• Advertisement
Sign in to follow this  

OpenGL Mipmaps with multiple texture arrays

This topic is 1049 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

So, I'm in need of using a couple texture arrays, rather than packing everything into a single array. I'm baking a lightmap/normals/etc into various images to use for most of the background objects, as the player doesn't interact with them, so there's little need to be calculating shadows/normals etc in real-time. However, I'm finding that using higher definition images when baking those results in much better results, especially as I'm scaling some of them rather large. But, since I don't want to use huge textures for everything, I thought I would just use two arrays. One for regular objects, that will have their shadows mapped, and normal/specular images separate, and one for the higher definition images, with the shadows/normal lighting baked in.

 

However, I'm struggling a little with the mipmap levels, I think. It's possible I'm erring elsewhere, but I'll try to explain:

 

When I draw two different objects, each using a different texture array, the image from the array with smaller sized images array loaded first becomes darker, especially when viewed from a distance (like it's being improperly mipmapped). Upon moving closer to the object, it becomes more visible, but still a bit wonky. If I drop the mipmap to 1 for the array that's loaded first, the darkness on the texture goes away (but, then I have no mipmapping on the textures in that array).

 

Here's a stripped down example, forgive the naming conventions, I'm terrible at coming up with things on the fly, and needed to rewrite these in a simpler format than they exist:

void Textures::LoadTextures()
{
    CreateTextureArray();
    LoadImages();
    CreateHDTextureArray();
    LoadHDImages();
}
void Textures::CreateTextureArray()
{
    glGenTextures(1, &TextureArray);
    glBindTexture(GL_TEXTURE_2D_ARRAY, TextureArray);
    glTexStorage3D(GL_TEXTURE_2D_ARRAY,11, GL_RGBA8, 1024, 1024, 18);
}
void Textures::CreateHDTextureArray()
{
    glGenTextures(1, &HDTextureArray);
    glBindTexture(GL_TEXTURE_2D_ARRAY, HDTextureArray);
    glTexStorage3D(GL_TEXTURE_2D_ARRAY,11, GL_RGBA8, 4096, 4096, 1);

}
void Textures::LoadImageToArray(sf::Image &imageIN, const char* imageName, int layerNumber)
{
    if(!imageIN.loadFromFile(imageName))
    {
        std::cout<<imageName<<" is borked..."<<std::endl;
    }

    glTexSubImage3D(GL_TEXTURE_2D_ARRAY, 0, 0, 0, layerNumber, imageIN.getSize().x,
                    imageIN.getSize().y,1, GL_RGBA, GL_UNSIGNED_BYTE, imageIN.getPixelsPtr());
}

When drawing:

void EnableTexture(Object &ObjectIN, GLuint &textureArrayIN, glm::vec3 &imageIndexIN)
{
    glActiveTexture(GL_TEXTURE0);
    glBindTexture(GL_TEXTURE_2D_ARRAY, textureArrayIN);
    glUniform1i(ObjectIN.gTextureArrayID(),0);
    glUniform1i(ObjectIN.gLayerNumID1(), imageIndexIN.x);
    glUniform1i(ObjectIN.gLayerNumID2(), imageIndexIN.y);
    glUniform1i(ObjectIN.gLayerNumID3(), imageIndexIN.z);
} 

the opengl references are a little light on information for glTexStorage3D, but from what I gather, I can use a value up to a certain fraction of the image size for the mipmap level (in this case 11, for the 1024, and I believe 13 for the 4096 textures.

 

My question is, when binding a new array, does that override the settings for the previous array (specifically mipmap sizes for both arrays)? Or, am I perhaps going about this entirely incorrectly?

Is there a step I'm missing when binding the array for use in the shader?

 

Additionally, I've read I can just use the single larger array, and use a mipmap portion when I need a smaller image. Is there any advantage to that, rather than this approach? I imagine there's some benefit, as I'm not rebinding the arrays, but if it's marginal, this seems like the simpler approach.

 

If I'm not providing enough information, or if some images would help explain, please let me know.

 

I appreciate any advice smile.png

 

Edit*

Here are a couple of images to maybe help explain. Forgive the overwhelming redness, it's just a hodgepodge of stuff thrown together for an example with none of the shaders balanced tongue.png

Edited by Misantes

Share this post


Link to post
Share on other sites
Advertisement

So you have 2 separate objects. 2 texture arrays, and what appears to be only mipLevel 0 appearing on one, and any other further away smaller mips come out as black?

