m81 m82 - CORRECTION I need opinions

[Ioannis Ioannou]

Hello fellows,

I rarely post since I do not have much time for the hobby these days, and I think I need (a lot ) more practice and experience in AP  - especially processing.  I need your opinion now on my last photo. I'm not looking for congratulations, but rather for ways to improve my processing skills, digging in details of PI.

Korinthia - Greece 2011-12-03, FSQ @ f/5 + QHY-IMG2Pro @ -20C + Baader filters,
L : 9x15 mins - Ha/R/G/B : 9x5 mins each
Link to full PNG version here


The sequence I followed is (skipping some obvious steps, like STF, dynamic alignments, dynamic crops etc):

Pre-processing in PI (the standard procedure, the only thing different is that that I use DefectMap instead of dark calibration)

4 master frames, L, R,G,B,Ha

For colors:

LRGBCombination to produce a rgb
Silvercup's script to combine Ha with rgb and get a good HaRGBCombination frame
BackgroundNeutralization with a preview at the left of M81 where seems to be the least IFN
ColorCalibration with the above preview and a preview of as much of M81 as possible for White reference
SCNR on green
HistogramTransformation with careful stretching

L:

HistogramTransformation careful stretching


Final image :

LRGBCombination to produce the L-HaR-G-B
CurvesTransformation to increase saturation
HDRMultiscaleTransform on layer 6 with AWT and deringing
HDRMultiscaleTransform on layer 5 with MMT
SCNR again on green
ACDNR soft mask targeting the noise inside the structures
ACDNR hard mask, targeting noise of the background, while maintaining as much as possible Holmberg IX and IFN (but apparently I need longer subs and better sky for these two)
HistogramTransformation

I tried in many stages to use MMT for individual or all frames, both for noise and for sharpening, but apparently I do not know how to use it yet (although I tried hard) : the result had a feeling of "plastic", or fake or over-processed.

My main concern is the way I handled the core of M81 with HDRMultiscaleTransform. For M82 there was great benefit. But I'm not sure for M81's core. The one on layer 6 seems to be correct, but the one on layer 5 maybe introduced artifacts or it is too much ?
I need opinions and possible advice for anything else that may had improved the image.

Thank you in advance

[Ioannis Ioannou]

Arrg, I do not know how I screwed it  , the link for the full version was pointing to an older photo of M81 only, the correct is this one (corrected above also)

[Alejandro Tombolini]

 I tried in many stages to use MMT for individual or all frames, both for noise and for sharpening, but apparently I do not know how to use it yet (although I tried hard) : the result had a feeling of "plastic", or fake or over-processed.

Hi Ioannis,
you can use MMT to improve details in the galaxies.
First build a range mask to focus in galaxy borders. Apply MMT protecting core and background
Second without mask apply SCNR (I saw some green)
Finaly build an starmask and apply MMT protecting stars.

I also can see a gradient. It could be necesary at the begining to apply DBE.

Regards, Alejandro.

[pfile]

i also like the effects of a masked Local Histogram Equalization (just a bit, slider less than 1/2 way) to bring out some details in the spiral arms.

[Ioannis Ioannou]

Thank you guys, I'll try your suggestions

Alejandro are you sure that it is gradient and not IFN ?
None of my mono fits show any signal of gradients

Forget the above, you must be correct, I compared it with one from the internet and seems there is gradient.
There are signs of IFN but I have problems at the left, most probably from Ha - I have to recheck my fits.