Another dumb newbie question about stacking

[DarkKnight (Kev)]

I'm sure most of you remember your early days when you had a plethora of unanswered questions and lots of bright ideas on how to do it better.

My question today is what is the advantage in stacking multiple images as opposed to getting one really good image, copying it, and stacking the copies?

It seems to me that multiple images, in theory, should be capturing the same data, so what is to be lost by using a single image multiple times?

[Merlin66 (Ken)]

Stacking the same image doesn't add any new data.
The individual images have random (shot) noise both in the target and the background. By stacking these images the noise is reduced (relatively) and the target enhanced but the additional signal.
You can experiment yourself and see the difference.

[DarkKnight (Kev)]

Thanks Ken.

Of course, random signal noise should be minimised in the comparison of multiple shots whereas it will be a constant in the single shot.

[ZeroID (Brent)]

And that is only part of the equation. As the seeing alters various details sneak through of the target slowly accumulating more information. Checka lunar video out one day and watch it shimmy and shake as the air above you wobbles and blows. Then stack the lunar video and watch it as the details firm up as the software reduces the seeing errors.
Quite interesting to do so.

[jsmoraes (Jorge)]

Stacking isn't only adding data. Stacking add data that aren't in the previous stacked image. If there aren't any new data, stacking will do nothing.
Stacking is more complex. It adds and subtracts also: noise for example.
Your idea make sense if you think about difference of 10 x 1 minute, total of 10 minutes stacked, and only one with 10 minutes of exposition. But as stacking is more complex than only add photons information ... 10 x 1 minute will show better result than only one 10 minutes.

[glend (Glen)]

Originally Posted by ZeroID

And that is only part of the equation. As the seeing alters various details sneak through of the target slowly accumulating more information. Checka lunar video out one day and watch it shimmy and shake as the air above you wobbles and blows. Then stack the lunar video and watch it as the details firm up as the software reduces the seeing errors.
Quite interesting to do so.

Brent it seems to me that the video frame (or lucky image) processing techique is very much treating each frame as if it was dithered. Bad seeing is somewhat like dithering applied by the atmosphere. The stacking parametres result in the smoothing of the image as the frames lying outside of the tightening alignment are discarded.

[rcheshire (Rowland)]

Statistical variation in images/pixels contributes to increased SNR with multiple shots. A single shot stacked multiple times is a summing of identical pixel values, whether mean median or otherwise. This is applicable to calibration frames as well.