Thanks for the reply as I have been looking for an explanation for some time.
After stacking dark frames in Pixinsight, the mean of the integrated image as reported by Pixinsight is always less than 16 times the offset. This is with unity gain. I would expect that 16 times the offset would be the minimum.
For example: At unity gain and an offset of 22, a stack of 100 x 1 second darks stack up with a mean of 329, why is the result not greater than 352.
At 0 seconds the mean is 315, not greater than 352. (There is a drop off below .1 seconds)
The trend is the same for the the 50 or so combinations of offsets and integration times I have looked at. (All less than 1 second)
I really like the camera and am not complaining about anything. I'm just trying to understand my results and all I can come up with is that the gain is not unity at a setting of 139.
Ray J