xml based postprocessing framework
Re: xml based postprocessing framework
Tested on Geforce GTS250 and GTX550 aswell as Geforce 7300 LE, all work perfect in both OpenGL and DirectX. Its likely yet another problem with the DirectX hlsl shaders only related to ATi. Ps: what ATi is it? Still an ATi Radeon or allready an AMD Radeon?
EDIT: Also working on Intel GMA X4500 IGP.
EDIT: Also working on Intel GMA X4500 IGP.
Re: xml based postprocessing framework
dual ati 4850's mobility in cfx (yes two laptop card in cfx)
the cards of that generation are ati cards that's the logo i have on my card but they have been rebranded to AMD since then
and honestly i don't see why it's not working only in dx and only on that version
i still have the older version wich work fine for both openGL and DX
honestly i now get why i'm always selected for beta test my hardware find every single little bug in a graphic code and makes it X10000
the cards of that generation are ati cards that's the logo i have on my card but they have been rebranded to AMD since then
and honestly i don't see why it's not working only in dx and only on that version
i still have the older version wich work fine for both openGL and DX
honestly i now get why i'm always selected for beta test my hardware find every single little bug in a graphic code and makes it X10000
Re: xml based postprocessing framework
I got error while compiling the code
ShaderMaterial.cpp
The error is
I got that error on Code::Blocks and Microsoft Visual Studio. What can I do to get rid of this error?
ShaderMaterial.cpp
Code: Select all
for(; !psParamIter.atEnd(); psParamIter++)
{
services->setPixelShaderConstant(
psParamIter.getNode()->getKey().c_str(),
&psParamIter.getNode()->getValue(), //This line causes a problem
1);
}
Code: Select all
ShaderMaterial.cpp|61|error: lvalue required as unary '&' operand
Re: xml based postprocessing framework
did you try the following:
if this does not work, at least the following should do it:
Code: Select all
&(psParamIter.getNode()->getValue())
Code: Select all
for(; !psParamIter.atEnd(); psParamIter++)
{
f32 value = psParamIter.getNode()->getValue();
services->setPixelShaderConstant(
psParamIter.getNode()->getKey().c_str(),
&value,
1);
}
Re: xml based postprocessing framework
It's working now, thanks
Re: xml based postprocessing framework
any idea why the dx mode of the new one is not working on my ati hardware?
Re: xml based postprocessing framework
there are still graphics cards with old/buggy drivers wich clamp to range 0.0- 1.0 floating point textures. this can be a problem for HDR right? maybe it is the problem.
Junior Irrlicht Developer.
Real value in social networks is not about "increasing" number of followers, but about getting in touch with Amazing people.
- by Me
Real value in social networks is not about "increasing" number of followers, but about getting in touch with Amazing people.
- by Me
Re: xml based postprocessing framework
but the opengl version works fine and i doubt a dual 4850 is likely to pose problems especial as i'm running drivers of this month
-
- Posts: 1638
- Joined: Mon Apr 30, 2007 3:24 am
- Location: Montreal, CANADA
- Contact:
Re: xml based postprocessing framework
Just to warn you, I've tested the shaders on Intel (The chip is ok, but I hate the driver ) in OPENGL (glsl)and there are other shaders that are failing. (about 4-5, don't remember the names). It look like that if a shader fail compilation, then the framework will fail. To make it work quickly, I've removed the failing shaders from the list and the framework was working fine again.
Granyte, It should work on your ATI in DirectX but all of thoses shaders need to be retested one by one and fixed, since when I tested the first time it was only the glsl shaders. So if nobody checked the hlsl shaders with AMD Shader analyser, then they have been tested to work only on NIVIDIA hardware.
For retest, The AMD shader analyzer is the way to go to test thoses. For Intel on DirectX (hlsl); this time it should run better as Intel only put energy in that driver...
Granyte, It should work on your ATI in DirectX but all of thoses shaders need to be retested one by one and fixed, since when I tested the first time it was only the glsl shaders. So if nobody checked the hlsl shaders with AMD Shader analyser, then they have been tested to work only on NIVIDIA hardware.
For retest, The AMD shader analyzer is the way to go to test thoses. For Intel on DirectX (hlsl); this time it should run better as Intel only put energy in that driver...
Re: xml based postprocessing framework
wierd unless they have been modified since the first version they should still work cause the first version was working on my computer it's integrated into my current project but the version 2 no shader at all works in DX
i mean even the invert won't work and shader analyser won't say anything wrong about it
i mean even the invert won't work and shader analyser won't say anything wrong about it
-
- Posts: 1638
- Joined: Mon Apr 30, 2007 3:24 am
- Location: Montreal, CANADA
- Contact:
Re: xml based postprocessing framework
Ha? Then it could be something else, can you make a debug version and check the console when you initialize the app? Normally, I have "shader compilation" messages coming out, perhaps something is failing and is not related to the shaders...
Re: xml based postprocessing framework
hm,
I tested all shaders with AMD shader analyzer, so i've no idea so far ....
I tested all shaders with AMD shader analyzer, so i've no idea so far ....
-
- Posts: 1638
- Joined: Mon Apr 30, 2007 3:24 am
- Location: Montreal, CANADA
- Contact:
Re: xml based postprocessing framework
Looked at your pictures, compilation of the shaders seem to went well. So there is something else. TBW is the demo is running fine in DirectX? Have you tried compiling it to IRRlicht SVN? (From Granyte screenshot he's using the SVN version)...
Granyte have you changed something in the code to make it compile?
Granyte have you changed something in the code to make it compile?
Re: xml based postprocessing framework
nop nothing at all
also i recompiled it against the SVN after the precompiled version showed that behavior in case my computer was running a more recent version of the dx executable or what ever but nop no avail even the recompiled version against my version of the dx sdk does not help still a full black screen
is there any one else with AMD hardware that can make a test to be sure it's not juste my cfx being stuborn (even if i tryed running without CFX)
also i recompiled it against the SVN after the precompiled version showed that behavior in case my computer was running a more recent version of the dx executable or what ever but nop no avail even the recompiled version against my version of the dx sdk does not help still a full black screen
is there any one else with AMD hardware that can make a test to be sure it's not juste my cfx being stuborn (even if i tryed running without CFX)