- Primary question: * Can someone give an short and exact definition of vertex shader and pixel shader? - And some secondary questions in case the answer to the first one doesn't explain these things: * What are the differences? Which of them is used for what? Why use them? * Do they involve only color choices etc. or can they affect things such as locations of objects? If so, are they preferable over normal matrix operations to use for translating world coordinates etc.? * What are the basic primitive operations a vertex shader can do? What are the basic primitive operations a vertex shader can do? * Where does HLSL come into the picture (is it used both for vertex and pixel shaders?)? HLSL like other languages I assume must be standardized by someone, who? Can I find a full documentation about the language on their homepage?

(From Wiki):

Vertex Shader:

Vertex shaders are applied for each vertex and run on a programmable vertex processor. Vertex shaders define a method to compute vector space transformations and other linearizable computations.

Fragment (Pixel Shader):

Those kind of shaders are used to compute properties which, most of the time, are recognized as pixel colors.

Pixel shaders are applied for each pixel. They are run on a pixel processor, which usually features much more processing power than its vertex-oriented counterpart.

HLSL is DirectX language for writing shaders, you used to have to write shaders out in asm style code, which is harder and more prone to errors, HLSL gives programmers a C like language to write shaders which is easier, more readable and better to understand.

Quote:Original post by all_names_taken
- Primary question:
* Can someone give an short and exact definition of vertex shader and pixel shader?

- And some secondary questions in case the answer to the first one doesn't explain these things:
* What are the differences? Which of them is used for what? Why use them?
* Do they involve only color choices etc. or can they affect things such as locations of objects? If so, are they preferable over normal matrix operations to use for translating world coordinates etc.?
* What are the basic primitive operations a vertex shader can do? What are the basic primitive operations a vertex shader can do?
* Where does HLSL come into the picture (is it used both for vertex and pixel shaders?)? HLSL like other languages I assume must be standardized by someone, who? Can I find a full documentation about the language on their homepage?

A vertex or pixel shader is a small program that runs on the graphics card.
A vertex shader operates on every vertex getting passed in. The application specifies vertices in 3D world space, and the vertex shader transforms them into 2D clip space. Things you can do in a vertex shader include vertex blending for animated meshes. You move the vertex based on some values (blending weights in this case)
A pixel shader operates on every pixel that is drawn to the screen. It takes a bunch of information like the currently set textures for things, vertex colour, etc and outputs a single ARGB value. Things you can do in a pixel shader include things like procedural fire effects and ripples.

Using vertex and pixel shaders gives you more control over what you're drawing, although not all graphics hardware supports them. Shader language looks a bit like assembly. There's two versions of each, one for OpenGL and one for DirectX. HLSL is a bit like C, and is only availaible for DirectX. I don't know about the OpenGL shader language at all, but the DirectX one can be found in the DirectX SDK. The OpenGL shader language is standardized buy the ARB, and the DirectX one by Microsoft.

Vertex and pixel shaders can only reference the one thing they're operating on. A vertex shader can't create vertices (But the geometry shader comming in D3D10 can), and it can't reference any vertex except the one that's getting transformed by the shader, and similarly for the pixel shader.

Ok thanks, looks like it's worth learning about it after all, even though it seems difficult to learn...