Understanding Sigma Clipping Parameters in Live Stack

Anything that doesn't fit into any of the other forums
Post Reply
mAnKiNd
Posts: 308
Joined: Sun Jun 11, 2017 4:55 pm
Location: Long Island, NY
Contact:

Understanding Sigma Clipping Parameters in Live Stack

#1

Post by mAnKiNd »

Hi Robin,

Would you be so kind as to share some guidance as to how best one can set up their sigma clipping settings in live stack mode such as:

- Initial Frame Count
- Sigma Threshold
- Sigma Low limit

I've noticed that setting the minimum values on the above shows a status that clips/rejects up to 40% of pixels, but I'm not sure what one should be aiming for here?

I know if you hover the mouse over each of the above parameters you get a text box with a brief explanation, but I still feel I require a bit more clarification to get the best out of it!

Many thanks
Minos
User avatar
admin
Site Admin
Posts: 13296
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Understanding Sigma Clipping Parameters in Live Stack

#2

Post by admin »

Hi Minos,

no problem...

In non-live stacking using sigma-clipping you would run through all the frames calculating the mean and sigma (standard deviation) of the value of each pixel, then go back through and stack but only keeping data from pixels where the value is within a certain number of standard deviations of the mean.

The 'Sigma Threshold' parameter is the simple one - it represents the number of standard deviations away from the mean that a pixel value can be to still be included. Set this too low and you will reject a lot of data, set it too high and you will not be blocking anomalous values. Somewhere in the 3 to 4 range is probably a good starting point.

The other two parameters come into play because Live Stacking doesn't have all the frames in advance to work out the mean and sigma values for each pixel.

Initial Frame Count is the number of frames at the start of the stack where no sigma clipping will be applied - all pixels in these frames will be used. This allows time for enough data to build up for SharpCap to at least estimate the values of the mean and sigma for each pixel. Once the inital frame count has been passed sigma clipping will be applied and also the mean and sigma values for each pixel will continue to be updated with new pixel data.

Finally, the Sigma Low Limit is an artificial limit on how small a value of sigma the stacking algorithm uses. This is measured in % of the full pixel range, so in 8 bit mode, setting this to 1% means that SharpCap will use a mininum sigma of 2.55 (255*1%) where the calculated sigma of a pixel falls below that level. This is particularly important in 8 bit mode where the value for a pixel may vary very little - over the first 10 frames it might come in as 20,20,20,21,20,20,20,20,20,20. These sort of values lead to such a low calculated sigma that any future values of 21 (or 19) would be rejected by the sigma clipping, which is obviously not what we are aiming to do. Setting a lower threshold on the sigma value to be used fixes that problem.

Hope this helps!

Robin
mAnKiNd
Posts: 308
Joined: Sun Jun 11, 2017 4:55 pm
Location: Long Island, NY
Contact:

Re: Understanding Sigma Clipping Parameters in Live Stack

#3

Post by mAnKiNd »

Thank you Robin, very helpful indeed!

I'll try your suggestions and get back to you here if there is anything further!
Cheers
Minos
RickBG
Posts: 56
Joined: Sat Feb 15, 2020 3:52 am

Re: Understanding Sigma Clipping Parameters in Live Stack

#4

Post by RickBG »

Hi Robin,

You addressed an 8 bit scenario, but what about a 12 bit? What setting would you recommend for sigma low limit? As you move the sigma low limit slider to the left the numbers begin to increase, therefore am I to assume these numbers represent the % of rejection and if so what would be a decent number to start with?
You indicated in your last paragraph that these values lead to such low calculated sigma, therefore it was advised to set the threshold to a lower value. I would think by doing this would further increase rejections because you are narrowing the threshold. Based on what you said in the second paragraph, "set this too low and you will reject a lot of data", makes me believe in order to resolve the issue stated below you need to increasing the threshold not decrease it. What are your thoughts?
User avatar
admin
Site Admin
Posts: 13296
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Understanding Sigma Clipping Parameters in Live Stack

#5

Post by admin »

Hi,

I think the wording wasn't medically clear in the final paragraph – what I meant by setting a lower threshold was that the application stops the calculated sigma value from going too low by setting it to the specified sigma low limit value if the calculated value drops below that point. Ideally you want the fraction of pixels rejected each frame to be very small until something like a satellite comes past. Certainly less than 1% would be ideal

Cheers, Robin
DiNakamoto
Posts: 9
Joined: Wed Nov 06, 2019 6:51 pm

Re: Understanding Sigma Clipping Parameters in Live Stack

#6

Post by DiNakamoto »

Hi Robin,

I wonder what the difference is between the "default" and "Sigma" stacking when used during Livestack, as far as the final image is concerned ?
Thanks !

--- Dave
Post Reply