Synthetic luminance
A synthetic luminance is built from existing channels instead of being captured with an L filter. It carries structure and sharpness; the colour image carries colour. Both are processed separately and merged with Replaces the brightness of a colour image with a separate luminance – structure and sharpness from the luminance, colour from the colour image. Open fundamentals →.
Purpose
| Data | Luminance from | Benefit |
|---|---|---|
| colour camera (OSC) | L* of the colour image | process structure and colour separately |
| L + Ha | max(L, Ha_LF) or weighted | nebula from Ha with better signal, star field from L |
| narrowband (Ha, OIII, SII) | max(Ha, OIII_LF, …) or weighted | each structure with its strongest signal |
Whya synthetic luminance?
In LRGBCombination the luminance determines the entire image brightness and structure. A luminance from one channel only suppresses structures that shine mainly in other channels – they would end up dark in the final image although the colour image contains them. A luminance built from all channels carries each structure with its strongest signal.
Principle
- Maximum takes the brightest channel at every point. Structures come from the channel in which they are brightest; stars are not added from several images.
- A weighted sum is softer and less noisy, but averages faint structures of one channel with the noise of the others.
- Before either, the channels must be brought to similar levels; otherwise
max()always picks the channel with the highest level, regardless of content.
Operation
Matching levels with LinearFit
- Copy the weaker channel (e.g.
OIII_LF). - LinearFit with the strongest channel as reference.
- Reapply the auto-STF – with the old STF the image often looks white. Statistics: median ≈ median of the reference.
Combining
Combines images pixel by pixel with formulas – for combining channels, blending luminances or recombining stars. Open fundamentals →, destination Create new image, greyscale, rescale off.
| Variant | Expression | Effect |
|---|---|---|
| maximum | max(Ha, OIII_LF, SII_LF) | each structure from the brightest channel; stars from the channel where they are narrowest |
| weighted | Ha*0.5 + OIII_LF*0.3 + SII_LF*0.2 | softer, less noisy; weights sum to 1 |
Recommendation: start with maximum. If the result looks noisy in faint areas (the maximum favours noise peaks), use the weighted variant with a higher share of the strongest channel.
Typical applications
Colour camera: L* from the colour image
ChannelExtraction on the colour image, CIE L*a*b*, L only. L* corresponds to perceived brightness and is exactly the component that LRGBCombination later replaces.
As the luminance comes from the same image, it contains no additional information; its benefit lies solely in separate processing. Two moments are possible:
| Route | Extraction | Stretching | Matching before combination | Use |
|---|---|---|---|---|
| A – simple | from the stretched colour image | not needed | not needed (identical to the colour image’s L*) | standard |
| B – more control | from the linear colour image | separately | required | the luminance should be stretched differently (deeper black point, more structure) while the colour image stays softer |
L and Ha
max(L, Ha_LF) after LinearFit of Ha to L: nebula from Ha (stronger signal), stars and star field from L (larger and brighter there). Weighted: L*0.6 + Ha_LF*0.4; more Ha brings more nebula structure but a harsher look.
Narrowband
max(Ha, OIII_LF) or with SII; stars come from Ha, where they are narrowest. With noisy OIII, the weighted variant with a higher Ha share.
What a good result looks like
Typical errors
| Observation | Cause | Remedy |
|---|---|---|
| Structures of one channel missing | levels not matched | LinearFit first |
| Luminance noisy in faint areas | maximum favours noise peaks | weighted variant |
| One structure colour too dark after combining | weighting too low | adjust the weighting; Measures at fixed spots whether a separate luminance and the colour image match in background, midtones and highlights, and adjusts the luminance with a curve – so the LRGB combination is neither pale nor oversaturated. Open fundamentals → |
| Image white after LinearFit | old STF | reapply the auto-STF |
When to use it – and when not
| Good for | Better another tool |
|---|---|
| processing structure and colour separately without an L exposure | a real L exposure exists → use it, or combine it with Ha |
| narrowband: bringing all channels into the structure | colour image without separate processing → stretch directly |
HaLRGB workflow · Narrowband workflow (HOO/SHO) · OSC workflow with a quad band-pass filter · PixelMath