Taking multi panel images

[sinbad40]

Hi all,

Just a quick question on taking multi panels.  Is it best to take a panel per night, or take a number of subs for each panel per night?  So if you have 8 hours capture during a night, and you are looking at doing say 4 panels, would you do a single panel a night or would you split each night and do 2 hours per panel.  Just trying to do some planning for winter.  My scopes are not fast at 6.3 - 7.5.

 

Cheers all

 

G

 

[ONIKKINEN]

I have done only a handful of 4 (and some 2) panel mosaics, but always tried to capture data for all the panels on the same night. That makes the mosaicing part significantly less painful since all the panels were imaged under more or less the same conditions so you will have fewer issues with visible seams between the panels.

So i would take a 4 panel mosaic by repeating each panel at some short interval, like 30 minutes per panel and onto the next one. That guarantees that you get a similar amount of data for each panel.

With more panels i think there will inevitably be an issue of different imaging conditions per panels, so i might not try to get all on the same night. Planning an 8 panel M31 and a 12 panel veil which will have to be chopped up for sure (not sure if either will actually go through). PixInsight's photometric mosaic script is pretty damn good though so I'm sure it will be fine in the end.

[ollypenrice]

I live at a dark site with very consistent skies so the many large mosaics I've done have generally been done 'one panel in one go.'

One could make a case for mixing them up at a less favourable site so as to get an 'averaged' result per panel. However, this panel would have multiple gradients to be removed and maybe the gradient software will be more effective at dealing with one consistent gradient?

How's that for not answering your question???

lly

[sinbad40]

6 minutes ago, ollypenrice said:

I live at a dark site with very consistent skies so the many large mosaics I've done have generally been done 'one panel in one go.'

One could make a case for mixing them up at a less favourable site so as to get an 'averaged' result per panel. However, this panel would have multiple gradients to be removed and maybe the gradient software will be more effective at dealing with one consistent gradient?

How's that for not answering your question???

lly

Yes, I think the location would have to be taken into account etc, so for me, subs on each panel multiple times in a night may be best , as the weather isn't the best in NW UK.

[Elp]

My latest 9 panel trial I typically did 3 x 1 hour per session but mostly because whatever non darkness there was lasted only 2.5-3 hours. I did manage all 9 at 20mins each though that is no where near enough even at f2. So next time I'll likely do the same, minimum 1 hour per panel or do one panel per night. Matching levels per panel is the challenge.

[ollypenrice]

1 hour ago, Elp said:

My latest 9 panel trial I typically did 3 x 1 hour per session but mostly because whatever non darkness there was lasted only 2.5-3 hours. I did manage all 9 at 20mins each though that is no where near enough even at f2. So next time I'll likely do the same, minimum 1 hour per panel or do one panel per night. Matching levels per panel is the challenge.

This does depend on how big you want the final image to be. I've done mosaics where the project was to make a zoomable mosaic in which small objects could be viewed at full resolution. However, if you don't want it to be zoomable, but just great to view at a moderate size, you need only a fraction of the data. You can hardware or software bin, depending on you camera technology.

Olly

[ONIKKINEN]

Since the topic has taken a turn towards the question on how long is enough for a single panel i will shamelessly plug one of my own images here:

This is a 4 panel mosaic of which 2 are 9 minutes and 2 are 10 minutes. So hows that for speed? Captured at f/5, but binned to an absurd level and the result is that just 9 minutes is enough for what i think is an ok image.

[Elp]

It's a great image. IC434 I find is an exception of a target because you can resolve it quite well in a short amount of time. I will always remember my first 2 minute sub with my modified Canon 600D on it and seeing both the horse head and the flame distinctly on the camera preview screen once the image was taken. I'm still yet to finish my IC434 which I started capturing in 2021/2022. Then I started a planned mosaic with the Samyang 135 soon after. Hopefully this winter I'll get around to at least a 2 panel or a 4 panel mosaic.

 

[ollypenrice]

10 hours ago, ONIKKINEN said:

Since the topic has taken a turn towards the question on how long is enough for a single panel i will shamelessly plug one of my own images here:

This is a 4 panel mosaic of which 2 are 9 minutes and 2 are 10 minutes. So hows that for speed? Captured at f/5, but binned to an absurd level and the result is that just 9 minutes is enough for what i think is an ok image.

Interesting! If we use a larger aperture telescope requiring 4 panels next to a smaller aperture telescope requiring only one - and a) we bin the 4 panel image down to the size of the single panel b) shoot the same total exposure, which will be better?

It's surprising that this topic isn't discussed more widely.

Olly

[ONIKKINEN]

4 hours ago, ollypenrice said:

Interesting! If we use a larger aperture telescope requiring 4 panels next to a smaller aperture telescope requiring only one - and a) we bin the 4 panel image down to the size of the single panel b) shoot the same total exposure, which will be better?

It's surprising that this topic isn't discussed more widely.

Olly

Focal ratio does have an advantage here, because you take fewer panels to reach a certain FOV. Comparing my f/5 200mm aperture scope with a RASA for example will have a clear winner in terms of speed (the RASA). If both are f/5 then the theory would say that the larger aperture will win.

I think in practice there are some "losses" in efficiency with mosaicing, namely having to account for panel overlap which will eat some of the speed up. There are also some considerations with how to image the panels, in one go or several nights which will surely affect the total speed of the project when the single panel scope can just image whenever. So i think this is one of these "in theory faster" type situations, where its not so simple in the end. Would love to see a comparison if someone were able to image a target with 2 different scopes on the same night, something like a small f/5 refractor and a typical newtonian and see which gets the better image.

[Elp]

I've kind of done this when I did Thor's Helmet late last year both in o3, also did f2 with the C6 afterward. F5.9 with my Z61 Vs f6.3 (or whatever the f ratio is) with my reduced C6. The comparison however showed the small refractor giving better signal (brighter image which is kind of expected), there was more detail in the C6 image but it was so dim the data wasn't used, even after resizing the image for a binned result. It's not really a fair comparison as they should be the same f ratio on a target which gives a decent signal.

[The Lazy Astronomer]

On 23/07/2023 at 17:59, sinbad40 said:

Hi all,

Just a quick question on taking multi panels.  Is it best to take a panel per night, or take a number of subs for each panel per night?  So if you have 8 hours capture during a night, and you are looking at doing say 4 panels, would you do a single panel a night or would you split each night and do 2 hours per panel.  Just trying to do some planning for winter.  My scopes are not fast at 6.3 - 7.5.

 

Cheers all

 

G

 

I've done a couple of smaller (2 - 4 panel) mosaics both ways with no difference observable in the final product (to my eyes, at least), but my preference is to try and shoot a bit of each panel every night to average out the different atmospheric conditions and hopefully end up with a similar FWHM for each panel. 

As long as you know how to properly utilise the mosaic building feature in your processing software of choice, even quite extreme gradients can be overcome.