 

Have you tried changing the black object to not use a texture array and verify it is mip mapping correctly? Did you try calling glGetError() at any point? I recently had some issues with my texture array code and mip map generation and it caused the images to be all black. I can't remember the issue but I think you need to have your glTexParemeteri() setup for mipmapping before uploading your mip maps.

Edited by dpadam450

Share this post


Link to post
Share on other sites

Yeah, all the images mip fine individually, have tested pretty thoroughly with that. I can render each perfectly fine and mip-mapped on their own within an array or on their own. It's only when I try to render both with arrays that the array first loaded gets screwy. glGetError() calls aren't reporting anything.

 

my parameters looks like this:

    glTexParameteri(GL_TEXTURE_2D_ARRAY, GL_TEXTURE_WRAP_S, GL_REPEAT);
    glTexParameteri(GL_TEXTURE_2D_ARRAY, GL_TEXTURE_WRAP_T, GL_REPEAT);
    glTexParameteri(GL_TEXTURE_2D_ARRAY, GL_TEXTURE_MAG_FILTER, GL_LINEAR);
    glTexParameteri(GL_TEXTURE_2D_ARRAY, GL_TEXTURE_MIN_FILTER, GL_LINEAR_MIPMAP_LINEAR);

    glGenerateMipmap(GL_TEXTURE_2D_ARRAY);

Which, I believe should be fine (again, it's fine for individual arrays). It is loaded before the arrays are created.

Share this post


Link to post
Share on other sites

Sounds like a potential state issue. What happens if you apply the same texture array to both objects? Have you tried rendering them as object 1 , then 2  and then 2,1 as well to see if it is always the second object, or if it is always that one specific object that won't work?

Edited by dpadam450

Share this post


Link to post
Share on other sites

Thanks, I appreciate the continued help smile.png

 

There are other objects in the scene from the same array, the issue persists with all of them, not just a single object.

 

When using the same texture array on both objects (using the 4096 array), things are rendered fine, excluding the fact that the texture isn't properly uvmapped.

Using the 1024 array is seemingly fine as well, though looks rather odd since it's trying to use the 1024 texture in lieu of a 4096 one, so I get a hodgepodge of various images in that array.

 

Rendering order has no effect, the problems persist regardless of which is drawn first, and rendering order doesn't affect which array has issues.

 

The only thing that really affects things is whichever array is loaded first is the one with issues. As though the mipmap level values get deleted on the first array when the second one is loaded. 

 

Should I be clearing or resetting anything between binding textures? Is there any issue with glTexSubImage3D, loading different sizes into that, bearing in mind I'm loading all of one size, binding another array, then loading all of the next size?

Edited by Misantes

Share this post


Link to post
Share on other sites

glGenerateMipmap(GL_TEXTURE_2D_ARRAY);

 

Is this called before or after the image loading calls? I forget if it should be before or after.

 

"The only thing that really affects things is whichever array is loaded first is the one with issues."

 

Just to be clear. If you bypass loading the second image, the first image will is completely fine?  And when you tested your texture arrays being applied, did you still load both, but only bind one during drawing? Since it obviously sounds like a loading issue, just wondering  if you still loaded them both.

 

Have you tried gDebugger to see if it says your texture arrays are mip-mapped. You can also use the texture viewer to see what they hold.

Edited by dpadam450

Share this post


Link to post
Share on other sites

glGenerateMipmap(GL_TEXTURE_2D_ARRAY);
 
Is this called before or after the image loading calls? I forget if it should be before or after.

I've been calling it after. A quick test of calling it before makes both arrays have the problem. I'm not 100% certain this is correct, but has served me thus far.

 

 


Just to be clear. If you bypass loading the second image, the first image will is completely fine?  And when you tested your texture arrays being applied, did you still load both, but only bind one during drawing? Since it obviously sounds like a loading issue, just wondering  if you still loaded them both.

 

Edit* my apologies, I've probably been confusing you with poorly used terminology, when I've said "bind" or "load" before, I was usually referring to when the array is bound at creation, not during the render function. It is the array that is created first that has the issues.

 

Yes, images in the first array are perfectly fine if I don't create the second array, and vice versa.

 

When I tested them, I did the following:

1- create and load first array only. Bind and apply images from first array to all objects when rendering. Result: everything fine, excluding uvmapping tongue.png

2-create and load second array only. Bind and apply images from second array to all objects when rendering. Result: everything fine, excluding uv-mapping.

3-create and load both arrays, try various bind/render orders. Result: whichever array was created first when textures are loaded, has issues.

