peterm1
Veteran
Bit of a long story so please bear with me.
For quite a while I have owned a Canon 50mm f1.2 in LTM but only ever tried it on my M8 once or twice. I expected to have problems with rangefinder calibration and/or focusing generally when shot wide open on a rangefinder camera so I was not in the least bit surprised when I did. The last time I tried this was a few months back and I just put it out of my mind as this really was never its intended use.
Instead I intended to mainly use it on one or another mirrorless camera where I would be shooting at portrait distances where, with focus peaking and image enlargement to assist accurate focusing this never really presented a problem.
Previously I only ever shot at closer portrait distances. Today for the first time I was fooling about and for probably the first time mounted the lens on a Sony NEX and tried focusing full open at f1.2 at infinity only to find that it just would not focus at infinity when shot wide open.
I had expected it to be a little off - as usually lenses are stopped down when shot at infinity and stopping down will often move the focus point backwards due to focus shift. In fact though, it is substantially off - subjects considerably closer than infinity are out of focus too. So for more abundant caution, I swapped out the adapter (I happen to have a couple of L39 - NEX adapters) and found much the same with the second adapter. I then decided to test my Canon 50mm f1.4 in LTM and found that it did not seem to have anywhere near the same problem on either adapter (it just reaches infinity and when stopped down a tad will work fine). Just to be sure, I repeated the test with both lenses with another NEX body as well - and got same result with both lenses. So the problem is not the adapter nor the camera, it is the f1.2 lens only. OK I recognize this is still a bit artificial as one does not normally shoot a fast lens wide open at infinity but remember this was a test.
Having decided that this lens is definitely front focusing at infinity I then decided to see how it performed at more normal portrait distances (where I would be shooting wide open or near wide open more regularly) by comparing the f1.2 with the f1.4 on the same adapter and camera body. I picked a target at about 3 meters (by eyeballing it - I did not measure it) only to find that when focus peaking lights up on the camera the f1.4 lens shows about 2.6 meters on the lens' distance scale for that subject while the f1.2 shows about 3.1 meters. In other words the f1.2 lens is front focusing by 0.6 meters or about 2 feet at that range compared with the f1.4 lens. Thats huge - no wonder the M8 was unable to focus this lens accurately
I will repeat the tests tomorrow with a target at a measured 3 meters and compare results in order to get a more accurate and repeatable result. I know that I should be able to fix this - the 50mm f1.2 is an easy lens to open up and the brass shim inside the lens is easy to remove and in theory easy to make thinner - I just need a flat diamond sharpening stone (I have one) and a bit of time and patience. Great - if all works as it should, with an hour or so of patient work I then I should be able to also use this lens on my M8 as well as on sundry mirrorless cameras as I presently do. It is worth the time, I judge.
My point, I suppose is how the hell does a Canon lens get this far out of calibration? I did some research on the internet and it seems this is not that uncommon for that era of Canon lenses. Other people in other threads reported essentially the same problem. I find it hard to imagine how a manufacturer of Canons standard could have let a lens leave the factory that far off. Or do the Canon bodies of that era have a different spec to Leica bodies and regularly require recalibration? There is nothing about the lens that would suggest its been hacked by a subsequent owner. I have opened it up before to clean the fog off the elements and recall removing then reinstalling the shim (which is presumably the factory installed one) as you would expect. (In fact what this problem suggests is that the shim is too thick not too thin or missing so of course it is still there!)
Anyone else had to deal with this?
For quite a while I have owned a Canon 50mm f1.2 in LTM but only ever tried it on my M8 once or twice. I expected to have problems with rangefinder calibration and/or focusing generally when shot wide open on a rangefinder camera so I was not in the least bit surprised when I did. The last time I tried this was a few months back and I just put it out of my mind as this really was never its intended use.
Instead I intended to mainly use it on one or another mirrorless camera where I would be shooting at portrait distances where, with focus peaking and image enlargement to assist accurate focusing this never really presented a problem.
Previously I only ever shot at closer portrait distances. Today for the first time I was fooling about and for probably the first time mounted the lens on a Sony NEX and tried focusing full open at f1.2 at infinity only to find that it just would not focus at infinity when shot wide open.
I had expected it to be a little off - as usually lenses are stopped down when shot at infinity and stopping down will often move the focus point backwards due to focus shift. In fact though, it is substantially off - subjects considerably closer than infinity are out of focus too. So for more abundant caution, I swapped out the adapter (I happen to have a couple of L39 - NEX adapters) and found much the same with the second adapter. I then decided to test my Canon 50mm f1.4 in LTM and found that it did not seem to have anywhere near the same problem on either adapter (it just reaches infinity and when stopped down a tad will work fine). Just to be sure, I repeated the test with both lenses with another NEX body as well - and got same result with both lenses. So the problem is not the adapter nor the camera, it is the f1.2 lens only. OK I recognize this is still a bit artificial as one does not normally shoot a fast lens wide open at infinity but remember this was a test.
