Subframe Selector / Values

[Petersnap]

Good Day All

I Know nothing about Pixinsight ,downloaded trail version and trying to figure out what the values mean when using the Subframe selector. (I assume this is used to distinguish good from bad subs)
I use a Dslr Canon 1100d ,with regard to S/R are these subs any good / and what base values should one aim to achieve?
I used the settings from a video I saw on the net.
Attached find a screen capture of the readings.

Thank you
Andre

[pfile]

i would start with the following:

 StarSupport: how many stars were detected. if there were clouds in some of your subs you'll see some subs with fewer stars
 FWHM: a measure of how fat the stars are in your images. reject outliers...
 Eccentricity: a measure of how oval your stars are. reject outliers.

SFS has all kinds of other statistics. i think it has documentation - if you click the little dog-eared page at the bottom of the script, you'll see it.

rob

 

[mschuster]

SNRWeight is usually a good measure of relative frame quality. The numbers are comparable for frames of the same target and the same filter. Frames with higher numbers are better, eg., less light pollution, more transparent skies, more exposure time, etc.

SNRWeight is not foolproof. You still should blink your frames. Strong gradients, bright star halos from high thin clouds, etc, can result in bogus SNRWeight values.

pfile, FYI, I don't use StarSupport to evaluate my under sampled frames. Stars are tiny, the dimmest are rejected as hot pixels. Better focus, better seeing, and better tracking result in smaller stars, higher hot pixel rejection, and smaller StarSupport numbers. So on my frames a smaller StarSupport often means a better frame. This assumes of course that light pollution, transparency, exposure time, etc. are not factors. I do use FWHM and Eccentricity.

Mike

[pfile]

i'm not talking about subtle differences in StarSupport… i'm talking like 1/2 the number of stars.

lots of times these bad frames will have other properties that seem great in SFS and then when you look at them it's clear the sub is clouded out; a (low) StarSupport outlier (at least in my system) is a pretty obvious indication that the sub is bad.

maybe those subs would show low SNRWeight… not sure.

rob

[mschuster]

OK, SS should be good for that. On my frames occasionally SNRWeight does not diagnose cloud and gradient problems. On my setup it does seem to work well for transparency problems (eg smoke, haze and extinction distance away from meridian) and light pollution (airglow, dawn).

Mike