So i captured a LOT of data Sunday night in hopes of doing my first planetary animation. I'm trying to not waste time processing captures that aren't worth the effort. Some captures have a ref frame with a score in the 70's while most frames are in the 50's while other captures never peak past 50 or so.....is there a cutoff # that you guys have found to be an indicator of a degrading frame?.....meaning there's no data stacking value and it would only reduce overall resolution?
On a side note, i notice that if i run the video through PIPP first, the quality numbers in AS3 are MUCH higher than if AS3 alone....I'm not using PIPP for any image processing......the boxes are all unchecked.....i just use it for centering the image and trimming the frames......makes the whole process much faster than allowing AS3 to do it all
Thanks for reading!
Glad you were able to hopefully make the most of Sunday evening's unquestionably good (but not excellent!) seeing conditions. Conditions like that are far too rare in NJ.
I don't quite follow what you mean by "a reference frame with a score in the 70's"? Do you mean the quality estimation score (next to the "Q: 100.0%") displayed after you run AutoStakkert's Analyse step? I was under the impression that only applied to the very first ("highest quality") frame.
My experience is that there is no hard-and-fast cutoff point for stacking. It depends on the nature of the seeing at the time of capture (assuming best focus, etc). Sometimes you can stack 75% of your frames and get a result which can be deconvolved to an incredible-looking result. Other times you can't get there even with only 10% of your frames.
The minimum number of frames I would consider (to keep the SNR high enough) is 1200, but even that may vary from person to person. Maybe your exposure is a bit longer and your gain a little smaller, perhaps letting you get away with a stack of 800 for a similar SNR.
It sounds like you may be in the situation where it may benefit to try multiple stack sizes (at one time) and see which percentage (or number of frames) is the best compromise between sharpness and SNR. There will definitely be a cutoff where it's not worth including any more frames in the stack - it's where that cutoff lays in each particular situation which is trial and error for the majority of us.
Humor me for a bit, please - I'm not that familiar with PIPP (I use it to create animations and occasionally change file formats). What is it that PIPP can do in your context that AutoStakkert cannot? I ask this because I can set my capture cut-out size to 400x400, which causes a 400x400 image to be saved to my SER. Then when I run that SER through AutoStakkert, the amount of time spent in the actual stacking phase is about 96% of the total time. That is, the analysis/reference/alignment (which includes any trimming) is only 4% of the overall total time.
Even with a Dobsonian, you should be able to set an appropriately-sized cutout - that achieves the centering and trimming during the acquisition phase. What am I not understanding, here, please?
Regardless - eager to see the results!