Tube Lens Test Final Results

[Lou Jost]

Adalbert, yes, it is good to focus on one magnification at a time, but it is also important to realize that there is nothing special about 200mm except that this is the tube focal length that gives the rated magnification. Since you are also already considering other magnifications, I just want to remind you that there are other (possibly better) ways to get sharper corners. Note that using a longer focal length does not take the objective away from its designed optimum, whereas as Rik says, changing the infinity design point can cause problems. Robert has shown that sometimes it works fine, but in cases that haven't been tested, nothing is certain. On the other hand, if the corners are bad and the center is sharp, using a slightly longer focal length is virtually guaranteed to sharpen the corners.

Edit: Of course this will also slightly reduce the overall sharpness; you'll need to experiment to see what balance you want to reach in your application.

[RobertOToole]

Hi Robert,

Your test: https://www.closeuphotography.com/tube-lens-test under: “Tube Lenses shorter than 200mm”

Thorlabs ITL200
https://www.closeuphotography.com/thorlabs-itl200/
Set-up:
• Short focus, 144 mm extension from sensor to lens
• Reverse mount
• 75mm from tube lens to objective
Image Quality:
• Overall: 9.6 out of 10
• Center: 8.5
• Corner: 8.5
• Chromatic Aberration Control: 10
.
.
.

RAYNOX 208MM (43mm x 0.50 mount)

Also know as: DCR-150, Macroscopic Lens Model M-150, Model CM-2000 1.5X, +4.8 diopter

Closeuphotography.com/raynox-tube-lens
Set-up:
• Short focus, 144 mm extension from sensor to lens
• Normal mount
• 50mm extenstion from tube lens to objective
Image Quality:
• Overall: 9.6 out of 10
• Center: 9.5
• Corner: 10
• Chromatic Aberration Control: 9.5

So, at 144mm is the quality in the corner for Raynox = 10 and for ITL200 = 8.5.
Independent of the typo at ITL, Raynox has achieved the full 10 points in any case.
The image quality in the corner is very important for me, therefore I’m asking for that :-)

BR, ADi

Yes, you are correct.

I don't know how Raynox did those designs. As closeup filters they are ok but as a tube lens they are just about impossible to beat in IQ.

[RobertOToole]

Also in my recent post about my findings with the Mitutoyo 10x M Plan,

http://www.photomacrography.net/forum/v ... hp?t=38289

Short focus did not improve image quality with the ITL200 but thankfully with infinity and in reverse mount the ITL200 performance was excellent.

FYI, I'm finishing up the 10x test now over the next couple of days. The results are very interesting so far

Robert

[Adalbert]

Hi Robert,
I’ m looking forward to seeing the results of your test (and comparing them with my assumption)!
So, if I had to guess I would say for the TL focused to the infinity first the ITL200,
then the DCR-150 and for the shorter length of the tube the both on the same place :-)
DCR-150 is really not bad and a good and cheap alternative to the ITL200.
I’m reassembling my Harley rail.
BR, ADi

[JohnyM]

Great test Robert, thank you for doing that

Sorry for digging up, but i just now started to think about making univerasal tube lens system (no bellows) and sticking to ~100mm tube would make it really eazy for me.

Could this phenomenon that Robert discovered be explained by this:

1) Tube lens like Raynox is heavily stopped down by microscope objective. So it's not responding to extension change as much as to be expected.
2) Shorter tube lenses are less stopped down, so they do not respond to this "short focusing" as well as longer ones.
3) Since path between objective and lens is no longer 00 , changing distance between objective and lens serves as spherical abberation compensation. And could possibly serve to introduce "cover glass" correction.
4) Perceived improvement is result of larger effective aperture?
5) This result is possible due to moderate objective aperture and will not carry to higher NA objectives, and will eventually bring worse results, unless distance between tube lens and objective will be "nailed". Eventually (with increasing NA and decreasing extension) that "nailed" spot will be inside the tube lens or past it.

Sadly, due to 3,4,5 this is not a good solution for a system with multiple objectives and single tube lens?

Offtopic:
Anyone tested Raynox 250 +1.7 TC vs Raynox 150? Robert ?

