The new version (currently in Beta) of BlurXTerminator by Russel Croman is going to fix many of your previous issues. He's training it on tilt, coma and focus issues and it's exicting that in true Google style software is going to be able to assist with hardware issues soon.
Saying that, in terms of bang for buck absolutely cannot fault the Askar! I had a Sharpstar 94 for years and only just sold it. They are great scopes.
Lovely images :)
Sincere thanks for the list. I listed Sync because I used that for years on Reddit, and I just consider it default. The updates were endless and free and the sheer quality of the app was notable.
So API is the issue? How is it that Lemmy instances have an API - have the created their own? I am non-schooled in this - just not used to it because I have been on Reddit for years.
Thought I'd share my first image in a while! NGC 6823 has sooo much going on and looks great in both RGB and Narrowband. I went for Foraxx here, with RGB stars. Desperately needed more O3 - I lost 80% of my subs due to high haze. I had to stretch it within an inch of its life!
Taken at the awesome Roboscopes facility in spain on Pier 3 (ASA 12N). I am currently in the process of sending my Tak 160 there for hosting, the skies are FANTASTIC and the managed hosting service the guys there provide is world class. So much better than alternatives.
Processed in Pixinsight and Photoshop
Stars RGB.
Manual calibration - WBPP gets Bias wrong :)
DBE on each channel, SPCC, BXT / NXT / SXT. I created a synthetic lum channel from the Ha O3 and S2 and processed that using GHS and curves.
I then GHS'd each NB channel to get them popping as much as possible, and combined them in Foraxx. Added the Lum using LRGB combine.
Then off to Photoshop for a few levels tweaks and the brilliant filter Camera Raw Filter (CTRL + SHIFT + A on a new layer). So many amazing tweaks to be had here. Use opacity to decide how much of the filtered layer to keep.
120 second subs for NB, 60 second for RGB. 3nm Antlia filters.
60 x 120s O3
180 x 120s S2
120 x 120s Ha
15 x 60s RGB
That is a thing of beauty!
Basically, no. It doesn't create data. It is trained on perfect images, which are then ruined in ways that we astrophotographers only know too well: haze, tilt, focus etc. etc.
The AI then tries to get back from the shit picture to the perfect picture, recording steps taken. So in essence, it applies these steps to your image as per its training. This has all be verified, there's a 2 hour Adam Block video where he compares his good data to NASA's amazing data and you can see BXT is making his data more true.
Topaz on the other hand creates stuff that is not there all the time. I have zero hesitation using and recommending BXT.