[Sticky] Tricky Mosaic workflow, alignment problem

Ralph

My pixel alignment for the Alignment Mosaic came out at 0.11  for 257 star pairs so its consistent with yours. 

In step 8 you didn't mention an MBB setting. I set 10% as it seemed sensible to smooth the panel edges a little in the Alignment SuperLum Mosaic.  Also, your image of the Alignment Reference Mosaic looked like you had removed a lot of the LP but you didn't mention this. Was that just for appearances or would you LP correct the Alignment Reference Mosaic before using it in step 12?  I didn't and it didn't seem to impact the workflow. 

Having gone through your workflow above I too had a 6 outlier subs that produced eye catching RMS numbers.  Here are a couple of screen shots only because the worst (57.29) was badly distorted after registration and had three satellite trails through it as you can see.  @vincent-mod  Any thoughts or reasons that might explain this?

Regards

Chris

@vincent-mod

Good spot Vincent. The early subs were taken with APT and the later with NINA. Would not have done that if I'd have thought it would have made a difference. Not best practice though as you say so noted.

You say the Flats need work. I could try going higher than full well (20K) on the sensor or get a dimmer flat panel to make the exposure longer. Is that the kind of 'work' that you meant or am I off the point?

@chrispeace Well, the "work" is that it seems the light gradient produced by the flats, didn't quite match those of the subs. So when I checked the correction of a frame with the flats (by selecting L-calibrated on top of the viewer), I saw a nice sub without much gradient going to one with a gradient. Which should be the other way around.

@vincent-mod

Got it thanks. I'll start with my light source. It's a electro-luminescent panel rather than LED so I cannot dim it and it could have a gradient I cannot see visually.

Chris

Having gone through your workflow above I too had a 6 outlier subs that produced eye catching RMS numbers.  Here are a couple of screen shots only because the worst (57.29) was badly distorted after registration and had three satellite trails through it as you can see.  @vincent-mod  Any thoughts or reasons that might explain this?

If you select "L-c-registered" after the registering, you can go through your frames to see how they registered (zoom in 100%). This is also nice when you registered using the mosaic-mode, you can then check already if the panels are correctly placed and don't have weird distortion. That's way faster then waiting on the entire registration. I didn't leave out any frames actually, seemed to work fine for the luminance.

I've send Mabula 2 frames I couldn't register, maybe he sees way faster what the issue was then me. 🙂

Ah so, according to Mabula the issue I'm having with the RGB is due to the undersampling. The main problem is that the stars in the 2x2 are so small it's going to be very tricky for him to get that to work in the algorithm. The boundary between stars and hot-pixels becomes so vague then that the registration engine might then cause more issues than it solves. So I think the best would be to keep RGB also at 1x1 for your setup, that should work.

hello all,little question about  mosaics.

What is the max  APP can handle when combining panels ?

i havve a 7 panels of the soul nebula,all more then 1 gig,when combining to all APP crash.

ps its not my pc,16 gig ram   etc...

greetings.  for the rest APP performse very well for me,the pollution tool is very great.

groet John https://www.astrobin.com/users/kajouman/

That will depend on memory used by both your system and APP. So even with 16GB things can go wrong, depending on your system settings. In theory there is no hard limit, but the system will have one.

@kajouman

Hi John

Whilst I’m not an expert on APP capabilities I would suggest you try building the mosaic in stages rather than trying to put all 9 panels together in one go. This was a suggestion I got when I was having my difficulties. So try connecting just two panels at a time and do that 4 times to give you 4 new (double) panels and one single panel. Then combine two double panels together to give you 2 new panels plus one single frame. Continue To grow your mosaic until you get every panel combined. 
Also,  look to see how much memory you have allowed APP to use from your total 16gb. It’s configurable and you may not have allocated sufficient memory priority to APP to get the job done.

Make sure your laptop is plugged into the mains as many run more slowly on battery power.

Finally if you have a solid state drive I’d work from that as it will be a lot quicker.

Good luck and do post me an image when you get this all together. 

Chris

@chrispeace

thanks,did this also.  in pixinsight i can build  all into one. very well merged.