Adding HA to OSC_RGB images

[timh]

Dabbling with adding HA from a mono camera into the red channel of RGB images taken with an OSC seems to work quite well as an approach to getting colour pictures with much added interest and detail. I guess that it could all just be done on just the MONO camera with a filter wheel and added RGBL filters - but having the OSC seem to simplify things.

Images below were all RGB taken under Bortle 6 skies using an AS1294 PRO MC and H II image data taken using an AS1294 PRO MM all at - 10C.
The HA images were combined - as a difference signal minimizing stars - into the red channel using pixelmath in PI.

Optolong 7 nm HII filter
Skywatcher 200m Newtonian with 0.9X reducer/ flattener at F 4.5 , f 910 mm
Processing - pre and post - PixInsight

M82 image was about 130 x 40s RGB exposures (some rejected by subframe selector) and 33 x 2 min HA exposures at gain 124

IC1805 image was only about 60 x 40s RGB exposures taken using an Astronomik UHC filter combined with 30 x 2 min HA exposures (combined at 50% intensity since the RGB signal was relatively low) also at gain 124.

Thought that it is interesting to see how the HA filter really lights up the HA regions in galaxy photos (in this case bursting out from the near the core) and shows far more nebula detail than I have seen hitherto using just the RGB plus UHC filter combination.

best wishes
Tim

[timh]

IC1805 seems to look a lot more natural with narrow band OIII into the image as well. This below is the former IC1805 HA- supplemented image but with a further 30 min of RGB AS1294 PRO MC UHC exposures (as above) and also 81 minutes of 3 minute OIII exposures (Astronomik 6 nm) added to the G and B channels. I suppose that the decision about the relative proportions of RGB, HA and OIII to mix in is somewhat arbitrary ? - Below the areas with both OIII and HA seem to appear in shades of pink through to white nebulosity. It may be that the OIII is rather too much with it added to both G and B channels ?

I am quite liking the combination of quite fast optics (F 4.5), OSC camera plus Astronomik UHC filter and MONO camera plus narrow band for doing astronomy under UK skies where it is often a case of getting outside quickly to take advantage of the odd hour or two slots when the skies are clear.


best wishes
Tim

[admin]

Hi Tim,

yes, there's definitely more 'punch' to that second image, which helps a lot. Great work all round, thanks for sharing

Cheers, Robin

[turfpit]

Tim

Very nice images. The Ha has certainly brought out some interesting detail at the centre of M81 (I must try that). For IC1805 the improvement will be a combination of the added OIII and the significant increase in integration time (totals probably worth stating). The less black image looks better (that may be a personal thing).

IC1805_comparison.JPG (78.65 KiB) Viewed 1894 times

It may be that the OIII is rather too much with it added to both G and B channels ?

I use Astro Pixel Processor for processing. One of the tools is an RGB Combine. One of the pre-sets is RGBHOO. This is the default way APP assigns to channels for this pre-set (R, G, B, Ha, OIII stacks combined):

• R 100%

• G 100%

• B 100%

• Ha, 100% to R

• OIII, 100% to G & 100% to R

which supports your decision to add OIII to G & B. There is no harm in experimenting though.

Dave

[timh]

Many thanks for the comments folks.

The standard astroprocessor pixel numbers for RGBHOO are indeed comforting because it otherwise did seem somewhat arbitrary.

I tried some further experimentation on NGC1491 (the fossil footprint nebula) which has a bright OIII core (about as bright as H I think) and found that adding 100% of the HA to the R channel and 70% of the OIII signal to each of the G and B channels (for equal exposure times) resulted in a shade of red that sort of 'looked right' --i.e a bit warmer and not too puce while making the core white (which would be correct for equal OIII and HA contributions at the core).

While RGBHOO is, I suppose, an attempt at a sort of hyper-realism which is why I like it (as opposed to using the Hubble palette) it is no doubt quite flawed however you do it? For example it is difficult to completely cancel out the stars when doing the subtractions and so some of the OIII signal will probably always be a contribution from reflection nebulosity ? But the pictures do look nice and also do have some real scientific content regarding the content and distribution of the ionized gases - which also says something about the history of the star dust from which the new stars are forming.

Tim

[timh]

Just a postscript question on the off chance that anyone should look this far down the thread again.

I liked the idea of using the H alpha filter to spot areas of active star formation in galaxies and indeed active galaxies per se. But - thinking about this - even at 100M light years ~(Z = 0.007 or so?) - narrow band H alpha filters must start to 'miss' because of red shift (4-5 nm). Does anyone else share this somewhat obscure interest and know of fixes (broader filters?) etc?

Tim

PPS and answering my own question I find that this is an active topic of discussion on other fora in case of interest
https://www.cloudynights.com/topic/6950 ... -overview/ and an SII filter maybe just about right for H alpha in say the coma cluster?