Canon lenses on r-d1 front focusing?

[Pinphot]

I've recently bought a couple of Canon rf lenses on ebay. A 50/1.8 and a 35/1.5 for very decent prices, (£45 and £120 respectively).
One problem i have is that both lenses appear to front focus, and to get around this problem, I have to defocus the rangefinder spot by focusing slightly towards infinity past what the rangefinder says is correct focus. All my other lenses, 21/4VL, 28/1.9VL, 50CF Summicron and 90 Tele Elmarit are absolutely spot on on the r-d1. Is this a known problem with Canon lenses on the r-d1? I've done a forum search and can find no reference to this issues. I have a feeling that the flange to focal plane distance is marginally too long and that the Canon lenses will not focus to infinity, but somewhere in front. When the lenses are focused properly though, they are both outstanding. The 35/1.5 being far better than I was led to believe by casting around on the net. Stopped down, it is as sharp as the Summicron. I can live with this issue, by remembering to defocus the lenses towards infinity past what the rf says is correct focus, but this is a pain and a bit hit and miss. Any opinions?

Thanks,

Mark

 

[Terao]

I haven't noticed it with my 50 f/1.2 wide open but then I haven't gone looking for it either. Have certainly got sharp centre focus with that lens at close range...
Certainly hope its not a common factor, I have 28,35, 50, and 100 Canon lenses on the way 🙂

 

[ferider]

I am absolutely sure this 35/1.5 front-focuses heavily ! Get rid of it ... Sell it to me 😀 Where do you find deals like this ?

(I have good experience with most Canon lenses. The 50/1.2 can be off sometimes, I had to have one re-collimated of the two that I used).

Roland.

 

[jlw]

I'm not sure about other focal lengths, but I've noticed this front-focusing behavior when using all my Canon 50mm lenses (1.4, 1.2, and converted 0.95) on my R-D 1. (The 85/1.5 and 100/2 don't seem affected, but that could just be because I haven't tested them as rigorously.)

My theory -- which seems to be borne out by specs I've read in Canon repair manuals -- is that there was a time period during which Canon set up its cameras and lenses using slightly different assumptions about film flatness etc. than did, for example, Leica or Cosina. The result seems to be that a Canon lens that focuses perfectly on a Canon may front-focus very slightly when used on an R-D 1.

(Of course, another possible cause of this behavior -- particularly with non-50mm lenses -- is a too-thick screw-to-bayonet adapter, which is something we've discussed extensively before in the R-D 1 forum. I suggest you check your adapters' thickness before you try anything else. I've never had any problems with genuine Leitz-brand adapters, but have bought several non-OEM adapters that turned out to be slightly too thick. Correct thickness between the mounting flanges is 0.98-0.99mm, which can be measured with a micrometer if you've got one.)

If it turns out that the issue is with the lenses themselves, you've got some decision-making to do. I'm not sure about the earlier Canon lenses, since I don't own many of them, but many of the later, mostly-black ones have their base extension determined by a "collimation shim" -- a precisely-dimensioned brass ring that sits between the lens head and its mating surface on the focusing mount.

After I had my 50/0.95 converted to M mount, for example, I discovered it front-focused slightly on the R-D 1. (The technician who had done the conversion had set it up on a Leica M2 and swore it focused perfectly on that.)

So, I "corrected" the lens for the R-D 1 by removing the lens head from the focusing mount, extracting the collimation shim, and thinning it very, very slightly (a few thousandths of a mm) by rubbing it against a sheet of fine abrasive paper on a flat surface. I'd remove a very small amount of material, reassemble the lens, and do some test shots until I sneaked up on the correct amount of thinning. Once I did this, the lens focused perfectly on the R-D 1.

The reason you need to think about this, though, is that if you ever decide to use the lens on a Canon again, it will no longer be adjusted exactly the way the factory intended. I've never actually tested any of my "thinned" lenses to see whether or not they're still within tolerance on a Canon RF body. It might well be that they're still fine, but I don't know for sure.

Good luck...

 

[foto_fool]

My Canon 50/0.95 TV was collimated on an M3 by the tech that made the conversion. It is perfect on my M6 and Hexar RF, but front-focuses badly on the R-D1.

I expect the M6 and Canon to outlive the R-D1 so I'm not going to re-shim the Canon. The ZM 50mm Sonnar is only a couple of stops slower and gives a similar look wide-open - and works perfectly on the R-D1.

 

[Pinphot]

Thanks for the replies chaps. Rangefinder mechanics appears a very arcane science. One would have thought that a range of lenses that worked on one platform would also work on another, in that the assumption would be that the cam movements etc would need to be universal, especially if another range of lenses eg Voigtlander or Leica seem to work well across various marques. Obviously, it's rather more complex than that and beyond my capability to understand. Were there ever compatibility issues using Leica made screw mount lenses with Canon rf bodies? I'm in two minds about what to do with these lenses. 35mm on the r-d1 is not a focal length that I am really bothered about, as 50mm in my film days was never a favourite focal length of mine, it was neither wide enough nor long enough for where i liked to stand. I only got the 35mm Canon as the price seemed too good not too, especially for the speed. 50mm on the r-d1 however, (75mm equivalent), is a lovely length, but is obviously served very well by my CF Summicron which suffers absolutely no problems at all with dicky focus issues.