4-create and load both arrays, using images from only one of the arrays during render. Result:  if using images from first array, problems persist. If using images from second array, no issues.

 

Basically,if more than one array is created, whichever array is created first is the one whose textures will have issues, regardless of render order, which object they're applied to, or seemingly anything else.

 

To note: if I set the mipmap level for the array with the 4096 version to 1, and load it first, things render more or less ok, as they're large scaled-up objects with large textures. The issues don't seem to persist, if only because the objects are never too far away or small enough for the lack of mipmapping to be noticeable.

 

I've never used gDebugger, but I'm downloading it now. I'll try it and out and see what it reports.

 

In case it helps clear things up at all and is easier to see, here is the issue with the larger array illustrated.

Edited by Misantes

Share this post


Link to post
Share on other sites

glGenTextures(1,&GL_Index);
    glBindTexture(GL_TEXTURE_2D_ARRAY,GL_Index);
    glTexStorage3D(GL_TEXTURE_2D_ARRAY, numDiffuseMipMapLevels, pixelFormat, TextureSize,TextureSize, textures.size());
  

mipLevelUploading = 0;
//loop through 'i' textures

glTexSubImage3D(GL_TEXTURE_2D_ARRAY, mipLevelUploading, 0, 0, i, diffuseTextureSize, diffuseTextureSize, depth, pixelFormat, GL_UNSIGNED_BYTE, texture->pixels);
glGenerateMipmap(GL_TEXTURE_2D_ARRAY);
    glTexParameteri(GL_TEXTURE_2D_ARRAY,GL_TEXTURE_MIN_FILTER,GL_LINEAR_MIPMAP_LINEAR);
    glTexParameteri(GL_TEXTURE_2D_ARRAY,GL_TEXTURE_MAG_FILTER,GL_LINEAR);

That's my code and this is called about 8 times back to back uploading all my texture arrays.

 

Maybe your glTexStorage3D call isn't factoring in mip maps properly?..........aha scrolled up,:

glTexStorage3D(GL_TEXTURE_2D_ARRAY,11, GL_RGBA8, 4096, 4096, 1); <-------------------13 mip maps, not 11

Edited by dpadam450

Share this post


Link to post
Share on other sites

MipMapCount = log((double)TextureSize) / log((double)2);
This will calculate the mip map level given an input "texturesize", so you don't have to hard code it.

Share this post


Link to post
Share on other sites

That's my code and this is called about 8 times back to back uploading all my texture arrays.

Hrm. Our code is looking pretty identical. I'm now questioning whether this is actually the issue. I'm not seeing any significant differences.

 

 

 


glTexStorage3D(GL_TEXTURE_2D_ARRAY,11, GL_RGBA8, 4096, 4096, 1); <-------------------13 mip maps, not 11

 

Unfortunately, during testing, I had already calculated the 13 out and used that for the larger array(though still hardcoded, I should fix that), but it doesn't affect things. I should have updated that here. Also, 11 should still be a valid value, and it affects the smaller array as well, which had the correct mipmap level.

 

We haven't talked about it, but I still need to rebind each of the arrays at render time, right before they're used, correct?

 

I still appreciate the help though. I'll make up a quick small program, trying to eliminate any other variables that might be coming into play here, and see how that plays out. I'm pretty stumped though, otherwise.

Edited by Misantes

Share this post


Link to post
Share on other sites

Alright, I feel incredibly stupid right now, but I figured out the issue.

I was setting the glTexParametri() functions and generating the mipmap levels just once, after all of the arrays were loaded (only setting mipmap levels for the last array created). Which, in hindsight, is incredibly obvious...but, I was overlooking the generateMipmap() function and thinking more that the glTexparameter settings were the same for both arrays, so was not repeating my function that included those calls for each of them.

 

'tis working now. But, thank you again, I truly appreciate the help, and your layout of things led me to the solution. :)

 

Cheers!

Share this post


Link to post
Share on other sites

All texture parameters need to be set for the current bound texture. You can also change them at run time.

 

Also, you may want to turn on anisotropic filtering for some nicer results. Something you can change at run-time with the click of a button to see the change. 2,4,8,16 you can set it to. Will drop your performance a bit. I use 8.0 most of the time.

GLfloat fLargest;
glGetFloatv(GL_MAX_TEXTURE_MAX_ANISOTROPY_EXT, &fLargest);
glTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_MAX_ANISOTROPY_EXT, fLargest);

Share this post


Link to post
Share on other sites

 

All texture parameters need to be set for the current bound texture. You can also change them at run time.

