Olympus Projection Eyepieces turn apos into achros (not clickbait)

[Macro_Cosmos]

What is going on here?
Am I doing something stupidly wrong or are these UIS-era projection eyepieces simply incompatible with modern UIS2 objectives?



Specification:
Nikon Z6, 24MP BSI, 35mm format, with OLPF.
Original photomicro L and OM>Z adaptor.
Universal condenser configured for Kohler illumination, superwide trinocular head.
100W halogen light source.
Target is a precision-made 400lp/mm resolution chart, coverslip adhered to the chart via matching RI liquid.
All post-processing was removed. No sharpening etc.

Virtually no CA with direct projection, all kinds of CA with a silly colour pattern when PEs are used. I tried to convince myself that I'm doing something stupidly wrong but I cannot think of an error.
Any advice on how to troubleshoot this? I will try and get more examples using different objectives.

Edit: So these were designed during the film days. A common issue with manual, vintage lenses is softness and aberrations due to modern sensor's window. The Z6 has a very thin one but with a microscope objective, even that piece of glass could be too much. Another reason is that these projection eyepieces are not actually apochromatic. Combined with a lowish magnification, the issue is exacerbated. I tried 20x and 60x, the eyepieces still bring chromatic aberration, but it was no where near as bad. What do you know? It was worse with a 4x objective.

[blekenbleu]

Virtually no CA with direct projection, all kinds of CA with a silly colour pattern when PEs are used.

Scarodactyl reported similar disappointment with Nikon PEs:
viewtopic.php?t=42625
Better results with direct projection and even afocal.

[rjlittlefield]

A common issue with manual, vintage lenses is softness and aberrations due to modern sensor's window. The Z6 has a very thin one but with a microscope objective, even that piece of glass could be too much.

That explanation doesn't work for me. The problem with thick window on the sensor appears at wide effective aperture, as seen by the sensor, when a wide cone of light has to traverse the window glass. Thick cover glass at the subject is a problem for objectives, for the same reason. But at the sensor objectives give a narrow cone of light, typically f/14 to f/20.

--Rik

[Scarodactyl]

That's a bummer. I thought it was really weird when I got those results on the Nikon PL eyepieces, and definitely would not have expected it to be that bad in the infinity era.
I wonder how it compares to a high quality teleconverter.

[Lou Jost]

"I wonder how it compares to a high quality teleconverter."

I use teleconverters as a matter of course, even with finite objectives, and good ones work very well.

[abednego1995]

How do they look when mag is normalized?

[Macro_Cosmos]

A common issue with manual, vintage lenses is softness and aberrations due to modern sensor's window. The Z6 has a very thin one but with a microscope objective, even that piece of glass could be too much.

That explanation doesn't work for me. The problem with thick window on the sensor appears at wide effective aperture, as seen by the sensor, when a wide cone of light has to traverse the window glass. Thick cover glass at the subject is a problem for objectives, for the same reason. But at the sensor objectives give a narrow cone of light, typically f/14 to f/20.

--Rik

Great point. This is consistent with my observations as well, those manual lenses get really good when stopped down just a tad.
I am dumbfounded and cannot come up with a good explanation.
- PEs were designed during the film era and the optics/formula/layout were not as good
This could work but it sounds too lazy to me.
I was told that Zeiss ones suffer from similar issues.
Any ideas? I do not have a UIS-era 10x objective, however, I know someone who does and I might be able to convince him to try it out.

How do they look when mag is normalized?

I did digitally zoom in. The XApo is the clear winner but the SApo is not any worse. There are faint hints of green LoCA. Such targets are challenging anyway, the CA surely will not present itself with normal imaging.

Here is the duo.

[Pau]

In Nikon scopes the projectives PL and PLI are dedicated for CF and CFI series respectively and according to the maker not interchangeable.
I think that we know the reason: CF series aren't really "Chroma Free" (full corrected in our jargon) but they need a small compensation done at the eyepiece, this is why their CFW eyepieces show a faint orange halo at its field limit (typical of compensating eyepieces).

