Imaging at 5X -- a tale of two objectives

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: Pau, rjlittlefield, ChrisR, Chris S.

rjlittlefield
Site Admin
Posts: 21125
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Imaging at 5X -- a tale of two objectives

Post by rjlittlefield »

Here are some lens tests you may find interesting.

First, the optics.

On the left, a Nikon CFI60 Plan 10X NA 0.25 objective, mounted on a Sigma 105 mm f/2.8 macro lens focused at infinity.

On the right, A Nikon CF BD Plan 5X NA 0.13 objective, mounted on a old Vivitar 200 mm f/3.5 telephoto lens.

Image

Now, ignoring potential differences between the 200 mm telephoto and a real Nikon tube lens, it's apparent that the 5X BD Plan is being used "as designed" at 5X, while the CFI60 10X is being dragged way outside its design space, operating at measured 5.15X on a nominal 105 mm "telephoto" that isn't even really a telephoto.

Comparing these two lenses, the results should be a foregone conclusion. The 5X lens will probably have less central resolution because of its smaller aperture, but should be pretty uniform, while the unfortunate 10X lens that's being forced to cover twice its usual field may have a really nice center, but its corners are going to be junk.

And indeed, when we look at the images, the corners for one lens are notably deficient, while the corners for the other lens look quite good. The following are actual-pixel crops from a Canon T1i, 15.1 megapixels on a 22.3 x 14.9 mm sensor.

Image

But here's the surprise: the good corners come from the CFI60 10X dragged down to 5.15X on the 105 mm macro.

Here are the corners at actual-pixels with proper order and labeling. Those are followed by the centers at actual-pixels, then a reduced full frame showing the crop locations. Zerene Stacker PMax, no post-processing except laying out for posting.

Image

Image

Image

As mentioned, this is a surprising result.

The obvious answer, of course, must be that the 200 mm Vivitar is just a duff lens.

Well, that's hard to test rigorously, especially since it's after dark right now and I can't just point the thing at some target a long ways off. Come to think of it, that wouldn't necessarily be a good test anyway, because maybe the thing would work OK by itself but not play nicely with an objective.

So how about we just put the CFI60 10X on the 200 mm Vivitar and see how that looks?

Here's the answer, corner and center, comparing CFI60 10X with BD 5X, both on the 200 mm Vivitar. In this set, the CFI60 is with a different part of the wing (different day, after some other tests). But I think the result is clear enough. The centers have similar sharpness, but with the CFI60 10X there's not much difference from center to corner, while with the BD Plan 5X the corner is notably deficient.

Image

Conclusion: the problem is in the BD objective, not the telephoto. I suppose I could nail this down even farther by putting the BD on a different telephoto too, but really, enough is enough!

No doubt different people will see different things in this exercise, but here's my set of take-aways:

1. The CFI60 Plan 10X NA 0.25 objective also makes a very good 5.15X objective, when paired with a 105 mm rear lens on a 1.6 crop-factor sensor. There's detail at the pixel level, corner to corner at 15 megapixels.

2. At least this particular CF BD Plan 5X NA 0.13 will make a fine addition to my collection of unused lenses. Is it a duff sample? I can't tell, except to say that I can't see evidence of abuse or disassembly. But it's pretty safe to say that this model of lens is not going to appear on my personal short list of lenses to recommend to other people. This is annoying because I bought the 5X objective expecting it to work well.

3. There's a big difference between these two objectives, despite that they're both Nikon CF infinity's. The newer design CFI60 objective beats the socks off the older BD Plan. Would this hold up with more samples and other lenses from the same lines? Can't tell, more testing required.

4. Tests beat preconceptions, sometimes quite badly.

Hope you find this interesting.

--Rik

RogelioMoreno
Posts: 2962
Joined: Fri Nov 20, 2009 11:24 am
Location: Panama

Post by RogelioMoreno »

Rik,

Interesting test!

Thank you.

Rogelio

georgetsmurf
Posts: 46
Joined: Tue Nov 30, 2010 5:55 pm

Re: Imaging at 5X -- a tale of two objectives

Post by georgetsmurf »

It would seem to me that the BD lens is a dud, maybe there is a loose element in there or something. Why not do the same comparative test using only the microscope objectives ie not attached to the camera lenses. In my macro work I have always found that adding other bits of glass-no matter how good the quality, to otherwise OK lenses is never a good option. Then again I suppose you could also compare-test just the camera lenses. :)

ChrisR
Site Admin
Posts: 8564
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

George these are infinite objectives so they're designed to be used with a tube lens, in a microscope, which would have a focal length of 200mm.
Without the tube lens, they're not great.
The objectives have different mounting threads, neither is rms, so a test on a macro photographer's microscope wouldn't be completely straightforward!

