SharpCap Forums

You've mentioned the 10/20 min stack would be a total integration time of 15 hours or so.

The deep image would be (6x1) + (36x0.67) for a total of 30 hours.

If that's right, yes the deep image should in theory have an SNR of 1.4 times the short subs.

The subs time calculation is for a given total integration time.

If possible, you could integrate 15 hours worth of the deep subs, and then the two sets would be more comparable.

Ahh - thank you zerolatitude - yes, at last it is now beginning to make some sense

Greg

Hi Robin

I have watched your excellent youtube video on this topic. Marvellous.

I wonder if it would be possible to extend the calculator on https://tools.sharpcap.co.uk/ to also allow the user to enter the camera info to take them all the way to the ideal sub exposure?

Hi,

the idea is rather to persuade people to buy the SharpCap Pro license, which will calculate that for you

cheers,

Robin

This chart is extremely interesting, Robin. It seems that if you are stacking then you should consider the stack dynamic range instead of the frame dynamic range. In this situation (which applies to my camera) I would be tempted to go with gain 100 because it has the same range as the other peak but has a fuller well depth. But that's wrong when stacking.

Do you have a formula for stack dynamic range? Seems vital.

Cheers
Tony

Hi Tony,

yes, the SNR of the stack is one of the key things - I don't think the formula is listed in the posts at the top of this thread, but it's relatively simple :

Stack SNR = number of subs * sub FWD / stack noise (in electrons)

The stack noise is the only thing there that is complex, but the formula for that is given in post #8 of this thread.

Hope this helps,

Robin

Um, that's not quite right. I was asking about dynamic range of a stack, not SNR. Did you mean

Stack DR = number of subs * sub FWD / stack noise (in electrons)

or

Stack DR = number of subs * sub FWD / stack read noise (in electrons)

I used the second formula and the results of the sensor analysis for my camera and I did not get for any value of 'number of subs' a point where one red outline is higher than the other red outline.

If we use StackFullWell = n * FullWell, StackReadNoise = sqrt(n * ReadNoise.^2) and StackDR = log2(StackFullWell ./ StackReadNoise)

then we get

StackDR(n) = 0.5 * log2(n) + StackDR(1)

so we don't get that bump in DR from stacking. Stacking just shifts the whole curve up by 0.5 * log2(n)

Hi,

so you need to use the full noise of the stack (including the shot noise of the background, which should dominate the read noise because that is how we choose the sub-exposure length).

Once you have picked a gain value, the camera read noise and sky background brightness allow you to calculate the recommended sub exposure length (t). Additionally, you get the FWD as a function of the gain value being used.

The total exposure time is T, so the number of frames captured is T/t. That means that the brightest thing you can capture in time T without loss of information to oversaturation is T/t * FWD (measured in electrons). We assume that the dimmest thing you can detect is of the order of the stack noise (in fact for a single pixel it would be a few times the stack noise, but for an extended object it may well be less than the stack noise...). Anyway, now we have the brightest and dimmest things that can be properly represented which gives us the dynamic range (being basically the SNR for a saturated signal).

Hope that's clearer,

Robin

My apologies if this has been answered somewhere.

Presuming that your guiding can handle it, and you care not concerned about spoilage of individual frames (satellites, airplanes, etc.), why would you ever trade dynamic range for higher gain and reduced read noise? As long as you are swamping the read noise with the appropriate exposure length, the total noise in the final image is determined by the total integration time, no? Why would you ever give up dynamic range?

Thanks.

Hi,

because although you might give away dynamic range in the individual frames, you can often win on dynamic range for the final stack because the you can take more frames in the same amount of time because you don't need such long exposures.

cheers,

Robin