Lights to stack % in integration

Great questions and I think some good points for us to improve upon, I've forwarded this to Mabula.

Hi @outer-space,

Thank you very much for your question. Your question addresses something that is already on our issue list indeed. We do want APP to continue when star analysis fails on some frames, some holds for registration and normalization. The logical step would be to deselect those frames automatically for further processing. Then the user can see in the end what happened and which failed. Maybe we can popup a warning after the integration, so the user is informed that some frames seemed to have issues. So we will change that soon.

So now to your questions:

A. I am unclear why Registration and Normalize go through the entire 240 subs although only 145 selected are relevant for integration.

Okay, this is the current situation. We had it in this manner, because I felt in the past you want to be informed as soon as possible if there is something wrong with the images. Sometimes it is caused by calibration even so it has no use continuing at that stage. But I realize that these days, more and more users want APP to perform all steps up until and including the integration results and see the problems then, because many load really large datasets (1000s of frames) and they set APP to run at night and want APP to be finished in the morning for instance. For both situations, I think something can be argued in favor.

There is another complication here. If frames are deselected after 3) only based on star analysis results in combination with the % slider in 6)Integrate, then you run the risk that images with bad sky background or noise levels do enter the integration, where ideally you want to exclude those rather than some of the frames with less ideal stars in my opinion and experience.

And finally, I I would let APP deselect them, the % slider will start working on only the selected frames. Check for yourself by selecting/deselecting frame and check where the integrate mark shows in the quality column. Leaving you less frames to be stacked in the end then the initial number of frames to be stacked before starting processing at all.

B. It is great that APP asks me to pay attention to failed subs, but I wish there were a 'auto-remove' option so its fully hands off. Would that be possible? The awesomeness about APP is the fact that I get great results with mostly default / automatic settings. Having this inspection of failed images as a manual step is a bit tedious (as it stops the program from proceeding to the next step). I am good if APP simply marks them for me, excludes it from the next processing step and I can inspect it after my final stacking is complete.

Yes I understand. At the moment, I feel like the best option would be to have a setting in APP's CFG menu, where you can control this behaviour. So either choose to auto-deselect frames that do not pass a processing stage successfully and warn the user after processing has stopped as planned (like after integration when the user clicked on integrate). Or do not auto-deselect and stop and warn after frames do not pass a processing stage (the current behaviour). I think that will suit everyone and that setting will of course be remembered per restart of APP.

C. In cases where the failed images are more than a handful (tends to happen due to unforseen cloudy conditions, something went wrong with focus etc. etc.). What does % lights to stack do? for e.g. if say, 60 of the 240 subs fail, then is the percentage calculated out of 180 or still 240? Ideally, I still get the 60% of the 180 that did not fail.

As explained above, that % slider works only on those frames that are actually selected. So yes, you would get 60% of the images that passed all processing stages successfully 😉

Hope this answers it?

Thanks for your great compliment 😊 

Mabula

Hello Mabula,

Yes, having the option to auto-deselect in the CFG  will be very useful. I'll look forward to that in the future releases. By the way I have  also used the workflow step-by-step manually (perhaps as originally intended?) and the way the UI flows is fully logical. 

Thanks for your thoughtful response here. I understand the design decisions. I must acknowledge that APP is quite user friendly and well designed. It is a very technical software and I am amazed at how robust it is in stacking. It always gives a superb result. And I need to do very little after that. I don't think I have ever been disappointed.  Even when I know I am giving it poor data, it simply works.

And it is so stable! high resolution images, very large number of frames, high(er) drizzle settings, comet registration, alignment, bad light pollution, vignetting,  all just works no matter what I throw at it. Even if I am missing calibration files, I am able to fix things and get a great result. I also find the recent releases have become quite fast

Oh and the post-stacking tools -- the light pollution removal is simply brilliant! Maybe magic is a better word to describe it. I can't believe the thing it does with the challenging data I give it. The resulting flat background -- I don't think its possible to get any other tool to do it that well.

Really great work!