Mar 28 2026 APP 2.0.0-beta40 will be released in 7 days.
It did take a long time to have the work finished on this and it will have a major performance boost of 30-50% over 2.0.0-beta39 from calibration to integration. We extensively optimized many critical parts of APP. All has been tested to guarantee correct optimizations. Drizzle and image resampling is much faster for instance, those modules have been completely rewritten. Much less memory usage. LNC 2.0 will be released which works much better and faster than LNC in it's current state. And more, all will be added to the release notes in the coming weeks...
Update on the 2.0.0 release & the full manual
We are getting close to the 2.0.0 stable release and the full manual. The manual will soon become available on the website and also in PDF format. Both versions will be identical and once released, will start to follow the APP release cycle and thus will stay up-to-date to the latest APP version.
Once 2.0.0 is released, the price for APP will increase. Owner's license holders will not need to pay an upgrade fee to use 2.0.0, neither do Renter's license holders.
Hey Guys,
I recently finished collecting the data for my new mosaic and I just cant get it to work... It makes not sense to me at this point.
General informations:
It's a 9 panel mosaic of the orion region with 2 different telescope + camera setups in monochrome LRGBHa. The images have more than enough overlapp with 25% (bigger sensor) and ~12% (smaller sensor) so the registration is no problem at all.
I always get the same problem with panel 6,8 and 9 which seem to be stretched more than the others resulting in a very ugly brightness and contrast difference - see the images down below:
lum:
red:
green:
blue:
ha:
The settings I used were:
Advanced normalization with the standard multiply-scale and bwmv scale, changing the method only mildly improved the overall result but not by alot...
I also used LNC 1st degree at 1 iteration, since this is the best setting which worked for pretty much every mosaic I ever created.
Here are the things I tried: preprocessing the images like stretching them a little bit before stitching, removing gradients and then stitching it, integrating the two datasets (per camera) seperately instead of the "overlapping stack" which I created per panel and filter (I hope you know what I mean) I also tried to use different reference images which are then used for the normalization process (I assume) and also normalizing the stacks before starting the mosaic processing in APP with the standard mode - it's more or less the same for both datasets when used combined or seperately which results in this pattern you see above.
Also worth mentioning is that these images were all captured under pretty good sky conditions (bortle 2/3) and most of the times under a new moon, the exception beein a few frames from panel 9, so they are basically LP gradient free. All panels were stacked with the same bias, flat and dark set which worked perfectly fine over multiple months.
It also doesnt matter if I integrate one filter at a time or when I load in all the frames at once with a multifilter session.
Do you have ANY Idea what I could do to fix that? Would be very sad to be stopped by that after spending almost 200hrs on the data acquision 🙁
CS
Jakob
Hello Jakob,
It seems that there is something going wrong with panel 6.
What reference images have you tried ? The 6 ? The one in the lower right corner ?
Have you also tried normalization in standard mode for the mosaic ?
Have made any kind of gradient removal before making the mosaic ?
@dav78 Hey!
As mentioned in my post I tried everything you mentioned...
I tried to make the bad and the good ones as reference. Not all of them but panel 6,9 were used for the bad ones and panel 2,3,5 for the good panels - it didnt change the outcome!
With the standard normalizazion mode I get very bad/mixed gradients on the overlapping points which are not really improved with LNC :/
Considering the reference image chosen, it can sometimes create this kind of artefact, that's why I asked you if you tried the lower right corner image as reference. It doesn't have many DSOs, so it can lead to a best result for the mosaic.
Have you made any process operation after stacking the images ?
Just tested it with image 1 (bottom right) as the reference image. Doesnt help 🙂
To your question: I tried alot, cropping, removing gradients, nomalizing, stretching it's always the same...
Hi Jakob @jkbsahner,
Thank you very much for reporting this issue. Thank you very much @dav78 for your assistance here.
Thanks for all the details and testing, I think I should have a look at your data myself to understand the issue and what is causing it somehow to fix this going forward.
Can you please upload all the data from all frames and mosaic panels? So all the lights and also the calibration frames? That would help me greatly.
https://upload.astropixelprocessor.com/
username: uploadData
password: uploadTestData
Please make a folder like : Jakob-mosaic-normalization-issue
and upload the data there and let me know once uploaded.
I will test as soon as possible.
Mabula
Ok. Yes that's quite strange.
I use to crop and remove gradient before going to the mosaic steps, it gives me the best results and I'm currently working with big mosaic (30 images per mosaics 😉 ).
There is maybe something wrong with image 6. Have you tried to integrate it once more ?
@mabula-admin Hey Mabula,
Thanks for the response!
Are you sure that you want the full dataset? That would include 50gb+ of data!
For now I will just upload the final stacked data, which I actually just restacked to see if I made any mistake - sadly enough it didn't fixed the problem.
CS
Jakob
Hi Jakob @jkbsahner,
Thank you very much, okay, I will try to process with your mosaic panels then myself in the coming days 😉 I first need to release 2.0.0-beta33 after which I will look at your data, it is on my issue list.
Mabula
@mabula-admin Hey Mabula,
That sounds good, I can upload the full dateset If you have no success with my already stacked data!
CS
Jakob
Hey Guys!
good news: I got it to work, not 100% perfect but usable for further processing.
Just after stitching and RGB combination:
And then with MSGR processing in Pixinsight:
I got it to work after I applied another LP removal iteration on each frame per filter and then stitch each filter individually.
Well done 👍
Panel 6 had to have a sky background level very different from the others panel.
Gradient removal seems to have solved this. What process/software do you use for gradient removal, APP, Pix, Graxpert ?
I used this:
https://www.nightphotons.com/guides/multiscale-gradient-removal/
It's in my opinion one of the most powerful processes you can do in pixinsight. Otherwise this image would be almost unfixable/unusable because the gradients are within the bright and dusty parts of the nebulae - taking a sample from these regions can kill the intire image and the overall look.
Hi Jakob,
Excellent, I am glad that you were able to complete your mosaic successfully now. Is there anything that you changed in the APP processing before making the RGB composite? Did you do light pollution removal on the individual frames with APP's tool or not?
I will try to use your data to improve things in APP of course 😉
Mabula
Hi Mabula,
I changed two things, I completely reintegrated the data to avoid any mistakes there and then I did a LP removal on each stacked mono intergration that seemed to work this time. It's not perfect as you can see from the initial RGB combination but usable and completely fixable with the MSGR processing in Pixinsight!
I can upload the new stacked data later to your server its a little cleaner and also with MBB enabled to reduce the "stepping" on the edges of the stacks.
CS
Jakob
Hi Jakob @jkbsahner,
Okay thanks for the feedback. Using MBB on the stacks for the mosaic panels is actually somethign we highly advise to reduce those stacking artefacts at the borders 😉 I am glad to know that at least this way, you could complete your mosaic 😉
Mabula
