Explaining Sigma

[manni_m]

Hello everyone,

I need some help in understanding the Sigma Clipping term in Pixinsight, or just in general the term "Clipping" for Linear fit, Percentile etc... 
I read the tutorials for image integration but it didnt helped though for this question, so if anyone knows a thread about this topic and i missed it, please let me know.

So I try to formulate my questions as trivial as possible:

Is it correct to assume, the clipping sliders control the Signal To Noise ratio (like an average of highest and lowest noisy pixel within either the upper band (e.g. high Sigma) or lower band (low Sigma) of an image?

So, would it be correct to say:

IF I would prefer a more noisy image as result I have to lower the Clipping parameters for high and low bands?

And on the other hand, if I would prefer a less noisy picture and loose, for example, some faint nebulosity as well, I raise the Clipping parameter?

So in the very end: Is the low clipping somehow controlling the "pixel dirt" in the background image ?



Thanks for any comments   
Manni

[NGC7789]

My understanding is that "sigma" is a standard deviation of the data set. More sigmas means more variation is admitted before it is clipped (rejected). So to have less variation (less noise AND less signal) you lower the sigma. If you want more variation (more noise AND more signal) you raise the sigma. Where or not these changes result in more or less signal to noise ratio depends on how much variation is contributed by signal versus noise.

[pfile]

hi - try reading this, it's the best documentation for stacking that exists anywhere:

http://pixinsight.com/doc/tools/ImageIntegration/ImageIntegration.html