I've done a search, but there doesn't seem to be anything dealing with the BatchPreprocess script that was recently released. I know there's been a lot of theory discussion about rejecting images, or the hangs/long processing times PI has with "bad images", but it begs the question since DeepSkyStacker seems to handle this very quickly and efficiently.. What can be done to have Star Alignement quickly decide to toss images if they are out of focus or have too many elongated stars? As noted, DSS does this very quickly and very well. I can start with 220 frames and within a few minutes it starts stacking the 125 it found that passed its test.
The reason I ask is, when I used Star Align directly I was able to set a slider for tolerance that reduced "discovered stars" from 40,000 to 1,000 or so (still too many, but it worked without running past 20 failed attempts on various frames.. but it required me to manually delete over 100 bad frames that DSS just ignored when it found them unacceptable). When I use the BatchPreprocess script, I don't see where the detection threshold would be set for my lights, and I want to use it because it takes in flats and bias and darks. I'm using the same source light frames as I managed to stack with just Star Align, but with the script it's going crazy (yes, I have lousy frames, but other programs find and stack them regardless.. I just want the better/cleaner process with flats/darks/bias that PI seemingly has). At least 1.7.6 fixes the huge memory leak that 1.7.4 seemed to have after 15 or 16 failed attempts on a frame (at one point I had 5 gig of ram used while trying to figure out the next run, now it seems to properly reset for each failed attempt and it stays near/under 1gig a shot).
I hope this wasn't clear as mud. Just looking for how to get BatchPreprocessing to adjust star detection thresholds as well as find out why PI can't parse images as well as other stacking software. I'd hate to have to keep using DSS to stack since PI would give better results (if I could get it to work well).
TIA..