 

Also, you may want to turn on anisotropic filtering for some nicer results. Something you can change at run-time with the click of a button to see the change. 2,4,8,16 you can set it to. Will drop your performance a bit. I use 8.0 most of the time.

GLfloat fLargest;
glGetFloatv(GL_MAX_TEXTURE_MAX_ANISOTROPY_EXT, &fLargest);
glTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_MAX_ANISOTROPY_EXT, fLargest);

Nice, thanks for the tip :)

Share this post


Link to post
Share on other sites
Sign in to follow this  

  • Advertisement
  • Advertisement
  • Popular Now

  • Advertisement
  • Similar Content

    • By khawk
      We've just released all of the source code for the NeHe OpenGL lessons on our Github page at https://github.com/gamedev-net/nehe-opengl. code - 43 total platforms, configurations, and languages are included.
      Now operated by GameDev.net, NeHe is located at http://nehe.gamedev.net where it has been a valuable resource for developers wanting to learn OpenGL and graphics programming.

      View full story
    • By TheChubu
      The Khronos™ Group, an open consortium of leading hardware and software companies, announces from the SIGGRAPH 2017 Conference the immediate public availability of the OpenGL® 4.6 specification. OpenGL 4.6 integrates the functionality of numerous ARB and EXT extensions created by Khronos members AMD, Intel, and NVIDIA into core, including the capability to ingest SPIR-V™ shaders.
      SPIR-V is a Khronos-defined standard intermediate language for parallel compute and graphics, which enables content creators to simplify their shader authoring and management pipelines while providing significant source shading language flexibility. OpenGL 4.6 adds support for ingesting SPIR-V shaders to the core specification, guaranteeing that SPIR-V shaders will be widely supported by OpenGL implementations.
      OpenGL 4.6 adds the functionality of these ARB extensions to OpenGL’s core specification:
      GL_ARB_gl_spirv and GL_ARB_spirv_extensions to standardize SPIR-V support for OpenGL GL_ARB_indirect_parameters and GL_ARB_shader_draw_parameters for reducing the CPU overhead associated with rendering batches of geometry GL_ARB_pipeline_statistics_query and GL_ARB_transform_feedback_overflow_querystandardize OpenGL support for features available in Direct3D GL_ARB_texture_filter_anisotropic (based on GL_EXT_texture_filter_anisotropic) brings previously IP encumbered functionality into OpenGL to improve the visual quality of textured scenes GL_ARB_polygon_offset_clamp (based on GL_EXT_polygon_offset_clamp) suppresses a common visual artifact known as a “light leak” associated with rendering shadows GL_ARB_shader_atomic_counter_ops and GL_ARB_shader_group_vote add shader intrinsics supported by all desktop vendors to improve functionality and performance GL_KHR_no_error reduces driver overhead by allowing the application to indicate that it expects error-free operation so errors need not be generated In addition to the above features being added to OpenGL 4.6, the following are being released as extensions:
      GL_KHR_parallel_shader_compile allows applications to launch multiple shader compile threads to improve shader compile throughput WGL_ARB_create_context_no_error and GXL_ARB_create_context_no_error allow no error contexts to be created with WGL or GLX that support the GL_KHR_no_error extension “I’m proud to announce OpenGL 4.6 as the most feature-rich version of OpenGL yet. We've brought together the most popular, widely-supported extensions into a new core specification to give OpenGL developers and end users an improved baseline feature set. This includes resolving previous intellectual property roadblocks to bringing anisotropic texture filtering and polygon offset clamping into the core specification to enable widespread implementation and usage,” said Piers Daniell, chair of the OpenGL Working Group at Khronos. “The OpenGL working group will continue to respond to market needs and work with GPU vendors to ensure OpenGL remains a viable and evolving graphics API for all its customers and users across many vital industries.“
      The OpenGL 4.6 specification can be found at https://khronos.org/registry/OpenGL/index_gl.php. The GLSL to SPIR-V compiler glslang has been updated with GLSL 4.60 support, and can be found at https://github.com/KhronosGroup/glslang.
      Sophisticated graphics applications will also benefit from a set of newly released extensions for both OpenGL and OpenGL ES to enable interoperability with Vulkan and Direct3D. These extensions are named:
      GL_EXT_memory_object GL_EXT_memory_object_fd GL_EXT_memory_object_win32 GL_EXT_semaphore GL_EXT_semaphore_fd GL_EXT_semaphore_win32 GL_EXT_win32_keyed_mutex They can be found at: https://khronos.org/registry/OpenGL/index_gl.php
      Industry Support for OpenGL 4.6
      “With OpenGL 4.6 our customers have an improved set of core features available on our full range of OpenGL 4.x capable GPUs. These features provide improved rendering quality, performance and functionality. As the graphics industry’s most popular API, we fully support OpenGL and will continue to work closely with the Khronos Group on the development of new OpenGL specifications and extensions for our customers. NVIDIA has released beta OpenGL 4.6 drivers today at https://developer.nvidia.com/opengl-driver so developers can use these new features right away,” said Bob Pette, vice president, Professional Graphics at NVIDIA.
      "OpenGL 4.6 will be the first OpenGL release where conformant open source implementations based on the Mesa project will be deliverable in a reasonable timeframe after release. The open sourcing of the OpenGL conformance test suite and ongoing work between Khronos and X.org will also allow for non-vendor led open source implementations to achieve conformance in the near future," said David Airlie, senior principal engineer at Red Hat, and developer on Mesa/X.org projects.

