Pixinsight - Using subframe selector properly - Forums

Topic starter

Nov 4, 2022

Hi!

I have a question about subframe selector.
When I run my raw fits in it, I get decent Eccentricity.
But when I run my calibrated image it is kind of marginal. I get good tracking/guiding, stars seems good to me, but the value I get are just bad.

Any opinion ? Should I run it twice? first on unprocessed file to remove the really bad frame, then calibrate my lights and go for another run of Subrame Selector to get my best frame and weight?

Thanks
Karl

webcubus

Nov 4, 2022

Here's my process with SS:
* Run WBPP and uncheck all the boxes under "Pipelines" just to calibrate and debayer
* Run SS against the debayered frames
* I then look mostly at Stars and Eccentricity to filter out cloudy subs or the occasional wonky tracking.
* Once I've got a list I want to take out, I move the corresponding light frames to a rejected folder.
* I re-open WBPP, check off the things I want to do in "Pipelines" and go. It automatically removes the light frames I moved.

If you have backfocus issues or sensor tilt, the stars around the edges can be distorted, which I suspect would mess with your eccentricity scores.

Can you post an example graph and maybe a sub or two where the eccentricity is reported as bad, but looks OK to your eyes?

Helpful Insightful Respectful Engaging Supportive

Paul Deeter

Nov 4, 2022

I've been using Blink to quickly weed out bad subs prior to subframe selector and WBPP. I use it on my lights and flats. Blink loads quickly and, in my opinion, is very user friendly.

Well Written Concise

Karl Theberge

Topic starter

Nov 4, 2022

Here's my process with SS:
* Run WBPP and uncheck all the boxes under "Pipelines" just to calibrate and debayer
* Run SS against the debayered frames
* I then look mostly at Stars and Eccentricity to filter out cloudy subs or the occasional wonky tracking.
* Once I've got a list I want to take out, I move the corresponding light frames to a rejected folder.
* I re-open WBPP, check off the things I want to do in "Pipelines" and go. It automatically removes the light frames I moved.

If you have backfocus issues or sensor tilt, the stars around the edges can be distorted, which I suspect would mess with your eccentricity scores.

Can you post an example graph and maybe a sub or two where the eccentricity is reported as bad, but looks OK to your eyes?

I will tomorrow when I process the 3 previous night

Karl

Reg Pratt

Nov 4, 2022 · Edited: Nov 4, 2022

Paul Deeter:
I've been using Blink to quickly weed out bad subs prior to subframe selector and WBPP. I use it on my lights and flats. Blink loads quickly and, in my opinion, is very user friendly.

User friendly yes but not time friendly when you're pulling in hundreds or thousands of subs per session lol.

Roger Gifkins

Nov 4, 2022

If you run subframe selector on the raw files it will include the hot pixels in the calculations and give an inrealisticly low value for FWHM and eccentricity. Once they are calibrated and cosmetically corected it will give the correct value. At least that is what happens for me. Cheers Roger

Helpful Concise

Karl Theberge

Topic starter

Nov 4, 2022

Roger Gifkins:
If you run subframe selector on the raw files it will include the hot pixels in the calculations and give an inrealisticly low value for FWHM and eccentricity. Once they are calibrated and cosmetically corected it will give the correct value. At least that is what happens for me. Cheers Roger

Make sense. But unless I am mistaken, SubSelector use a star map and based on that star map, only real stars were selected.

But still, make sense.

Karl Theberge

Topic starter

Nov 6, 2022

@webcubus Finally I decided to not upload phicture and graph because @Roger Gifkins was right on point.
I was evaluating too many hot pixel and that gave me biased result.

Even using the Maximum Hot pixel removal in the Subframe Selector, I was getting wierd data.

My data is not that bad tho, and when Im done Drizzeling it, I get round stars. Some of them are bloated but since I use RGB stars from shorter exposure on long exposure SHO, I get ok result.

Thanks for the help and answer

Karl

Related discussions

RFA PixInsight Stacking Problem, Calibration Problem, or... who knows what problem

Issue: I receive the failure below when running WBPP. Results in what seems to be vignetting. The below image has had Graxpert, ABE, SPCC and stretched. Saved as jpg. I don't detect vignetting in the subs. Maybe my calibration frames are the prob...

Discusses calibration issues affecting image quality and processing results.

Aug 2, 2024