Telecentric combo at 0.8X to 1.69X

[Guppy]

Hi

For the information of photographers who are interested in object-side telecentric lenses.

The Zhongyi Mitakon Creator 85 f/2.8 1-5X Super Macro is nearly telecentric.
Concerning telecentrie
viewtopic.php?f=27&t=43090
Concerning image quality etc
viewtopic.php?f=25&t=39495

The Nikon Toolmaker Measuring Microscope Objectives are available for 1x, 3x, 5x,10x, 20x and 50x they are object-side telecentric
Here is a picture of the connector with the
Nikon Toolmaker EDF 20031 TM OBJECTIVE LENS 3X-A


I could not find any advantage over entocentric lenses in focus stacking artifacts, only the perspective changes!

Kurt

[patta]

Nice, some more information about those telecentrics!

Issue with mirrorless lenses... how to control the aperture when it's mounted inverted? It is all electronic! Must just try, let's see who can loan one for some abuse...
The Mitakon 85mm, that is quite interesting too. Toolmaker microscope... that's a beautiful toy, but needs a larger apartment!

Another telecentric lens for sale at reasonable price is a 55mm from (Computar?) but I haven't tried it.
Also many used industrial C mount lenses on the market, but likely for small sensor and low res.

Telecentric is useless.. hope not! I've wasted one month on it

[Lou Jost]

"Issue with mirrorless lenses... how to control the aperture when it's mounted inverted? It is all electronic!"

This is not a problem if you can borrow a mirrorless camera to set the aperture. Attach the desired lens, focus at infinity, set the aperture as desired, oress the depth of field preview button, then twist off the lens without turning off the camera. Viola, your aperture is set.

"Telecentric is useless.. hope not!"

The main reason we on this forum use telecentric lenses is to avoid stitching artifacts. For this, it is wonderful!

[rjlittlefield]

May I suggest another method for acheiving telecentricity? Find a lens for a mirrorless system that advertises image-side telecentricity, and reverse it either on extension tubes or any tube lens, with no added apertures anywhere.

With respect to telecentricity, I am curious for more details about how this works in practice.

The reason for my curiosity is that I just now went through all the lenses that one can select in the Photons to Photos Optical Bench, and I found not a single one of them that is accurately telecentric on the image side.

Even the best of the bunch:

Olympus 17mm f/1.24 IF (Olympus M.Zuiko ED 17mm F1.2 Pro)
Olympus 30mm f/3.56 IF macro (Olympus M.Zuiko ED 30mmF3.5 Macro)
Olympus 291mm F/4.08 IF (Olympus M.Zuiko ED 300mm F4.0 IS Pro)

were off by several degrees for the most extreme edge cones.

I know that part of the design concept for micro four thirds is that lenses will be telecentric on the image side. But the Photons To Photos simulations suggest that real lenses are more like telecentric-ish, than accurately telecentric. There would be no way for a user to know, so it is an attractive target for fudging the descriptions.

Exactly which lenses have you used this way, and how have you exercised or tested the telecentricity?

--Rik

[Guppy]

Hi Lou

For stitched images, telecentric lenses are certainly an advantage.

Regarding resolution in the image with telecentric lenses:
Resolution is often not as high with telecentric lenses as it is with entocentric lenses, depending on the magnification.
The Nikon MM 1x Toolmaker (is telecentric on the objective side) has a qualitatively uniform image circle at 1:1
that is well suited for crop format 2.0 (diagonal 21.6mm).
On a camera with a pixel pitch of 4.8µm, the resolution is 90L/mm.
With a smaller pixel pitch of 2.1µm, the resolution barely increases, to about 100LP/mm.

For comparison:
The LAOWA 100mm F2.8 CA-Dreamer Macro 2x at 1:1 on a camera with pixel pitch of 4.8µm resolves the object at about 95LP/mm.
On a camera with a pixel pitch of 2.1µm the resolution is about 170LP/mm.
Here you can see clearly that the resolution of the LAOWA lens is higher by a factor of 1.7.
With an image scale of 1:1, it is therefore very important for measurement results and comparisons that the pixel pitch of the camera is known.

