Add new comment

Okay I did some tests with a

Okay I did some tests with a black frame by doing averages on various non-overlapping windows and the biases across channels are consistent.