Having decided that this lens is definitely front focusing at infinity I then decided to see how it performed at more normal portrait distances (where I would be shooting wide open or near wide open more regularly) by comparing the f1.2 with the f1.4 on the same adapter and camera body. I picked a target at about 3 meters (by eyeballing it - I did not measure it) only to find that when focus peaking lights up on the camera the f1.4 lens shows about 2.6 meters on the lens' distance scale for that subject while the f1.2 shows about 3.1 meters. In other words the f1.2 lens is front focusing by 0.6 meters or about 2 feet at that range compared with the f1.4 lens. Thats huge - no wonder the M8 was unable to focus this lens accurately
I will repeat the tests tomorrow with a target at a measured 3 meters and compare results in order to get a more accurate and repeatable result. I know that I should be able to fix this - the 50mm f1.2 is an easy lens to open up and the brass shim inside the lens is easy to remove and in theory easy to make thinner - I just need a flat diamond sharpening stone (I have one) and a bit of time and patience. Great - if all works as it should, with an hour or so of patient work I then I should be able to also use this lens on my M8 as well as on sundry mirrorless cameras as I presently do. It is worth the time, I judge.
My point, I suppose is how the hell does a Canon lens get this far out of calibration? I did some research on the internet and it seems this is not that uncommon for that era of Canon lenses. Other people in other threads reported essentially the same problem. I find it hard to imagine how a manufacturer of Canons standard could have let a lens leave the factory that far off. Or do the Canon bodies of that era have a different spec to Leica bodies and regularly require recalibration? There is nothing about the lens that would suggest its been hacked by a subsequent owner. I have opened it up before to clean the fog off the elements and recall removing then reinstalling the shim (which is presumably the factory installed one) as you would expect. (In fact what this problem suggests is that the shim is too thick not too thin or missing so of course it is still there!)
Anyone else had to deal with this?
Musashi1205
Established
Hi, I had mine open a few times to clean the fog.
Obviously one owner lost the shim or took it out on purpose, maybe because it was frontfocusing.
Anyway, mine was backfocusing a lot, infinity was reached when the lens said it’s at approximately 3 meters on the distance scale.
I use a helicoid adapter which works well with my Canon Ltm 35mm f2 and also the 50mm f 1.4 with an A7RIII, so it’s the 1.2 having the issue.
I tried with a thin cardboard and paper cut out to shim it. It’s rough but much closer. It’s focusing just a little over infinity when the distance scale is at the infinity point.
Maybe you could try it with paper shims before you file off the original brass shim?
Obviously one owner lost the shim or took it out on purpose, maybe because it was frontfocusing.
Anyway, mine was backfocusing a lot, infinity was reached when the lens said it’s at approximately 3 meters on the distance scale.
I use a helicoid adapter which works well with my Canon Ltm 35mm f2 and also the 50mm f 1.4 with an A7RIII, so it’s the 1.2 having the issue.
I tried with a thin cardboard and paper cut out to shim it. It’s rough but much closer. It’s focusing just a little over infinity when the distance scale is at the infinity point.
Maybe you could try it with paper shims before you file off the original brass shim?
AlwaysOnAuto
Well-known
A quick simple way to see which way you need to go with the shim is to take it out completely and put the lens back together and try it that way.
peterm1
Veteran
I had the same thought yesterday, but I will tell you about that shortly.
First I should say that I wanted to set up a more controlled situation to do some more extensive testing at 3 meters - I have found this distance to be useful in the past as (a) it approximates portrait distances (b) the requirement for focusing accuracy is more critical at shorter distances and (c) most rangefinder lenses will have 3 meter (or 10 feet which is reasonably close enough) marked on their distance scale so it's easy to scale focus the lens in order to compare with the focus according to the rangefinder when using an M camera - especially an M digital where results are immediately available for review.
So.... I set up a target at 3 meters which I carefully measured.
First, I tested a few other RF lenses on the M8 to see if they were focusing correctly (I expected the RF on the M8 to be reasonably accurate as I had it adjusted by a technician a couple of years back and have seldom used it since. I am just now getting keen to use it once more. Never the less I wanted to be comprehensive and check). I also wanted to check using both native M lenses and adapted LTM ones to ensure that any focus discrepancy was not in the thickness of my Leica badged LTM - M mount adapter used - I thought this to be unlikely but I have had one previous such made adapter which was off tolerance so better safe than sorry.
