Coloured fringes round my stars

[almcl]

For a while I've been concerned about coloured fringes appearing round the stars in my lights. DSS seems to do quite a job of preserving these into the stacked image, too.

The accompanying image (which is cropped from the top left part of an unprocessed light) shows the effect which generally appears as a red fringe on the outer side of the star and a blue one on the inside.

I'm imaging with my 8" reflector, a coma corrector, an Astronomic Clip filter and an astro modded (but Baader filtered) Canon 700d.

Is there a specific name for this, what appears to me to be chromatic aberration and, perhaps more importantly, is there any way of mitigating it?

[Stub Mandrel]

Have you tried doing an RGB align? They all look spread the same way.

[D4N]

Is the spacing for your coma corrector right?

/Dan

Sent from my iPad using Tapatalk

[steppenwolf]

This does look like chromatic aberration but a reflector doesn't cause it. It could be the coma corrector but the colours are 'displaced' and this is the key to correction. If you open the image in an image processing program like PhotoShop and select the 'Channels' palette, you can shift the channels around. I moved the red to the right a couple of pixels and the blue to the left 1 pixel. To get a better idea of how this had worked, I adjusted the colour balance to 'equalise' the intensity of the three colours (RGB).

If you want more info. on the exact procedure in PhotoShop, let me know.

[jambouk]

Is the scope perfectly collimated?

James

[D4N]

From the op I assume this is one corner of a frame and that the blue is always to the middle. It may be possible to fix that by splitting the channels and scaling then separately (shrink the red and a large the blue) but it would be tricky. Better to try and find the cause and fix it.

/Dan

Sent from my iPad using Tapatalk

[laser_jock99]

I don't know what corrector you are using but budget coma correctors can cause chromatic aberration.

A well designed, high quality, ED glass coma corrector will likely cost as nearly much as a small ED refractor.

[almcl]

Thanks all for the speedy replies!

Stub: brilliant, many thanks - I've just tried restacking with align RGB and it's made a great difference.

D4n: Not sure if the SW coma corrector is at the correct distance, it's attached with the SW supplied official Canon connector (at vast expense) so I hope so, but is there a definitive test? (your 2nd post is spot on, too!)

Steppenwolf: Thanks for that demo, I only have Photoshop at work at the moment (since two days ago) but think I can figure out how to do that in my usual image processing kit at home.

Jambouk: I wouldn't say 'perfectly' collimated but I do check it each outing with either a Cheshire or a laser (yeah, I am sad enough to have both and it isn't far out.

And just to show how helpful this exchange has been here's a quick reprocess (whole image this time) using the info supplied (it's M33):

[jambouk]

I'd be over the moon with that image

[Leveye]

That has to be from the CC. I've never come across that issue with a reflector before. Steppenwolf has the brilliant fix!

[Stub Mandrel]

If RGB align fixed it, then I wonder if the coma corrector is at a slight angle?

Nice pic by the way!

[almcl]

I wonder if the coma corrector is at a slight angle?

That's a thought, Stub.

There's provision for adjusting the angle of the focuser on the SW 200, although I've never thought to try it.

Wonder how you measure the inclination...?

I'll have to do some research. Wonder if one of the Astronomy Shed videos covers this?

[glowingturnip]

could be atmospheric refraction if it was low to the horizon ?  different colours get refracted at slightly different angles through the atmosphere and can cause fringing like that.

+1 for the RGB align or steppenwolf's method.  I routinely separate my channels in pixinsight, star-align them to each other and then recombine.

[ollypenrice]

Al,

I'm intrigued by whether or not each corner is the same or whether the colour shift displaces as you go round each corner. I'm inclined to suppose that it does? I can't think of any obvious way that an optical issue could shift one colour right across the chip in the same direction.

So the top left that you posted has red displaced to the left. What about the bottom right? I'll guess that red is displaced to the right? On a one shot colour chip it can't be software channel alignment so it has to be optical in origin. 

It would be interesting to take the linear image, split it in into three colour channels, take those channels into Registar and register red and blue onto green. When recombined R and B would be resized and recurved to sit on top of green. If you'd like to try this just send me the linear RGB and I'll run it through Registar and recombine.

Olly

[almcl]

Most kind, Olly, thanks.

Have sent a pm with a link to a couple of unprocessed subs.

The fringes do seem to 'change side' but not quite symmetrically. I wonder if the camera's ability to record the Ha part of the spectrum (it is one of Juan's Astro modded ones with the Baader filter added) is part fo the cause?