M33 - Reprocess using a couple of new PI tools

[gnomus 2,843]

Back in October 2016, I did a short-ish M33 - 1 hour each of RGB, 2 hours of Lum and a paltry 80 mins of Ha.  The thread can be read here: https://stargazerslounge.com/topic/278949-m33-with-super-luminance/?tab=comments#comment-3054617

Whilst I was on holdiay, a new version of PI was released.  This contained a new process called PhotometricColorCalibration.  The idea, as best as I can understand it, is: plate solve your image; compare it with a colour image from a 'reference' catalogue; and adjust the colour calibration to try to match that reference.  The reference catalogue used is  the AAVSO Photometric All-Sky Survey or APASS.  There has been some debate on the PI forum about whether or not it might be better to use a different catalogue.  However, I only skim read this - I can only get a few posts into a typical PI forum thread without wanting to poke sharpened matchsticks into my tender areas.  Anyhow, I thought I'd look for an RGB image to try out this new process, and M33 came up.  

I was also alerted, by the one and only Barry Wilson, to another process called 'LocalNormalization' .  I don't know if this is a brand new process in the latest release, or if it has been around for a while but I hadn't noticed it before.  Barry gave me a brief run down on how to use it.  You choose your best sub as a reference and then run the process on all of your other subs, just prior to ImageIntegration (ie after Star Alignment) - or at least that is what I did.  LocalNorm files are produced and you add these to the ImageIntegration files (a bit like you add the Drizzle files).  Barry suggested that this process might reduce the demands on (or even the need for) DBE - and certainly the stacks did seem reasonably 'flat'.  

I should warn that I have only a vague idea of what I am doing with these processes, and it may well be that I am using them incorrectly.  Please consult with a PI Expert (or 'Sensei', as I believe they prefer to be called) before trying these things for yourselves. 

The new version does seem less 'muddy' to my eyes.  I have a vague recollection of seeing some M33 images with nice 'ruddy' cores, and in the earlier version I was probabaly chasing that.  I am still trying to aim for subtle ....... largely in an attempt to compensate for my personality, which is anything but.  

This is my Esprit 120; QSI 690 on the Mesu 200: 6 x 600s each of RGB, 24 x 300s of Lum and 4 x 1200s of Ha.  If I was doing this again I would probabaly try to double-down (at the very least) on everything:

 

Edited August 14, 2017 by gnomus

[wimvb 4,530]

Great image. A little less colour saturation and better taming of the stars, it seems. I like it.

[gnomus 2,843]

Thank you Wim.