Here are the results - all shot at the widest aperture of each lens where any discrepancy will show up most strongly. Where I say below that the image is "sharp" this was assessed using the M8's rear LCD which, as users would know, is only approximate because the M8 LCD does not have high resolution. I could get more accurate results by exporting images to the PC and checking on screen but this would need a more extended test, better note taking and more time. As I get closer to a result I may do this though to get maximum sharpness in the final outcome.
TEST RESULTS:
- Summicron 50mm f2 V3 Rangefinder matches lens scale at 3 meters. Image sharp (This lens was calibrated with the M8 when I had its rangefinder last adjusted so this outcome is welcome but not unexpected).
- Summarit 50mm f1.5 When the rangefinder indicates it should be sharp the scale indicates a little less than 10 feet. i.e. approximately correct but not quite. Image seems usable though perhaps a little off maximum sharpness as would be expected given the above.
- Canon 50mm f1.4 When the rangefinder indicates it should be sharp the scale shows a fraction over 3 meters. Image sharp.
-Canon 50mm f1.2 When the rangefinder indicates the image should be sharp the scale shows a fraction over 3 meters. i.e at first blush it seemed to be almost correct. But in reality on checking the image, that image is completely blurred and totally unusable. WAY off.
- I then tried focusing the Canon 50mm f1.2 by scale (ignoring the rangefinder) and shooting successive test images till I began to get sharp(ish) images. I found I had to turn the scale to about 7 meters before the images shot at 3 meters were reasonably sharp ???!!! (Of course the rangefinder said the lens was totally out of focus when I did this). In other words the lens must be front focusing significantly compared to the scale. Of course it also corresponds to why the lens could not focus all the way to infinity (which is what caused me to notice the problem in the first place) - as when the lens was rotated to infinity and could rotate no further because a hard stop had been reached, the actual point of focus was still somewhere short of this.
- Yesterday as noted above, I had already tried what you suggested and removed the shim completely to test focus without it - but did not make notes (my bad). The result was so anomalous it threw me into a headspin. It seemed as if it was STILL front focusing even with no shim.
I had expected the lens to now back focus without a shim installed, which would mean that I would need to thin the existing shim or make a new thinner shim and install that in place of the original. But this is not what I seemed to find. Having said that it was late at night, and I had had a vodka (or two) and so decided to re run that test today but have not yet done so. It is possible that I made a mistake or misinterpreted the result. If however the result is replicated later today when I rerun it this test it would rather suggest to me that given that when I tested the lens on the M8, the camera's rangefinder and the lens scale almost coincided but the image was well out of focus that the problem is not with the shimming but that someone - a previous owner maybe has fritzed about with the optical elements. (I have cleaned the lens of haze but only the rear elements and these are definitely correctly placed. It's the first thing I checked, besides both rear element groups are contained in a metal ring which only allow you to screw them in one way so you cannot fritz it even if you try).
I do not know what the problem could be in this case as the optical block diagram does not suggest that it is in anyway possible to (say) install any front element backwards or something silly of this sort. In that case I will be stumped unless someone here has some ideas. (The good news of course is that the lens does work OK on M4/3 or NEX cameras with adapters except at infinity where it falls short at wider apertures.)
peterm1
Veteran
I just had an epiphany.
A blinding flash of realization that points to something SO obvious I am kicking myself. I was grasping at straws and going down for the third time when suddenly I understood - with this magnitude of misfocus, the problem is, the problem must be, the problem only can be....... the helicals are incorrectly assembled and misaligned. My only excuse (and not a very good one) is that I have been so fixated on the shimming of the lens that the answer has taken a while to sink in. It is not the shim - it's the helicals. DOH! Someone unqualified dodderer has taken the lens apart to lubricate the helicals and has not properly restored them to their original position vis a vis each other. It's so obvious that I am kicking myself for not recognizing it before. (Although it is only obvious once the magnitude of the error has become apparent and that was not really apparent until this morning - that's my other "back up" excuse and I am sticking to it).
I know how easy it is to disassemble a lens' helicals and then fail to properly mark where they separated. If that happens, reassembly is a pain requiring potentially hours of trial and error to get it back together with the lens able to focus to infinity. I know because I have made the requisite mistake that causes this problem (and in my case I then managed to fix the problem I caused - smug look
).
This was with an old Triotar 135mm f4 which needed a CLA because it was locked up tight due to it not being serviced for years. My Canon lens on the other hand, is focusing silky smoothly and that should have been a clue that it most likely had been CLA'd which in turn should have warned me the helicals are the most likely source of the problem.
