Adjusting focus on a new Jupiter-3+

[Alizarinus]

I recently bought a used Jupiter-3+ by Lomography. The lens has an amazing fingerprint, like my old Canon 50/1.5 LTM, but with a great contrast and definition. Unfortunately, it also mis-focuses badly on my Leica III-g, - wide open, the point of focus is about 6cm too close at 1 meter, and at infinity, all far away buildings look soft and out of focus, while some of the closer objects appeared in sharp focus. I know that sonnars have a focus shift, and maybe this lens is optimized for something like f8, but i just love how it renders wide open, and wanted to try to calibrate it myself. When I unscrewed the front part from the focusing module, I saw that it had an extra shim, - less than a millimeter thick. I took it out and reassembled the lens, thinking that this just might get it into the ballpark, but as I was setting up a focus test shot, I realized that the aperture ring has shifted - at wide open, it indicated something closer to 2.8. Is there anything that can be done to re-adjust the aperture ring? It goes back to normal when I reinstall the extra shim. Most of my other LTM lenses focus more or less accuratelly on this camera. Any help will be greatly appreciated!

 

[TenEleven]

First of all Jupiter 3+ are calibrated to hit their best focus from f2.8 onwards, by f1.5 focus looks usable at the default calibration, so it seems a bit odd. Do you know that the Leica III is in good calibration? Have you tested it with other lenses and gotten good in-focus results?

Also from calibrating my Jupiter 3 and the other lenses I worked on I feel it's safe to assume that taking out the entire shim will be in all likelihood way, way too much and you are going to overshoot your focus target by a significant margin. From what you describe I'd guess that you need about 0.03-0.05mm less shim. I bet the entire shim is significantly thicker than that, so you can't just take it out and hope that things will be alright.

You could either reduce the shim you have - which I would not recommend because if you take off too much you are now stuck with a lens that will not focus and will have to now make a shim anyway or sell it "for parts". Instead you should try to manufacture your own shim from either thick card stock, plastic sheets or something like that. Use the existing shim as a stencil. Keep experimenting until you find the right depth.

A tripod, a 30x or more loupe and a piece ground glass will help to find the appropriate shim thickness here, because the chance of getting it right first try are basically nonexistent. You could also use a digital camera, but again the camera would have to be verified to be hitting infinity at the infinity position with another known good LTM or M lens.

Lastly, the aperture is off because the lens (optical block) is *screwed* into the focus mount portion. So of course by removing the shim you are changing the degree of rotation of the lens block to the lens body. This can be alleviated by taking out the three grub screws of the aperture index ring and tapping new holes at the correct position. The grub screws are delicate so it's best not to use force and try to just "jam" them in there without tapping holes.

 

[Alizarinus]

All of these points make sense! I can't be 100% sure about the calibration of my LTM Leica, but I do have at least one or two lenses that focus accurately on it, - all of them Canon (50/1.5, 50/1.8, and esp 50/1.4). I have a Sony A7 camera with a macro LTM to Sony E adapter, and so I marked the spot on the macro adapter where the canon 50/1.4 was in register at infinity, - it's just a tiny bit away from the hard stop. When I check the J-3 on the Sony and the same adapter, with or without this tiny correction, it does not reach infinity at all, not until I set it to f8.

The good news is that there are two shims inside the J-3, - one thick and the other very thin, - .47mm. I think you are correct that I will need to replace the thin shim with something even thinner, maybe half the thickness. Another good tidbit is that the screws that fix the aperture ring do not tap through the inner ring, - there are 3 indentations, - no tapped holes. I should be able to adjust it,- fingers crossed.

I think the cause of the problem is the focus block, - the distance index on it does not reach the top of the lens when mounted, while all other lenses go past the top/center. Anyway this is typical Jupiter-3 story. It seems like this optically awesome new version is just as poorly made as it's aluminium brothers 🙂

The plan is to calibrate the lens to the Leica as it is right now, and use a piece of tape in place of aperture markings. After I get the camera serviced, I will check the lens callibration again, and then adjust the aperture ring.

Thank you for your advice!!!

 

[TenEleven]

First of all the three versions of the Jupiter-3+ I bought have had no such issues. (One for me, two for friends - although I would have liked a backup. It's a great lens.) Since Lomography does not sell these anymore I am going to assume you got it through other channels, therefore I think the issue here is more likely unscrupulous sellers either self-assembling parts lenses or just hocking lenses that previously failed QA.

