Why do stars shift colors when zoomed in, but not so much when zoomed out?

[Maury Markowitz]

I just realized I have not seen an explanation for this.

Scintilation causes the image of stars to move about on the sensor, and shifts across the color sub-pixels. But why does this cause the color of the whole object to change when it is zoomed in or unfocused, more than when it is properly imaged? For instance:


Source: https://www.youtube.com/watch?v=lq02auKvSzw&t=16s


or:


Source: https://www.facebook.com/watch/?v=1045730969127340


Compare to, say:


Source: https://www.facebook.com/watch/?v=788211875463574


It appears the effect is "magnified" by being spread out, but based on how I understand the sensor to work, I would expect the opposite.

 

[MonkeeSage]

Possibly related to chromatic aberration?

In optics, chromatic aberration (CA), also called chromatic distortion, color aberration, color fringing, or purple fringing, is a failure of a lens to focus all colors to the same point.[1]​[2]​ It is caused by dispersion: the refractive index of the lens elements varies with the wavelength of light. The refractive index of most transparent materials decreases with increasing wavelength.[3]​ Since the focal length of a lens depends on the refractive index, this variation in refractive index affects focusing.[4]​ Since the focal length of the lens varies with the color of the light different colors of light are brought to focus at different distances from the lens or with different levels of magnification.

Source: https://en.wikipedia.org/wiki/Chromatic_aberration

 

[Z.W. Wolf]

It's very simple.

- At any one moment only certain wavelengths of light are reaching the camera lens.

why does this cause the color of the whole object to change when it is zoomed in or unfocused, more than when it is properly imaged?

It doesn't.


Properly focused, Sirius does produce pure color images.


See:
https://www.metabunk.org/threads/un...tar-in-southern-sweden-2011.9224/#post-214671

 

[AmberRobot]

The atmospheric refraction that leads to twinkling of stars is wavelength dependent so the speckles of different colors can be different sizes and shapes, leading to a star appearing to change colors during scintillation. When you defocus a star and get a pupil image you're spreading the light over more pixels, which should make it less likely to saturate and be easier to record the dominant color.

See this paper for a good description of scintillation:

https://academic.oup.com/mnras/article/452/2/1707/1067913

 

[Maury Markowitz]

Actually, in retrospect, the potential answer appears simple:

When people use their phone and "zoom all the way in", the image processing is simply copying pixels.

So it's actually still a sub-pixel image and thus shifting colors, and when they apply the software zoom it's just copying that outwards?

 

[AmberRobot]

It appears the effect is "magnified" by being spread out, but based on how I understand the sensor to work, I would expect the opposite.

If a well focused star saturates the pixels then it will appear white. If you take a photograph of a sunset, the Sun will often appear yellow while the sky around it is red. This is not because the sky is redder than the sunlight, it's just that the Sun is so bright it is saturating the red and green sensors on the pixels, but not the blue.

So, spreading out the light across more pixels makes it less likely that any given pixel is saturated and thus easier to see the varying colors then letting the light all into a small number of (most likely) saturated pixels.

I don't know how the cameras in these videos specifically work, but most still cameras if set on auto-exposure will attempt to even out the entire frame to a medium gray. That means if you're taking a photograph of a bright object on a dark background it will likely overexpose the bright object. If I go outside and try to take a photo of the Moon with my camera on auto I get a washed out image of the Moon. I have to shorten the exposure (and/or modify the aperture and ISO) to get a well exposed image of the Moon.

 

[Mick West]

Scintilation causes the image of stars to move about on the sensor, and shifts across the color sub-pixels. But why does this cause the color of the whole object to change when it is zoomed in or unfocused, more than when it is properly imaged? For instance:

The comparisons you use are not really valid. You need the same star, same time, same camera, but with varying focus.

And scintillation is nothing to do with the color sub-pixels. The colors come from atmospheric refraction bending different colors different amounts, like a bunch of wobbly prisms.