1

I attempted to photograph the Big Dipper last summer. While I am very happy with the overall result, I recently found that the color of the stars has been greatly reduced during stacking. The following photos show about the same section of the sky.

Single Light Frame:
enter image description here

Unprocessed Stack output:
enter image description here

Final Image:
enter image description here

As you can see, the violet color of the two stars is greatly reduced and almost lost in the final image. For the difference between the latter two, this can be explained by an (apparently) too strong brighntess increase of the stars (I should have limited this to the fainter stars).

I stacked 75 Light Frames (ISO 1250, f/2.8, 4 seconds), 25 Dark Frames, 30 Flat Frames and 30 Bias Frames. In terms of post-processing, I mainly adjusted levels, curves, removed a gradient, increased star brightness and selectively increased saturation for some stars.

What could be the cause of the loss of color when stacking?

Some more information in respond to the comments: The two bright stars are Mizar and Alkor. Micar (the lower, larger star) is Spectral Type A1V, Alcor is A5V+M3-4V.

The violet-ish color is present on every light frame.

Improve this question
6
  • You answered your own question... any color at/beyond maximum reproducible brightness becomes white. – Steven Kersting Dec 17 '20 at 16:36
  • @StevenKersting I was rather asking about why color reduces when stacking, not during postprocessing – Jonas Dec 17 '20 at 16:45
  • The star light isn't violet. That comes from somewhere else. – Thorbjørn Ravn Andersen Dec 17 '20 at 18:17
  • 2
    The color is possibly from color fringing in the lens. The software may take care of that for you automatically. – BobT Dec 17 '20 at 22:23
  • 1
    In some other answer in these parts it was pointed out that stars can also be small enough to fall onto a single sensel so they can be red, green, or blue (or missing one of the channels) depending on where they fall on the sensor. Stacking images would average their position and mitigate this, removing the color bias. – xenoid Dec 18 '20 at 9:39
2

In some other answer in these parts it was pointed out that stars can also be small enough to fall onto a single sensel so they can be red, green, or blue (or missing one of the channels) depending on where they fall on the sensor. Stacking images would average their position and mitigate this, removing the color bias.

Improve this answer
1

Suppose that Mizar had a colour value 20:5:20 so the red and the blue were stronger than the green giving you a violet star.

Stack up 10 copies: 200:50:200. Same ratios.

Stack up 20 copies 400:100:400. But your rgb values are limited to a max of 255 So you end up with 255:100:255

Stack up 50 copies and you end up with 255:250:255.

Now the stacking software is not this simple minded, but this illustrates the problem.

Improve this answer
1

The image is more accurate, not less accurate. The star color is not str color at all. It is an aberration.

Improve this answer

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.