What's up with my Canon 200 mm f2.8 USM

Looks like some spikey eccentricity and some chromatic aberration. Is this the middle, or a corner? Which of the two cameras mentioned took this pic?  With the ZWO camera, how to you attach the lens to the camera?  With the (budget) ZWO adapter?

What was the ambient temperature?  Maybe pinching of the lens elements from the cold?

Maybe an issue with the coatings, as in coating is failing. You have 2 PSF for each source one fainter than the other, apparently. The only way to get a grip on this is to study an out-of-focus image, inside and outside focus to try and figure out what the issue is. Most likely isn' t fixable and you'd need a new lens.

How do you hold the lens to the tripod? I have seen similar effect when tightening too much the collar ring around the lens. I even see it in my 127mm refractor when the rigs are too tight.

The lens is too heavy to be held by the camera body, so the EOS adapter may be flexing

Have not used mine in a while but have never seen such a thing.  Here is one taken with that lens and an old SBIG STT 8300. I did have a lens support as well as using an F5 aperture stop instead of the iris. I would never use it at 2.8 - too much edge distortion. You might make a simple aperture stop and try that to see what effect it has.

https://www.astrobin.com/9pz60c/

Mark, I'd thought of that but I can't find anyone near me (Detroit) who offers the service and the Canon website is of absolutely no help!! Would be interested how you find someone who can do such a service.

John Noble:
Mark, I'd thought of that but I can't find anyone near me (Detroit) who offers the service and the Canon website is of absolutely no help!! Would be interested how you find someone who can do such a service.

I have sent a lens to canon for repair and I went through the vendor I purchased the lens from and in my case it was Adorama.  This was a while ago so things may have changed.  Your canon warranty card should have information however on how to get in touch directly with canon.

I've bought and sold stuff to KEH and I know they do lens repair, I had them rebuild a Canonet film camera ages ago for me.

The only Canon lens that is a flat field with practically zero color distortion is the 100mm f2.8L Macro.  But you should be getting better performance from your 200mm, something clearly is not right.  Hope this helps.

.  https://www.keh.com

Steven Bellavia:
That lens has a very flat field for a 4/3 sensor.  At least my copy does.

These are images taken with that lens and 294MM camera:

https://www.astrobin.com/cibuek/

https://www.astrobin.com/r9f979/

https://www.astrobin.com/wz7vsn/

I am happy to share the raw data.

For narrowband I stop down to f/3.7
For L-RGB I stop down to f/4.5
In both cases, I use thread-in step-down rings

Oh sure I agree with you I was speaking for full frame.  If you have a 4/3 or APS-C you miss/avoid a lot of it. I am not disparaging the lens really because all camera lenses behave this way particularly at different stops and the 200mm L is a particularly good lens that is mostly flat, I used to own it.

People have in their minds that all lenses should be flat field and crispy sharp across all focal lengths to the very edge of the FOV and at all f-stops and that is not really the number 1 and 2 consideration of every lens design.  We all know the famous quote by Henri Cartier-Bresson (1908-2004) - “Sharpness is a bourgeois concept”. There are some Leica lenses that are highly sought after because of their distortion. But I digress.

Check this out and you'll see what I mean.  I put the 50mmL in there for comparison.   It's not that the 50 is a bad lens, it is a beautiful lens the draws your attention to the center of the frame with it's vignetting distortion so to speak.
https://www.imaging-resource.com/lenses/canon/ef-200mm-f2.8l-ii-usm/review/
https://www.imaging-resource.com/lenses/canon/ef-100mm-f2.8l-macro-is-usm/review/

https://www.imaging-resource.com/lenses/canon/ef-50mm-f1.2l-usm/review/

 The Zeiss Jenna Biotar has a particular type of distortion that makes it highly desirable for some photographers. It is a circular bokeh distortion.
http://forum.mflenses.com/userpix/20152/big_5_carl_zeiss_jena_biotar_75mm_f1_5_exakta_nex715_1.jpg
 
I've worn this out by now probably.

Lens for AP should work fully open with resonable un-aberrated (distortion doesn't play a part here as can be corrected during processing so we're talking chromatic aberrations, coma and astigmatism (3rd and 5th order)) field across a APS-C sensor. If they don't there is no point in using them, I believe (just get a slower lens, chepaer and lighter too).

Graeme Coates:
I couldn't focus mine adequately with 3.75um pixels (QHY163M) so had to make a belt/pulley based stepper system to automate focussing - the critical focus zone at f4 is 40 microns for 3.75um pixels, but nearer 160 microns for 9um pixels. If an inspection of the lens looks OK, I would suspect focus issues first of all.

A f/4 lens (let's assume a perfect one) has an Airy disk diameter of ~5.4 um. If we use 90% of encircled energy criterion then is more like 6.5 um. Real lens are hardly diffraction limited so in practice it is going to be slightly bigger (or a lot if the lens is soft) than that. This means that, at 2 pixels per sampling (3.76 um per pixel), your depth of focus (depth of focus = ~2*f/*2*3.76um) = 60 um or 0.06 mm. I'd put it at as mere minimum requirement but in a real f/4 lens it is likely that you can accomodate a 50% larger focus margin and still be on the sharp side of focus (for a real lens). On my Nikkor f/2.8 I can routinely achieve that sort of precision with the manual focus ring.

The diffraction like abberation is probably some pinching, the axial abberation causing the blue/green to focus at different point than red is normal for that lens unless you stop down to F/4 or F/5 depending on your sensor size.