      View full story
    • By _OskaR
      Hi,
      I have an OpenGL application but without possibility to wite own shaders.
      I need to perform small VS modification - is possible to do it in an alternative way? Do we have apps or driver modifictions which will catch the shader sent to GPU and override it?
    • By xhcao
      Does sync be needed to read texture content after access texture image in compute shader?
      My simple code is as below,
      glUseProgram(program.get());
      glBindImageTexture(0, texture[0], 0, GL_FALSE, 3, GL_READ_ONLY, GL_R32UI);
      glBindImageTexture(1, texture[1], 0, GL_FALSE, 4, GL_WRITE_ONLY, GL_R32UI);
      glDispatchCompute(1, 1, 1);
      // Does sync be needed here?
      glUseProgram(0);
      glBindFramebuffer(GL_READ_FRAMEBUFFER, framebuffer);
      glFramebufferTexture2D(GL_READ_FRAMEBUFFER, GL_COLOR_ATTACHMENT0,
                                     GL_TEXTURE_CUBE_MAP_POSITIVE_X + face, texture[1], 0);
      glReadPixels(0, 0, kWidth, kHeight, GL_RED_INTEGER, GL_UNSIGNED_INT, outputValues);
       
      Compute shader is very simple, imageLoad content from texture[0], and imageStore content to texture[1]. Does need to sync after dispatchCompute?
    • By Jonathan2006
      My question: is it possible to transform multiple angular velocities so that they can be reinserted as one? My research is below:
      // This works quat quaternion1 = GEQuaternionFromAngleRadians(angleRadiansVector1); quat quaternion2 = GEMultiplyQuaternions(quaternion1, GEQuaternionFromAngleRadians(angleRadiansVector2)); quat quaternion3 = GEMultiplyQuaternions(quaternion2, GEQuaternionFromAngleRadians(angleRadiansVector3)); glMultMatrixf(GEMat4FromQuaternion(quaternion3).array); // The first two work fine but not the third. Why? quat quaternion1 = GEQuaternionFromAngleRadians(angleRadiansVector1); vec3 vector1 = GETransformQuaternionAndVector(quaternion1, angularVelocity1); quat quaternion2 = GEQuaternionFromAngleRadians(angleRadiansVector2); vec3 vector2 = GETransformQuaternionAndVector(quaternion2, angularVelocity2); // This doesn't work //quat quaternion3 = GEQuaternionFromAngleRadians(angleRadiansVector3); //vec3 vector3 = GETransformQuaternionAndVector(quaternion3, angularVelocity3); vec3 angleVelocity = GEAddVectors(vector1, vector2); // Does not work: vec3 angleVelocity = GEAddVectors(vector1, GEAddVectors(vector2, vector3)); static vec3 angleRadiansVector; vec3 angularAcceleration = GESetVector(0.0, 0.0, 0.0); // Sending it through one angular velocity later in my motion engine angleVelocity = GEAddVectors(angleVelocity, GEMultiplyVectorAndScalar(angularAcceleration, timeStep)); angleRadiansVector = GEAddVectors(angleRadiansVector, GEMultiplyVectorAndScalar(angleVelocity, timeStep)); glMultMatrixf(GEMat4FromEulerAngle(angleRadiansVector).array); Also how do I combine multiple angularAcceleration variables? Is there an easier way to transform the angular values?
  • Advertisement