You could also try to figure out what is causing it. I don't know APP, PI, or if there's any guarantee that they would force-stack frames that DSS refuses to. Or if that would even be a good thing.
That said, there's probably no harm in trying a trial version, depending on how hamstrung it might be. You might like the overall package better too.
But back to DSS. I tend to aim for 90-110 stars. Some targets that is not possible, of course. Can you tell why frames are being rejected by looking at the file list?
Try changing the option to stack immediately after registration. Just register, and then look at the info it gives you. Run compute offsets if those columns are not already filled in. Also review what stacking mode you are using (standard, intersection, mosaic, etc), and the alignment mode.
DSS should score and rank the subs by stars found, sky brightness, FWHM, and whatever else. I think it should stack if at least 8 common stars are found. Note that when you run compute offsets, subs that are not computed are ones DSS does not like. Now, is it because there aren't enough round stars, or common stars? What do the offsets look like on the rejected subs? Is it possible you had a lot of field drift? If that is the case, you might be able to try selecting a different reference frame - one that is more in the middle of any drift over time.
Apart from buying new software or trying a trial version, Sequator is a fairly simplified free stacker. ASTAP is a more complicated free stacker. You might see if they are more forgiving of your subs. But again, there's bound to be a reason that DSS is doing what it is doing.
I honestly cant tell what is causing it since the differences are so extreme..
While only using the best subs that I hand picked....
I normally aim for about 50 because I found a question about the number of stars needed that was answered by the dev. of DSS. Apparently it maxes out at about 50 stars for alignment purposes.
I went back over the registration to use about 90-110 stars like you suggested and tried to stack after registration, also tried computing offsets several times but they wouldn't compute, the boost in the number of stars let me stack 3 frames instead of 2 on the Monkey Head Nebula so its possible it helped it find a few more stars on some of the worse subs.
I like to use intersection for for stacking and automatic for alignment. I like to see how well the data is pulling through pretty early in the acquisition so I might only be stacking <10 frames or 50+.
There was essentially no drift at all except for a very gradual drift from PE and some shifts from the wind. Before beginning my imaging session I polar aligned to <1" on both the altitude and azimuth.
The last quote on this reply worked to force them all to align and stack, pretty shocked but also happy it worked.
Not to probe unintentionally too deeply - but 3 minute unguided exposures at 1200mm FL? Isn't that asking rather a lot from the mount?
If you "pixel peep" in DSS, are your stars genuinely small and round in the majority of your frames?
Oh definitely... I'm pushing it hard. I spent a lot of time working on the mount to get it to do as well as it does unguided.. The biggest problem right now is that I basically have a sail in the wind with my telescope. I am fairly certain that I am dealing with a good bit of PE in some frames but for the stars to be as tight as they are without guiding I am pretty happy.
I wouldn't be surprised if I backed off the exposure and boosted the ISO soon to get the perfect round stars I really should be aiming for. Whenever I get set up for guiding I'll obviously back it off to 200ISO again and expose until I hit the right spot in the histogram.
I will post a couple pictures in the reply after this one to show the extreme of what DSS liked, what is pretty good, and what the stack turned out like after getting all the subs to stack.
What I've found is that DSS is telling you something about your image that needs to be fixed, and changing the stacking tool is not going to fix it. The trick is figuring out what the issue is.
Can you post a couple of examples of the images it likes, and doesn't seem to like?
A couple of thoughts... When you lower the % of subs to stack, it's likely changing which sub it's using as the "reference frame". As a diagnostic, not a fix, you might sift through your subs and try to pick a better frame as a reference. Another diagnostic might be to try stacking with Sequator (yes, a different tool, but it's free); it's a lot less fussy about star shapes, which I think is the root issue.
Bottom line, you do need to figure out how to capture subs without any trailing of the stars. That's probably the root problem. I made a breakthrough some years ago when I got DSS to stack more than *one* sub. The issue was star shape, and the answer was a combination of a bunch of things, including better focus, better polar alignment, and ultimately, adding an autoguider. Each improvement increased the number of subs DSS would accept. That's a crude way to evaluate star shape, but ultimately it's the one that matters.
I am still not positive what the root issue was but after testing every sub as the reference frame I was able to get them all to stack with 1 of them, using any of the others would exclude some of the frames.
I think you are right that it is probably the star shapes being the root issue, but it doesn't make sense why it rates the extremely bad frames so high.
Personally I cant wait for the autoguider, I just cant completely justify it yet. I wanted to focus on working through all the issues I would encounter one by one first before adding in small set of equipment and software. By doing that I am very happy with how familiar I have become with my setup so far and its helping me look forward to guiding that much more.
It only stacks what it can align. It's not stacking based on quality if you have pictures stacked to check.
Pick a different photo that's in the "mean" positon of all the best photos as your reference frame.
Thank you for this, I was wondering why the offsets were so drastically different or nonexistent. This really helped to narrow down why I couldn't force DSS to stack them all. Like I said in the reply above this I tested each frame as a reference frame and calculated the offsets for each one and finally found one frame where DSS calculated the offsets for all the frames.
This is also great to know and keep in mind for future reference and it makes a lot of sense. It will really help me to select reference frames in the future better. I always chose the frame with the best star shape and size previously.