|
looking forward to your results Greg
Since read noise was so low in the binning test, decided to try software binning to take the 1x1 result to 2x2 in Nebulosity. Ended up with an SNR of 24.7 which is practically identical to the hardware binning. Thinking about it, the main advantage of hardware binning is the possibility that the single read is less noisy than the combined read noise from four pixels merged in software - but if the read noise is well below the shot noise anyway, software binning produces the same result as hardware binning.
This ties in with comments on Greg's parallel thread where it is suggested by Rick and Rob that full resolution colour data could be gathered 1x1 and then software binned if required - looks like a good strategy if you have read noise well below the shot noise.
|