Viewfinder effects under 10x setup

[hero]

So, I noticed something quite odd about my setup, which uses the Raynox 150 reversed at infinity focus and a Mitutoyo 10x. I'm using a full frame body.
When I look through the viewfinder, I can see the edge of the reflex mirror at the top edge of the viewfinder. I can even see part of the sides of the mirror; specifically, the tab that holds the mirror to the reflex assembly. However, when I actually take the photo, there's no issue; the image circle covers the full sensor and there's no vignetting. I don't see this with my regular lenses.

Any explanation for this phenomenon?

[rjlittlefield]

Any explanation for this phenomenon?

The exit pupil of the objective + Raynox is far forward, maybe something like 250 mm from the sensor. This causes light rays that are headed toward the edges of the sensor to strike the plane of the mirror quite far out. The ones that are farthest out apparently miss the polished mirror entirely, producing a shadow or strange reflections in the viewfinder. When you take a picture or use Live View, the mirror flips up so that far-out rays can make it all the way to the sensor.

Your other lenses don't have this problem because their exit pupils are much closer to the sensor, so edge rays strike the mirror plane farther toward the center. If you had a long prime that was not telephoto design, it would have the same behavior as the Raynox.

Long ago, a couple of my film SLRs had a version of this problem that appeared even in the captured pictures. The problem there was that the mirror stop protruded into the film chamber far enough that with long bellows setups it would make a shadow on the film. No problem with short setups or telephotos, only with long extensions.

--Rik

[ChrisR]

Just being curious, what's the camera, Hero?

[enricosavazzi]

So, I noticed something quite odd about my setup, which uses the Raynox 150 reversed at infinity focus and a Mitutoyo 10x. I'm using a full frame body.
When I look through the viewfinder, I can see the edge of the reflex mirror at the top edge of the viewfinder. I can even see part of the sides of the mirror; specifically, the tab that holds the mirror to the reflex assembly. However, when I actually take the photo, there's no issue; the image circle covers the full sensor and there's no vignetting. I don't see this with my regular lenses.

Any explanation for this phenomenon?

This is a known problem, especially with film SLRs, when using very long lenses, photomicrography lenses and similar setups with a lot of distance between optics and film plane. These cameras are built with somewhat smaller mirrors than necessary to cover the whole viewfinder area, so with these optics you see one edge of the mirror, usually the lower edge (seen inverted at the top of the viewfinder of course). With lenses placed closer to the film plane, there is little or no noticeable vignetting by the edge of the mirror.

This only affects the viewfinder image, not the recorded image. Once the mirror lifts up before the exposure, the problem disappears.

I used to have this problem with Olympus OM-2n and other OM cameras.

[Pau]

I used to have this problem with Olympus OM-2n and other OM cameras.

The OM 2n has a relatively large mirror. Things were worse with the first camera I hooked to a microscope: the Yashica TL Electro X ITS, a pretty advanced camera for its time...I'm getting old

[hero]

Just being curious, what's the camera, Hero?

I am using the Canon EOS 5D Mark III. Not the ideal body for this use case, but I'm still feeling things out with the Mitutoyo.

[Alan Wood]

Hero

If you tether your Canon EOS 5D Mark III to a computer using the free EOS Utility software, the problem will disappear because the image from the sensor appears on your computer screen.

Alan Wood

[hero]

Any explanation for this phenomenon?

The exit pupil of the objective + Raynox is far forward, maybe something like 250 mm from the sensor. This causes light rays that are headed toward the edges of the sensor to strike the plane of the mirror quite far out. The ones that are farthest out apparently miss the polished mirror entirely, producing a shadow or strange reflections in the viewfinder. When you take a picture or use Live View, the mirror flips up so that far-out rays can make it all the way to the sensor.

Your other lenses don't have this problem because their exit pupils are much closer to the sensor, so edge rays strike the mirror plane farther toward the center. If you had a long prime that was not telephoto design, it would have the same behavior as the Raynox.

Long ago, a couple of my film SLRs had a version of this problem that appeared even in the captured pictures. The problem there was that the mirror stop protruded into the film chamber far enough that with long bellows setups it would make a shadow on the film. No problem with short setups or telephotos, only with long extensions.

--Rik

I had to sketch a diagram to think about the geometry of what you've described.... and I must say, while I believe I understand it now, this is not at all intuitive!

As the distance of the exit pupil to the image plane is proportional to the effective f-number of the system, for this latter quantity to be fixed, a distant exit pupil would have a proportionally larger diameter than a nearby pupil. So my initial question was, "why then do I not see this phenomenon if I stop down a 'regular' lens?" Then I realized two things:

First, that the mirror is not situated at the image plane; moreover, it is the fact that the mirror is diagonally situated relative to the sensor that explains why only the bottom edge of the mirror is observed, because it is this edge that is farthest from the sensor.

Second, although an ant (or microbe) on the sensor would see the same apparent exit pupil (unable to distinguish a large distant pupil from a small nearby pupil) for a given effective f-number, it is not generally the case that the position of these exit pupils are the same to the ant. If the ant is at the center, then yes. But if the ant walks toward the edge of the sensor, the pupil's position will look different in these two cases (never mind the change in shape). The closer pupil is more oblique relative to the ant's position. In other words, if we were to project the distant pupil onto the plane of the closer pupil, it will be off axis. If far enough away, even at the same effective f-number, that distant pupil's rays could miss the bottom edge of the mirror because the apparent location of the pupil at the edge of the image plane still appears to be mostly directly overhead, and the mirror is not actually sqrt(2)*24 mm x 36 mm in size.

Finally, putting this all together, staring at my crude diagram, and reading your explanation a few more times, it makes sense! I think Canon needs to qualify their viewfinder coverage percentages in their specifications and marketing materials

[rjlittlefield]

while I believe I understand it now, this is not at all intuitive!

I agree in both regards: the problem is not at all intuitive, and your explanation shows you now understand it completely.

I like your analogy of how an ant on the sensor sees the pupil.

--Rik

[ChrisR]

the Yashica TL Electro X ITS, a pretty advanced camera for its time...I'm getting old

Surely we all remember the image-clippig trapezoidal mirror on the Zenit-B ?
That Yashica was about the first to do away with swinging needles in the viewfinder, and use LED's, yes? - rocket science! All those letters in the name made it sound like a new Ford of the era.