6 replies to this topic

[moleir00]

Hello everyone! I’d like to understand what I'm doing wrong, or what I can do to improve my overall LRGB processing, 

 

Context: I’m new to processing LRGB images, as I spent the last few years doing only SHO, since I live in a Bortle 9 city. Now I acquired some data under a Bortle 4 town and I’ve been trying to process it for over a week, but I never seem to get it right - either the image is supersaturated, or missing colors, or missing contrast, or the bright stars are too bright…every time I get something right, I mess up two other things 😦

 

In the link below you will find two folders: one named Original, with every frame exactly as they left WBPP, if you'd like to give it a try from zero; and another folder called “Half Baked”, which contains the image in the last stage I get it to before I don't know what to do anymore. After this point, I just can’t effectively combine the two images without making them look horrible. I’m aware of the higher noise on the bottom part of the HaRGB image, but applying BlurX or NoiseX seems to mess up my colors, so I don’t know what I did wrong unfortunately.

 

https://drive.google...?usp=drive_link

 

Also, here are the details of my acquisition and calibration:

 

Data: 2 Panels Mosaic of SH2-1

 

L = 107 x 300s (8h55min)
Ha = 83 x 180 (4h09min)
Blue = 29 x 180s (1h27min)
Green = 27 x 180s (1h21min)
Red = 33 x 180 (1h39min)

 

Total = 17h36min

 

Calibrated and Integrated with Bias & Flat frames (no Darks, as I use the ASI2600MM, but I am reconsidering it.)

 

Thank you very much!!

[scopewizard]

Very quick process of the RGB images only.

 

 

Attached Thumbnails

• 

[scopewizard]

My process from last post.

 

selected each file and rename accordingly R,G,B

Combine use RGB

Stretch

BackgroundNeutral

ABE

Slight Curves adjustment.

Denoise

 

Reduced size to fit post.

 

Nothing else. really quick

 

Click on image to see larger size

[moleir00]

Thanks for looking into it! However, my painpoint is actually calibrating/combining everything, given there are two panels for each filter (meaning 10 images in total), which I haven't had to do before.

 

if you refer to the folder named "Half Baked", you'll see I went through the processes you described, but I can't seem to find a way to make everything stitch together without creating aberrations/differences in luminance/etc.

[scopewizard]

Adding HA to LRGB is not that simple.

 

This process is simplify in this video. Download the scripts in the link.

 

https://www.youtube....h?v=fSVIdDSviR4

 

Many other videos are available on Youtube.

[dswtan]

I took a look. I did find something weird, but it may not be ultimately important -- your G and R channels seem to be swapped.

 

The G_mono_PANEL files actually show "R" in the FITS header, and vice versa (R-mono_PANEL show G in the header).

 

Now, I'm not new at LRGB and/or +Ha -- at least, for galaxies. Also mosaics. But I'm less familiar with processing reflection nebulae. That said, I thought this might be good practice. I was wrong. :-)

 

My 2c: I see your data capture stats and they look good at 17hrs -- though is that maybe really 8.5hr per panel? Even so, I would have expected to see more structure in the stretches. I tried a number of methods. 

 

Your "half-baked" L looks pretty good to me. At least, I couldn't do much better (starless can show a bit more, though it's hard to deal with bright stars), and comparables on Astrobin.

 

Your "half-baked" HaRGB doesn't show much structure at all. I found basically the same in retracing your steps, but by setting up the mosaic first (in APP) and using my own workflow defaults.

 

I couldn't see anything really in your Ha -- though looking at Astrobin, that might make sense since it seems the target just has a big red largely featureless blob around Pi Scorpii. There is some structure in the deepest images though.

 

Conclusion? I'm not sure. The R-G swap shouldn't make a real difference, though might be confusing. Maybe the challenge seems to be in the capture somewhere, or my limits with such data. Unless that is just what to expect with this target. 

 

It might be good to find someone who has shared or is willing to share their data for this target. Collaboration opportunity perhaps?

[dac4215]

I have never tried an LRGB image before, but weeks of "cloudy nights" brings on boredom. I thought I would give your data a go.

Not very good at using pixinsight for mosaics so I used astropixel processor to put the L R G B panels together. Decided to not attempt the Ha.

Cropped and star aligned the L R G B

ADBE on the L R G B 

Channel combination on R G B

SetiAstro find background, Background Neutralization, Color calibration

STX on rgb

Statistical stretch on rgb and L

Nxt

From here it was using masks, curves, selective color script, and JR astrophotography macros in affinity photo.

Used star stretch and combined stars and starless

LRGB combination to add in the L

Acdnr

Touches to the color

At the end I realized that I had left out Blurx, but decided that I wasn't going to start over. If it was my data I would go back and start over, use Blurx, and try to make improvements on the stars. I probably left out a few steps, I was trying different ideas. Good data, but I would also want to add more time to it.

 

 

 

Attached Thumbnails

•