seperable blur, 8 samples per pass, so 16 blur samples per pixel, and i render to a P-Buffer bound as a texture so there is no readback of the framebuffer (a main killer)
all in all, i only ever render the scene 2 times, 4 times if i want excesive visual quality, i never readback the framebuffer and everything is done on the GPU.
read the article on Gamasutra, excelent as it provides some PS/VS code to help you along.
This topic is closed to new replies.
Advertisement
Popular Topics
Advertisement
