Help with shifted color channels?

[Josh Lake]

A friend of mine has some great data on the Leo Trio, but when printed large, he noticed that stars had red and blue edges. Sure enough, it's in the raw data and I'm not really sure what to do about it.

First, what could the cause be? He's had other images with perfectly round, white stars. I thought it might be because of a tilted camera, but I was interested in other ideas too.

Second, can I fix it? Is there some clever morphological work I could do to fix the edging for this image? I guess I'd have to apply it to the red and blue and leave the green as is.

[georg.viehoever]

Is it from a refractor (as opposed to a reflector), or is any lens element in the optical path? If yes, I would guess it is uncorrected chromatic dispersion http://en.wikipedia.org/wiki/Dispersion_(optics) . Even the best lens systems cannot fully fix it (although it may be unnoticable in real good ones).

You may be able to remove it by aligning R,G,B using StarAlignment.

Georg

[Enzo De Bernardini]

Hi,

You can try ChannelMatch process (Geometry > ChannelMatch) to apply offsets to each channel. George solution it's a good idea too.

Best regards,

Enzo.

[Josh Lake]

Me again. I worked through another image and found the same issue: Channel Match works to correct the color on the bright stars, but the faint stars (which were good) turn bad.

Any idea what could cause this? Perhaps the red filter is tilted slightly, causing larger stars to have the shift but not smaller ones? Could that even happen? Thanks for any help, I'm lost on this one. I'd prefer to fix it at the hardware level than try to adjust it with PI later!