Leica/FSU lens focusing incompatability

[Paul Connet]

There has been a good deal of talk about the reasons why most FSU lenses do not focus properly on Leica or Canon bodies. Usually when this subject comes up, someone mentions a technical explanation presented on Dante Stella's site.

(http://www.dantestella.com/technical/compat.html)

Dante's discussion centers around the idea that the Soviets took a path of least effort when they designed the Fed and Zorki LTM camera bodies, and while taking most of the camera mechanical design from the early Leica II, they designed their LTM lenses using the Contax standard focal length of 52.3mm instead of the Leica standard focal length of 51.6mm, and this is supposedly the source of the focus problems when trying to use a FSU lens on a Leica/Canon body or trying to use a Leica/Canon lens on a FSU camera body.

I personaly find that argument to be questionable and would like to hear from members of the RFF what their feelings are. Nowhere have I run across a statement from a known expert in Soviet camera design to either counter or confirm Dante's argument.

What say you?
Paul C.

 

[Kim Coxon]

I had 2 "new" black J3's from a seller in Siberia. Both worked perfectly throughout the range on both Canons and a Bessa R body. I have set up several J9's that do the same and very few people have problems with J8's and the Indy's. IMHO, the problem has more to do with quality control and the amount of home repairs carried out before many of these came on the market. Many lenses were also "tuned" to the body, hence the "passport" system stating what the lens was set to. This would also "throw" them out.

Kim

 

[Paul Connet]

The Deviation in Focal Length from what a Leica (or Canon) RF is calibrated for will cause the focus to diverge from the setpoint. The difference in the lenses focal length from 51.6mm will decrease/increase the walkoff. I've tested 7 lenses on three cameras with film so far, so a mechanical fault in the cam is not likely. Milling the cam to compensate for such a slight difference would be hard. I'm experimenting with grinding down the lens barrel to space in the rear group closer to the front group. (no shim to space the rear modules that I've seen). If that works for the "basket case lens", Ill try it one one that walks off slowly.

Brian,
No disrespect intended, but I suspect your reworking of the lenses to correct for one defect is creating another. I most certainly do not know how the lens designers compensate for focal length variations, but I do know they exist. In Erwin Putts book on the Leica lenses he shows a table that lists the code number which is marked on certain lenses to indicate the actual focal length of the lens. The code numbers for 50mm M-lenses run from 00 to 22 which covers the range of 50mm to 52.2mm. Interestingly the Leica 75mm, 90mm &135mm lenses have their own code numbers to indicate the actual focal length of a particular lens.

I have 2 such coded Leica lenses, a 50 Summicron with code 19 which means it has a focal length of 51.9mm, and a 90 TE which has a code of 10 which means it has a focal length of 91mm. Obviously the Leica designers either found that those focal length variations could be accomodated by lens assembly adjustments, or that the focal length variations were within acceptable tolerances for use on any camera body. Certainly my M6 doesn't know that my 50 Summicron is actually a 51.9mm.

My own experience with FSU lenses used on Leica or Canon bodies has been mixed with some good and some not so good. Usually after a good home CLA they are better than when received. A good lube job takes out the slop. Using Leica or Canon lenses on the FSU bodies is even better in my experience because the bodies can easily be adjusted for both close focus and infinity.

Bottom line is that I am in agreement with Kim, that most problems stem from QC during manufacture, or after tinkering by a user.

I hope someone chimes in with the real answer.

Regards, Paul C.

 

[Kim Coxon]

Hi Paul,
There is a big difference with Leica lenses and the FSU 50's. To simplify things, the soviets mount the lens unit direct to the cam mech in most 50's. You can separate them otherwise you wouldn't be able to focus any focal length but 51.6 but in these cases you must have some sort of seperate mech for the Cam. I suspect the reason Leica put a code number on is so that the focal length can be matched to the right cam so the focus does follow. In the simplest case, the I22, the basic lens tube was the cam which is why eary versions had to be matched to the body. The FSU lenses including the J9 only have one cam mech with no adjustment so the lens focal length must be made to match this to avoid problems. One of the big reasons why you can't mix and match FSU lens components although I suspect some have tried in the past which is another reason for the problems. So the defect is there and what Brian and I do is to minimise that defect to make a usable lens. In other words we are trying to undo what others have done in the past.

Kim

Brian,
No disrespect intended, but I suspect your reworking of the lenses to correct for one defect is creating another. I most certainly do not know how the lens designers compensate for focal length variations, but I do know they exist. In Erwin Putts book on the Leica lenses he shows a table that lists the code number which is marked on certain lenses to indicate the actual focal length of the lens. The code numbers for 50mm M-lenses run from 00 to 22 which covers the range of 50mm to 52.2mm. Interestingly the Leica 75mm, 90mm &135mm lenses have their own code numbers to indicate the actual focal length of a particular lens.

