The purpose of stacking

[x6gas]

I am just getting started in astrophotography using a DSLR (Nikon D40) at prime focus for DSOs and I have some basic questions. I've read a lot on this forum (and elsewhere) about stacking (and about darks, flats, flat darks etc.) but I just want to check a few things about stacking vs longer exposure - just to make sure I am on the right track. I apologies that this is probably really basic stuff for many of you (but I'm just a tyro!)

So, as I understand it, signal to noise ratio improves with a longer exposure; stacking also improves signal to noise ratio. Someone better shout if I've got that wrong!

Question 1: As far as signal to noise goes (and without calibration / noise subtraction etc), is a 60 minute exposure broadly equivalent to stacked 6 x 10 minute subs?

Question 2: Before I started looking into this, I assumed that when you stacked subs, the brightness (luminosity?) of the resulting image would be broadly the same as a long exposure equal to the sum of the subs (i.e. that the brightness of each pixel would be summed). This doesn't appear to be the case, so could I just check that the result of stacking is an image with ~the same brightness as the individual subs which you can then image process to give the same brightness as a longer single exposure (and that you can do this because of the better signal to noise ratio).

Aside from the difficulty of alignment / tracking and the increased risk of undesirable things coming into the field of view during a long exposure, and I guess the advantage of shooting the same target over different imaging sessions and stacking the results, I am really struggling to understand the benefits of stack for DSO images...

Thanks in advance!

[brianb]

I am really struggling to understand the benefits of stack for DSO images...

1. Tracking errors - if you take a large number of images and there is a tracking error that affects a few of them, you can toss out the bad frames & just use the good ones. With a single long exposure you'd have lost the whole thing.

2. Drifting clouds, aircraft lights, satellites etc ... same benefit. In fact kappa sigma stacking removes flaws in a single frame with absolutely no loss to image quality over the frame as a whole.

3. Dynamic range. If you take a single long exposure, the brightest points get overexposed. With many stacked frames, there's still useful information in the brighter parts of the image that aren't overexposed. Think of the ADU range in a stack of 100 images being 0 to 6,553,500 instead of 0 to 65,535.

Downside of stacking is that the readout noise is summed over the frames instead of being just a single number. That means that, though stacking does improve signal to noise ratio, it doesn't do so quite as much as if it was just a matter of counting incoming photons. How much the difference is does depend on the readout noise (and therefore the camera) ... with dedicated low noise astro CCDs, probably you need about 120 subs to get the same s/n as a single exposure 100 times as long. With DSLRs used at high ISO settings, I reckon you'd need 400 x 1 min subs to equal 1 x 100 min sub ...

... but don't forget that the 100 min sub is almost certain to be flawed ... and the brighter parts will be burned out through gross overexposure.

[Merlin66]

Yes is the answer!

A 60min exposure will have the same s/n as a stack of 10 x 6 mins and the total intensity recorded will be the same.

The issue is that it's easier to get a good 6min exposure (guiding errors etc etc) than it is for a 60 min exposure.

Stacking can assist in s/n due to the random pixel noise, but it can't give better than you would have achieved with a long exposure.

[leenewtoastro]

top tip

didn't know that about kappa sigma

[dph1nm]

How much the difference is does depend on the readout noise (and therefore the camera) ... with dedicated low noise astro CCDs, probably you need about 120 subs to get the same s/n as a single exposure 100 times as long. With DSLRs used at high ISO settings, I reckon you'd need 400 x 1 min subs to equal 1 x 100 min sub ...

This depends on the ratio of the read noise to the sky noise as well. If sky noise dominates the shorter exposure then there is little practical s/n advantage in doing one long exposure. I doubt from my back garden you would need 400x1min shots to match 1x100 min one!

NigelM

[x6gas]

Thanks for the really helpful replies. This is helping to clear things up in my head!

In the words of Columbo, "There's just one more thing that's bothering me..."

A 60min exposure will have the same s/n as a stack of 10 x 6 mins and the total intensity recorded will be the same.

So does that mean that the stacked image should look pretty similar (in terms of the intensity of the objects imaged - i.e. brightness of the stars on the image) to the longer exposure BEFORE you've done any image processing?

Sorry to be asking dumb questions here, but I am just trying to work out the aims of each part of the process so that I can better understand this!

[Merlin66]

Yes, there's no reciprocity failure like the ol' film days.

Each photon is faithfully collected... then you add them together for each exposure so all things being equal you end up with the same number ie Photons/ ADU whatever you want to call them.

[x6gas]

Thanks Merlin - much appreciated. So not to labour the point, the information on the DeepSkyStacker website is a bit misleading. To quote (from DSS site How to create better images -> Why combine):

"Why combine?

The answer is simple: only to increase the Signal to Noise Ratio (SNR).

Is the resulting image more luminous? No.

Is the resulting image more colorful? No.

The goal of combing many images into one is only to increase the SNR. The resulting images are neither more luminous or more colorful but they contain much less noise which will let you stretch the histogram a lot more which will give you more freedom to bring back colors and details."

I guess what the DSS site is trying to say is that the result of the stacking is no more luminous than the equivalent single exposure (not that the result is not more luminous than the individual subs. The statement on the DSS website didn't help my confusion levels!

I should say that I like DSS from what I've seen so far and hats off to the author for writing it and making it freely available. Not bashing them at all!

[Merlin66]

Yes, the end result is the same....

BTW in spectroscopy I've done some samples of stacked exposures v's single exposures and the stacking doesn't improve on the the longer exposure s/n .

ie the s/n in a 10 min exposure is better than a stacked 10 x 1min exposure....

There are different methods of measuring s/n but I can assure you that stacking compared with a single longer exposure doen't improve the s/n.