Amount of Lum subs to RGB ?

[Danjc]

I have read various ways that people tend to go and appreciate the detail is in the lum. 
So if I image Galaxy x is the way forward to go for example 

lum 3 X 120s,

red 1 x 120s,

green 1 x 120s,

blue 1 x 120s

so 3111 for equal lum to RGB ?

Cheers, Dan.

 

[DaveS]

I'm not sure there's a "Right" answer, but if there is I'd like to know as well. My own thinking is that as Luminance carries most of the info then as much as possible, but I've been mainly NB up to now as it's only recently that I've had my imaging rig set up properly in my new dark-ish location.

My latest effort has 3 hours each RGB, and 5 hours L. Not too bad but could do with another 5 hours L.

[discardedastro]

It will depend on the target, I think.

Most video compression/encoding codecs use resolution subsampling at a ratio of 1:4 or 1:2 for colour. For instance, for a group of four "L" pixels, we just encode one pixel's worth of RGB data and mark that as applying to all four of those pixels. Most people can't tell the difference between this and RGB 1:1 data, even on high end monitors/TVs.

But this does assume, for video codecs, pretty even distribution of input light source - i.e. we generally have videos that have all sorts of colours in them.

For astro targets if your target is highly emissive in R, for instance, you'd be much better off capturing more R than say G or B. But this does require you to appropriately process and combine (because your higher-SNR R data will still suffer from being combined with low-SNR G data, for instance).

So the general advice I've seen is usually to keep your RGB exposure times matched for a quiet life, and use R+G+B=L as a good rule of thumb for exposure time. But there are undoubtedly better ways to do it...

[wimvb]

The exposure time isn't the only variable in this equation. Sky quality and light pollution matter also. If you have lp from street lights, RGB filters can help you decrease its impact, because most filters are designed to block lp wavelengths. Luminance filters let anything through.

If you can image lum at 120 seconds without overexposing the stars, you can safely increase the exposure time for rgb to get better snr.

L 4 x 120 s

RGB 1 x 240 s each or even 1 x 300 s each

Shoot RGB until you can extract a colourful image from the data. Shoot L on nights with clear and stable skies to get as much detail as possible.