[Georg]

Hey Robert
May I ask u which extensions u use for the raynox 150 setup? For both settings
This should also work fine with a full frame camera?
Will use a sony A7R4
Thanks for your help

[Lou Jost]

I've founmd in my tube lens tests that best full frame settings are not generally the same as best APS settings, and that was true of the Raynox. The arrangement that gave the best central sharpness also gave worse FF corners. That arrangement would look like the best arrangement on APS, but on FF it was worthwhile to sacrifice a bit of central sharpness to get decent corners.

[Georg]

I've founmd in my tube lens tests that best full frame settings are not generally the same as best APS settings, and that was true of the Raynox. The arrangement that gave the best central sharpness also gave worse FF corners. That arrangement would look like the best arrangement on APS, but on FF it was worthwhile to sacrifice a bit of central sharpness to get decent corners.

You‘re right Lou

The more I‘m doing my research for the best system to use the mitus on FF the more I get confused ... Dozens of ways to mount and tweak every combination of tubes extensions ...
I will go for the 200mm/180mm camera lens and in front the mitu
And the bellows raynox 150/250 system - think the bellow is quite useful in stepping down the magnification and otherwise I can also revers the raynox lenses to step down...

This forum is really useful but can make u crazy in the next second

[mjkzz]

Just saw a post about "short focus" with infinite objectives over Facebook, I also watched a video on Youtube about this "technique", I had reservations back then but I liked the result, now since it is being spread over to Facebook and Youtube, it is kinda important. To me, "short focus" is deviating from the very fundamental design, ie, the tube lens should be focused to infinity.

So, does this "short focus" technique have any theoretical backings at all? Or is it a hit and miss, trial and error thing? I can see people will claim something due to some particular settings and this kind of thing can get widely spread out and become a "I heard this on the internet" thing.

[mjkzz]

Basically, what I am saying is that, if this does not have any theoretical backings, I am worried that people over Facebook or Youtube might overly GENERALIZE this idea and spent a lot of effort just to "oh, this configuration is even better".

[rjlittlefield]

Regarding theoretical backings, several thoughts come to mind.

1. Any objective is just a lens that happens to be optimized for use as listed on the label. That does not mean it will not work well for other purposes also, particularly if "working well" is defined by flexible criteria.

2. Any lens will continue to work pretty well in some region around its design point. The size of that region depends on aspects that are impossible to predict precisely.

3. In general, the more specialized the lens -- think high magnification and large NA -- the less well it will tolerate deviations from design point. 4X NA 0.1 is very tolerant of being used in "off label" ways, but 50X NA 0.55 is not.

4. Using a lens very far off label is like playing the lottery or mining for gold: most of the time it's not good for much besides amusement, but sometimes you get lucky. In either case there is value, for some people more than others.

That last bit of theory is mostly psychology, not optics, but it's important anyway. If somebody wants to spend time making good pictures, the efficient approach is to stay with known good setups. But if somebody wants to spend time testing other setups to see if they can get lucky, that's OK too. My best advice is to be aware of that difference and act accordingly.

--Rik

[Scarodactyl]

One factor is that microscope objectives are not really optimized for the absolute best image because there are other concerns. These objectives will be used on a nosepiece with a common tube lens, so the setup has to work well with every one of their objectives, and the tube lens has to be focused to infinity to allow room for coaxial illumination, filters and so forth. So it kind of makes sense that at times moving outside the parameters, particularly at lower NAs where spherical aberrations are less fatal, could give you a bump in performance.

[Lou Jost]

To Rik's advice, maybe I would add that there is nothing special about any particular tube length anyway (except as a way to specify a nominal magnification), so I am not sure it is fair to say that the objective is being used away from its design point.

[rjlittlefield]

Addressing Lou's point, I agree that an infinite objective that is simply used with a shorter tube lens does remain at its design point.

However, my understanding of the "short focus" technique that mjkzz mentions is that the rear lens is not focused at infinity. That does drag the objective away from its design point. Even so, the objective may continue to perform well. For example Robert OToole's tests on his website note that for the Mitutoyo 5X + Raynox DCR-150, "Short focus (less extension) results in less magnification, better sharpness and less CAs".

A good case can be made that the Raynox DCR-150 as tube lens in forward orientation is always being used away from its design point, and certainly using the Raynox+Mitutoyo 5X combo with less than infinity extension also moves the Mitutoyo away from that objective's design point.

