Creating a Mosaic with HA and Calibration Frames - Forums

Topic starter

Sep 25, 2024

I want to create a 6 panel mosaic with the ASIAIR for the Triangulum Galaxy with a One Shot Colot camera (ASI533MC pro). I want to also add HA data to the mosaic using the Antlia ALP-T Dual Band filter.
whats the best way to set this up? Do I shoot all the panels to get RGB data then shoot all the panels with the ALP-T filter for the HA? Or do I shoot the first panel with RGB and then shoot the HA for the first panel right after then continue with the 2nd, 3rd panel etc.?
Also when do I shoot the calibration frames for each panel?

Engaging

Tony Gondola

Frequent contributor

Sep 25, 2024

I'm not sure about what you are looking at for integration time but for a six panel mosaic, I would think you are looking at several nights worth of imaging to get all the data you'll need. I would get the RGB panels and Ha panels in separate sessions as it's just more efficient that way. The calibration frames would be the sets you would normally produce for a night's work.

Well Written Helpful Respectful Concise

Uhhjoe82

Topic starter

Sep 25, 2024

Ok, yeah I’m definitely looking at several nights. Does each panel need its own calibration frames or can I use a master flat, dark, bias and flat-dark for all the panels at once?

Tony Gondola

Frequent contributor

Sep 25, 2024

All the RGB panels can have the same calibration files. The dual band stuff would need it's own calibration files but again, the same for all panels.

Oskari Nikkinen

Sep 25, 2024

Ok, yeah I’m definitely looking at several nights. Does each panel need its own calibration frames or can I use a master flat, dark, bias and flat-dark for all the panels at once?

On the topic of several nights needed for 6 panels, that's not always true. If you bin your data after stacking you increase its signal to noise ratio by the bin factor. As an example if you binned x6 you'd get 6x SNR. You should decide on a working resolution early on and plan the mosaic with that in mind, and you'll save some time and dont need 30 hours worth.

As an example here is a 6-panel image with only 15 minutes per panel, shot under a full Moon with an Antlia Triband filter (30nm+ passes, not very good for anti-Moonlight):

Here is a 4-panel mosaic with 9 minutes on 2 panels and 10 minutes on the other 2:

As a cherry on top here is the Andromeda galaxy with just one 60s sub per panel (4 panels):

I think we can agree these 3 images look better than we would expect them to at such short integration times. So in short, if you want to make the project go along faster just bin and enjoy the signal to noise ratio boost. You could easily finish the whole project in a couple of nights, even one night if you are willing to accept the loss in resolution.

Helpful Respectful Engaging Supportive

Tony Gondola

Frequent contributor

Sep 26, 2024

That's great advice which I hadn't thought of. For the most part these images are presented on screen where unless you zoom in, ultra resolution is wasted.

Well Written

Alex Nicholas

Sep 26, 2024 · Edited: Sep 26, 2024

My Process with it is usually:

* Darks can be shot whenever, as long as you have the same Gain, Offset, Temperature Setpoint and Duration as your light frames, it doesn't matter when you take them... My current darkframe library is nearly 6 months old, and still calibrates the data brilliantly.

1: Shoot all the RGB Lights for all panels.
Note: I will be sure to not rotate the camera or remove filters at all during this time.

2: Shoot the RGB Flats/Flat darks.

3 Shoot all the Ha/OIII Lights for all panels
Note: Again, I will be sure to not rotate the camera or remove filters at all during this time.

4: Shoot the Ha/OIII Flats/Flat darks.

5. Calibrate/Integrate all the individual panel data sets for both RGB and Ha/OIII.

6. Stitch your mosaics for both Ha/OIII and RGB separately.

7. Run your favorite gradient correction procecss on the two stitched mosaics.

8. Use dynamic alignment to align the two merged mosaics together, then crop as required.

Then, go about processing the two images however you normally would.

I typically extract the Ha from the Ha/OIII data, then process then run deconvolution and noise reduction (I use bxt and nxt, but you can do this however you choose), then extract the stars from the Ha data and discard them, then run deconvolution and noise reduction on the RGB image, extract the stars and process them separately. I will then split the channels for the RGB image, add the Ha to the R in pixel math, then recombine the Ha+R, G and B channels, add the stars back in, then image solver and SPCC, and then process the image as per usual, as if it was a straight RGB image...

Thats just me.

2 panels, 21hrs per panel, Ha/OIII and SII...

Helpful

Uhhjoe82

Topic starter

Sep 26, 2024

Oskari Nikkinen:
Ok, yeah I’m definitely looking at several nights. Does each panel need its own calibration frames or can I use a master flat, dark, bias and flat-dark for all the panels at once?

On the topic of several nights needed for 6 panels, that's not always true. If you bin your data after stacking you increase its signal to noise ratio by the bin factor. As an example if you binned x6 you'd get 6x SNR. You should decide on a working resolution early on and plan the mosaic with that in mind, and you'll save some time and dont need 30 hours worth.

As an example here is a 6-panel image with only 15 minutes per panel, shot under a full Moon with an Antlia Triband filter (30nm+ passes, not very good for anti-Moonlight):

Here is a 4-panel mosaic with 9 minutes on 2 panels and 10 minutes on the other 2:

As a cherry on top here is the Andromeda galaxy with just one 60s sub per panel (4 panels):

I think we can agree these 3 images look better than we would expect them to at such short integration times. So in short, if you want to make the project go along faster just bin and enjoy the signal to noise ratio boost. You could easily finish the whole project in a couple of nights, even one night if you are willing to accept the loss in resolution.

Those are incredible images. Thanks for sharing. I would like to get the highest possible resolution with the most details as possible. So, I don’t think I will bin, but I’m sure I can shoot 6 panels of RGB in 2 nights and 6 panels of HA+OIII in 2 nights. I’m using a Celestron edge hd 9.25, so it is a larger light bucket and I should get plenty of data to work with from a total of 4 nights. Hopefully

Uhhjoe82

Topic starter

Sep 27, 2024

Alex Nicholas:
My Process with it is usually:

* Darks can be shot whenever, as long as you have the same Gain, Offset, Temperature Setpoint and Duration as your light frames, it doesn't matter when you take them... My current darkframe library is nearly 6 months old, and still calibrates the data brilliantly.

1: Shoot all the RGB Lights for all panels.
Note: I will be sure to not rotate the camera or remove filters at all during this time.

2: Shoot the RGB Flats/Flat darks.

3 Shoot all the Ha/OIII Lights for all panels
Note: Again, I will be sure to not rotate the camera or remove filters at all during this time.

4: Shoot the Ha/OIII Flats/Flat darks.

5. Calibrate/Integrate all the individual panel data sets for both RGB and Ha/OIII.

6. Stitch your mosaics for both Ha/OIII and RGB separately.

7. Run your favorite gradient correction procecss on the two stitched mosaics.

8. Use dynamic alignment to align the two merged mosaics together, then crop as required.

Then, go about processing the two images however you normally would.

I typically extract the Ha from the Ha/OIII data, then process then run deconvolution and noise reduction (I use bxt and nxt, but you can do this however you choose), then extract the stars from the Ha data and discard them, then run deconvolution and noise reduction on the RGB image, extract the stars and process them separately. I will then split the channels for the RGB image, add the Ha to the R in pixel math, then recombine the Ha+R, G and B channels, add the stars back in, then image solver and SPCC, and then process the image as per usual, as if it was a straight RGB image...

Thats just me.

2 panels, 21hrs per panel, Ha/OIII and SII...

This is great. I really appreciate you posting the order of you setup and process. This really helps. I’ll make sure to try and follow this when I give this a shot.

Respectful Supportive