[Brightcolours]

Makes me want to try this now, although a bit late at the moment! Thinking more I may have been a bit quick to blame a "bad lens", although in the use I've had so far from my 50 Art I hadn't come across anything like that. The nearest is some effect in the corners, but I've not tried with high contrast (very bright highlights) type shots so that might expose weaknesses I haven't seen yet.

 

I probably should get more sleep...

 

Edit: added the test shots I did earlier. No diamonds here, but as said these are lower contrast. And relative defocus may play a part in it?

What exactly causes the cut off in the shapes at top/bottom near the top/bottom edge? 

In one of the images you see the bottom being cut off when the highlights are near the bottom and the top being cut off at the top of the image, and the other image shows the top being cut off near the bottom of the image and the bottom being cut off near the top of the image. 

[popo]

I think you mixed up the image name - the greenish ones show the background highlights.

The purple ones the front highlights.

 

We might be using front and back with different references here. I was describing the focal point relative to the target. e.g. I focused in front or behind the target. I guess that would be opposite to describing the target relative to focal point e.g. the target is in front or behind the focal point.

 

 

What exactly causes the cut off in the shapes at top/bottom near the top/bottom edge? 

In one of the images you see the bottom being cut off when the highlights are near the bottom and the top being cut off at the top of the image, and the other image shows the top being cut off near the bottom of the image and the bottom being cut off near the top of the image. 

 

I think I mentioned it already. To me it looks like the rectangular area of the 5D2 mirror box doesn't allow all the real lens element to reach the sensor. The cutoff may depend on if it is focused front or back.

[joze]

popo / just to understand you right: you think that the mirror box vignettes front or back-focused spots? Here's a sketch of what I understand:

1. The lens is focused to on object O1 such that we get a sharp image on the sensor (blue rays)
   
2. The object O2 (red rays) is out-of focus, in fact it's in front of the focus
   
3. The image I2 of O2 lies then behind the sensor
   
4. The top rim rays get vignetted by the mirror box
   

Is this what you believe?

 

If this theory is true, I see the following consequences:

• out-of-focus objects which are BEHIND the focused plane don't get vignetted --> true, at least to my test images which would confirm your theory
  
• highlights in front of the focus would always generate this ugly diamond shape for a given body / aperture rendering the body / aperture combination pretty much inappropriate for this use case (if you like roundish bokeh). But maybe most of the real world use cases are such that you've the highlights behind the sharp object anyway. I'd consider that a major design flaw of the body.
  
• what seems strange to me is that JoJu's body (Nikon) shows the same behaviour as mine (Canon). If it was the body I'd have expected to see one body having the problem and the other not (unless the mirror box is built conformant to some standard).
  

I've to say that this is the most plausible explanation I've seen here.

 



 

If there's an asymmetriy of this kind of mirror box vignetting -- eg. top field points are vignetted differently from bottom field points -- the hypothesis could be easily proven by turning the camera over the top and reapeating the shot.

[joze]

but wait / popo / then the diamonds shouldn't rotate with the azimuthal position in the field, should they?

[JJ_SO]

Meanwhile I'd like to see the drops while they are flying. In slow motion. Pity I don't have a high speed movie cam.

 

I don't see the reason in a mirror box, if that was close to possible the rhomboid shapes would be in the upper or lower border - not in the sides where they concentrate in the corners. Plus, if it was independent from movement, you'd see the same shapes when you photograph a water surface with sunlight on it.

 

Call me from the Jupiter moon after rethought  

[Brightcolours]

but wait / popo / then the diamonds shouldn't rotate with the azimuthal position in the field, should they?

That is an entirely separate issue. The mirrorbox cutting off "bokeh highlight disks" can be seen in popo's posted attachments. They have nothing to do with the "diamond" shaped astigmatism like spots in the other photo(s).

[Scythels]

>Onion ring bokeh is just a sign that an aspherical element is not totally smooth

 

That's the jist of it, but the lack of smoothness results in hot and cold spots across the lens' surface.  Which is a lack of acuity.  Which is an alteration of the point spread function from ideal... in a concentric manner...which is sagittal... which is astigmatism.

 

It's less "not totally smooth" and more "not ground finely enough - visible peaks at detectable scales." 

 

>Could you please expand on this thesis

 

There are several issues which will alter the PSF of the lens.  Astigmatism will alter it - if tangential resolution is lower than sagittal you will have spikes that form concentric 'wings' - if sagittal is lower, you get spikes that point towards the optical axis.

 

Mechanical vignetting will alter it outside the center of the lens by shaping it to reflect the apparent iris.

 

Pinched optics will alter it by causing astigmatism-esque changes to the PSF, but the changes will not be symmetric.  Standard QC results in lenses being centered, so it is highly unlikely that an element was not centered before it even became coated, let alone be allowed into a produced lens.  When a lens is pinched, the stress alters its centering (please note - centering in this case is slightly different to decentering for a photographic lens, which is a system of lenses.  In this case, it means the center of curvature matches the physical center, whereas for an optical system it's a matter of alignment) which will produce asymetrical astigmatism.

 

----

 

> And pinched optics?? Really? No.

 

You're welcome to guess what the problem is, just as I am.  But please try not to be so dismissive of others - not everyone has the same knowledge base, and it's possible that your guess is wrong   Then again, mine could be as well.

[Brightcolours]

Well, I can hardly be supportive of such an odd proposition. It is just nonsensical 

 

Now, I do make mistakes myself (switching terms). In above posts where I typed astigmatism, I meant coma.

[Scythels]

Brightcolours, if I corrected every error you made in posts, I'd have a second full time job.  Be wary of hubris.

[Brightcolours]

Well, I would have appreciated if you would have pointed out to me that I was switching terms. Would have been less confusing to everyone.