Thanks everyone for the comments. I use PixInsight for image processing. Concerning my workflow with the L-Enhance filter, I first process the image to completion as a standard RGB image. I do not split the channels for RGB processing and I do not use any narrow band settings, as the L-Enhance filter is not a NB filter. So, for example, for color calibration using PCC, I use the broadband setting as working mode. My typical workflow after stacking is:
Crop to remove stacking artifacts
ABE or DBE
HT, or a blend of HT/MS
CT for contrast
Saturation using CT with a masked background
For the NB image, I use the finished RGB image. It seems to work better in terms of NB color than what I previously did, which was to use the RGB image immediately after taking it nonlinear. I first use Starnet to remove the stars; the stars image to be processed separately later. For the starless image, I split the channels, deleting the B channel. I use HT and/or CT to approximately match the G channel’s brightness to that of the R channel. For channel mapping, I leave R as R, G as G and map a blend of R and G to B, usually 50-50, sometimes more R. I then recombine the channels and use CT’s individual channel curves to get the look I want, followed by an RGB saturation curve. The stars only image is only slightly processed, using SCNR and a little saturation increase. I then recombine with the starless image using PM.