If the focus block (or rather the mount portion) does not screw into the LTM mount properly that indeed could cause all sorts of problems.

To clarify, do you mean the focus scale (aka the very bottom part that mates with the camera) or the distance scale ring (which rotates to focus) - if the distance scale is off then the lens has likely been disassembled and inexpertly re-assembled. If it is the focus scale (aka threaded) part - then it seems likely to me that the mount was mis-machined and someone has indeed sold you a lemon.

 

[Alizarinus]

This seems to be the case, - it is a bit of a lemon, and I kind of knew that it would be, since it was listed as "open box", but came without the above mentioned box, or the adapter, or any other extras. I took a chance on it because, as you mentioned, they are almost impossible to find. I do not plan on returning it. It is ok optically, - no centering issues. Anyway, here is a photo of it on my Leica IIIg, - the index mark on the part that attaches to the camera does not reach the top/center, but when I mount any other of my lenses, they go way past the center. I ordered some plastic shims online, and got a finger-operated mini-drill, - will try to calibrate 🙂 I kind of estimated by what turn distance it needs to be compensated, - we'll see how it goes.. It's probably going to be a pain, but I will start by setting it to focus on infinity on my Sony A7, and then, I will do a film test for the minimal distance - 1m in this case.

 

[Sonnar Brian]

It's amazing how a small change in the shim produces large shifts in focus.
Back focus for a 51.6mm lens at 1m is 54.40742303- what your lens should be.
Back Focus for a 51.6mm lens at 0.94m is 54.59702837- what it actually is.

I would try reducing the shim by ~0.2mm. I use the formula for focal length solving for back focus to calculate required shim thickness. Solve for the distance indicated by the RF, and by the actual focus. Then you get the required change in thickness. Focal length is assumed to be 51.6mm for the J3plus.
b=1/((1/f)-(1/d))

​

 

[Räuber]

My J3+ has issues focusing to infinity wide open too. I need to stop it down to F5.6 to have perfect sharp focus in the center at infinity. Additionally it seems it is decentered too. So my experience is not as flawless with the new Jupiter. But it is nothing a good CLA can not fix. I use my lenses on mirrorless cameras anyway. So no issue with focus shift.

Since I've the chance I need to ask. Can you tell me the serial of your J3+? It would be interesting to see how many copies they made before closing down the production. The highest serial I have seen so far suggests it is about 1300 copies.

 

[dexdog]

Since I've the chance I need to ask. Can you tell me the serial of your J3+? It would be interesting to see how many copies they made before closing down the production. The highest serial I have seen so far suggests it is about 1300 copies.

Rauber, I have three Jupiter 3+lenses, the highest serial is 00001588

 

[Alizarinus]

It's amazing how a small change in the shim produces large shifts in focus.
Back focus for a 51.6mm lens at 1m is 54.40742303- what your lens should be.
Back Focus for a 51.6mm lens at 0.94m is 54.59702837- what it actually is.

I would try reducing the shim by ~0.2mm. I use the formula for focal length solving for back focus to calculate required shim thickness. Solve for the distance indicated by the RF, and by the actual focus. Then you get the required change in thickness. Focal length is assumed to be 51.6mm for the J3plus.
b=1/((1/f)-(1/d))

​

Thank you for the formula! Last night, I decided not to wait for the shim material to be delivered, and cut two shims from a polyester-based drafting film, - "dura-lar", which is about .1mm thick, so the shim is .2mm. I added these new shims and took out the thin original shim, which is .47mm, but kept the thick original shim in place. With this mod, on my Sony A7, the J-3+ is now matching my Canon LTM 50/1.4 at infinity / wide open. I will do a film test tonight and report back. Fingers crossed.. The serial on the lens is 00000402. It's probably a factory reject.

 

[Alizarinus]

My J3+ has issues focusing to infinity wide open too. I need to stop it down to F5.6 to have perfect sharp focus in the center at infinity. Additionally it seems it is decentered too. So my experience is not as flawless with the new Jupiter. But it is nothing a good CLA can not fix. I use my lenses on mirrorless cameras anyway. So no issue with focus shift.

Since I've the chance I need to ask. Can you tell me the serial of your J3+? It would be interesting to see how many copies they made before closing down the production. The highest serial I have seen so far suggests it is about 1300 copies.

The serial on the lens is 00000402. Seems like an early production factory reject. I should have bought one in 2016...

 

[pyeh]

Rauber, to add to your data set, I have one, which I bought in February 2016. Its serial number is 434.