I guess that PL and "CF photo" photoeyepieces of that era also will do

This is also shown at the graph posted by peter-h at
https://www.mikroskopie-forum.de/index. ... #msg166258
quoted by Scarodactyl at
https://www.microbehunter.com/microscop ... e64#p80756
and here previously discussed, for example at viewtopic.php?p=234051#p234051

Could be possible that Olympus UIS had the same approach (small non declared CDM compensation) not present in UIS2?
Could you try other brands' non compensating projective photoeyepieces like Nikon PLI and Zeiss P 2.5?
If you shot afocally through one of the visual eyepieces, do you see the same issue?

[Scarodactyl]

In Nikon scopes the projectives PL and PLI are dedicated for CF and CFI series respectively and according to the maker not interchangeable.
I think that we know the reason: CF series aren't really "Chroma Free" (full corrected in our jargon) but they need a small compensation done at the eyepiece, this is why their CFW eyepieces show a faint orange halo at its field limit (typical of compensating eyepieces).

Though this is appealing I don't think this is the correct explanation. My tests were done on nominally fully compatible parts and only direct projection yielded good results--certainly the pl eyepieces were not correcting anything in the bd plans' image. Nikon cfw and cfwn viewing eyepieces do have that faint ring but they also add some chroma, not by design but as a limitation. They also sold a much more expensive better-corrected eyepiece variant but they are rare since nobody bought them.
The quoted graph shows some chroma in the tested nikon objective but it's not an apo or even a fluorite so it is expected that some amount would be present.
The pl and pli eyepieces are quite cross-compatible incidentally, they went out of their way to make them so (the cfuw/cfuwn and cfiuws even have their notches in the same positions so new ones fit on old heads). The infinity 45mm parfocal cf and cfi60 lines still coexist and share parts. From my limited testing the pli eyepiece I tried seemed to give cleaner results with the finite objectives too--certainly the intermediate pl-2.5xa did though I am not sure what system it was designed for. I wish I still had all the parts in front of me and could run more tests, maybe I'll have a chance again in the future.

[viktor j nilsson]

They also sold a much more expensive better-corrected eyepiece variant but they are rare since nobody bought them.

I've managed to snag a pair of these 10x CFD eyepieces, would be very interesting to compare them in an afocal setup against the CF PL 2.5x.

[Macro_Cosmos]

Could you try other brands' non compensating projective photoeyepieces like Nikon PLI and Zeiss P 2.5?
If you shot afocally through one of the visual eyepieces, do you see the same issue?

Not the easiest to test since the projection eyepiece sits in the U-SPT, I doubt the other two fit. The Zeiss ones look big, like actual eyepieces.
No (visible) chromatic aberration if I shot through the 10x eyepiece (SWH10x-H 26.5mm FN), I will try directly through the port sometime next week.
More comparisons to come.

[Scarodactyl]

Yeah, the PE eyepieces are what, 25mm? You could fit the 23mm version of the pli 2.5x but I don't think they made a 23mm version of the 2x.
I guess you can always direct project onto a smaller sensor.

[abednego1995]

How do they look when mag is normalized?

I did digitally zoom in. The XApo is the clear winner but the SApo is not any worse. There are faint hints of green LoCA. Such targets are challenging anyway, the CA surely will not present itself with normal imaging.

I meant, how would they look if the apparent magnification is the same for all the images? e.g. adding 5x to the direct projection and comparing with the PE5x. And where are we looking at within the FN? I presume the comparison is somewhere in the middle, but if it's on the periphery I'm not so sure.

[Pau]

They also sold a much more expensive better-corrected eyepiece variant but they are rare since nobody bought them.

I've managed to snag a pair of these 10x CFD eyepieces, would be very interesting to compare them in an afocal setup against the CF PL 2.5x.

Do these CFD show a faint yellow-orange halo at the limit of the visual field?

If not, this could definitively invalidate my hypothesis

[viktor j nilsson]

They also sold a much more expensive better-corrected eyepiece variant but they are rare since nobody bought them.

I've managed to snag a pair of these 10x CFD eyepieces, would be very interesting to compare them in an afocal setup against the CF PL 2.5x.

Do these CFD show a faint yellow-orange halo at the limit of the visual field?

If not, this could definitively invalidate my hypothesis

Did a quick test. 10x 0.30 CFN plan fluor objective. Cover slip and oil on slide.







I think it's safe to say that Nikon intended the CFN objectives to be fully corrected. And that the orange fringe on the CFW eyepieces is not a design feature, but a sign of them being optically inferior.

Thanks for pushing me to do this, hadn't really realized how good the CFDs were.