Low contrast stimulus issues

[wkinerk]

Hi,

I'm trying to run a contrast detection experiment and I've run into some issues with the luminance profiles of low contrast sine wave stimuli. I'm running Psykinematix on a new 13" MacBook Pro with a NVIDIA 320M video card connected to an external CRT monitor (19" NEC FE992) with 1280x1024 resolution at 75Hz.

In achromatic mode, there appears to be a "step" type anomaly at mean luminance. In this mode and using a 10 bit native card, is the output still limited to 256 grey levels? Is there a way to verify the output is finer than 256 levels? I tried running the calibration module at 516 levels, but was unable to choose a number of levels greater than 256.

I also tried running the experiment in the 10.8 bit stealing mode. The central 3 cycles of the luminance profile improves, but beyond that there are erratic overshoot/undershoot irregularities around mean luminance. Is there a known fix for these anomalies? I'm concerned that these may affect the visibility of the stimulus.

I've attached 1.75% contrast (sigma=0.5) luminance profiles obtained with ImageJ in achromatic mode and 10.8 bit stealing mode. Any help you can provide with this would be greatly appreciated.