I have 2 such coded Leica lenses, a 50 Summicron with code 19 which means it has a focal length of 51.9mm, and a 90 TE which has a code of 10 which means it has a focal length of 91mm. Obviously the Leica designers either found that those focal length variations could be accomodated by lens assembly adjustments, or that the focal length variations were within acceptable tolerances for use on any camera body. Certainly my M6 doesn't know that my 50 Summicron is actually a 51.9mm.

My own experience with FSU lenses used on Leica or Canon bodies has been mixed with some good and some not so good. Usually after a good home CLA they are better than when received. A good lube job takes out the slop. Using Leica or Canon lenses on the FSU bodies is even better in my experience because the bodies can easily be adjusted for both close focus and infinity.

Bottom line is that I am in agreement with Kim, that most problems stem from QC during manufacture, or after tinkering by a user.

I hope someone chimes in with the real answer.

Regards, Paul C.

 

[doubs43]

I just sent Brian images that I took with a J-9 on a Leica body plus one taken with a 9mm Elmar wide open at f/4. The J-9 images are at f/2 & f/4. The J-9 should make a fine portrait lens for older women in particular as it's soft in a nice way. The Elmar is clearly sharper on the Leica body.

The camera was tripod-mounted and the distance to the sign was exactly 6 feet.

The images are cropped from a much larger picture and sharpened a bit. They were originally scanned at 2400 dpi.

From left to right the pictures are: J-9 @ f/2; J-9 @ f/4 & Elmar @ f/4.

Walker

 

[jlw]

I have never understood Dante's comment. Why does focal distance have anything to do with focusing accuracy?

Because lenses of different focal lengths require different amounts of extension to focus from infinity to close-up. You've seen this on a large scale when you consider a wide-angle lens and a tele. Set a wide-angle lens to infinity, measure its length, then set the focus to, say, 2 meters and measure again; you'll notice it gets only slightly longer. Try the same experiment with a 100mm lens and you'll notice the length increases much more.

This is because the amount of focusing extension is proportional to the focal length.

(A pretty good formula for how much the lens needs to be extended to focus on a given distance is:

x' = f^2/p-2f

where x' is the amount of focusing movement required, p is the object distance from the film plane, and f is the focal length of the lens.)

So, lenses of different focal lengths need to be extended by different amounts to focus on a given close distance. But obviously, the coupling mechanism of your camera's rangefinder can only be set up to handle one "extension profile."

So how does the camera designer handle interchangeable lenses, with their different extension requirements for focusing? By picking one particular focal length, designating it as "normal," and designing the rangefinder to correspond to the in-and-out movements of that focal length.

To accommodate lenses of other focal lengths, the designer incorporates some sort of differential mechanism to "translate" the lens' focusing movements so that the movement of the coupling arm matches the amount it would move for the "normal" lens. In the case of a wide-angle lens, which requires less movement, the lens' in-and-out movement is proportionally amplified so the arm moves as much as it would for the normal lens. In the case of a tele, which requires more movement, the lens' in-and-out movement is proportionally reduced so the coupling arm only moves as far as a normal lens would move it to achieve the same focused distance.

One common way of doing this is via a ramped cam, of the sort you can see on the back of a 40mm Summicron-C lens. This cam is affixed to the focusing ring of the lens so that it turns as the lens is focused. The camera's rangefinder coupling arm rests against the cam, so it's the cam's position that controls the in-and-out movement of the coupling arm. Because the cam is ramped, it moves the coupling arm back and forth more than the lens itself moves back and forth -- compensating for the lens' shorter focal length compared to the designer's chosen "normal" focal length.

Various RF-coupled lenses have different ingenious ways of doing this proportional translation, and there's an article on Dante's site that discusses some of them.

The important thing to remember, though, is that these coupling mechanisms all are set up to work in proportion to the defined "normal" focal length -- in other words, the designer's chosen normal focal length is always the reference. If the body's rangefinder was designed for one reference focal length, but the lens' coupling cam is cut for a different reference focal length, then the cam will provide the wrong amount of coupling-arm movement as the lens moves in and out.

That's why focal length makes a difference in rangefinder coupling. The difference between 51.6mm and 52.3mm, or whatever, may not sound like much -- but it matters, because an optical rangefinder is exquisitely precise (which is why they're so expensive.)

How precise? Well, you've probably done the experiment (and if you haven't, try it now) of putting your camera on a tripod, carefully lining up the rangefinder on an object at a specific distance, then putting a piece of transparent tape on the coupling cam; after you've done this, you'll notice that the rangefinder is now slightly "off," because of the thickness of the tape. That tape is actually only about 0.04mm thick (I just measured some) yet it makes a visible difference in how the rangefinder lines up!

