Do I even need to shoot green?

[Horwig]

As it says, should I concentrate on getting maximum R and B integration, and then combine them for a lower noise synthetic G? After all, if I see any outstanding green in my image I kill it dead with SCNR

Light blue touch paper and retire to a safe distance...

[Pompey Monkey]

Good question!

I'm not sure if SCNR on the green channel completely wipes out my green data, but let's hear what the experts have to say....

[swag72]

I have used this approach on a number of occasion with LRGB data........ I use the Steve Cannistra method to create a synthetic green.

I have't done a comparison between an image using this image and getting a complete set of data, but anecdotally with the false green channel I found the colour balance and stars quite tricky to get right. If I had time then I'd get a complete set, but whether its better to cane the red and blue and use a synthetic layer or have the same amount of date but spread it across the 3 channels I'm not sure. Common sense tells me that surely it would be better to get all three channels and then pull it all back into line with some good luminance.

[ollypenrice]

Splutter, expand, resemble Colonel Blimp, turn red, make choking noises, become apoplectic, whack silver knob of cane on table and bellow

Whaaaat?????

Green data is real data. It needs to be measured, not invented! SCNR reduces green noise and subdues green but if you take too much out your colour goes to pot. Your blues become magenta. No, I'm having none of this. Good colour has a colour-three-dimensionality about it which is hard to explain, but even very good HaOIIIOIII images imitating RGB don't have it. They have a kind of missing colour dimension.

lly

[Horwig]

19 hours ago, Horwig said:

Light blue touch paper and retire to a safe distance...

Well worth asking silly questions sometimes, just to hear Olly's measured, reasonable reactions

There's nothing like experimenting with real data, so here's the Pacman in RGB:

And with synthetic Green:

I think that answers my question, but interestingly, star colours are not a million miles different.

Huw

[ollypenrice]

32 minutes ago, Horwig said:

Well worth asking silly questions sometimes, just to hear Olly's measured, reasonable reactions

There's nothing like experimenting with real data, so here's the Pacman in RGB:

And with synthetic Green:

I think that answers my question, but interestingly, star colours are not a million miles different.

Huw

Nobody ever comes to me for a measured response! The blue stars in the forgery  are OK but the red stars are green. In fact they're a bit green in the RGB but a small shift in CB would soon fix that.

A gentleman buckles down to work and shoots a green channel. (Naturally he can send his butler in search of a good single malt to keep him company in his endeavours.)

Blimp (Colonel Retrd.)