 

[dexdog]

The serial on the lens is 00000402. Seems like an early production factory reject. I should have bought one in 2016...

I have three of these lenses, did not mean to, it just kinda happened in 2021. The first one came in the original box, lens caps and instruction manual, but no M-mount adapter. It was perfect at infinity and at the one meter/wide open test. I bought another one relatively cheap on eBay that would not focus to infinity, so contacted the seller about returning it for a refund. He made me an offer that I keep the existing lens, and he would send me another one at no charge. The idea was that I could use parts from the two lenses to get one working lens. I decided to accept his offer, with an encouraging word from Sonnar Brian that both were likely fixable.

The first lens was a pretty easy fix, someone had taken it apart and put the mount back together incorrectly (it is a 4-start thread, only one of which is the correct one.). I tweaked the shims and got it dialed in fast.

The second lens was more of a challenge because one of the 2 screws holding the focusing helical together had fallen out entirely (yeah, that should not happen). The reason that the screw fell out was because the hole where the screw was supposed to fit was too big, and furthermore was not even tapped with threads! I tapped the hole to a larger size and modified a screw I had in the parts bin to make a replacement. I had to shim the second lens a few times to get it right, but got it dialed in, too. The glass in all three lenses was pristine. So, out of three, one was perfect, one was worked on by someone who did not know what they were doing, one was probably drilled wrong at the factory and passed on down the assembly line anyway. The perils of buying lenses from eBay!

 

[TenEleven]

Thank you for your confirmation dexdog, yeah that pretty much confirms my hunch that a bunch, surely not all, of these eBay lenses are basically factory QA rejects (for various reasons) that got "absconded". In a way - isn't it funny how (Zeiss) history repeats?

It's a crying shame that Lomography stopped selling these before they even had a chance to become popular.

 

[dexdog]

It's a crying shame that Lomography stopped selling these before they even had a chance to become popular.

Agree, they are great lenses if you get a good one. The perfect lens is no. 92, the misassembled one is 1144, the one with bad screw hole is 1588

 

[Alizarinus]

So, after a bunch of film tests, and a lot of shimming, I managed to calibrate it to focus more or less ok at 1m / f1.5, but the infinity is just a hair short of sharp focus - will have to stop down for that… and the aperture scale is now on a piece of masking tape, for now, anyway. Thanks again for all the help!

 

[TenEleven]

So, after a bunch of film tests, and a lot of shimming, I managed to calibrate it to focus more or less ok at 1m / f1.5, but the infinity is just a hair short of sharp focus - will have to stop down for that… and the aperture scale is now on a piece of masking tape, for now, anyway. Thanks again for all the help!

Glad to hear you managed to get it into at least somewhat workable condition. And I think that is pretty much how mine is calibrated.

The Sonnar design in particular does not "like" being shot at infinity wide open, most lenses (Gauss, Tessar, etc.) struggle a bit with this test, but the Sonnar in particular clearly does better at mid-range at wide or full aperture. There is some variation, but this is a general trend I observed. Stopping down even 1/2 the way to f/2 usually helps a lot if you want to shoot night-scapes on film without a tripod.

Agree, they are great lenses if you get a good one. The perfect lens is no. 92, the misassembled one is 1144, the one with bad screw hole is 1588

Yeah, it is humorous that one of the things that stood out to me with the official Lomography ones was how consistent the quality and calibration was. All three lenses were basically identical. Maybe they went a bit too hard on the QA and that's why they had to give up and there's so many rejects? Hah.

 

[Sonnar Brian]

My first J-3+ is serial number 00200, and probably the first sold in the US.
I suspect there was a QA problem on the focus mount, as there were on some of the original J-3's. That's probably why so many "seconds" ended up on Ebay.

 

[TenEleven]

My first J-3+ is serial number 00200, and probably the first sold in the US.
I suspect there was a QA problem on the focus mount, as there were on some of the original J-3's. That's probably why so many "seconds" ended up on Ebay.

Yeah, I have taken mine apart to re-grease it and the internal construction of the helical is not identical to the original Jupiter-3. It is compatible and you could screw in a any Jupiter-3 or Sonnar 50/1.5 in there, but the construction is not identical IIRC. This, on top of making everything out of brass likely has caused them some QA headache since that is a lot of change to do all at once for a first run.

They probably bit off more than they could chew.

 

[Alizarinus]

I made this aperture scale sticker for the re-calibrated jupiter-3+ It was printed on a sheet of address label paper. Attaching the PDF file for printing. Arts-and-crafts strikes again...