C27 Crescent reprocessing as HaRGB

[Padraic M]

I managed to capture some nice Ha data of the Crescent Nebula over two clear, if moon-lit, nights in the last two weeks. City skies, which Ha tolerates well.

I also had some RGB data from last year. The starfield was a bit psychedelic. iirc, it was also from two full-moon nights, but in otherwise very dark skies.

I've processed both sets together, using the two integrated stacks from APP, separated the RGB channels, and added Ha to the red channel as a new HaRGB image.

Stretching and tweaking in Siril, with final adjustments in Gimp.

The Ha data is SW Esprit 80/ASI1600MM Pro, and the RGB is SW 150PDS/ASI533MC Pro, so very different fields of view and completely different rotations. After a bit of trial and error, I managed to get APP to register the four channels perfectly, cropped and combined. Siril photometric colour correction and some judicious stretching. It's still got a lot of stars, but my experience with star reduction/removal hasn't been great so far. Might leave that for another day.

Siril registration really is great. I think the key to getting it to work was increasing the number of stars for Star Analysis to 1000, and using one of the RGB channels as reference for registration.

Capture times were:
Ha 37x 600s (6h10m)
RGB 70x 300s (5h50m)
Total of 12 hours.

 

Full Ha frame below just for reference.

[Ouroboros]

@Padraic M Well done! Sounds like that was hard work.  I’m not quite sure what you’ve done exactly, because I only do OSC. Hoping to get into all that at some point.   The crescent nebula has been a particular focus of mine recently. It’s certainly a fascinating object.

Your stars have set off your Ha data very nicely.  Just a thought but it might also be good if you could capture some RGB star data to combine with your Ha data at the same scale.  You’ve picked up some nice Ha nebulosity in the larger Ha image.  

[Padraic M]

Thanks @Ouroboros. This was the first time doing this process and there is lots to learn. I'm surprised how much difference there is in the FOV given the pixel scale difference and the bigger sensor in the 1600MM. I'd like this to be a regular feature as I have the Esprit/1600MM setup in Dublin using narrowband to beat the light pollution, and the 150PDS/533MC is in Kerry under really dark skies, where OSC works well. I have both systems working reasonably well at this stage, so the thoughts of swapping the cameras from one to the other would fill me with dread! I'll need to do some more experimentation and see what works best.

[Ouroboros]

 

7 hours ago, Padraic M said:

Thanks @Ouroboros. This was the first time doing this process and there is lots to learn. I'm surprised how much difference there is in the FOV given the pixel scale difference and the bigger sensor in the 1600MM. I'd like this to be a regular feature as I have the Esprit/1600MM setup in Dublin using narrowband to beat the light pollution, and the 150PDS/533MC is in Kerry under really dark skies, where OSC works well. I have both systems working reasonably well at this stage, so the thoughts of swapping the cameras from one to the other would fill me with dread! I'll need to do some more experimentation and see what works best.

Interesting. Do you have to rescale/resize the images taken with different devices, or does the software do that automatically?  And what about if there’s a difference in field curvature between different image sets? Is that taken into account too? 

[Padraic M]

Yes, you can combine different image sets from different sessions on different rigs. Uncheck the 'same camera and optics' option and it will automatically rescale the images. There's also a 'distortion correction' option that addresses differences in curvature etc. between the two rigs. It's very smart.

Here's a test shot showing how the smaller FOV red, green and blue channels are registered in the centre (given that one of them was the reference frame), with the larger FOV Ha frame rotated and overlaid and added to the red channel. The size of the central RGB square is approx 3000x3000 pixels which is the sensor size of the 533MC so that suggests that all subs are resized to match the reference frame.

[Ouroboros]

Thanks, @Padraic M.  That’s nice and clear. If I ever get round to doing that kind of thing I’ll have to work out how to do it with Pixinsight.