Question

I’m trying to apply a video file as a texture in OpenGL ES on iOS 5.0+ using CVOpenGLESTextureCache.

I’ve found Apple’s RosyWriter sample code, and have been reading through it.

The question I have is:

How are the textures finally being delivered to the uniforms in the fragment shader?

In the RosyWriterPreviewView class, I follow it all the way up to

glBindTexture(CVOpenGLESTextureGetTarget(texture), 
               CVOpenGLESTextureGetName(texture))

after which some texture parameters are specified.

However, I don’t see the texture uniform (sampler2D videoframe) ever being explicitly referenced by the sample code. The texture-sending code I’ve become used to would look something like:

GLint uniform = glGetUniformLocation(program, "u_uniformName");

with a subsequent call to actually send the texture to the uniform:

glUniform1i(GLint location, GLint x);

So I know that SOMEhow RosyWriter is delivering the texture to the uniform in its fragment shader, but I can’t see how and where it’s happening.

In fact, the sample code includes the comment where it builds up the OpenGL program:

// we don't need to get uniform locations in this example

Any help on why this is & how the texture is getting sent over would be great.

Answer 1

In the RosyWriter example, I think the reason they’re able to get away without using glUniformi() at any point for the videoframe uniform is that they’re binding the input texture to texture unit 0.

When specifying a uniform value for a texture, the value you use is the texture unit that texture is bound to. If you don’t set a value for a uniform, I believe it should default to 0, so by binding the texture to unit 0 always they never have to set a value for the videoframe uniform. It will still pull in the texture attached to unit 0.