Messier 83 - Southern Pinwheel - Still learning but getting there

[ionia23]

Okay, time to get brave. There's a laundry list of things wrong with this image, but this is the first time I managed something that is a little aesthetically pleasing. Since I've had nothing but cloudy nights for the last two months I've been investing my time in image processing, building a workflow, trying to understand the tools.

Yes, the Object of Interest should always be in the center

Object: Messier 83
Location: Tucson, AZ
Seeing: Beats me, urban skies but at least magnitude 5, no clouds, no wind.
Imaging date: 04-19-2021

Imaging details
------------------
Telescope: Orion Skyview Pro 8" (203mm , 1000mm focal length)
Camera: Orion Starshoot G4 One-Shot-Color
Capture Software: Orion Camera Studio
Capture format: Orion's weird version of FITS.
Images: 8 x 30 seconds (4 minutes total exposure time)
10 dark frames and 10 flat frames subtracted while still in FITS
Debayering: Orion Camera Studio to YCBR, no normalization or color adjustment.
Export all to TIFF images since DeepSkyStacker can't debayer these properly due to an X offset of 1. Exported images converted to RGB from CMYK
Debayering pattern:
M G M G
C Y C Y
G M G M
C Y C Y

Stacking Details:
Deep Sky Stacker 64 bit
Stacked using Median/Intersection/ 3X Drizzling
Output image saved as integer/FITS.

Startools:
Forgot the log file like a knucklehead.
Loaded as Linear as I didn't want to deal with Compose mode. Confuses me.
Autodev for analysis. Uneven field off to one side. I have a bear of a time removing these in Wipe. Cancel changes.
Crop out stacking artifacts. Only two pixels on each side. Not bad.
Bin all the way back down to the original image dimensions 752x582 . No Square Stars.
Wipe - Ultimately fruitless trying to remove the uneven lit field. Saved changes to date and did the "synthetic flat" trick in Photoshop with subtract mode, 85% opacity. Better.

HDR - Tried suggestions about lowering the detail. Very surprising how all of the sudden small details started enhancing.
Deconvolute
Shrink - I used the Tighten mode, defaults.
Color - Had an absolutely nightmare of a time trying to get the colors right. They still aren't.
Entropy detail, defaults.
Super Structure: I used DimSmall and backed off the airy disc just a tad
DeNoise: I might have overdone it at this stage and lost some detail I shouldn't have lost.
Synth: I love this module. I configured the settings to match the telescope used in the capture and backed off the aggressiveness of the diffraction spikes a bit. Synth does some amazing work "beautifying" detail.

Saved, exported, etc.

I know, a lot to learn, plenty of mistakes, could be better, but I feel good about this.

[Mike in Rancho]

Good start. Sounds like a bit of a challenge with your equipment, but making the best of it.

Although yeah I did keep trying to scroll down to see the bottom of the galaxy..

If you can get some practice on available data and get used to Compose mode. It's weird at first, but that goes away pretty fast, and before long it'll be weird to process in color/non-compose.

I think in a recent thread Ivo noted that Entropy and Flux are more for nebula images?

[ionia23]

Good start. Sounds like a bit of a challenge with your equipment, but making the best of it.

Although yeah I did keep trying to scroll down to see the bottom of the galaxy..

If you can get some practice on available data and get used to Compose mode. It's weird at first, but that goes away pretty fast, and before long it'll be weird to process in color/non-compose.

I think in a recent thread Ivo noted that Entropy and Flux are more for nebula images?

Yeah, yeah, rub it in about centering the Object of Interest

I'm going to try and go back to editing in Compose. Let me ask this, though. I'm using a One Shot Color. In Compose mode I won't even see any color data untill I get to the Color module (with the exception of in part of the Wipe module).

What's the advantages of doing it in compose vs. opening as linear?

[Mike in Rancho]

Yeah, yeah, rub it in about centering the Object of Interest

I'm going to try and go back to editing in Compose. Let me ask this, though. I'm using a One Shot Color. In Compose mode I won't even see any color data untill I get to the Color module (with the exception of in part of the Wipe module).

What's the advantages of doing it in compose vs. opening as linear?

Well, we'll need Ivo for a real answer of course...

Off the top of my head - you get to work in luminance and see the detail, color will be appropriately dealt with later as it is special secret parallel processed in the background; I believe it knows to take advantage of the bayer array when creating that synthetic luminance; it also knows (or assumes) that the color channels have not been messed up by white balancing or channel neutralizing, which sounds awfully helpful so it must be somehow ( ); and based on that video interview Ivo did (link posted here somewhere) I think it also helps with the tracking - i.e. ST's signature advantage.

It's probably also laid out properly and in better detail in the module descriptions.

[ionia23]

You may find this amusing. I decided the best way to use Compose is to just "get used to it". So on all my image processing I'm opening up the data as the linear with bayer option. Takes some serious getting used to, but it makes handling color adjustment much easier in the long run.