More pages: 1 ...
11 ...
21 ...
31 ...
41 ...
51 ...
61 ...
71 ...
81 ...
91 ...
101 ...
111 ...
121 ...
131 ...
141 ...
151 ...
161 ...
171 ...
175 176 177 178 179 180
181 182 183 184 185 ...
191 ...
201 ...
211 ...
221 ...
231 ...
241 ...
251 ...
261 ...
271 ...
281 ...
291 ...
301 ...
311 ...
321 ...
331 ...
341 ...
351 ...
361 ...
365
Query Failed
backslash
Saturday, October 14, 2006
nice demo
65fps 1024x768 6xAA X1900XTX
30fps 1600x1200 6xAA
eXile
Thursday, October 12, 2006
Nice one

Gendolf
Thursday, October 12, 2006
1024x768 34 - 37 fps X800GT
rP
Wednesday, October 11, 2006
Awesome! Beautiful effects!
17fps 1152x864 GF6800GT
jur
Tuesday, October 10, 2006
work on 6600gt but slow. 640x480 ->14fps (windowed). good for demo, but not for games for now.
Humus
Monday, October 9, 2006
It should now work on Radeon 9500 and up. I expected that X800s would be able to run it since it supports enough instructions, but I didn't test before. Turns out it ran in software due to exceeding the number of registers. I've now added a multipass path to the demo, so now it should even run on Radeon 9500.

Gendolf
Monday, October 9, 2006
Not work on Sapphire X800GT 256mb (agp)

Humus
Monday, October 9, 2006
It's a 8bit cubemap, so yes, that's 8bit depth, but it's linear, unlike regular shadow mapping, so that compensates a bit for the lack of precision. Plus most lights are quite short range.
lVec is the unnormalized value, so the dot returns the length squared. For the comparison I square the shadow instead of computing the square-root of dot(lVec, lVec).
More pages: 1 ...
11 ...
21 ...
31 ...
41 ...
51 ...
61 ...
71 ...
81 ...
91 ...
101 ...
111 ...
121 ...
131 ...
141 ...
151 ...
161 ...
171 ...
175 176 177 178 179 180
181 182 183 184 185 ...
191 ...
201 ...
211 ...
221 ...
231 ...
241 ...
251 ...
261 ...
271 ...
281 ...
291 ...
301 ...
311 ...
321 ...
331 ...
341 ...
351 ...
361 ...
365