Sign in to follow this  
xtapol

OpenGL GLSL lighting performance

Recommended Posts

I'm not a newbie to OpenGL, but I am a newbie to *modern* OpenGL, and I'm seeing some bizarre performance characteristics in my specular lighting shader. I have a multipass renderer, and a test scene with a plane, a sphere, and 8 lights each rendered individually in two passes - one for diffuse lighting, one for specular lighting. I recently reworked the renderer, but my problem also occurred in exactly the same way when I was rendering all lighting in one pass, so I know the multipass structure is not the cause. The problem occurs when viewing the sphere from only certain angles and only when zoomed in on one or more specular highlights. Suddenly the framerate drops from ~8000 FPS (vsync disabled) to 50 or 60. This is 100% reproducible and I can have the camera in one position and be getting 8000 FPS, and then rotate a few degrees and hit the problem, even though it's still rendering the same number of specular highlights. I've tracked it down to this specular component in my fragment shader. When I comment this out, the framerate no longer drops: -- halfV = normalize(halfVector); NdotHV = max(dot(n,halfV),0.0); color = att * gl_FrontMaterial.specular * light.specular * pow(NdotHV, gl_FrontMaterial.shininess); -- The really strange part is that the performance drop seems to come from RETURNING a value, not from the calculations themselves. If I don't do the calculations but simply return vec4(1, 1, 1, 1), I still get the performance drop. If I do the calculations but return vec4(0, 0, 0, 0) instead of the result, there is no problem. Here are two screenshots. The first runs at a framerate of ~8000 FPS, the second, different only in camera rotation, drops to 50-60. Again, this is reliably reproducible (GeForce 9600M GT on OS X). http://img532.imageshack.us/i/screenshot20100424at507.png/ http://img519.imageshack.us/i/screenshot20100424at507.png/ Can somebody help me track this down? What is happening here? Are there any good OS X tools to help debugging/profiling GLSL code?

Share this post


Link to post
Share on other sites
Sure, here it is. There are some extraneous varying parameters from when this was doing 4 lights at a time, but they are not populated by the vertex shader.

As a side note, I've tried removing all the branching statements, but it didn't help.

--
varying vec3 normal, lightDir[4], halfVector[4];
varying float dist[4];
varying vec4 ambient;

vec4 pointLightContribution (int lightNum, gl_LightSourceParameters light, vec3 n)
{
float NdotL = max (dot (n, normalize(lightDir[lightNum])), 0.0);
vec4 color = vec4(0, 0, 0, 0);

if (NdotL > 0.0)
{
vec3 halfV;
float NdotHV;
float att;

att = 1.0 / (light.constantAttenuation +
light.linearAttenuation * dist[lightNum] +
light.quadraticAttenuation * dist[lightNum] * dist[lightNum]);

if (gl_FrontMaterial.shininess > 0.0)
{
halfV = normalize(halfVector[lightNum]);
NdotHV = max(dot(n,halfV),0.0);
float p = pow(NdotHV, gl_FrontMaterial.shininess);

color = att * gl_FrontMaterial.specular * light.specular * p;
}
}

return color;
}


void main()
{
vec3 n = normalize(normal);
vec4 color = vec4(0, 0, 0, 0);

color += pointLightContribution (0, gl_LightSource[0], n);
gl_FragColor = color;
}

Share this post


Link to post
Share on other sites
Quote:
The really strange part is that the performance drop seems to come from RETURNING a value, not from the calculations themselves.


If you're concerned that it is returning from your function that is the issue, can you just cut/paste inline the function to prove it?

Quote:
Are there any good OS X tools to help debugging/profiling GLSL code?


If you've got deep pockets you could spring for Mac version of gDEBugger, which is an awesome debugger yet costs like $800 dollars :(
You can play with the free trial version for a few weeks if you want to use it to look at your problem.



Share this post


Link to post
Share on other sites
I just tried inlining the whole function (by putting it in main()... I assume that's what you meant). Same results.

I didn't mean returning the value from the function was the problem, but returning a non-zero vec4 from the shader itself into gl_FragColor.

Share this post


Link to post
Share on other sites
Its possible that by setting the final fragcolor to 0,0,0,0 that GLSL is just optimizing out your entire shader and not doing anything (as having 0 alpha would do nothing to the image), so this might not be a very good idea for a test.

I'm having trouble viewing your first image (it loads halfway and then the image dissappears and says "The image “http://img532.imageshack.us/img532/2552/screenshot20100424at507.png” cannot be displayed, because it contains errors." So I'm not sure whats happening there.

Can I ask what resolution you are running this test at?

Share this post


Link to post
Share on other sites
Hmm, yeah. Seems like something broke during the image upload. I uploaded another copy, this one looks ok:

http://img227.imageshack.us/i/screenshot20100424at507.png/

My display is 1440x900, but the renderer is not quite fullscreen - there's a sidebar with some Cocoa controls and things. The renderer resolution is probably somewhere in the ballpark of 1200x900.

Share this post


Link to post
Share on other sites
Does your slowdown always correspond to when looking at the top of that sphere? It looks kind of funky, like the texture is super-compressed or something.