 

[ZorkiKat]

There has been a good deal of talk about the reasons why most FSU lenses do not focus properly on Leica or Canon bodies. Usually when this subject comes up, someone mentions a technical explanation presented on Dante Stella's site.

(http://www.dantestella.com/technical/compat.html)

I personaly find that argument to be questionable and would like to hear from members of the RFF what their feelings are. Nowhere have I run across a statement from a known expert in Soviet camera design to either counter or confirm Dante's argument.

What say you?
Paul C.

Paul

I too find his arguments questionable. It has been some time since I read them, and perhaps I should read them again. But my general impression is that he made a sweeping comment about all FSU LTM lenses based on one or two samples.

I find it hard to believe too, his statements which made it appear that the Soviet lens designers and makers oversimplified the LTM system and blindly adopted the
Contax lenses' (after which most of the Industar and Jupiter lenses were patterned) camming without considering Leitz's.

The Soviets were making Leica copies (the FED) long before they were making Contax (Kiev) camera and lens copies. In those 10 or so years of headstart, they likely learned enough to understand how the Leitz system worked. It would not be acceptable to summarily dismiss all their efforts into something as simple as what Mr Stella had said.

Most focus issues, as others have already noted, stem from incompetent servicing of the lenses. The most problematic lenses, namely the J-3 and J-9, often suffer from what appears to be focusing incompatibility due to improper assembly/disassembly or unauthorised replacement of parts. The latter often happens when sellers attempt to sell a "good" lens which had been made from contributions from other lenses. A new lens, so to speak, from good parts coming from other lenses.

I've several Jupiter-9 in Leica M39 mount. Many did not focus right when they came, but eventually did so after reassembling or reqreasing them. Only one (the optic block was th problem) did not.

The Jupiter-9 -and the Jupiter-3 as well- had complex focus helicals. The J-9 consists of three barrels interacting with each other and how these barrels are
assembled determine how accurate focus movement, camming and RF coupling would be. There is only one way that these barrels can be assembled correctly and perhaps a dozen more incorrect ways, having multi-entry barrels and such. It is hard to really see if they went in incorrectly, and only the lens's performance can tell. I've found many whose focus cam /coupler jut out too far or too short - this is an indication of incorrect repair.

The same can be said of the 50mm Jupiter-3. In all cases, sometimes the way the lens helicoids is lubricated can affect how it focuses. I would stay clear of lenses whose helicals move very smoothly. The Soviets did not care about the "Leitz Precision" in terms of fitting helicals. They found that using a certain type of grease which both lubed and dampened the movements could make the helicals do the job with less precision. Replacing the lube with something more slick and smooth may well cause the lens movements to falter. The best working Jupiter 9 LTMs I've got have rather sticky focusing barrels. I left them that way.

I would suspect too that the FSU cameras mentioned there used for the tests were not calibrated right. One contributor described his calibration method which sounds (IMO) very different from the accepted procedures of Soviet Leica rangefinder calibration methods. The way he altered the camera's RF sensor tip bothered me.

When I started getting my J-9s and FED-1's I thought that there was something wrong with the lenses as they did not focus right even with the FED. That was before I heard of the necessity to adjust Soviet Leicas both at infinity and 1 metre focus settings - something which wasn't done with Leica or Canon clones. When I adjusted the FED (and my other Soviet Leicas), the focus started gettting right. Not just with the J-9, but with Jupiter 8 and 3, and as well as other large aperture or long-focus lenses from Leica or Canon (if they fit- rf coupler is the issue, but that's another story).

Finally, what even makes Mr Stella's arguments hard to take is the fact that I have several Jupiters which FOCUS RIGHT in my Leica III, Leica IIIc, Canon IVSB, and yes, a Leica M3! I've seen them work right in a Bessa R2 as well. And this is not just one lens we're talking about- these are several lenses of the same types (which would make a justifiable population sample) which confirm compatibility of
FSU LTM lenses and Leicas.

My Leica M3 has a Jupiter-3 mounted on it almost permanently. Since I 'lost' (well sold,) my Summicron, and replaced it with a $20 Jupiter -8, I felt that I had a replacement which was as good or even better. The I-61 L/D, though 1 stop slower worked fine as well. But when I got a J-3, my camera seemed to be happy with it and decided to marry them off! 🙂

Jay

About three years ago, I posted some test exposures made with a Leica M3 and several Soviet lenses at

http://www.pbase.com/zorki/jupiter&page=2

the photos posted were scanned from prints, so there's some loss of quality. The fact that the other photos were from Tri-X negatives pushed to EI 1600 did not help the issue much, but nonetheless show what a Jupiter-3 can do in 'real life' situations.

