>Psychtoolbox>PsychGLImageProcessing

gloperator = CreateGLOperator(windowPtr [, imagingmode] [, shaderhandle] [, opname] [,..shader options])

Creates an image processing operator for use with the Screen(‘TransformTexture’)
function (to be passed as ‘transformProxy’ argument) and returns
a handle ‘gloperator’ to it. The operator should be destroyed via Screen(‘Close’, gloperator);
when its not needed anymore. It gets automatically disposed when its associated onscreen window is closed.

‘windowPtr’ is the window handle of the parent onscreen window.
‘imagingmode’ (optional) to set a specifc combination of imaging flags
that affect how this operator works. Most useful flags are:
kPsychNeedDualPass if this operator should contain exactly two
operations, e.g., if it defines a separable convolution which consists of
two 1-D convolution passes, or kPsychNeedMultiPass if this operator should
contain more than two operations. Also useful are kPsychNeed16BPCFloat or
kPsychNeed32BPCFloat if you expect results of the operator to be signed
or to require floating point resolution. You can use ‘mor’ to combine
multiple of these flags, e.g., mor(kPsychNeedDualPass,kPsychNeed32BPCFloat)
to create an operator suitable for high precision processing with
possibly signed results and two processing passes.

If you want to immediately assign a single image processing operation,
you can do so by assigning an OpenGL GLSL shader ‘shaderhandle’. You can
give this shader an unique name ‘opname’ (for debugging purpose) and
provide possible additional arguments for it.

If you want to add multiple operations to the operator, you can use the
AddToGLOperator() command.

GLOperators can also get assigned to the builtin stimulus post-processing
pipeline if they should affect all created visual stimuli. See ‘help
PsychImaging’ - the section about ‘AddGLOperator’