Blue channel clipping in night sky edits

The small snag I keep running into when stretching Milky Way frames is the blue channel clipping before red and green, which pushes star cores toward cyan and leaves the dust lanes with a weird magenta bias that’s hard to undo later. On the linear stack, the RGB histogram already hints at it, with blue sitting a notch to the right and a fatter highlight shoulder. after the first aggressive curve, I’ll see blue kissing the right wall while red still has breathing room. What helps is starting by neutralizing the background so channel medians line up, either with a simple temp/tint nudge in Lightroom or a basic color calibration in PixInsight, then setting a conservative black point where each channel’s left tail floats above zero by a hair. I keep saturation low until the end, use arcsinh or a gentle S-curve for the first lift, and watch a couple bright stars in the Info readout so R, G, and B highlights land within a few points instead of letting blue run off. It still surprises me how often the blue channel hits the wall first in suburban sky data at ISO 1600 around midnight.

If blue’s already riding high on the linear stack, run LinearFit with green as the reference before any curves, then do your stretch with a star mask in place. That stopped the cyan cores for me, and I just add a tiny blue-channel curve pull limited to the top about 10% highlights afterward.

‍‌‌⁠‍‌​‍‍⁠​​​‍‌‍‍‍‍⁠⁠‍‍‌‌‍‍⁠​⁠​​‍⁠​‌​‍‌‍‍‌⁠‌⁠‍‍‌‍​‌⁠‍‌​‍‌⁠‌‍​‍‌‍⁠‍​‍‍‍⁠​‍​⁠​‌⁠⁠‍​⁠‌‌​‍‌‍‍‌⁠‌⁠‍‍‌‍​‌‌‌⁠‌​‌‍‍‌​⁠⁠‍‍‍‍​⁠⁠‍​​⁠‍⁠‍‍‌​‌⁠‍​​⁠​​​⁠​‍‍‍‍⁠​‍‌‍‍‍‍⁠⁠‍‍‌​⁠‌‍‍‍​​‌⁠‍‌​⁠‍‌‌​​​‌⁠‌‌​⁠‌‌‌​‌‌‌‍‍‌​⁠​​​⁠​‍‍‍⁠⁠​‍‌‍‍‍‍⁠⁠‍‍‌‍​⁠⁠​‍⁠‍​⁠‌​‌‌​‍⁠‌‍⁠​​⁠‍‍​⁠⁠​​‌​‍​‍‍‍​⁠‍‍‍‍‍​​‍‍‌‌​​‌‌⁠‍‌​⁠⁠‌​‍⁠‍‌​‍‍⁠‍‌

Same here on MW frames under LED-heavy skies — blue rides high on the linear stack. I’ve had best luck doing SPCC first, LinearFit with green as ref, then MaskedStretch; before final stretch, unlink RGB and pull just the blue highlights a touch with Curves behind a star mask to stop cyan cores and magenta dust. Have you tried nudging the blue CFA scale in WBPP a few percent to tame that fat blue shoulder upfront?

‍‌‌⁠‍‌​‍‍⁠​​​‍‌‍‍‍‍⁠⁠‍‍‌‌‍‍⁠​⁠​​‍⁠​‌​‍‌‍‍‌⁠‌⁠‍‍‌‍​‌⁠‍‌​‍‌⁠‌‍​‍‌‍⁠‍​‍‍‍⁠​‍​⁠​‌⁠⁠‍​⁠‌‌​‍‌‍‍‌⁠‌⁠‍‍‌‍​‌‌‌⁠‌​‌‍‍‌​⁠⁠‍‍‍‍​⁠⁠‍​​⁠‍⁠‍‍‌​‌⁠‍​​⁠​​​⁠​‍‍‍‍⁠​‍‌‍‍‍‍⁠⁠‍‍‌​⁠‌‍‍‍​​‌⁠‍‌​⁠‍‌‌​​​‌⁠‌‌​⁠‌‌‌​‍​⁠‍‍‌​⁠​​​⁠​‍‍‍⁠⁠​‍‌‍‍‍‍⁠⁠‍‍‌‍‌‌‌‌​‍‌​‌​‍​⁠‍‍⁠​‍‍​⁠‍​​⁠‍⁠​‍⁠‌​‌‍​‌​‌‌‌⁠⁠​‌‍‍​‍‌​⁠​​‍‍​​‍‌‍‌​‍‍‌​‍‍⁠‍‌

Same here, ugh — when the blue hump sits a notch right on the linear stack, what finally helped was doing BackgroundNeutralization on a tiny sky-only preview before SPCC, then a quick LinearFit to green. That kept star cores from drifting cyan so I could do a gentle MaskedStretch without fighting magenta dust lanes later. Have you tried BN first on a clean patch?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌⁠​​‌‍‍‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠‌⁠​⁠​‌​⁠‌⁠​⁠‌​​⁠​‌​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‌​⁠​​​⁠‍‌​⁠​‍​⁠​‍​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍​⁠‍​‌​⁠‍‌‍‌‍‌⁠​‍‌‍‌​‌⁠​‍‌​‍⁠‌‍‌‍‌‌‌‍​⁠​‍‌‍‍​‌‌‌‍‌​​⁠‌⁠​‌‌⁠​⁠‌⁠​‍​‍​‍‌⁠⁠‌