The Nikon EDF 20031 TM OBJECTIVE LENS 3X-A has a larger image circle, it is suitable for full format.
With a camera with pixel pitch 4.8µm,
the resolution in the center of the image is 290LP/mm and in the corners 270LP/mm, which is very good!
With a camera with a pixel pitch of 2.1µm, the resolution in the center of the image, minimally more, is about 300LP/mm.

For comparison:
The Laowa 25mm F 2.8, 2.5-5X Ultra Macro resolves 320LP/mm in the center of the image at 3:1 on a camera with a pixel pitch of 4.8µm.
With a camera with a pixel pitch of 2.1µm, the resolution is 400LP/mm.
Thus, with a smaller pixel pitch, the factor is 1.25 and is lower than with 1:1.
At 5:1, the difference in resolution on the object side, between camera with pixel pitch 4.8µm and 2.1µm is even much smaller and hardly noticeable in the image.

Kurt

[Lou Jost]

Exactly which lenses have you used this way, and how have you exercised or tested the telecentricity?

Rik, I have always used these reversed MFT lenses in lens combos on FF, usually at magnifications greater than 2x, so I only was looking at the central portion of the MFT lenses. I have used many different MFT lenses, but I have only stitched a few, and measured the Zerene scale factors for a few. I don't remember which ones...I'll try to look.

Resolution is often not as high with telecentric lenses as it is with entocentric lenses, depending on the magnification.

The MM 1x is maybe not the best one...the other MM objectives are better.

[rjlittlefield]

Rik, I have always used these reversed MFT lenses in lens combos on FF, usually at magnifications greater than 2x, so I only was looking at the central portion of the MFT lenses. I have used many different MFT lenses, but I have only stitched a few, and measured the Zerene scale factors for a few. I don't remember which ones...I'll try to look.

Thanks for the clarification.

Adding some detail about my concerns, here is an illustration from a different source. This originates from an article at PetaPixel , titled "Telecentric Lens Design: Did Nikon and Canon Follow Micro Four Thirds’ Lead?"

... telecentric lens design, meaning optics that are designed in such a way that the light rays will hit the sensor more directly without too much straying off—a design that was already adopted by Four Thirds system lens mount in 2003!

Yes, Olympus and Panasonic, the companies that started the Four Thirds DSLR system. Olympus released the Olympus E-1 DSLR in 2003 alongside their first fully-realized telecentric design lens, Zuiko Digital 14-54mm f/2.8-3.5. These technical concerns and approach to optimize lens mount and subsequent optical design were fully implemented by Olympus and Panasonic’s Four Thirds system 16 years ago. 16!

Olympus adopted telecentric optical design with their Four Thirds mount, and this lens, 14-54mm f/2.8-3.5, which was released in 2003.

As illustrated here, edge cones for the upper "telecentric" lens are a lot less sloped than those of the classic design older lens. But they're still way short of the 1-part-in-10000 scale change per focus step that is needed for totally clean focus stacking, and is easily achieved by adjusting the entrance pupil to be visually at infinity.

It does seem likely that these new "telecentric-ish" designs, with their rear elements that are large compared to their image circles, will work really well in combination with an external aperture to make them fully telecentric. Experimental evidence in support of that idea would be great.

--Rik

[4odonates]

Hi Rik! As you might recall I own a Canon 5D Mark II DSLR camera (full frame) and a Canon EF 100mm Macro lens (f/2.8 L series). I also own a Raynox DCR-250 close-up lens. How far in front of the macro lens does the Raynox need to be in order to make the lens telecentric?

I would also appreciate suggestions for cobbling together the lens and Raynox. I already own two nested 67-52mm and 52-43mm step-down rings that I use to mount the Raynox DCR-250 on my Canon 100mm macro lens.

Thanks for your help!
Walter

[rjlittlefield]

Hi Rik! As you might recall I own a Canon 5D Mark II DSLR camera (full frame) and a Canon EF 100mm Macro lens (f/2.8 L series). I also own a Raynox DCR-250 close-up lens. How far in front of the macro lens does the Raynox need to be in order to make the lens telecentric?

In general I cannot give a number, for the reasons discussed in the questions at viewtopic.php?p=117579#p117579 and my answers in the next post at viewtopic.php?p=117615#p117615 .

