Canon 85 1.8 RF LTM (8x 100kb) - collimating and RF matching

[menos]

I have a small issue setting up a Canon 85 1.8 LTM lens and would like to ask for some input.

The lens I received yesterday looks very clean and upon inspecting the optical cell and mount, seems to be in original condition without fiddling.

Mounted on a properly adjusted Leica M9, the Canon 85 1.8 front focusses by a large amount (about −100mm).

Infinity is beautifully reached with sharp detail already wide open on the same M9.

My plan was, to change the original shim to a newly made shim and gradually reduce thickness, so infinity is acceptably sharp @ f2.8 - f4, but close focus is spot on, as I intend, to use this as a portrait lens mainly, while shots at distance in good light should be possible as well.

original shim thickness is t=1.44mm

exchanged shim thickness in the end has been t=1.33mm

@ t=1.33mm the lens still front focussed at minimum focus distance by about 10 − 15mm, but infinity focus is completely off, not reaching acceptable detail @ f8.

I am now about, to find other solutions, to match this Canon beauty to Leica M and ask for input.

My strategy, to get there, as thought out during breakfast is this:

- make a new shim of t=1.40mm and collimate the lens, so infinity is acceptably reached @ f2.8 via the mechanical infinity stop
- adjust close focus by shortening the RF coupling, to reach spot on focus at close range via rangefinder

two ways, to shorten RF coupling:
- polish RF coupling surface
- exchange the small roller Ø, that transfers the change from the internal RF cam to the RF coupling feeler

Are these rollers available somehow, as used in several Canon RF tele lenses with this construction (my Canon 100/2 LTM shares the same design of the RF coupling) ?

Is it common, that Canon RF lenses are adjusted to a different mount to film distance than the Leica standard?
This is not the first Canon tele, that shows this severe front focussing, I try on a digital Leica?
Is there any technical material, I could read up?

Here are a few pictures of Yesterday night:

the optical cell with shim and retaining ring removed (left), focussing mount (right)
The retaining ring is screwed on the upper threat, seen on the optical cell. The shim is held between body of the optical cell and retaining ring.

focussing mount

collimating shim (left) and retaining ring with set screw (right)

original shim measurements

I need help, to get this beautiful lens spot on matched to the M9 - I have seen the rendering of it and immediately fell in love!
It clearly is a candidate for the bokeh thread here on RFF!

Another question - this is a 5 lens design, looking like a Sonnar - is it a Sonnar design?

Frank Mechelhoff has a few words regarding this lens on his website with lens group drawings, compared to a double Gauss and the newest Leica 90 Summicron APO (interesting 🙂 ).

I unfortunately cannot find much more about this lens on the net :-(

 

[Sonnar Brian]

Let's back up one step: before going further, verify the thickness of the m-Mount adapter that you are using. The adapter thickness is critical for using Telephoto and Wide-angle lenses on the M-mount camera. That is because the movement of the optics is NOT 1:1 with the cam, as it is on normal lenses.

If the thickness of the Adapter is not the issue, then the problem is more likely the exact focal length of the lens is off. I would verify that all of the optics are in place. One element out of position will change the focal length, and the actual focus across the range will not agree with the RF.

Once focal length and adapter thickness is verified, you should be able to adjust the main shim for focus across range.

The Canon lenses may slightly front-focus on an M9 as the assumptions made for film thickness do not apply to the sensor.

 

[menos]

Brian, thanks for the info.
Yes, the LTM-M adapter is verified as spot on.

It is t = 1.00 mm

It sits perfectly on the lens mount (I have quite some experience now with quality control of different Chinese brand adapters, as I am sitting at the source so to speak, having gone through endless combinations of different adapters (a tip to the posters of regular "should I buy this cheap brand adapter from eBay" - don't!).

The particular adapter, I am using for the 85 1.8 is the the one, I am using on the Canon 100/2, which focusses spot on, even at close focus.
As an anecdote: i have handpicked this particular adapter from a batch of about 20 (!!!) adapters of same brand and type from different shops, which all (!!!) except this very one had their mounting thread cut wrong, so the adapter would not screw fully onto the lens, preventing the RF coupling from lining up with the cameras RF roller 😱

I have found two other points of interest here. Maybe, you could verify this:

1) reference measurement mounting flange to RF coupling

I measure the RF coupling flange extension (coupling thong with the Canon) from the lens mount (the travel, the RF roller in the camera will be pushed into the camera body, measured from the mount).

Infinity (mechanical stop):

Canon 85:
6.58 mm

Leica 35 Lux ASPH:
6.50 mm

So the rangefinder coupling of the Canon 85 1.8 will push the RF roller of the body 0.08 mm deeper into the camera body, when focussed @ infinity.