 

[jlw]

I find it hard to believe too, his statements which made it appear that the Soviet lens designers and makers oversimplified the LTM system and blindly adopted the Contax lenses' (after which most of the Industar and Jupiter lenses were patterned) camming without considering Leitz's.

The Soviets were making Leica copies (the FED) long before they were making Contax (Kiev) camera and lens copies. In those 10 or so years of headstart, they likely learned enough to understand how the Leitz system worked. It would not be acceptable to summarily dismiss all their efforts into something as simple as what Mr Stella had said.

I didn't interpret his article as saying they had oversimplified the system. What I thought he said was that after the Soviet Union got into the Contax business (via its acquisition of the Zeiss-Ikon production line machinery after the fall of Nazi Germany) its planners made a decision to adopt the same focal-length standard for all their camera lines.

This type of centralized decision-making would be very typical for a planned economy, and it would be very sensible from the viewpoint of the planners since it would allow more flexible sourcing of parts. The fact that the lenses wouldn't be fully compatible with Western cameras wouldn't be a great concern in this case, since they were intending the lenses mostly for their own camera bodies.

I'm not saying that is what happened, only that it would have seemed to be a very logical way to handle the situation from the standpoint of industrial planning.

 

[Paul Connet]

I didn't interpret his article as saying they had oversimplified the system. What I thought he said was that after the Soviet Union got into the Contax business (via its acquisition of the Zeiss-Ikon production line machinery after the fall of Nazi Germany) its planners made a decision to adopt the same focal-length standard for all their camera lines.

This type of centralized decision-making would be very typical for a planned economy, and it would be very sensible from the viewpoint of the planners since it would allow more flexible sourcing of parts. The fact that the lenses wouldn't be fully compatible with Western cameras wouldn't be a great concern in this case, since they were intending the lenses mostly for their own camera bodies.

I'm not saying that is what happened, only that it would have seemed to be a very logical way to handle the situation from the standpoint of industrial planning.

I am happy to see an expanded discussion of this topic as it seems mostly we don't actually know what the Soviet camera people did or did not do with regard to focal length matters.

Jay makes a good point about the early start on LTM copies of the Leica even before the war and that they probably did discover the design intricacies necessary to have functional cameras. We know that the very early FED 1 had custom matched body/lens with non standard dimensions, (film to lens flange, etc.) But somewhere along the way they did standardize using Leica dimensions at least for the body. So why not the lens?

We see many references to the 52.3mm focal length of a lot of their normal lenses. Even if they did adopt that length as "standard" for all normal lenses, we also know that they adopted the Leica lensflange to film plane dimension of 28.8mm. It would be no more difficult to design the lens rf cam shape to accomodate 52.3mm than it would be to accomodate 51.6mm. Obviously Leica has done so with their coding system to accomodate the varying focal lengths of their 50's, 75's, and longer tele's.

I still stand with Jay that the Soviets knew how to design cameras, and IMO most problems stem from QC either in the factory or post manufacture by tinkerers.

Regards, Paul C.

 

[Peter Klein]

On Dante's site, there is some additional material by a gentleman who tried a large number of lenses, both FSU and Leica, on both FSU and Leica bodies. His results were consistent with Dante's hypothesis. But I grant you that there are many variables, so it is possible to believe anything one wishes.

My own tryouts of Soviet lenses, and the majority of those done by people I know, have all turned out the same: Either the lens focuses correctly close-up but not at a distance, or vice-versa. Either way, the lens acts like it is slightly longer than the RF thinks it is. That indicates to me that there is an most likely some systemic incompatibility. It's also possible that there's bad QC, bad repair, bad assembly, switching of parts, etc. Or that newer lenses are made to the Leica standard and older ones aren't. Or some combination of all of the above.

Me, I'm a practical cat. In the final analysis, it doesn't matter *why* the focusing errors occur. The important thing is that they *do* occur, and I think they occur more often than they don't. The bottom line is that if you want to use an FSU lens on a Leica-standard camera (including Bessas and Canons), you have got to be either lucky, or a good lens tinkerer, or both. And you may be disappointed anyway. Or, if you only shoot outdoors at f/4 and smaller, it may not matter.

The good news is that a good FSU lens and body can be adjusted to each other. And then they work at all distances. Good folks like Oleg and Fedka can get you such a combination.

You want to tinker, have fun. Me, I like to confine my lens experiments to how they render photographs when they focus on what I intended. Or as the late DeForest Kelley (a.k.a. Dr. McCoy) might say, "Dammit, Jim, I'm a photographer, not a lens technician!" 😀

--Peter