I am using gratingstim to create sin grayscale drifting gradients of various orientations. I recently gamma corrected the monitor (see attached graph) by custom script and noticed tha grating changed appearance. In particular, the black bars seem to become narrow, while white bars wider. I cant really attach the screen shot because it will look normal (sinusoidal) on different monitor.
Does anybody experienced this? Is it normal for gratings to change after gamma correction?
Yes that is exactly what you would expect from successful gamma correction. The mid-greys are brighter on a corrected monitor to undo the effects of the curve you posted, and that means the bright bars will look wider.
Sorry for digging up this question. I have the same problem, after successful gamma correction the grating seems distorted as mentioned here. However, I want my grating not to be distorted. Any suggestion in how to correct the grating so that an observer will see a perfect sinusoid?
Correction only makes sense, if wanted. “looks like” may differ from “measured” contrast. After correction, just calculate the stimulus/sinusoidal:
if you linearized (gamma=1.00), and want the sinusoidal “look good” (which is not linear in the sense of luminance contrast), calculate it with intended gamma, e.g. 2.20.
if you corrected to gamma=2.20 let’s say, and want the sinusoidal having linear contrast, calculate it with the inverse gamma of 2.20.
After doing some experiments, I can confirm this distortion is a pure illusion due to the Weber-Fechner law. Its only our eyes that see black bars to become narrow and white bars wider. But that is not what is happening in reality.
I measured the display with a luminanciometer and obtained a perfect smooth sinusoid.
t