Yet in this case two "wrongs" fortuitously add up to one "right", leading to Robert's conclusion and making this particular combo a "known good" setup.

But even if the rear lens remains at infinity focus, you have to ask whether that lens is being used away from its design point. The answer to that question has to be "yes, away from the design point", except for purpose-made tube lenses like the ones from Nikon, Mitutoyo, and Thorlabs. The reason is aperture placement. Lenses that have an internal aperture are designed for that to be the limiting aperture, and when used with an objective out front, that is not the case. The effect is that, away from image center, the image formed by the rear lens is using different portions of the rear lens glass then the lens was optimized for. In particular, lenses with internal apertures are designed around the fact that light from the right side of the field will enter primarily the right side of the lens, but when used with an objective, light from the right side of the field primarily enters the left side of the lens. In mjkzz's other thread, using a 50mm f/1.4 rear lens, those degraded portions of the image near the vignette are formed by a narrow pencil of light that came through the f/1.4 edge of the lens. When used as designed, those portions of the rear lens would never be used at all except when the lens was set at f/1.4, in which case their aberrations would be heavily diluted by other light coming through better behaving parts of the glass, and even so you would not expect the final image to be particularly sharp.

Back to my numbered list, I think the Raynox as tube lens demonstrates at least three of those points very well: it's being used off-label, but works well anyway, despite that it's cheap and easily available as a new manufacture part. This is exactly the sort of situation that I am happy to promote as a "known good" setup for use with infinity focus. Modifying it to be short focus -- less than infinity -- that part I'm not so happy to recommend because it depends heavily on which objective we're talking about. What works well for NA 0.14 may work not nearly so well at NA 0.55, where the increase in spherical aberration is 238 times more important. (SA scales as the 4th power of NA.) Still, I doubt that most of the folks on Facebook will be dealing with such picky objectives, and I'd hope that the ones who do will also be knowledgeable enough to recognize when the edge of the envelope has been pushed too far.

--Rik

[mjkzz]

Regarding theoretical backings, several thoughts come to mind.

1. Any objective is just a lens that happens to be optimized for use as listed on the label. That does not mean it will not work well for other purposes also, particularly if "working well" is defined by flexible criteria.

2. Any lens will continue to work pretty well in some region around its design point. The size of that region depends on aspects that are impossible to predict precisely.

3. In general, the more specialized the lens -- think high magnification and large NA -- the less well it will tolerate deviations from design point. 4X NA 0.1 is very tolerant of being used in "off label" ways, but 50X NA 0.55 is not.

4. Using a lens very far off label is like playing the lottery or mining for gold: most of the time it's not good for much besides amusement, but sometimes you get lucky. In either case there is value, for some people more than others.

That last bit of theory is mostly psychology, not optics, but it's important anyway. If somebody wants to spend time making good pictures, the efficient approach is to stay with known good setups. But if somebody wants to spend time testing other setups to see if they can get lucky, that's OK too. My best advice is to be aware of that difference and act accordingly.

--Rik

Agree with all.

But one more point, every time I put my Canon 100mm f/2.8 with an objective on it, the camera will set focus to nearest because my adapter has electronic contacts, and sometimes, I forget to focus it to infinity, yet all seem to work without noticeable degradation (to me), it means any objective is just pieces of glasses bending light. And the point is this happens without being lucky! It just works all the time.

However, a big however, I am pretty sure an infinite objective is designed/optimized for being infinite objective, optical engineers spend a lot of effort to get it right. So an infinite objective are not just a group of glasses bending light, they are group of glasses bending light in certain way, it has certain intrinsic properties. But because of it is an infinite objective, a particular type of objective having certain properties, theory says you can push around the focal length of tube lens but keep focused to infinity, to some degree to fit our "hacking" needs (for rich men/women, they design their own objectives to fit their needs)

Rik's #3 is probably a good example of such -- when things get so specialized and optimized, those pieces of glasses are arranged in such way that deviation from their design point will be problematic. And my other thread about pushing a 10x 0.28NA down to 2.5x experiment shows this: the out of focus area has a lot of swirly patterns, definitely not the same if setup normally -- there is a limit. And yes, those out of focus area look like some cheap 4x objectives I got on Taobao

Pushing around a system without respecting its design constrains is like shooting in the dark, like Rik's #4 and when it is becoming a "thing" on the internet . . . I am worried