I'm wondering if there's something about the tube lens that's different in the microscope which that BD lens is meant for, but I have no idea what.
I happen to have much the same set of glassware so in theory, one day, should be able to make a similar comparison. Time and space may prevent that happening, and a general lack of equipment/competence/skill/knowledge would probably fuzzy the result. But I'll try.

Joaquim F.
Posts: 189
Joined: Wed Apr 28, 2010 3:06 pm
Location: Tarragona, Spain
Contact:

Post by Joaquim F. »

Rik, very good test and with a dose of suspense! :)
Totally agree with your results.
The Micro-Nikkor AF 105/2.8 D behaves in the same way with the CFI 10X.
The rms mount CF Plan 5X/0.13 is not good with any tube lenses I've tried. Is usable only in the center of the image and up to about 20mm of coverage, beyond is a real "rainbow" :(
I plan use it like a raynox 202 (similar magnification and borders quality with my lens) for little bugs placed in the image center... and for nothing more.

Greetings

georgetsmurf
Posts: 46
Joined: Tue Nov 30, 2010 5:55 pm

Post by georgetsmurf »

Chris said
George these are infinite objectives so they're designed to be used with a tube lens, in a microscope, which would have a focal length of 200mm.
Without the tube lens, they're not great.
Chris. Do you think maybe the specific tube-lens at 200mm focal length maybe necessary to correct what looks like resolution problem? Just a thought. It seems unusual for Nikon to produce something this dodgy otherwise.

Chris S.
Site Admin
Posts: 3549
Joined: Sun Apr 05, 2009 9:55 pm
Location: Ohio, USA

Post by Chris S. »

Rik, just to be sure, you did block the built-in light path that surrounds the BD lens, right? I hesitate to bring this up, as you almost surely thought of it.

For anybody not familiar with BD lenses (I myself have not had one in hand, but understand them from diagrams like this--scroll down to figure five), they have a sleeve surrounding the optical portion of the lens, and a reflecting lens aimed at the subject. While this was intended for providing subject illumination in the lens' original application, if left unblocked in a macro configuration such as used here, it could very effectively direct light onto the front of the telephoto lens, which might result in substantial flare. Probably not the case here, as Rik likely blocked this light path. But if not, I could easily see it producing the type of poor performance demonstrated.

Cheers,

--Chris

Charles Krebs
Posts: 5859
Joined: Tue Aug 01, 2006 8:02 pm
Location: Issaquah, WA USA
Contact:

Post by Charles Krebs »

georgetsmurf,
It seems unusual for Nikon to produce something this dodgy otherwise.
Keep in mind that we are using this in a manner never intended by Nikon. This 5X is at least one "series" older than the current version, and was meant for Optiphot microscopes that typically had a had 20mm FN (field number), but could be used with "ultra-widefield" 25mm (FN) tubes. So at the very most Nikon needed a "quality" 25mm image circle. With an APS sized sensor we are already larger than that. (And there is the "unknown" of what would occur with an official Nikon tube lens... my suspicion... purely a guess... is that it probably would not change these results dramatically).

It would be sort of like complaining that a view camera lens intended for the 4x5" format produced poor edge quality when used on an 8x10" camera!

So in a sense we're looking for "over-achievers" :wink:, objectives that perform well at image circles larger than was ever needed for their intended use. Some seem to be capable of doing so, others do not. Other than trying each individually, there does not seem to be a universal answer, even in the same vintage and line of objectives.

ChrisR
Site Admin
Posts: 8564
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

The telephoto make/quality doesn't seem to be too critical, judging by what I've seen and read. That inclines be to believe there's nothing very special about tube lenses.
I've taken a number of pics with various infinite objectives on a couple of telephotos, and must say that the "poor" objectives are capable of producing pictures which would please a lot of people a lot of the time. Mug shots of bugs are a lot less critical than wing pics with fine detail all over them. Perhaps used in a microscope with an eyeball, the differences aren't noticeable?

