LRGB vs RGB Imaging

[Euan]

I've been having this discussion with myself ever since I got a set of Baader RGB filters (makes me sounds mad doesn't it)....

What gives the best results, LRGB or RGB imaging?

The main reason is that I lock off focus before the start of a session, and at the moment until I replace the helical focuser on the Borg I can't automate focusing. The only parfocal(?) light pollution filter from Baader is the Neodymium (as far as I can see) and it hasn't really got the pedigree I would like in a good LP filter, I would really prefer an IDAS equivalent.

The major downside I've found with RGB is time. To just get 2 x 15min subs of each colour on my last image took 6 subs = 1.5 hours. Two subs is no where near enough to tackle the noise, I really need between 6-8 per colour for a good start, pushing up the exposure time needed for an average image to 4.5 - 6 hours, which is an eternity under UK skies.

I'm starting to think I might be better trying the old faithful LRGB approach, although reading various forums on the subject many many people get great results from straight RGB imaging.

With RGB being such a pain because of the amount of subs needed, I may just need to go for a 1.25" IDAS and have separate L and RGB sessions.

Anyone had success with RGB only?

[ollypenrice]

The only imager I know of who regularly works in RGB is Peter Shah. As you know, his images are beautiful but how he does it is a mystery to me! I'm sure he'd advise, though. One factor may be that he uses a very fast astrograph.

I like LRGB, even if I don't often bin the colour when doing things for myself. We do use binned colour when guests would like to work faster and that's fair enough. What's the theoretical virtue of RGB only?

Olly

[johnh]

I have found RGB works really well for clusters even dim ones and globulars and for small bright planetary nebula which don't need dynamic range. Peter Shah images at F3.8 so his RGB will be very bright but L would be advised for the fainter fuzzies.

John.

[ollypenrice]

I have found RGB works really well for clusters even dim ones and globulars and for small bright planetary nebula which don't need dynamic range. Peter Shah images at F3.8 so his RGB will be very bright but L would be advised for the fainter fuzzies.

John.

Good point, John. I'd forgotten clusters and do indeed just use RGB when I image them. Open clusters, anyway. I must go back and see what my globular RGB looks like on its own.

Olly

[Tim]

15 mins seems a long exposure for a R G or B filter Euan? Can you get away with that????

[Euan]

15 mins seems a long exposure for a R G or B filter Euan? Can you get away with that????

Both these were done this way:

2 x 15 mins per colour: http://stargazerslounge.com/imaging-deep-sky/119019-borg-77ed-first-light-short-m42.html

4 x 15 mins per colour: http://stargazerslounge.com/imaging-deep-sky/116237-ic5146-cocoon-nebula-first-mono-rgb-attempt.html

The major downside is the noise obviously, but the more I think about it, the more I reckon my colour depth is loads better than it was when I had the QHY8

I've read quite a lot of on various forums where people are advocating pure RGB, I'm going to stick it out for the moment and see what I can get. If it all gets too much to try and pull off, I will grab an IDAS and switch to LRGB

The real test is going to be M31 which I can do a direct comparison with the QHY8 from last year, which was even done with a faster scope

[Beyond_Vision]

Wouldn't LRGB only be an advantage if you are binning the colour?

[Euan]

Wouldn't LRGB only be an advantage if you are binning the colour?

I should have said that in the first post, when I talk about LRGB I mean 1x1 L and 2x2 RGB

[Whippy]

FWIW, I've stopped binning colour as it makes the stars a pain to get right. What I have been doing is racking up the luminence as usual and grabbing about an hour each of Red, Green and Blue unbinned and that seems to work.

Tony..

[ollypenrice]

This is all very interesting since I have been complacently satisfied that LRGB with unbinned colour was the Rolls Royce way to go, though note my admiration for Peter's pictures stated up front. I don't claim to be right and have never tried to destruction test LRGB against RGB. However, based on experience it is not high on my list of things to try. Am I wrong?

Olly

[Euan]

What I have been doing is racking up the luminence as usual and grabbing about an hour each of Red, Green and Blue unbinned and that seems to work.

I hadn't even considered that actually, I could probably give that a go, although finding the skies to get to an hour per colour is a challenge in itself at the moment

[Whippy]

Tell me about it, the last time I did that it took me 5 sessions! From my experience, you can get away with much more agressive processing with the RGB as it's not giving you any detail (several rounds of noise reduction for example). I liken it to colouring in tracing paper.

Tony..

[ollypenrice]

I reckon you can be pretty brutal with colour layers. In lab colour mode you can put a heavy Gaussian blur onto the a and b (colour) channels without losing resolution held in the Lightness layer.

Generally speaking I find that an hour of colour will hold up to a really big dose of luminance without being overpowered. 5 hours, say. Certainly going for more and more Lum is my own method once the colour is in the can.

Olly

[RobH]

I shoot star clusters in RGB only, and all of my starfields are RGB only.

I use luinance for the faint fuzzies (as JohnH suggests) and just ditch the stars.

All my subs are unbinned, with the occasional exception of some faint NB, for example, the jets on M82, where I need to go deep with the 6 inch scope and don't want to spend a week on one filter!

Cheers

Rob

[Rob_Jn]

How about synthesising B from R&G? There is a tool for doing this in Noel's Actions but I've never used it.

Rob.

[ollypenrice]

Interesting, Rob. Why do you ditch your lum field stars? Bloat? Colour burn out? Must have a go with this.

Olly

[johnh]

Olly ditching RGB is what I am trying to do now with nebula especially NB/Ha-O3 which gives wrong star colour - a lot of guys on the IceInSpace forum do this. You put the image in PS Noise - 'Dust and Scratches' with radius 12pixel threshold 120 and reduce both a bit at a time in about 8 iterations and the stars will slowly be eaten up till they are gone without having any effect on the nebula, then layer in the RGB later. Bright stars will leave a residue so need cloning away carefully.

John.

[RobH]

Interesting, Rob. Why do you ditch your lum field stars? Bloat? Colour burn out? Must have a go with this.

Olly

I find that I get much better star colour using just RGB for the stars Olly.

Also, often, if you're using long subs for luminance to capture faint objects, the brighter stars may be clipped. Layering them over RGB tends to wash out the colour, and there is also star bloat to deal with.

Cheers

Rob