Using webgl I need to perform 3 passes to render my scene. Each pass runs the same geometry and shaders but has differing values for some uniforms and textures.
I seem to have two choices. Have a single "program" and set all of the uniforms and textures for each pass. Or have 3 "programs" each containing the same shaders, and set all the necessary uniforms/shaders once per program, and then just switch programs for each pass. This means that I will do one useProgram call per pass instead of man setUniform calls for each pass.
Is this second technique likely to be faster as it will avoid very many setuniform calls, or is changing the program very expensive? I've done some trials but with the very simple geometry I have at the moment I don't see any difference in performance because setup costs overwhelm any differences.
Is there any reason to prefer one technique over the other?
Just send different values via glUniform if the shader programs are the same.
Switching between programs is generally slower than change value of uniform.
Anyway Uber Shader Program (with list of uniforms like useLighting, useAlphaMap) in most cases aren't good.
#gman
We are talking about WebGL (GLES 2.0) where we don't have UBO. (uniform buffer object)
#top
Summing try to avoid rebinding shader programs (but it's not end of the world) and don't create one uber shader!
When you have large amouts of textures to rebind, texture atlasing should be the fastest solution, so you don't need to rebind textures, don't need to rebind programs. Textures can be switched by modifying uniforms representing texCoord offsets.
Modifying such uniforms can be optimized even further:
You should consider moving frequently modified uniforms to attributes. Usualy their data source are provided using attribPointers but you can also use constant values when they are disabled. Instead of unformXXX() use attribXXX() functions to specify their constant values.
I think best example is light position. Normaly you'd have to specify uniform values for it every time light position changes to ALL programs that make use of it. In contrast, when using 'attributed' uniforms you can specify attribute value once globaly when your light moves.
-pros:
This method is best suited when you have many programs which would like to share uniforms, as we know we can't use uniform buffers in WebGL, it seams to be the only reasonable solution.
-cons:
Of course available size of such 'attributed' uniforms will be much smaller than using regular uniforms, but it still can speed things up a lot if you do it to some part of your uniforms.
Related
I have some vertex data. Positions, normals, texture coordinates. I probably loaded it from a .obj file or some other format. Maybe I'm drawing a cube. But each piece of vertex data has its own index. Can I render this mesh data using OpenGL/Direct3D?
In the most general sense, no. OpenGL and Direct3D only allow one index per vertex; the index fetches from each stream of vertex data. Therefore, every unique combination of components must have its own separate index.
So if you have a cube, where each face has its own normal, you will need to replicate the position and normal data a lot. You will need 24 positions and 24 normals, even though the cube will only have 8 unique positions and 6 unique normals.
Your best bet is to simply accept that your data will be larger. A great many model formats will use multiple indices; you will need to fixup this vertex data before you can render with it. Many mesh loading tools, such as Open Asset Importer, will perform this fixup for you.
It should also be noted that most meshes are not cubes. Most meshes are smooth across the vast majority of vertices, only occasionally having different normals/texture coordinates/etc. So while this often comes up for simple geometric shapes, real models rarely have substantial amounts of vertex duplication.
GL 3.x and D3D10
For D3D10/OpenGL 3.x-class hardware, it is possible to avoid performing fixup and use multiple indexed attributes directly. However, be advised that this will likely decrease rendering performance.
The following discussion will use the OpenGL terminology, but Direct3D v10 and above has equivalent functionality.
The idea is to manually access the different vertex attributes from the vertex shader. Instead of sending the vertex attributes directly, the attributes that are passed are actually the indices for that particular vertex. The vertex shader then uses the indices to access the actual attribute through one or more buffer textures.
Attributes can be stored in multiple buffer textures or all within one. If the latter is used, then the shader will need an offset to add to each index in order to find the corresponding attribute's start index in the buffer.
Regular vertex attributes can be compressed in many ways. Buffer textures have fewer means of compression, allowing only a relatively limited number of vertex formats (via the image formats they support).
Please note again that any of these techniques may decrease overall vertex processing performance. Therefore, it should only be used in the most memory-limited of circumstances, after all other options for compression or optimization have been exhausted.
OpenGL ES 3.0 provides buffer textures as well. Higher OpenGL versions allow you to read buffer objects more directly via SSBOs rather than buffer textures, which might have better performance characteristics.
I found a way that allows you to reduce this sort of repetition that runs a bit contrary to some of the statements made in the other answer (but doesn't specifically fit the question asked here). It does however address my question which was thought to be a repeat of this question.
I just learned about Interpolation qualifiers. Specifically "flat". It's my understanding that putting the flat qualifier on your vertex shader output causes only the provoking vertex to pass it's values to the fragment shader.
This means for the situation described in this quote:
So if you have a cube, where each face has its own normal, you will need to replicate the position and normal data a lot. You will need 24 positions and 24 normals, even though the cube will only have 8 unique positions and 6 unique normals.
You can have 8 vertexes, 6 of which contain the unique normals and 2 of normal values are disregarded, so long as you carefully order your primitives indices such that the "provoking vertex" contains the normal data you want to apply to the entire face.
