Log in

View Full Version here: : MaxIm DL5 Stacking question.


bloodhound31
05-11-2013, 12:45 PM
How does it change the result of stacking 40 light frames and 20 dark frames (all of the same exposure) using Sum, Average or Median?

Baz.

RickS
05-11-2013, 02:32 PM
Average and sum will give you effectively the same result except that a sum could overflow if Maxim is using integer arithmetic (it probably doesn't but you might as well use average to be safe).

Median will give you some rejection of satellite trails, cosmic rays (and hot/cold pixels if you dither) that you don't get with average but it also gives you a lower signal to noise ratio (about 20% less).

An average combine with a suitable rejection algorithm should usually give you the best of both worlds.

bloodhound31
05-11-2013, 02:42 PM
Thanks Rick. Can you please help me understand what you mean by a suitable rejection algorithm?

Baz.

RickS
05-11-2013, 03:09 PM
Maxim has a Sigma Clip algorithm that does an average combine with rejection. That's probably worth trying. Also SD Mask is supposed to work well with smaller numbers of subs. You'll want to tweak the parameters to do as little rejection as possible to maximize your SNR.

The ideal solution would probably be to use something more capable for processing ;) Maxim is good for image acquisition but I don't think it's that strong for processing. I gave up on it very quickly when I discovered CCDStack and PixInsight.

Cheers,
Rick.

2stroke
09-11-2013, 04:37 PM
Check this thread out it will help you alot with maxim http://www.iceinspace.com.au/forum/showthread.php?t=111969 Ricks comment is fair though maxim is pretty great to, CCDStack has alot more option though its not for the beginner and lol if it wasn't for the video tut Making every pixel count i wouldn't touch it. I use mean for lights and clip the darks for a median. Really CCDstack rocks best with data rejection and i think maxim has better alignment options from my limit experience.

Merlin66
09-11-2013, 05:39 PM
I've been using AstroArt for the last ten years or so...
It does all we're talking about (and more)......
(For spectroscopy the "sum" addition of lights I feel gives me the SNR I'm looking for)