Leo Triplet in LRGB

[powerlord]

Shot last night on my SW 200PDS with asi1600. About an hour each channel - 3 mins for R,G and B, 2 mins for L.

Edited April 26, 2022 by powerlord
removed magenta

[powerlord]

Oops, sorry. Could u move to imaging please @admin ?

[ollypenrice]

Why the shorter subs for L? I always shoot for longer in L and then remove it where it has too much signal, notably in stellar cores.

Olly

[powerlord]

it was a bit too high up the histogram with 180. But it was start of night, ZWO filters have some light pollution rejection in the RGB I believe which might explain it ? So when I can get away with it, I shoot all the same, but don't think ever shoot L longer than RGB prob because of that ? maybe I should then ?

[ollypenrice]

On 28/04/2022 at 10:05, powerlord said:

it was a bit too high up the histogram with 180. But it was start of night, ZWO filters have some light pollution rejection in the RGB I believe which might explain it ? So when I can get away with it, I shoot all the same, but don't think ever shoot L longer than RGB prob because of that ? maybe I should then ?

If you want to chase the faint stuff you will need to do so, I think. The Hamburger Galaxy in the Triplet, for instance, has a long but very faint tidal tail. https://www.astrobin.com/335042/

The way to get this is to use longish subs and have plenty of them. If this data saturates the stellar cores, why worry? In a Layers-based program you just make an RGB-only version with a gentle stretch for the stars. Ignore the galaxies. Get the background up to the level and colour balance of the over-exposed LRGB image, paste the LRGB over the RGB (which has an identical background) and erase the LRGB stars. (You can do it the other way round if you like, putting the LRGB underneath and erasing the top layer RGB wherever there's anything faint and fuzzy and interesting in the LRGB layer beneath.)

There are lots of images in which one stretch doesn't do all that we want to do. Some programs use masks to make selective modifications. I think using layers and the eraser is much easier, which is why I work mostly in Photoshop.

Olly

Edit; BTW, I think the core and spiral details in your M66 are particularly impressive.

Edited April 30, 2022 by ollypenrice

[ONIKKINEN]

4 hours ago, ollypenrice said:

If you want to chase the faint stuff you will need to do so, I think. The Hamburger Galaxy in the Triplet, for instance, has a long but very faint tidal tail. https://www.astrobin.com/335042/

The tidal tail was starting to form in an image i took a while back with just 2h of 60s OSC subs containing only 40 electrons of median signal in the green channel:

Not anywhere close to being drawn out to a presentable image of course but i did notice it was there so i think you could get away with short subs as well as long ones. Just takes a long integration to draw it out from the sea of noise i think. Not sure how to estimate how long but judging from my image around 10-20h would probably form the tidal tail nicely with similar quality short subs.

[ollypenrice]

20 minutes ago, ONIKKINEN said:

The tidal tail was starting to form in an image i took a while back with just 2h of 60s OSC subs containing only 40 electrons of median signal in the green channel:

Not anywhere close to being drawn out to a presentable image of course but i did notice it was there so i think you could get away with short subs as well as long ones. Just takes a long integration to draw it out from the sea of noise i think. Not sure how to estimate how long but judging from my image around 10-20h would probably form the tidal tail nicely with similar quality short subs.

It does take time and I think CCD cameras benefit far more from long subs than do CMOS.

Olly

[powerlord]

Thanks, good advice, I'll try next time. I try to setup as early as I can, which prob means I set exposure shorter than I could get away with for rgbl. I'm in bortle 6.

Ill wack that L time up next time and see what I can get.

[symmetal]

On 28/04/2022 at 08:54, ollypenrice said:

Why the shorter subs for L? I always shoot for longer in L and then remove it where it has too much signal, notably in stellar cores.

Olly

Once the noise from the sky background has swamped the camera read noise, making the read noise component insignificant, there is no advantage in having a longer exposure. You just need to take more subs to get the required total integration time. 10 stacked subs of 2 mins will have the same SNR as 4 stacked subs of 5 mins, if the 2 min subs have the read noise effectively swamped. The 5 min subs though will have more blown cores. 

I generally take 1 min L subs and 3 min subs each of RGB to get similar sky background levels where the sky background noise has swamped the read noise. Total integration time to start with is usually 3 hrs L and 1 hr each of RGB. This usually avoids having to do multiple layering in PS to recover the cores. 🙂

This method is mainly used with CMOS cameras for LRGB as they have lower read noise than CCD. It's true for CCD cameras too but exposures may be too long to be practical  for filters other than L.

Alan

[ollypenrice]

1 hour ago, symmetal said:

Once the noise from the sky background has swamped the camera read noise, making the read noise component insignificant, there is no advantage in having a longer exposure. You just need to take more subs to get the required total integration time. 10 stacked subs of 2 mins will have the same SNR as 4 stacked subs of 5 mins, if the 2 min subs have the read noise effectively swamped. The 5 min subs though will have more blown cores. 

I generally take 1 min L subs and 3 min subs each of RGB to get similar sky background levels where the sky background noise has swamped the read noise. Total integration time to start with is usually 3 hrs L and 1 hr each of RGB. This usually avoids having to do multiple layering in PS to recover the cores. 🙂

This method is mainly used with CMOS cameras for LRGB as they have lower read noise than CCD. It's true for CCD cameras too but exposures may be too long to be practical  for filters other than L.

Alan

Talking only about CCD, here I've been told this dozens of times. My experience, however, does not support it. The very long sub exposures, I have found, do go deeper than the shorter ones with the same total exposure. Other experienced imagers I've talked to have found the same - notably Tom O'Donoghue and Sara Wager. The most conclusive experiment concerned M31. I was going after the extended outer parts shown on star charts but not usually seen in images. A deep set of 15 minute subs didn't nail it. When I switched to 30 minutes, there it was. I've read all the arguments against this but... I'm a pragmatist.

Olly

[pipnina]

My Leo triplet LRGB I made earlier with my new mono cooled cam had about 15 hours of L, in 1 min subs. I only barely got the tail of the hamburger and it was faint. Meanwhile with 4 min subs (once I got my guidecam) and using only RGB (about 120min total) I was starting to see little hints of IFN showing through around M81.

I think longer subs help, but for me the benefit of reducing the burden on my computer storage is as good a reason as any to go as long as possible! That 20h total LRGB of the Leo Triplet created a project folder of about 500GB in size once pixinsight was done... 25% of my SSD's total capacity.

[powerlord]

would still be great if someone could move to imaging ? @Stu ?