1m marking on lens barrel:

Canon 85:
3.52 mm

Leica 35 Lux ASPH:
3.39 mm

So the rangefinder coupling of the Canon 85 1.8 will push the RF roller of the body 0.13 mm deeper into the camera body, when focussed @ 1 meter. The measuring tolerance here is higher than at infinity, as the distance markings on both lenses might differ from each other, but it shows a measurement of approximately ~ +0.10mm, coinciding with the very precise measurement, taken @ mechanically locked infinity.

My feeling here is (with the experience of 50mm lens corrections on shims between optical cell and mount), that a change in RF coupling of 0.10 mm is actually a very big amount off, translating in a big focus issue.

2) RF coincidence @ infinity with reference camera body
The Canon 85mm focus mount does not coincide @ infinity with my other calibrated lenses.
It does indeed "over leap" the RF coincidence slightly @ infinity (proving, that the RF roller is indeed pushed deeper into the camera body, as would be a perfect coincidence.

My strategy is now, to re shim the Canon with a new shim. I want to start at the original t = 1.44 mm (best acuity of images @ infinity, even wide open) and will polish it down, so the lens has a satisfying sharpness @ infinity, to correct as much, as possible with the shim.

I then want to polish the RF thong of the Canon lens, until close focus sits perfect (not going under a reference measurement of Leica lenses, I measure against this, namely the 35 Lux ASPH, bought new and my Noctilux f1, which I use as a reference for all camera bodies and LTM lenses.

Is my thinking wrong, as to correcting the lens focal point by shimming as possible and bringing the rest by correcting the RF coupling of the lens?

 

[menos]

Brian, I would have indeed issues, to verify the focal length.
On inspection though, the lens cell looks as new - there are zero track marks of opening.
All screws do have, what appears to be a original looking sealing.
Nothing rattles or looks strange.

Test images with the lens do not show any obvious issues (sharpness across the frame appears to be even - did I say already, that this and the Canon 100/2 are incredibly sharp, even wide open?).

Other, than measuring the lens on an optical bench, I see no chance in verifying the focal length - is there a "house wife trick"? ;-)

 

[Sonnar Brian]

Go for adjusting the main shim for close-up and wide-open. at worst, stop down to F2.8 for infinity.

On the focal length- best to leave it alone. Get the best focus with the main shim adjusted, report back.

 

[menos]

Go for adjusting the main shim for close-up and wide-open. at worst, stop down to F2.8 for infinity.

On the focal length- best to leave it alone. Get the best focus with the main shim adjusted, report back.

Brian, I tackled this Today, after having a good breakfast with fresh coffee and brain juice…

I have been taken the lens first for most precise measurements to a better equipped facility 🙂

I remeasured the focus coupling distance @ infinity with different lenses.
My Noctilux is in fact 6.50mm, as is the 35 Lux.

The Canon 85 1.8 is not 6.58, as measured Yesterday with lesser method (didn't had a proper measurement table with depth measuring caliper and gauges).

Soooo … I didn't touch the RF coupling of course. I then first checked again through a huuuuge box of 90/28 adapters and found one Japanese made adaptor, which in fact slightly improved the infinity coincidence with the calibrated M9 body.

The focussing mount of the Canon 85 and my Noctilux (+ the new from Solms Leica M lenses) coincide now perfectly @ infinity 🙂🙂🙂🙂😀😀😀😀😀🙂

I then set up a more precise testing range with measured tripods and aligned focus targets, to check close focus.

I polished the main shim to a final thickness of t = 1.31mm, which is the perfect measurement, to focus spot on with the lens at close range.

Infinity focus is not sharp at f2.8, so there seems in fact to be going something on with either the lens' focal length or …

I will now take the lens for a several day test ride and use it in the wild, to learn about it's behavior.

There is a last resort, bar from the correction of focal length (there is in fact, what seems, to be another shim in front of the 4th element, which I suppose could be used for adjusting focal length.

I don't intent to touch the optical cell without further precise measuring equipment.

Another idea is, that there is in fact an adjustment for the mechanical infinity stop of these Canon lenses.
One could let the lens focus the precise bit further (ignoring infinity coincidence with the Leica RF, to have the image crisp at infinity)

This is just a loose concept, of what potential next steps to take, should I move on.

What is your opinion? Does moving the mechanical ∞ stop sense?

I will post some pictures from the lens, as soon, as I back from shooting (Today is supposed to be a heavy storm in the Shanghai area, so this possibly has to wait).

Best,

Dirk

 

[Sonnar Brian]

It sounds like the focal length is off, or the interna cam that translates he 85mm movement to the RF cam is off somehow.