OK - now I have to consider spending the time and effort to do it properly. Or I must ante up the oofs, the moolah, the dosh, the dough, the sponduliks, the gelb, the ducats, the plunder, the shekels, etc to pay someone to do it for me.
DAMN! Don't I feel like a bally twit!
A blinding flash of realization that points to something SO obvious I am kicking myself. I was grasping at straws and going down for the third time when suddenly I understood - with this magnitude of misfocus, the problem is, the problem must be, the problem only can be....... the helicals are incorrectly assembled and misaligned. My only excuse (and not a very good one) is that I have been so fixated on the shimming of the lens that the answer has taken a while to sink in. It is not the shim - it's the helicals. DOH! Someone unqualified dodderer has taken the lens apart to lubricate the helicals and has not properly restored them to their original position vis a vis each other. It's so obvious that I am kicking myself for not recognizing it before. (Although it is only obvious once the magnitude of the error has become apparent and that was not really apparent until this morning - that's my other "back up" excuse and I am sticking to it).
I know how easy it is to disassemble a lens' helicals and then fail to properly mark where they separated. If that happens, reassembly is a pain requiring potentially hours of trial and error to get it back together with the lens able to focus to infinity. I know because I have made the requisite mistake that causes this problem (and in my case I then managed to fix the problem I caused - smug look
). This was with an old Triotar 135mm f4 which needed a CLA because it was locked up tight due to it not being serviced for years. My Canon lens on the other hand, is focusing silky smoothly and that should have been a clue that it most likely had been CLA'd which in turn should have warned me the helicals are the most likely source of the problem.
OK - now I have to consider spending the time and effort to do it properly. Or I must ante up the oofs, the moolah, the dosh, the dough, the sponduliks, the gelb, the ducats, the plunder, the shekels, etc to pay someone to do it for me.
DAMN! Don't I feel like a bally twit!
peterm1
Veteran
BTW this lens is worth fixing. While low contrast (made even less contrasty by the poor state of the coating on the front element of my specific lens) it does make nice images and low contrast is easily fixed in post - as I have done here.
Taken in my dojo.
Modern Samurai in Training by Life in Shadows, on Flickr
Taken in my dojo.
Modern Samurai in Training by Life in Shadows, on FlickrLightshow
Member
When I received my Canon S 50/1.2 the aperture ring was off, it was wide open at f2, as much as it bothered me I left it like that for a year or two, I finally broke down and opened it up to fix it, all better now.
I also received a Helios 40 with an element flipped and was touching the adjacent in the center, there was obvious damage to both elements from the trip over the pond. It was "professionally serviced" prior to being listed for sale.
So I think there are many many many lenses not put together correctly out there.
I also received a Helios 40 with an element flipped and was touching the adjacent in the center, there was obvious damage to both elements from the trip over the pond. It was "professionally serviced" prior to being listed for sale.
So I think there are many many many lenses not put together correctly out there.
peterm1
Veteran
I don't think I ever finalized this story. I took the lens into my camera guy who carefully thinned down the spacer ring in the lens. It took three tries but eventually it came good and is now spot on when mounted on my Leica M. Using a grinding plate he hand ground 2/10ths of a millimetre off the ring.
In these circumstances this is actually a heck of a lot to remove and it raises a mystery. Did the lens leave the factory like this all those years ago and if so how did they get it so wrong, and how did it not get picked up in Canon's QA process. Or was the spacer ring somehow swapped for a wrong ring (one which was 2/10th of a mm too thick ???!!!) sometime during its life. The lens seems to have had quite a lot of use so it seems likely that the wrong ring was inserted sometime later in its life - otherwise it could not have been used on rangefinder cameras and would not have acquired its marks of honourable use. Still it's a mystery even so as to how a wrong spacer ring could somehow wind up inside a lens even if it happened later in life. A harrowed and overly busy technician perhaps?
In these circumstances this is actually a heck of a lot to remove and it raises a mystery. Did the lens leave the factory like this all those years ago and if so how did they get it so wrong, and how did it not get picked up in Canon's QA process. Or was the spacer ring somehow swapped for a wrong ring (one which was 2/10th of a mm too thick ???!!!) sometime during its life. The lens seems to have had quite a lot of use so it seems likely that the wrong ring was inserted sometime later in its life - otherwise it could not have been used on rangefinder cameras and would not have acquired its marks of honourable use. Still it's a mystery even so as to how a wrong spacer ring could somehow wind up inside a lens even if it happened later in life. A harrowed and overly busy technician perhaps?