Synthetic Green - What's all that then?

[Sp@ce_d]

I've never really put much thought into this before. If it's possible to create a synthetic green, why do we bother taking green subs at all then? Wouldn't it save time & money not to bother with a G filter?

Having read the article by Prof Greg Parker in this months AN, I had a look around the data on the DSS survey (& others). They don't seem to use G so why should we? Is it because they are imaging for science and therefore not after the "Aesthetic" images most of "us" Imager's prefer? I had a quick play following his guide & must admit my colours haven't turned out to my eyes liking ....yet.

So, your thoughts please....

Do you use synthetic green, how do you do it, why aren't we all doing it!

I mean, when I'm processing, 9 times out of 10 I end up running SCNR to remove it anyway!

[Jessun]

Eagerly awaiting responses from the learned on this. Really good topic.

/Jesper

[martin_h]

I sometimes use a synthetic green when Narrowband imaging and want a more natural colour image, using the 3 traditional NB filters would give me a red, a deep red and a blue/green, so not a pretty sight. So using Synth green would give me Ha(red), synth green, OIII(blue).

For the wideband imaging I hardly ever use a synth green - its good but not that good. I collect my green subs last - just incase but always try to ure natural green.

[steppenwolf]

For LRGB I only use synth green if I am running out of imaging time and have little chance of collecting the data.  I tend to take my L and R channels first as I consider these to be the most important (I am into emission nebulae) and the blue, although fiddly, I consider to be important as it is furthest from the red but I aim to capture this channel when the object is at its highest altitude for reasons of dispersion. So, green is the last data set I capture and this can be an issue because of time.

However, for NB where anything goes as the colours are false anyway, I sometimes use a synth 'green'  (middle) channel as a variation.

My own take? Real L + R + G + B is the best route as the data is truly individual and each channel plays a genuine part in the whole.

[Sp@ce_d]

Oh it sounds like you use it more often than I expected. Well, it certainly is "interesting" trying to use it and it seems to behave different to "real" G. I followed Greg Parkers article & my results are rather browny red to bright orange! (Hmm Sara, swag72 would like it..  )

Also had a separate play trying Noels Action, blending 50% red to Blue and Pixel Math in PI, with similar results.

Any tips on technique or pointers to a good tutorial?

[johnrt]

Good tutorial here if your are using Pixinsight. http://lightvortexastronomy.blogspot.co.uk/2013/03/tutorial-nebulae-pixinsight-guide-for.html

I recently put an image of the Pacman Nebula together using the first method described with pleasing results.

[Sp@ce_d]

Good tutorial here if your are using Pixinsight. http://lightvortexastronomy.blogspot.co.uk/2013/03/tutorial-nebulae-pixinsight-guide-for.html

I recently put an image of the Pacman Nebula together using the first method described with pleasing results.

Ah thanks.. yes I've been more successful with the NB images, I guess as it's more your own interpretation, I like the HST palette.

The thing that got me thinking about it more was trying to produce a natural (L)RGB image using just R,B and generating a synth G. It seems a lot harder (well impossible in my case so far) to get a natural star colour for instance.

[Ewan]

Interesting Topic Spaced_d, i have seen quite a few posts using synthetic courz & didnt understand why when you can capture the real thing.

[ollypenrice]

Fine for NB but scandalous for true colour broadband. 

 Shocked, of SE France.

[Sp@ce_d]

Fine for NB but scandalous for true colour broadband. 

 Shocked, of SE France.

ha.. An extra helping of le vin rouge by any chance?  Moon must be up then Olly.. can't tell here for the clouds. .. you'll be pleased to know my dabbling with it hasn't given me any acceptable results anyway so I gave up.. (the synthetic green that is )

[ollypenrice]

ha.. An extra helping of le vin rouge by any chance?  Moon must be up then Olly.. can't tell here for the clouds. .. you'll be pleased to know my dabbling with it hasn't given me any acceptable results anyway so I gave up.. (the synthetic green that is )

Sober as a judge. (A French judge...  )

No, you have no new information if you don't use real green. If you do use it you do have new information.

However, is RGB the best option for astrophotography? Good question, since green is such a rum colour for us. Being rum we have to get it right, though, not guess it...

Olly

[AndyUK]

I'm certainly not "learned", but I concur with others in this thread - When NB imaging, I always take Ha first, then OIII, as that way I can get at least get an H(G)O "real" colour image... and then next time I'll take the SII (maybe).  However, an H(G)O doesn't seem to work that well on all DSO's - When planning, I'll download a hubble palette version from Google images, open it in PS, split out the H and O channels and then put it back together again to see roughly what it looks like!

For LRGB, after getting the L, my process is to first get a night of R and then the next two sessions will be half night of G followed by half night of B - My imaging is invariably done "on the rise" so this helps to ensure that the B frames suffer less atmospheric extinction...

Just FYI, although there are probably more, there are two Synthetic Green PS processes I'm aware of - The obvious is in Noel's actions (which is SO easy!), but Steve Cannistra has another (here) - it's not as easy as Noel's but the result with a NB H(G)O is markedly different.  Sara (Swag72) uses the Cannistra synth G to provide a very different false colour palette with some of her NB images when she combines it with a HOO.  I personally like the effect, but false colour imaging isn't everyone's cup of tea!