Combining Ha and O3 for Luminance

[niteman1946]

Hi Group,

I use Ha for luminance, and fully process this data for the detailed image.  I then combine Ha, O3 and S2 and produce a blurred color image that is added to the luminance.   This follows (mostly) kronjm's tutorial of Feb 24, 2013 for processing.
http://lightvortexastronomy.blogspot.com/2013/02/tutorial-pixinsight-workflow-for-hdr.html

I am currently working on NGC6960 which is predominantly Ha, and have extracted a nice snappy result (see left image).  However, there is some O3 detail that gets lost when its color component is added to the Ha luminance.

Previously with M1, I had good results using PixelMath to combine the Ha and O3 for luminance.  However, no such luck with this effort (see right image).

It may be difficult to see, but the Ha only (left) image has a nice dark backgound and bright nebula.  The Ha plus O3 (right) image is somewhat flatter, although it has incorporated the O3 signal.  I've processed the Ha plus O3 image the same way as the Ha only, but have been unable to pull out the same amount of contrast.

The PixelMath equation was simply adding the two data sets together.  Rescale results was checked.  I also tried the same formula with Rescale unchecked, with no discernable results.

I would appreciate any suggestions on how to better combine the mono versions of Ha and O3

Thanks,

Mark


 

[pfile]

perhaps try a linear fit of the OIII image to the Ha image before the pixel math combination. the signal strengths usually differ wildly between OIII and Ha. of course the OIII is going to get boosted up and the noise will be boosted as well…

rob

[niteman1946]

Thanks Rob,

Hard to say if it made any difference - maybe ever so slightly.  But I like the strategy.

My O3 needs quite a bit more integration, so I will revisit this when (if) I get more photons.

Mark

[niteman1946]

Any other suggestions from the group?

Mark

[Ioannis Ioannou]

Not an expert in the field, just my 2 cents.

I'm missing the point why to create and use a pseudo-Luminance while you are doing HST or bicolor and why you expect not to loose O3 since you use Ha as L.

My suggestion: did you tried to just RGB combine your frames ? No L, no strange pixelmaths, nothing.
Either HST (R=Sii G=Ha B=O3 and maybe curves->Hue to make the colors more "appealing" ) or bicolor (R=Ha G=B=O3)

[oldwexi]

Hi,
would never use Ha as Luminance because in RGB it gets you pink nebulae
as L is  always distributed not only to the red channel also in the same amount to the green and blue channel.

Especially with Veil nebula where there are a lot(!!) of fine and great OIII details, they are only a little fainter than
the Ha, using the Ha as Luminance kills, coats, covers, the OIII data.
So Ha in one channel, OIII in one channel (or distributed to 2 channels, and SII in another channel
keeps in a Color Images the Color details.

generally using Ha as Luminance in my experience is simply not the best processing style.
H-Alpha has to go mainly in one channel and not in all channels by using it as Luminance...

Gerald

[niteman1946]

Hi Ioannis and Gerald,
thanks for your responses.

I'm not sure I fully understand some of your suggestions, but freely admit it's probably my limitations, and how well I explained my process.
I spent a couple of days taking your recommendations into consideration and revisiting the process -  and ultimately came back to my method (well, kronjm's method):
1.  Add linear monochrome Ha and O3 using PixelMath.  I did find that checking "rescale result" did boost the contrast, as opposed to my original statement.  This image was fully processed to obtain a very detailed monochromatic picture.  See left image.
2.  Use PixelMath to add the linear monochrome Ha and O3 into a color image.  The color combination was my choice.  This color image was fully processed and finished off with a blur.  See middle image.
3.  Use LRGB to combine the mono detailed image and the color blurred image.  See right image.

The difficulty I was running into was with the monochrome addition of Ha and O3 (first step).  Not sure if I got everything I wanted, but ultimately found it was not too bad.

Mark

Here's a link to the finished product.
http://astrob.in/67035/0/