Also, try shrinking the size of the window when you're getting your slowdown and see if that helps speed things up. Could identify if you're really fill limited (pixel shader) or if there's something else going on.

Share this post


Link to post
Share on other sites
No, it doesn't correspond to the top of the sphere, and it happens when the weird part of the texture isn't visible as well. That was just a quick Maya export for test purposes, I think the texturing issue is just funky UV coords. The same issue happens with other models, including simple cubes, textured or not.

I just did a test - the problem does still occur with a much smaller window (400x400 maybe), and the frame rate drops to the same 50-60 FPS level.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

Sign in to follow this  

  • Forum Statistics

    • Total Topics
      627745
    • Total Posts
      2978904
  • Similar Content

    • By DelicateTreeFrog
      Hello! As an exercise for delving into modern OpenGL, I'm creating a simple .obj renderer. I want to support things like varying degrees of specularity, geometry opacity, things like that, on a per-material basis. Different materials can also have different textures. Basic .obj necessities. I've done this in old school OpenGL, but modern OpenGL has its own thing going on, and I'd like to conform as closely to the standards as possible so as to keep the program running correctly, and I'm hoping to avoid picking up bad habits this early on.
      Reading around on the OpenGL Wiki, one tip in particular really stands out to me on this page:
      For something like a renderer for .obj files, this sort of thing seems almost ideal, but according to the wiki, it's a bad idea. Interesting to note!
      So, here's what the plan is so far as far as loading goes:
      Set up a type for materials so that materials can be created and destroyed. They will contain things like diffuse color, diffuse texture, geometry opacity, and so on, for each material in the .mtl file. Since .obj files are conveniently split up by material, I can load different groups of vertices/normals/UVs and triangles into different blocks of data for different models. When it comes to the rendering, I get a bit lost. I can either:
      Between drawing triangle groups, call glUseProgram to use a different shader for that particular geometry (so a unique shader just for the material that is shared by this triangle group). or
      Between drawing triangle groups, call glUniform a few times to adjust different parameters within the "master shader", such as specularity, diffuse color, and geometry opacity. In both cases, I still have to call glBindTexture between drawing triangle groups in order to bind the diffuse texture used by the material, so there doesn't seem to be a way around having the CPU do *something* during the rendering process instead of letting the GPU do everything all at once.
      The second option here seems less cluttered, however. There are less shaders to keep up with while one "master shader" handles it all. I don't have to duplicate any code or compile multiple shaders. Arguably, I could always have the shader program for each material be embedded in the material itself, and be auto-generated upon loading the material from the .mtl file. But this still leads to constantly calling glUseProgram, much more than is probably necessary in order to properly render the .obj. There seem to be a number of differing opinions on if it's okay to use hundreds of shaders or if it's best to just use tens of shaders.
      So, ultimately, what is the "right" way to do this? Does using a "master shader" (or a few variants of one) bog down the system compared to using hundreds of shader programs each dedicated to their own corresponding materials? Keeping in mind that the "master shaders" would have to track these additional uniforms and potentially have numerous branches of ifs, it may be possible that the ifs will lead to additional and unnecessary processing. But would that more expensive than constantly calling glUseProgram to switch shaders, or storing the shaders to begin with?
      With all these angles to consider, it's difficult to come to a conclusion. Both possible methods work, and both seem rather convenient for their own reasons, but which is the most performant? Please help this beginner/dummy understand. Thank you!
    • By JJCDeveloper
      I want to make professional java 3d game with server program and database,packet handling for multiplayer and client-server communicating,maps rendering,models,and stuffs Which aspect of java can I learn and where can I learn java Lwjgl OpenGL rendering Like minecraft and world of tanks
    • By AyeRonTarpas
      A friend of mine and I are making a 2D game engine as a learning experience and to hopefully build upon the experience in the long run.

      -What I'm using:
          C++;. Since im learning this language while in college and its one of the popular language to make games with why not.     Visual Studios; Im using a windows so yea.     SDL or GLFW; was thinking about SDL since i do some research on it where it is catching my interest but i hear SDL is a huge package compared to GLFW, so i may do GLFW to start with as learning since i may get overwhelmed with SDL.  
      -Questions
      Knowing what we want in the engine what should our main focus be in terms of learning. File managements, with headers, functions ect. How can i properly manage files with out confusing myself and my friend when sharing code. Alternative to Visual studios: My friend has a mac and cant properly use Vis studios, is there another alternative to it?  
    • By ferreiradaselva
      Both functions are available since 3.0, and I'm currently using `glMapBuffer()`, which works fine.
      But, I was wondering if anyone has experienced advantage in using `glMapBufferRange()`, which allows to specify the range of the mapped buffer. Could this be only a safety measure or does it improve performance?
      Note: I'm not asking about glBufferSubData()/glBufferData. Those two are irrelevant in this case.
    • By xhcao
      Before using void glBindImageTexture(    GLuint unit, GLuint texture, GLint level, GLboolean layered, GLint layer, GLenum access, GLenum format), does need to make sure that texture is completeness. 
  • Popular Now