Accepting Charlie's point entirely about image circle, I do wonder how quickly the image degrades, away from the center. With some objectives (planapo finite 4x 0.2 comes to mind) the resolution dramatically falls off the proverbial cliff at some radius, whereas others are more gradual.

What bothers me is that with contrived subjects (like my black dots on white paper), CA starts to show much sooner (smaller radius) than I would have expected, in some objectives. I'm reluctant to condemn the thing though, until I see/learn more.
Just as a an illustration, for example, I no know nothing about microscopes. Where there's a tube lens, is there any procedure where the distance between the tube lens and the eyepiece gets adjusted? That would be to ask if they're used in a not-quite-infinite set up, to improve an image? I have no idea!
Too many unknown unknowns!

Charles Krebs
Posts: 5859
Joined: Tue Aug 01, 2006 8:02 pm
Location: Issaquah, WA USA
Contact:

Post by Charles Krebs »

ChrisR,
Where there's a tube lens, is there any procedure where the distance between the tube lens and the eyepiece gets adjusted?
No, none that I've seen. The tube lens is generally in the binocular (or trinocular) head itself at the point where it attaches to the stand. One eyepiece collar is "fixed", and that determines the "focus" of the tube lens. The second eyepiece collar does have some adjustment, but that is there to allow eye "matching" so that both eyes are in focus at the same time.

ChrisR
Site Admin
Posts: 8564
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

What about tube lens to objective distance?
If I understand the theory there's no limit how far that could be. But that feels wrong :?

Charles Krebs
Posts: 5859
Joined: Tue Aug 01, 2006 8:02 pm
Location: Issaquah, WA USA
Contact:

Post by Charles Krebs »

Chris,
What about tube lens to objective distance?
Yes, within a certain range. That is why, for example, the diagram showing the Nikon tube lens shows an acceptable range of 100-200mm between tube lens and objective. I don't know what the limiting factors are.

The real driving force behind the switch to infinity optics on microscopes was to allow all sorts of nifty components to be inserted between the objective and viewing eyepieces. As more sophisticated research lighting methods (especially fluorescence illumination) were developed, it became clear that there needed to be more room to add all that stuff. Even if they had been able to squeeze it all into a scope with finite objectives, from an optical standpoint there was less image degradation if these components were inserted into a "parallel" beam path instead of a converging one.

So while we don't make use of that new-found "infinity-space" here, it's pretty much a foregone conclusion that in the future (as in the recent past) any new and improved light microscope optics will be of this type. (... at least until they think of something even better :wink: )

Other than the new room to work with (crucially important these days), I've been told that there really is not any inherent reason that this type (infinity) of objective should provide better image quality (compared to finite). (But naturally optical design, optical glass, and manufacturing are always advancing).

ChrisR
Site Admin
Posts: 8564
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Hmm, why 100 - 200mm?
Would it make a difference to the aberrations if we increased the distance between our "tube" lens and the objective?
Real tube lenses are quite "thin". How about, say, trying that with a 240mm process lens? ;)

Charles Krebs
Posts: 5859
Joined: Tue Aug 01, 2006 8:02 pm
Location: Issaquah, WA USA
Contact:

Post by Charles Krebs »

ChrisR,

Don't know what the limitations are. After exceeding a certain distance you would probably get vignetting. While Nikon gives a 100-200mm range, Mitutoyo states that : "A distance smaller than the specification does not affect an optical performance. Contact us for detailed information."

See page 26 (and 24, 25) here:
www.krebsmicro.com/MitutoyoE4191-378.pdf

Real tube lenses are quite "thin". How about, say, trying that with a 240mm process lens?
The Nikon one is 29mm deep and the Mitutoyo one (MT-1) is 32.5mm. I suppose it depends what you consider "thin". :wink:

rjlittlefield
Site Admin
Posts: 21125
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Chris S. wrote:Rik, just to be sure, you did block the built-in light path that surrounds the BD lens, right? I hesitate to bring this up, as you almost surely thought of it.
It's always good to check -- sometimes I overlook stuff that in retrospect seems obvious. But yep, completely blocked.

Light leakage down the illumination path doesn't really do much harm to resolution or CA. It just adds haze -- unfocused light over the whole image.

With the path not blocked, the effect is catastrophic. These are with identical manual exposures.

Image

By the way, the comparison images published in first post were untouched with respect to contrast and brightness. In the corner, the BD retained excellent contrast for coarse detail; there just wasn't much fine detail to go with it.

--Rik

Post Reply Previous topicNext topic