M31, with and without Ha enhancement

[Bob Andersson]

Hi folks,

Here's M31 without additional Ha:

and with some Ha (Hydrogen alpha emission) blended in:

For your additional viewing pleasure I've made the 4096x4096 pixel versions available here and here respectively. All four images are copyrighted so please do not edit and/or repost without permission.

This is old data from 2012 but it's taken me until now to successfully apply my natural colour workflow (description). The relative RGB values at each pixel have been preserved during stretching with no boost to colour saturation at any stage of the workflow, with the white balance for the entire image determined by M110. Whether that was a valid white balance choice is for wiser heads than mine. How much prominence to give the Ha data was a matter of personal preference: in the end I chose what I hope is a subtle blend which doesn't overpower the rest of the galaxy and limited the Ha to the spiral structure.

I provided links to the full-sized versions not because they are pixel perfect but because they do offer an ability to see some of the hot blue stars associated with the structure of M31. I hope you enjoy the flight.  8)

Bob.

5 x 1000 seconds in each of R, G and B plus 10 x 1500 seconds Ha. TEC 140 plus ML16803.
Processed in PI & PS.

[kolec]

Very, very nice! Only colour of stars ! Maybe I have no good calibrated display

Kolec

[Bob Andersson]

Thanks Kolec.  8)

Only colour of stars

As in too red, too blue or too much? As best as I could manage it the proportions of red, green and blue at each pixel in the image are in exactly the same proportion as in the original unstretched data but there's always a question mark over how accurately one can reconstruct the colour of the more overexposed star cores and some of the bigger blue ones certainly do look a bit intense unless they are extremely hot! The same white balance decision which determined the colour of the galaxies was also applied to the stars and I think it has worked out reasonably well when one looks at the faint blue stars that actually belong to M31.

But with all that colour to play with I sometimes feel like a novice driver put in charge of a Ferrari - a car crash just waiting to happen! 

Bob.

[bianry]

Very nice! Smooth and with pleasing choice of colors.

You think you could get a bit more structure out of the core with HDRMT? Just curious.

Like it!

Mats

[sazam91]

Amazing!!!!

[Bob Andersson]

Hi Mats,

I did try both HDMRT and LHE but wasn't happy however I tweaked the settings. In the end I used two passes of Noel Carboni's "Local Contrast Enhancement" PhotoShop action for the core and just a single pass for the rest of M31, blending the two together using a feathered selection.

My aim for this image was to try and show as much as possible without the end result looking too processed. Hence I was content with the detail in the core being visible but not too in your face. Similarly with the Ha where I added my Ha data as a PhotoShop Screen layer with just a 31% Fill factor - it could have been a lot brighter and more would have been visible but on purely aesthetic and hence subjective grounds I didn't want the Ha to overpower the equally valid adjacent cyan nebulosity.

Thanks for all the positive feedback, folks.  8)

Bob.

[Bob Andersson]

Hi folks,

I think I should make it clear that whatever choices I subsequently made the images above could not have been produced were it not for PixInsight which was responsible for the data from the moment it left the camera until and including stretching. I routinely use PI later in the workflow as well so these images were something of an exception in that regard.

Correction: Actually the stars, which had been separated from the nebulosity, did pass back through PI later in the workflow. I rarely use it these days but the Deconvolution tool provided a handy way of making the stars less prominent while also retaining the small and fainter stars that actually belong to M31.

Bob.

[Shekhar Phatak]

Very beautiful!

Shekhar

[Torsinadoc]

Very nice. I am looking at your script.  It appears that you will use the script right before the initial stretch (after ColCal, SCNR, etc), stretch with maskedstretch then recombine and process as normal?  Are you boosting saturation later in the process?

Thanks
Alan

[Bob Andersson]

Hi Alan,

That's correct. Apply the script to an unstretched image, then stretch the "V" component and then recombine.

Using this workflow I find no need to boost saturation later on. Indeed, for purely aesthetic reasons, it is sometimes tempting to reduce the saturation by blending with a "normally" stretched image but I usually resist! That said, I've been working on some old Hubble data recently and with the extended range broadband filter sets (435W/550M/814W) I definitely did need to reduce the saturation - Arp 244, for example, and that's after desaturation! It's important to get the white balance of any image correct before applying the script and stretch and this seemed especially so for M31.

Bob.

[Torsinadoc]

great. Ill give it a try.  Appreciate you time in producing this script

[JoLo]

That is one beautiful M31, nice job.

Joe