EDIT: My understanding of how it works:
I have seen demos on WebGL that
color rectangular surface
attach textures to the rectangles
draw wireframes
have semitransparent textures
What I do not understand is how to combine these effects into a single program, and how to interact with objects to change their look.
Suppose I want to create a scene with all the above, and have the ability to change the color of any rectangle, or change the texture.
I am trying to understand the organization of the code. Here are some short, related questions:
I can create a vertex buffer with corresponding color buffer. Can I have some rectangles with texture and some without?
If not, I have to create one vertex buffer for all objects with colors, and another with textures. Can I attach a different texture to each rectangle in a vector?
For a case with some rectangles with colors, and others with textures, it requires two different shader programs. All the demos I see have only one, but clearly more complicated programs have multiple. How do you switch between shaders?
How to draw wireframe on and off? Can it be combined with textures? In other words, is it possible to write a shader that can turn features like wireframe on and off with a flag, or does it take two different calls to two different shaders?
All the demos I have seen use an index buffer with triangles. Is Quads no longer supported in WebGL? Obviously for some things triangles would be needed, but if I have a bunch of rectangles it would be nice not to have to create an index of triangles.
For all three of the above scenarios, if I want to change the points, the color, the texture, or the transparency, am I correct in understanding the glSubBuffer will allow replacing data currently in the buffer with new data.
Is it reasonable to have a single object maintaining these kinds of objects and updating color and textures, or is this not a good design?
The question you ask is not just about WebGL, but also about OpenGL and 3D.
The most used way to interact is setting attributes at the start and uniforms at the start and on the run.
In general, answer to all of your questions is "use engine".
Imagine it like you have javascript, CPU based lang, then you have WebGL, which is like a library of stuff for JS that allows low level comunication with GPU (remember, low level), and then you have shader which is GPU program you must provide, but it works only with specific data.
Do anything that is more then "simple" requires a tool, that will allow you to skip using WebGL directly (and very often also write shaders directly). The tool we call engine. Engine usually binds together some set of abilities and skips the others (difference betwen 2D and 3D engine for example). Engine functions call some WebGL preset functions with specific order, so you must not ever touch WebGL API again. Engine also provides very complicated logic to build only single pair, or few pairs of shaders, based just on few simple engine api calls. The reason is that during entire program, swapping shader program cost is heavy.
Your questions
I can create a vertex buffer with corresponding color buffer. Can I
have some rectangles with texture and some without? If not, I have to
create one vertex buffer for all objects with colors, and another with
textures. Can I attach a different texture to each rectangle in a
vector?
Lets have a buffer, we call vertex buffer. We put various data in vertex buffer. Data doesnt go as individuals, but as sets. Each unique data in set, we call attribute. The attribute can has any meaning for its vertex that vertex shader or fragment shader code decides.
If we have buffer full of data for triangles, it is possible to set for example attribute that says if specific vertex should texture the triangle or not and do the texturing logic in the shader. Anyway I think that data size of attributes for each vertex must be equal (so the textured triangles will eat same size as nontextured).
For a case with some rectangles with colors, and others with textures,
it requires two different shader programs. All the demos I see have
only one, but clearly more complicated programs have multiple. How do
you switch between shaders?
Not true, even very complicated programs might have only one pair of shaders (one WebGL program). But still it is possible to change program on the run:
https://www.khronos.org/registry/webgl/specs/latest/1.0/#5.14.9
WebGL API function useProgram
How to draw wireframe on and off? Can it be combined with textures? In
other words, is it possible to write a shader that can turn features
like wireframe on and off with a flag, or does it take two different
calls to two different shaders?
WebGL API allows to draw in wireframe mode. It is shader program independent option. You can switch it with each draw call. Anyway it is also possible to write shader that will draw as wireframe and control it with flag (flag might be both, uniform or attribute based).
All the demos I have seen use an index buffer with triangles. Is Quads
no longer supported in WebGL? Obviously for some things triangles
would be needed, but if I have a bunch of rectangles it would be nice
not to have to create an index of triangles.
WebGL supports only Quads and triangles. I guess it is because without quads, shaders are more simple.
For all three of the above scenarios, if I want to change the points,
the color, the texture, or the transparency, am I correct in
understanding the glSubBuffer will allow replacing data currently in
the buffer with new data.
I would say it is rare to update buffer data on the run. It slows a program a lot. glSubBuffer is not in WebGL (different name???). Anyway dont use it ;)
Is it reasonable to have a single object maintaining these kinds of
objects and updating color and textures, or is this not a good design?
Yes, it is called Scene graph and is widely used and might be also combined with other techniques like display list.
