2022-05-11: APP 2.0.0-beta 1 has been released !!!
Download links per platform:
Multi-Session normalization (background values) problem
I'm having a problem with background values computed by APP for a multi session runs. When I run a set of images from the same night as a single session, the background values computed by APP show smooth trends. However, if I attempt a multi-session run with two flats (one for each session), the same set images now has wildly varying background values ... even for consecutive frames which have very little actual background brightness difference.
The erratic background values only seem to occur for a multi-session run with separate flats. Below are some are screenshots illustrating the problem. Happy to provide data files if useful. Thanks for talking a look!
Single Session w/ all images from same night ... seems OK:
Single Session w/ images from two nights (added the first four files) ... seems OK:
Multi-Session w/ images from two nights and different flats for each night (darks and bias are the same) ... note the large variation in the background values such as frame #13 vs. #12 and #14 as well as the abrupt change after frame #18:
This might have something to do with the fact APP isn't separating the results per session in the graph, Mabula wants to change that to make it more consistent.
Thanks for getting back to me. It's more than just a graphing issue ... it affects the final integration results. If you compare the data tables in screenshots, you can see how background values computed by APP in the multisession pic are very different than the single sessions above even for consecutive images in the same session / same flat. The incorrect normalization is also visible when comparing images that are calibrated (very little difference) vs. normalized (very different).
Currently, I'm unable to use the multisession combine feature of APP because of this ... I'm only able to combine multiple nights as a single session / same flat for all images. I had this happen once before and thought it was a fluke. Please let me know if there is something I should try or if I can upload my data for you to take a closer look.
Interesting, I'll forward this to Mabula to see what he thinks about it. Thanks for sharing the potential issue!