I've read through several other posts, I have Warren Keller's book and I follow Kayron's workflow also on LightVortex and I'm still struggling with DBE on the Horsehead Nebula. On my Lum, RGB and even on my Ha the results seem very 'bitty', like DBE is being way too aggressive. The light pollution was pretty bad though, but I would have thought that the Ha would not be too affected by that.
I've had quite a few attempts, I use a sample radius of 15, I use no more than 6 samples per row, avoiding stars and faint nebulosity (difficult here) and then applying subtraction. So I'm mostly in the corners, around the edges and a few samples on what looks like neutral background dotted around. I tweaked the tolerance and minimum sample weights a bit just to get more boxes in the difficult places.
Running DBE separately on the RGB channels and then combining was not pretty, I got a better result from linear fitting G&B to R, then combining and then applying DBE. That is what is attached here.
If anyone has time and doesn't mind here is a link to my RGB data in my Google Drive, I would appreciate it greatly if someone can point me in the right direction or at least tell me that this is the best I can hope for with this data. After integration, only Dynamic Crop and MURE has been applied.
R = 101 x 60s G0 O10
G = 115 x 45s G0 O10
B = 105 x 45s G0 O10
https://drive.google.com/drive/folders/1PjdZHEhRXLYGQktrRulgMRu7_uV3UDfl?usp=sharing