Thanks again,

Mark

 

[jlw]

In general lenses are pretty much freely interchangeable among compatible RF cameras, but occasionally you do get into fringe situations caused by the fact that different manufacturers have made different assumptions about the tolerances needed.

Here's a link (click here) to an article by Dante Stella about a similar situation between Leica and Konica M-mount film cameras, showing that the occasional problem isn't digital-specific.

 

[Terao]

I've just tested my scuzzed up 50 f/1.8 & mint 50 f/1.2, both wide open @ ISO800. No evidence of front focus, fine detail about three metres away is super-sharp and in focus. Stuff like addresses on envelopes and wine bottle labels, box art, etc. Both lenses were mounted in a genuine Leica LTM-M adapter (a 50/75 not that that matters). This adapter works equally fine for my non-Canon LTM lenses (couple of Voigtlander wides and a Chiyoko Super Rokkor 45mm f/2.8) so its not a case of the R-D1 backfocusing a little and cancelling out any Canon-specific front focus...
Of course YMMV but this thread had me worrying 🙂
I'll happily take that 35mm f/1.5 off your hands 🙂

 

[ferider]

My Canon 50/0.95 TV was collimated on an M3 by the tech that made the conversion. It is perfect on my M6 and Hexar RF, but front-focuses badly on the R-D1.

I expect the M6 and Canon to outlive the R-D1 so I'm not going to re-shim the Canon. The ZM 50mm Sonnar is only a couple of stops slower and gives a similar look wide-open - and works perfectly on the R-D1.

John,

you might want to have that RD1 looked at, instead of the lenses.
If it's not a very new C Sonnar, it should be off 6cm wide open and close up ...

Sorry & best,

Roland.

 

[rolleistef]

wow a 35/1.5 for 120£! It normally costs more than 4/500£!! that's a one-in-a-life bargain!

 

[Pinphot]

Lens fixed for r-d1 usage

Lens fixed for r-d1 usage

Again, thanks for the replies chaps.
I took a decision as regards jlw's post and ground down the shim, which was a much easier job than I thought it would be. Admittedly, I was groping around in the dark regarding disassembly, but somehow or other I got it right.
I unscrewed the black ring directly adjacent to the edge of the rear element which allowed the body of the lens to be removed from the focusing mount. The brass shim in question came away with the lens body, being around the tube holding the rear elements. I removed the shim and ran it over some very fine emery paper a few times and then reassembled the lens and shot some photos of different subjects at various distances and focus settings wide open, (generally from minimum to maybe 7 or 8 feet), just so I had a variety to average out and take into account focus error on my part. I then opened these in photoshop and found that on average the lens was still front focusing, but not quite so badly. I continued this exercise disassembling, grinding, reassembling 3 or 4 times until the test shots showed an average around optimal focus, it now seems that I have a lens which focuses as close to perfect on my r-d1 as I should wish when discounting error on my part.

This lens is really nice, it's not razor sharp wide open, but stopped down to f2.8-f5.6 is absolutely brilliant. The lens has a very smooth rendering which appears to give a very usefull extension to dynamic range on dig'. In any instance, it's as good as a series 1 summilux, (in a diiferent way), without all the flare.

I'm sorry to disappoint all those people who've been contacting me desperately wanting this lens, but I now really think I'm going to keep it. By the way, did I mention it also came with a Walz UV filter and VL bayonet adapter😀

I was very lucky to get this lens. I put the term 'canon rangefinder' into e-bay last week and it popped up as a buy it now item at £120 with 9 days left to run, I think it had only been up a matter of minutes, before I snapped it up. I had visions of every rangefinder user in Britain desperately trying to log in as quickly as humanly possible at that moment in time to get it!! That said, there are a few very minor scratches on the glass, the body is well used, and the focusing action needs a relube, but now that I seem to have fixed it for use on my r-d1, I'm very happy.

Thanks,

Mark

http://www.thebppa.com/Mark-Pinder
http://www.unp.co.uk/photographers/photographer.php?title=mark_pinder

 

[jlw]

Glad this worked out for you, since even I probably wouldn't have had the nerve to thin down a valuable 35/1.5! (even though it could always be shimmed back to the original specs; very thin brass stock is available at hardware shops for this purpose, and it's so thin it can be cut easily with a hobby knife.)

Good guess on the disassembly procedure, too. In general the mostly-black Canon lenses are easy to disassemble, thanks to a very sensible form of modular construction. Somewhere on RFF there's a thread with pictures showing how a 50/1.4 comes apart; sounds as if the 35/1.5 is similar.

Incidentally, now that you know how to remove the optical section, it should be much easier to relubricate the focusing mount, since you'll be able to take the optics completely off and remove them to a safe distance.

 

[Terao]

Hmm, currently reading up on the Leica thread mount (not much to read granted!). Is it a imperial/metric conversion issue? This page mentions that the Canon screw thread is M39 x 1mm. LTM is M39 x 26 threads per inch Whitworth. 1 inch is 25.4mm so if those specs are correct then the Canon thread is ever so slightly out of spec. Enough to cause front-focus? No idea, but given how others have mentioned how critical LTM-M adapter thickness is perhaps it is. Sounds similar to the old Nikon vs Contax mount debate. I guess that fast lenses wide open are just getting critical enough dof-wise for it to become an issue, particularly on a flat sensor R-D1 where you can pixel-peep to your heart's content. I'm sure that in the film/scanner days such a minor deviation from spec wouldn't have showed up so easily...
Of course Wiki could be wrong...