I've been working on improving my OpenGL ES 2.0 render performance by introducing batching; specifically one creates a RenderBatch, specifying a texture and a shader (for now) upon creation. This sets the state into a VAO to allow for inexpensive state switching. I started the implementation looking something like this:
batch = RenderBatch.new "SpriteSheet" "FlatShader"
batch.begin GL_TRIANGLE_STRIP
batch.addGeometry Geometry.newFromFile "Billboard"
batch.end
batch.render renderEngine
But then it hit me: my Billboard file has vertices that are meant to be scaled and translated for specific instance usage. So I added a transform argument to the addGeometry call.
batch.addGeometry(Geometry.newFromFile("Billboard"), myObject.transform)
This solves the problem of scaling, translating, and rotating the vertices, but it does so by first looking up the vertex information, transforming it by the transform matrix, and then inserts it into the batch data. While this works it seems inefficient; it is CPU intensive and doesn't take advantage of the GPU's transformation power. However, it works, so not that big of a deal. (Would be nice to have a better way to do this though)
However, I've run into a roadblock: texture coordinates may need to be different for each instance as well, and that means I would have to pass in a texture transformation matrix, and now this is feeling hacky.
Is there an easier way to handle this kind of transformation to existing data using shaders that does not limit the geometry/models given and is easily extensible to use normal maps, UV maps, and other fancy tricks? Thanks!
It seems to me that what you are talking about are shader uniforms. Normally you would set up the vertex data and attributes for each batch in a VBO and a VAO. Then, in your render method, you switch to the correct VAO and set up the shader uniforms. These normally include a model-view-projection matrix to transform vertices into clip space, which necessarily would change nearly every frame, the correct texture to use, etc.
This is efficient because the unchanging vertex data is held in GPU memory, the VAO takes care of cheap state switching, and only the uniforms, which generally change often, are sent to the GPU each render call.
If you are batching multiple objects that require separate model view projection matrices, then you have a few options:
you have to perform a separate draw call for each batch that requires a separate model view projection matrix
use an array of model view projection matrices as a uniform and have an attribute for each object that provides the correct projection matrix index to use
you have to transform the vertices using the CPU and refill the VBO with the updated data
The first method is the preferred solution, it will be efficient and simple. The slow part of rendering lots of draw calls is generally getting the data from the CPU to the GPU, if you already have the vertex data in VBOs then the overhead of a draw call per object is not going to be a big deal. This also solves the problem of how to provide different uniforms per object based on object properties. In each objects render method, the relevant properties are set up as uniforms before the draw call is made. If each object requires different data sent to the GPU, then how else could this work?
This is a trade-off situation. Costs of state changes due to insufficient batching compared to costs of transformation on the CPU. There is no single best solution, but it depends on how much of your scene is static, how much is dynamic and how it is laid out.
A common solution is to put static objects, whose transformation relative to each other never changes into a single VBO, or few VBOs (if they use different textures, vertex formats, etc), completely transformed. This is done once before rendering. Not each frame. Dynamic objects (players, monster, whatever) are then rendered individually, with transformation done in the vertex shader.
You can still optimize for state changes by roughly ordering the drawing of the individual objects by textures and programs.
I've got a question regarding ComputeShader compared to PixelShader.
I want to do some processing on a buffer, and this is possible both with a pixel shader and a compute shader, and now I wonder if there is any advantage in either over the other one, specifically when it comes to speed. I've had issues with either getting to use just 8 bit values, but I should be able to work-around that.
Every data point in the output will be calculated from using in total 8 data points surrounding it (MxN matrix), so I'd think this would be perfect for a pixel shader, since the different outputs don't influence each other at all.
But I was unable to find any benchmarkings to compare the shaders, and now I wonder which one I should aim for. Only target is the speed.
From what i understand, shaders are shaders in the sense that they are just programs run by alot of threads on data. Therefore, in general there should not be any diffrence in terms of computing power/speed doing calculations in the pixel shader as opposed to the compute shader. However..
To do calculations on the pixelshader you have to massage your data so that it looks like image data, this means you have to draw a quad first of all, but also that your output must have the 'shape' of a pixel (float4 basically). This data must then be interpreted by you app into something useful
if you're using the computeshader you can completly control the number of threads to use where as for pixel shaders they have to be valid resolutions. Also you can input and output data in any format you like and take advantage of accelerated conversion using UAVs (i think)
i'd recommend using computeshaders since they are ment for doing general purpose computation and are alot easier to work with. Your over all application will probably be faster too, even if the actual shader computation time is about the same, just because you can avoid some of the hoops you have to jump through just through to get pixel shaders to do what you want.
As a learning experience, I'm writing an Immediate mode managed DirectX 9 application.
I'm manually calculating Vertex normals across all triangles in a scene to allow smooth Gouraud shading.
This works as expected, but I'm guessing this is not the most efficient approach. Is it possible to get the GPU to do this for me?
You could in theory generate the vertex normals inside the vertex shader. That involves computation every single time you render a mesh using that shader though, so why not generate them in advance.
If you mean you want to generate them in advance of rendering, but use the GPU instead of the CPU, I would say that it's not worth the bother of speeding up something you are only going to do once. Besides, I'm not sure if DX9 has a way to get computed vertex information back from a shader (DX10 does).
All in all, the best thing to do in most cases is the traditional: compute vertex normals in the program that saves the data files that contain the meshes - do it as a pre-computation step. Usually you have them if the mesh came from a 3d package like Max or Maya, because there is artistic information in the normals, unless you know the whole mesh is supposed to be perfectly smooth (or faceted), it's not computable in the general case.