Image processing workflow

[tooth_dr]

Cant believe I'm having to ask this, but looking a bit of confirmation on this workflow as its been over a year since I processed data and I'm struggling to remember.

I'm processing OSC, LUM and Ha data separately first, them combining.

 

WORKFLOW

OSC / LUM / Ha data:

Separately stacked in APP -> registration/alignment of the three files in APP -> equally cropped in APP and saved as FITS giving me linear a set of aligned/registered FITS

 

REGISTERED OSC FITS:

SPCC in PI -> DBE in PI -> BXT in PI -> SXT in PI -> Processing in PS (levels, saturation, etc, my preferred program for this for time being)

 

REGISTERED LUM FITS:

DBE in PI -> BXT in PI -> SXT in PI -> Processing in PS

 

REGISTERED Ha FITS

Stacking in APP -> DBE in PI -> BXT in PI -> SXT in PI -> Processing in PS

 

COMBINE LUM/HA/OSC in PS and adjust to taste

 

ADD in the stars....

 

When do I use NXT and when does HLVG get applied to the colour data?

Currently working on some vdb 152 data here and it looks promising enough to get a decent image!

 

Cheers! 

Adam

 

Quick stack:

 

 

 

 

 

 

Edited Wednesday at 22:12 by tooth_dr

[Ouroboros]

Chuckle. I don’t really have a work flow. More of a multi-channel flounder through a pan dimensional parameter space until  I wash up at the end with something I call a processed image.

Part of the problem (although in some ways it’s a nice problem to have) is that so many new scripts and processes have been launched recently for PI that it’s difficult to know which to choose.

The only reasonably consistent work flow I have is while the image is still linear.

I do straight OSC or OSC with dual Ha/OIII filter.

Dynamic crop.

BlurXterminator. Correct only.

Image solve and then SPCC

BlurXterminator again this time with sharpening. 

Sometimes I do a background extraction before separating the stars and sometimes afterwards. I usually try Graxpert, DBE, Seti Astro’s script and pick the best.

Stretch usually with Generalised Hyperbolic Stretch.  But sometimes I find one of the automatic stretch scripts work better than I do.

Once the image is non-linear I set out in a fairly random walk through curves, saturation, sharpening etc etc etc until I arrive at some kind of destination.

PS. I throw in some noiseXterminator at various points too.

D’you know …… I don’t think “work flow” really works as a concept in image processing anymore. The idea applies to a many times repeated task which image processing isn’t. 

Edited 21 hours ago by Ouroboros

[tooth_dr]

7 minutes ago, Ouroboros said:

Once the image is non-linear I set out in a fairly random walk through curves, saturation, sharpening etc etc etc until I arrive at some kind of destination.

Yep this is how I go about it too!  But just trying to establish a linear workflow before it all gets out of hand.

[Elp]

I don't use PI but the workflow sounds about right. I normally run NR right at the end very subtly, leaving some noise in the image as it looks unnatural without it. If HLVG is green NR, you can adjust your RGB curves peaks so they're aligned which will colour balance out excessive green due to the two pixels per one of RGGB bayer matrix on OSC cameras.

Edited 20 hours ago by Elp