+- Frictional Games Forum (read-only) (https://www.frictionalgames.com/forum)
+-- Forum: Technical Support (https://www.frictionalgames.com/forum/forum-57.html)
+--- Forum: Technical Support - Amnesia: The Dark Descent (https://www.frictionalgames.com/forum/forum-31.html)
+--- Thread: Parallax Occlusion Mapping... (/thread-4289.html)
RE: Parallax Occlusion Mapping... - sgi - 09-22-2010
Yes! That solved the problem. Just use
RayLinearIntersectionSM2(aHeightMap, vHeightMapPos, vEyeVec);
instead of
RayLinearIntersectionSM3(aHeightMap, fSteps, vHeightMapPos, vEyeVec);
in
core\shaders\deferred_gbuffer_solid_frag.glsl, line 147
and all problems disappear.
The only difference between two functions is fixed step instead of adaptive step.
RE: Parallax Occlusion Mapping... - Deu sex - 09-22-2010
(09-22-2010, 06:51 PM)sgi Wrote: Yes! That solved the problem. Just use
RayLinearIntersectionSM2(aHeightMap, vHeightMapPos, vEyeVec);
instead of
RayLinearIntersectionSM3(aHeightMap, fSteps, vHeightMapPos, vEyeVec);
in
core\shaders\deferred_gbuffer_solid_frag.glsl, line 147
and all problems disappear.
The only difference between two functions is fixed step instead of adaptive step.
Men
you rocks. YOU FUCKING ROOOOCKS !!!
i will try it right now !!
RE: Parallax Occlusion Mapping... - Deu sex - 09-22-2010
-_- it doesnt work, textures with parallax possibilities are all dark.
RE: Parallax Occlusion Mapping... - sgi - 09-22-2010
ok, I check now... they are dark if shader compile failed (there is an error in shader).
...
No - everything seems to be fine to me. What video-card/drivers do you have?
RE: Parallax Occlusion Mapping... - Blackhand - 09-22-2010
This worked for me, well done. It looks like there is only a minor performance penalty for enabling parallax, none of the big slowdowns/crashes I was getting earlier.
RE: Parallax Occlusion Mapping... - Mina Darsh - 09-22-2010
Wow, amazing, it works yes, however, does this mean the game switches back to Shader Model 2 as well?
RE: Parallax Occlusion Mapping... - Blackhand - 09-23-2010
@Hionimi: From what I can see, no. It looks like the two different functions just perform a calculation, one I assume optimized for SM2 and one for SM3.
RE: Parallax Occlusion Mapping... - CHANCEPHOENIX - 09-23-2010
IT WORKS for me too !!!
i thought at first that i was missing something switching from SM3 to SM2 , but ...
I CAN SEE no DIFFERENCE ... AND IT RUNS GREAT.
btw. modified pom is the one from sgi :
![[Image: 2w3nmfm.jpg]](http://i53.tinypic.com/2w3nmfm.jpg)
anyway , THANX for making this work ....er .... and please do sumthing to activate dynamic lighting/shadows(esspecially for the lantern) IF it"S possible
RE: Parallax Occlusion Mapping... - Zanderat - 09-23-2010
I have the dark shader issue, too. ATI 5770. Oh, Well. Getting closer.
RE: Parallax Occlusion Mapping... - sgi - 09-23-2010
All who have dark surfaces, try to change line 82 (in RayLinearIntersectionSM2 function) from
avEyeVec /= lSearchSteps; //Split up the eye vector into the number of steps
to
avEyeVec /= float(lSearchSteps); //Split up the eye vector into the number of steps