Some reversed lenses should be shot through thick glass

[Lou Jost]

I've written before about the need to match a lens with the sensor stack thickness that it is designed for, based mostly on this link which contains actual tests:

https://www.lensrentals.com/blog/2014/0 ... ed-lenses/

Leica digital cameras have a very thin glass filter in front of their sensors, while MFT cameras by Panasonic and Olympus have a thick 4mm glass filter in front of their sensors. Their lenses are designed with this glass element in mind. Film-era lenses and enlarger lenses are designed for no glass at all in front of the sensor (film). Mismatches will have the worst effect when the lens has a large (fast) aperture. So using a fast film-era lens wide open on an MFT camera might not be ideal.

As I've been thinking about how to achieve optimum diffraction-limited results with my MFT cameras, I've just realized that this must also be a factor when a lens is reversed. If an MFT lens is reversed and used wide open, there needs to be a thick glass filter (4 mm thick !!) between it and the subject for best results. And on the camera side, we're stuck with sub-optimal results when reversing an MFT lens on an MFT camera because on the designed-for subject-side the lens does not expect the 4mm thick filter that sits on the sensor.

If the lens is image-side telecentric, though, as the early four-thirds (not micro four thirds) lenses were, the rays are mostly parallel to the camera axis, so maybe the presence of a flat glass element is unimportant for such lenses.

Coupled lenses (a reversed lens mounted on a normally-oriented lens) can fix some of those problems. We could use a body-appropriate lens on the camera, and a film-era lens or possibly an image-side telecentric lens as the reversed lens. Alternatively we could add a thick filter stack between the reversed lens and the subject.

I think we have have been overlooking this issue (I cannot recall anyone ever suggesting that we need a "cover slip" in front of our reversed digital lenses, though we are all aware of this need for microscope lenses that expect glass in the optical path between subject and lens) and it might improve our photography if we are aware of it. Particularly for fast lenses used in reverse. Maybe later today I will have time to test this.

Anyone else who wants to test this is welcome to add their pictures to this thread.

[Lou Jost]

Here is some excellent follow-up information. Note especially the graphs which show that a 2mm glass sensor filter can be enough to destroy the corner resolution of a fast lens that expects no filter.

https://www.lensrentals.com/blog/2014/0 ... it-matter/

[ChrisR]

Somewhere -?- I've read of recent sensors having slimmer stacks for this and other reasons.

Schneider's "digital" lenses and image-side telecentricity are also associated in my head somewhere.

[Lou Jost]

Both those solutions would certainly help! If lens-makers are thinking about problem, we should too, at least when we are using fast lenses. Probably makes no difference for slow lenses, just as the presence or absence of a cover slip is not important for low-NA microscope objectives.

[Lou Jost]

I found a Part 3 to the series of posts above:

https://petapixel.com/2014/07/08/sensor ... i-summary/

Here is the bottom line, taken from that page:

"We should expect the following to give problems:

1. Using a lens designed for film on an SLR would give a 1.5 to 2.5mm stack difference, and we should notice a performance drop-off when using wide-aperture lenses with shorter exit pupil distances.
2. Using a lens designed for an SLR on a micro 4/3 camera would give a 2mm stack difference, and we may notice a performance drop-off on wide-aperture lenses with shorter exit pupil distances.
3. Using a lens designed for film on a micro 4/3 camera would give a 4mm difference, and if the other factors (exit pupil distance and wide aperture) are present we will almost certainly notice a performance drop-off.
On the other hand, using a Nikon lens on a Canon camera, or either of those on an NEX or Fuji camera shouldn’t give major problems since all of those sensor stacks are similar."

[zzffnn]

So using Schneider Componon S 50mm F/2.8 reversed and extended on micro 4/3 camera (Olympus E-PM2), I am getting 4mm stack difference.

I remember ChrisR kindly told me that I should use around F/5.6 in that case. That aperture is not top wide, I assume? Would F/2.8 be too wide to see a difference?

Will thin and thick subjects be equally affected? How about high contrast subjects vs low contrast subjects? I do not have macro experience to judge experiments myself, so I would rather rely on members' guestimates here.

I read the articles very quickly, if I understood them correctly, even reversing a m4/3 lens on m4/3 may produce a difference since sensor side of the lens is expecting 4mm, while subject side of the lens is expecting none.

[Lou Jost]

I don't know, but judging from the graphs in the articles, you should be safe at 5.6, especially considering that you are probably using it on extension so the effective aperture is even smaller.

However these graphs apply only to the center. It is possible that you could have corner problems due to the unexpected 4mm glass slab in front of the sensor.

Yes, reversing an MFT lens directly on an MFT camera (rather than mounting it reversed on a proper MFT lens) could be a worst case scenario. Now there is an unexpected 4mm slab of glass between the front element and the sensor, and an expected 4mm slab of glass is missing between the rear element and the subject.

But I should stress that we don't really know yet at what aperture these effects become important in real life.

[Pau]

This could be good for using reversed 4/3 lenses as macro lenses to shot through the glass of an aquarium or for amber included insects...

[ChrisR]

So using Schneider Componon S 50mm F/2.8 reversed and extended on micro 4/3 camera (Olympus E-PM2), I am getting 4mm stack difference.

I remember ChrisR kindly told me that I should use around F/5.6 in that case.

That was for the SC 50 generally, I didn't think anything about sensor pack thickness!


Lou if you look at Schneider's telecentric lens site, it has some pictures of doubly telecentric beasties. If have two 5mm Rodagon D's I always fancied trying that with..

If you look at Bjorn Rorslett (sorry, accents missed off) site, he makes a distinction between lenses which behave well on digital vs film. The slab thickness could be a factor. Some of his comments don't work for me and others , so - pinches of salt here and there.

[Lou Jost]

I wonder, if you couple two image-side-telecentric lenses front-to-front, and set them both to the same aperture, would that make it telecentric on both sides? Intuitively it seems like it must. Maybe Rik will check that intuition and confirm or throw cold water on it.

I have two identical allegedly image-side-telecentric Oly 50mm macro lenses I bought for coupling, I may try that if I have time....

[ChrisR]

You need a central aperture - see
https://www.schneideroptics.com/ecommer ... x?CID=1439


They'll sell you one for $4500.
11mm image circle

[Lou Jost]

Yes, I see that now. If you look into the coupled lenses from one side, the limiting aperture will probably be the one that is farther away rather than the nearer one that would make it telecentric if it were alone.

[Lou Jost]

I've just tested a fast coupled MFT lens pair at 3x with EA= approx 1.4, shooting a butterfly wing. I used a 75mm Oly f1.8 facing forward, and a 25mm Pana-Leica f1.4 reversed, both wide open. There was definitely an improvement when I shot through two filters versus when I shot with no filters. But I have not controlled the lighting well enough between the tests-- the filter rings cast a lot of shade. So I will try using transmitted light and I'll pile 4mm of cover slips on the wing.

[rjlittlefield]

if you couple two image-side-telecentric lenses front-to-front, and set them both to the same aperture, would that make it telecentric on both sides? Intuitively it seems like it must. Maybe Rik will check that intuition and confirm or throw cold water on it.

Sorry, it's cold water. As ChrisR says, you need a central aperture. If you have two separate apertures, then either a) one aperture limits and the other doesn't, in which case one side of the combo loses its telecentricity, or b) both apertures limit, but on opposite sides of the ray fans, in which case you lose telecentricity on both sides, and pick up some vignetting at the same time.

--Rik

[Lou Jost]

Yep, that makes sense. Oh well.

So a central aperture it is then. I think I'll save the $4500 and make one out of paper.