However, I do own lenses that sound identical to what you have, so I can give some more information. Most of that information is already presented in the post that starts this thread (at viewtopic.php?p=116283#p116283), and in my other post linked above. So for starters, I suggest to go study those postings carefully. It will probably take several readings before everything makes sense.

Adding some interpretation that applies to your specific situation...

I do not believe there is any way to use the lenses you have, to get a high quality telecentric field that covers your entire fullframe sensor. However, by using your camera in crop-sensor mode, you should be able to get results that are similar to mine from the T1i.

In the first post of this thread, I show a configuration that runs at "roughly 1.69X, 13.2 mm across my T1i sensor. The very corners of the field are astigmatic at f/11, but there's a high quality center about 10mm x 6.6mm. At f/16 the corners come in pretty good but the whole image is less sharp."

That configuration is with the Raynox mounted on its spring-loaded adapter. If instead you use your two screw rings, then the Raynox will mount slightly farther forward on the rear lens, which will require tweaking the rear lens ring to move the rear lens entrance pupil forward by the same amount. That will keep the system entrance pupil positioned at infinity, for Raynox plus rear lens combined. As shown in the first post of this thread, with the Raynox on its spring-loaded adapter, the critical position of the lens ring puts the indicator in the middle of "0." at the "0.39 m" ring setting. I have just now repeated a visual test using the Raynox on screw rings, and the critical position moves the indicator only to the "3" of the same "0.39 m" setting. The critical settings may be slightly different with your particular lenses, but the difference should be small.

It may be worth repeating how I do the visual test. It goes like this:

1. Get a pair of binoculars or a spotting scope, or a camera with telephoto lens, and manually focus on a distant object. Once focused, do not change this focus setting.
2. Mount the macro lens on a Canon body, turn on the camera, set manual aperture f/16, stop down by pressing the DOF preview button, and remove lens from body while keeping the camera turned on and DOF button pressed. This will remove the lens from the camera while keeping the lens stopped down.
3. Mount the lens on a tripod using its own tripod ring.
4. Mount the Raynox on front.
5. Looking through the Raynox by eye, turn the rear lens ring until the aperture diaphragm moves backward to near infinity.
6. Switch to looking through the Raynox using the binoculars or spotting scope or camera with telephoto, again look at the edge of the aperture diaphragm. It will probably be slightly blurred, using this equipment that is more sensitive to focus than the unaided eye is.
7. So then, tweak the position of the lens ring so as to bring the edge of the aperture into perfect focus. This will place the entrance pupil at infinity, or at least close enough for practical purposes.

You asked:

I would also appreciate suggestions for cobbling together the lens and Raynox. I already own two nested 67-52mm and 52-43mm step-down rings that I use to mount the Raynox DCR-250 on my Canon 100mm macro lens.

The rings by themselves will work as described above. But you'll get better results by moving the Raynox farther out using extension tubes or a helicoid. If you have some 52mm tubes, those can be inserted between your two rings. Or you could adapt down to m42 and extend with m42 tubes or helicoid as shown in my post linked above.

Again, the key aspects are to put the system entrance pupil at infinity, stop down as far as you can tolerate, and be aware that under some conditions only the center of the field may be telecentric. (See viewtopic.php?t=39724 for more about that last aspect.)

I hope this helps!

--Rik

[4odonates]

Thanks for the detailed reply, Rik!

I skimmed/read some of the related posts/threads. You know how super-cooled water freezes -- at first it's liquid then an instant later it's solid ice. Well, that's what happened to my brain when I read those posts -- one minute I'm lucid and in an instant I'm experiencing brain freeze!

I'm not stupid, but I don't have your awesome expertise. Sometimes I come to understand something by doing it repeatedly until something clicks and I get it. That's why I'm still hoping you (or someone) can tell me how to make it work without all the tedious set-up. Maybe I could understand your directions if I could see a video tutorial showing what you described.

So let me take one more run at this. In addition to the Canon EOS 5D Mark II, I own three Fujifilm X-Series APS-C cameras: X-T1; X-T3; and X-T5. I also own a Fringer Canon to Fujifilm adapter so I can mount my Canon EF 100mm macro lens on the Fujifilm cameras. I tested the rig and it works beautifully. Using this setup, can you (or anyone) tell me how far to mount the Raynox close-up lens in front of the Canon macro lens when it's mounted on a Fujifilm camera body?

