gain and exposure values for planets

Since it's no longer freezing in the observatory I can take he gloves off and to experiment a bit. One thing that's always confused me with planetary imaging is what the best gain/exposure combination is.

So here's the results of messing about with Mars the other night and some different settings:


I was a little surprised by the result. I'd expected the lowest gain (and hence lowest noise) to give the best result but it didn't. I'm putting that down to the seeing being unsteady enough that a 100ms exposure got blurred in the seeing.

So it seems that a sane value for gain, for my setup anyway, is around 50%. And I should pick an exposure value that gives a decent histogram, or suits the seeing, or gives a high framerate (for Jupiter).

Actually that's my next experiment. Online wisdom seems to suggest getting a histogram in the 70-90% range. I'm not sure about the wisdom of that though if you are using a higher gain to get into that range. It's the same data -you're just 'stretching' it using gain. So would my 10ms exposures, with a lower gain and a lower histogram give a better final image?

That's interesting Albert - I have always assumed that gain = noise and kept it as low as possible with longer shutter speeds as a result, but perhaps I should try turning things up a bit?

Interesting experiment Albert, and well illustrated. I'm not sure exactly how different gain levels affect final noise in the processed image. If you use low gain when capturing to keep noise down, don't you also lose some tonailty in the darker pixels? Equivalent to blowing out detail with too much gain? I'm not sure about this, the data may well be there depending on the sensitivity and bit depth of the sensor, but I have noticed that using low gain vs medium or high on a capture where all else is equal does not usually give me a better final result.
However in saying that I have not done any real thought out experiments on it, just a few test shots as an after thought on the night.

I do find though, that the gain level I select is pretty much dictated by the exposure time which in turn is dictated by the 'speed' of the seeing, but judging the exact gain level to use by eye on the night is difficult.