>Psychtoolbox>PsychDemos

Demonstrate how to do display devignetting aka per-pixel gain correction.

Usage: VignettingCorrectionDemo([docolor=0][, precision=max]);

‘docolor’ if set to non-zero will compute and apply per-color channel
gains instead of the default luminance gain.

‘precision’ selects the precision vs. speed tradeoff: 2 = max precision,
min speed, 1 = balanced, 0 = max speed and minimum precision.

Because gain correction is memory and compute intense, per-color
correction is slower than global luminance only correction. Higher
precisions for the definition of the gainmatrix are slower, and higher
display resolutions are slower.

The demo runs until a key is hit twice.

On modern graphics cards, Psychtoolbox allows to automatically apply a
luminance- or color gainfield to all presented stimuli. Each pixel
location can get individually gain corrected. This is useful to correct
for spatial luminance or color inhomogenities of display devices, e.g.,
video projectors (due to lens vignetting effects), or flat panels (with
their contrast and color view dependency).

You need recent graphics hardware for this to work and to work at a
decent speed.

See “help PsychImaging” under subsection ‘ColorCorrection’ and “help
PsychColorCorrection” under subsection ‘GainMatrix’ for explanation of
the math, the parameters and other notable things of interest about gain
correction.