[Solved] strange curve after register 2 mosaic channels

Hi Stefan @stefanastro,

The curved result si dynamic distortion correction in the alignment process. It seems a bit unstable on that mosaic.

What i suggest is that you approach this differently and try to make the mosaic of Ha and OIII at the same time.

Load the Ha stacks as lights and apply the Ha filter tag. Also load the O3 stacks as lights an apply the OIII filter tag.

Then analyse stars with 1500-2000 stars and register them as a mosaic. It should prove more stable and will give you an Ha and OIII mosaic which will be perfectly aligned immediately.

Let me know if this helps.

Mabula

@mabula-admin 

Okay if i see I have not to load the mosaic OIII and Ha in the light. But the panels? So 2 stacked OIII and 2 stacked Ha panels? And do the other steps you told me.  I'll try it out.

It's time to sleep now. Goedenacht 😀 

Hi Stefan @stefanastro,

Indeed, load the stacked OIII panels and the stacked HA panels and register them as a mosaic 😉 That is by far the best way to approach this. Did you try, did it work?

@mabula-admin 

Hi Mabula, I haven't test it out yet. I'll try today. Also the reinstall of APP.

@mabula-admin , Hi i test it with my 6 panels of M42 from the winter 2022/23. Not each panel was great because of dispersion, thx to the clouds 😣 . But each panel was between 12 and 16 frames.

And is it me or is the overlaps not good? I did stop scale 10 and LNC 1e,1 and MBB 5%.

I looks like i see black lines between some panels, specially on the outside. 

There was some noise but i did the steps  you told me.

After post-processing in Photoshop with noise-xterminator and other filters it looks better. Ill retry it this year again with more frames and less noise 😀 

Hi @stefanastro,

Okay, it looks like the overlap between the mosaic panels is very small, less than 10%, right? That makes it very hard to get things right normally.

For the mosaic, it will probably be better to use LNC 2nd degree, did you try that?

Mabula

@mabula-admin the overlaps is 10% yes.

I'll try that Lnc on 2 tomorrow. Thanks

@mabula-admin i test it out and it was better but still not beautifull mosaic because of the dispersion. I remove them and my next step is to create a giant mosiac of 80 panels to get the hole orion constellation.

Each panel at least 36 frames of 5min with L-enhance and also 5 x 36 with the SII-HB.  All 36 frames have to be clear from dispersion.

For RGB stars i use L-pro for 2 min 90 frames.

That takes a lot of time but a nice challenge, any tips maybe for mosaic?

Is it better to create 4 parts of mosaic of the 80 panels mosaic? So it is 8x10, if i do the first 2x10 panels and stack them extract them and do this with the next 2x10 agian and the next etc etc.

I get 4 mosaic parts of 2x10 and load them in the light and do the process it will be faster i think than if i load all 80 panels together in 1 time? 

Hi Stefan @stefanastro,

That sounds like a really ambitious project 🙂 wish you the best of succes !

Is it better to create 4 parts of mosaic of the 80 panels mosaic? So it is 8x10, if i do the first 2x10 panels and stack them extract them and do this with the next 2x10 agian and the next etc etc.

I get 4 mosaic parts of 2x10 and load them in the light and do the process it will be faster i think than if i load all 80 panels together in 1 time?

No, in fact this is a big problem in terms of optical distortion correction. It can work by dividing the whole mosaic in smaller parts, but you run the risk that it will be difficult to get accurate registration in the final result.

Normally you really want to calculate the entire mosaic in 1 go, to get the most robust end result.

Mabula

@mabula-admin 

Okay thanks.

So I'll stack each panel and load all 80 integrated panels and do the mosaic process.

Well first I need clear sky in the Netherlands 😀 

Great, keep me up-to-date and let me know if you need assistance 😉 !

Mabula

@mabula-admin Hi,

First a happy new 2024, while it's 2 weeks ago. I start with the mosaic, but i have to do 90 panels, i calculate it takes 560 hours.... But i saw Arenda from the store ganymedes that she had a new scope the William optics pleiades, a F3.8 and focal lenght is 260mm. I checked how many panels i need and it takes 49 panels and faster than my WO GT F4.7. I bought that new scope. But last week we had good weather so i test the L-pro and the new L-quad to see the difference. 

But i also start qith the mosaic to capture 1 panel of 2h with L-quad 300sec. And M42 to test the filters. 

I process them first and than take all stuffs of from my GT81 and place it on the Pleiades. Finger cross for more clear skies for my mosaic.

And i got 1 question, can stack the intergrated image, like an Ha of the M42 (from the GT81 scope) in a new intergration process of the M42 that i take with my new scope? Both have a different FL. GT81 with reducer is 382mm and pleiades is 260mm.

Hi Stefan @stefanastro,

Wish you all the best for 2024 !!!

Wish you great luck with the new telescope ! Yes, you can combine the data from both telescopes together, for sure! Make sure in 4) register that you disable same camera and optics 😉

Mabula

@mabula-admin Thanks. 

Anyway, do you want some light frames of my Antlia SII-HB Dualband filter for APP? To find out how to extract the 2 colors.

Or my Askar OIII-SII

Hi Stefan @stefanastro,

Yes please ! Thank you 🙂

Can you upload them here:

https://upload.astropixelprocessor.com/

username: uploadData

password: uploadTestData

Please make a folder with your name and description like : StefanRang-SII-HB  and StefanRang-OIII-SII

and upload the data into those folders and let me know once done, thanks!

Mabula

@mabula-admin 

I upload the Askar OIII-SII and the Antlia SII-Hb data with master frames in the map.

Good luck and keep going with updating your amazing program. I hope to see the option SII-HB extract SII in tab 0 ;). Even the Ha-OIII works.

Thanks Stefan @stefanastro !