BTW, I knew the Canon camera body you use with your 100mm macro lens has an APS-C sensor (rather than my full-frame 5D Mark II) -- I was just hoping it wouldn't matter. But if I use one of my Fujifilm crop sensor cameras instead of my Canon full frame camera, can I use the same numbers as your rig?

Thanks for your patience, and thanks for your help!
Walter

[rjlittlefield]

That's why I'm still hoping you (or someone) can tell me how to make it work without all the tedious set-up.

Short of buying a pre-built telecentric lens, I don't know how to avoid careful setup.

The technique of looking through the front and moving the entrance pupil to infinity is actually the least tedious method I know. It is a linear process -- start at the beginning, go through all the steps one time, and the job is done.

The alternative is to use tweak and test -- repeatedly adjust the focus ring and shoot test images until you get no change in magnification with distance. That may be conceptually simpler, but for me it takes a lot longer to reach the same level of accuracy.

You noted:

Maybe I could understand your directions if I could see a video tutorial showing what you described.

My video production skills are minimal, and that poses a significant stumbling block. But in quick scan of YouTube I don't see that anybody else has done it, so maybe it would be worth the attempt. More thinking required.

can you (or anyone) tell me how far to mount the Raynox close-up lens in front of the Canon macro lens when it's mounted on a Fujifilm camera body?

You can swap in a different body without changing the telecentricity of the optics. (Looking into the front of the lens, the entrance pupil will still appear to be at infinity.)

So yes, you can use a Fuji camera with mount adapter, instead of my Canon, as long as all the lens configuration in front is the same.

--Rik

[4odonates]

A "how to" video would be great! Perhaps you can collaborate with Allan Walls to create the first video by LW Productions. Seriously. Suggested topic for discussion during your weekly phone call with Allan tonight.

OK, back to my telecentric rig. I just want to be sure we’re on the same page. I think you’re saying that one of my Fujifilm APS-C sensor cameras and Canon 100mm macro lens (mounted on camera body using Fringer adapter) plus the Raynox DCR-250 mounted at the same distance in front of the 100mm lens as your rig will work identically to your telecentric setup? I hope so!

As always, thanks for your help!
Walter

[rjlittlefield]

OK, back to my telecentric rig. I just want to be sure we’re on the same page. I think you’re saying that one of my Fujifilm APS-C sensor cameras and Canon 100mm macro lens (mounted on camera body using Fringer adapter) plus the Raynox DCR-250 mounted at the same distance in front of the 100mm lens as your rig will work identically to your telecentric setup? I hope so!

Yep, sounds like the same page.

--Rik

[4odonates]

I took the liberty of annotating one of your photos, Rik. Did I correctly identify the parts? Some parts are followed by a question mark (?) -- meaning either I need confirmation or more information. Is the middle part a helicoid, and if so, how long is its range? Thanks!
Walter

[rjlittlefield]

Did I correctly identify the parts?

Close, but not quite perfect.

At the first question mark, the label should be 42mm - 52mm step-up, not -down. Such rings are traditionally labeled with their male thread first, then either -up or -down depending on whether the female thread is larger or smaller.

At the second question mark, the particular tube that is shown is just a fixed-length tube, not an adjustable helicoid. However, it could be a helicoid, and that would allow to maintain telecentricity over some continuous range of magnifications.

The third and fourth question marks are correctly labeled.

While we're talking about 42 mm threads, I need to mention one caution: thread pitch. 42 mm threads are commonly supplied as either pitch 1.0 mm [the classic Praktica-Pentax lens mount thread] or pitch 0.75 mm [known as T-mount, T1 or T2]. In principle it does not matter which you use, as long as you're consistent. All the extension tubes I know that are labeled just "42 mm" or "M42" have pitch 1.0. Unfortunately 42 mm adapter rings are now commonly available as pitch 0.75, and sometimes even the vendors can't keep them straight. I have recently ordered two batches of M42 rings that were clearly advertised as pitch 1.0, but came as pitch 0.75. Some people say that the fit is loose enough to be interchangeable, but that is not the case in my experience. For me, mismatching 42mm threads do screw together for something over 1 turn, but then they jam instead of proceeding to a clean stop at a square shoulder. At this time I do not know any foolproof way of receiving the proper threads.

--Rik