It focuses close-up and wide-open now. How far is it off at infinity? What is the maximum focus distance now?

 

[menos]

It sounds like the focal length is off, or the interna cam that translates he 85mm movement to the RF cam is off somehow.

It focuses close-up and wide-open now. How far is it off at infinity? What is the maximum focus distance now?

Brian, I have to test this in the open.
With my preferred focus target for infinity - a high building @ ~ 4km the lens reaches acceptable sharpness only from f8 on.
F5.6 is still very soft at the horizon.

The Noctilux f1 interestingly renders the horizon pin sharp @ f1 (while being spot on close up).

The Canon now is focussing spot on from 1m on.
so far, I only shot indoors Today and plan, to collect some data over the next days.

This is, what it does so far - even just for this purpose, I would be extremely happy ;-)

It shares indeed some character with the 100/2 Canon - remarkably contrasty and sharp, even from wide open. This is not a "soft lens".

 

[Sonnar Brian]

As it is configured now: sharp close-up and wide-open, falls short at infinity- the focal length is too long for the RF cam. Moving the rear element in closer by reducing the secondary shim will reduce the focal length. You might find it necessary to increase the main shim after doing this.

 

[menos]

As it is configured now: sharp close-up and wide-open, falls short at infinity- the focal length is too long for the RF cam. Moving the rear element in closer by reducing the secondary shim will reduce the focal length. You might find it necessary to increase the main shim after doing this.

Brian, thanks so very much for your help on this !!!
Is there any reasonable way, to relate the necessary corrections to measurements and numbers?

I tried desperately, to find for example calculations for necessary corrections of the main shim, but failed in a sea of RF settings related threads.

I would love, to be able, to calculate needed corrections, to prevent trial and error.

I have stashed the original main shim of the lens and will have somewhen during next week a few spare shims been made by a supplier (if I only had my own little machining shop in the backyard 😉 )

As soon, as I get the new shim, I should also have some feel, at which proper distance infinity stops wide open.

 

[Nokton48]

Lovely Result! Very Sharp!

Brian, I have to test this in the open.
With my preferred focus target for infinity - a high building @ ~ 4km the lens reaches acceptable sharpness only from f8 on.
F5.6 is still very soft at the horizon.

The Noctilux f1 interestingly renders the horizon pin sharp @ f1 (while being spot on close up).

The Canon now is focussing spot on from 1m on.
so far, I only shot indoors Today and plan, to collect some data over the next days.

This is, what it does so far - even just for this purpose, I would be extremely happy ;-)

It shares indeed some character with the 100/2 Canon - remarkably contrasty and sharp, even from wide open. This is not a "soft lens".

 

[Sonnar Brian]

You can compute the required change in shim thickness by using the focal length formula.

1/Focal_length= 1/front_distance+ 1/back_distance

http://en.wikipedia.org/wiki/Focal_length

SO: you need the exact focal length of the lens, measure a distance to focus on using the RF, and measure amount of error. Use the formula to compute the back focus between the two, and the difference is the amount that you need to change the shim to bring the back-focus in agreement with the RF.

I should code this up in FORTRAN.

 

[menos]

Lovely Result! Very Sharp!

Thanks ! I will give your comment to her ;-)

I feel, from the little, I have shot now with this lens, that indeed it is close to the Canon 200/2.

It really is very detailed, when focussed properly.
I have to use it much more though, to find a verdict. My heart tells me, I like the lens a lot! 😀

You can compute the required change in shim thickness by using the focal length formula.

1/Focal_length= 1/front_distance+ 1/back_distance

http://en.wikipedia.org/wiki/Focal_length

SO: you need the exact focal length of the lens, measure a distance to focus on using the RF, and measure amount of error. Use the formula to compute the back focus between the two, and the difference is the amount that you need to change the shim to bring the back-focus in agreement with the RF.

I should code this up in FORTRAN.

Brian - thanks a lot! That's, what I needed.
I will shoot the lens first a bit and will then get some measurements.
I will post some pictures here, once sone ;-)

 

[menos]

Just posted this in the Bokeh thread - I think, with the Canon, one can maximize this swirly bokeh effect with different light and a more close up subject:

"Canon 85/1.8 LTM @ f1.8 ~4m"

 

[menos]

I will post photos, taken with this lens now in this thread, I startet for the black Canon 85/1.8 LTM here:

http://www.rangefinderforum.com/forums/showthread.php?p=1688474#post1688474

I hope, more users of this lens will add …

 

[jmanivelle]

Late reply here !
A friend has a Canon LTM 85 1.8 I'd like to buy but its rear element is heavily damaged (looks like scratches !). It was adjusted some years ago and focuses well on my M8.
So if someone knows of a Canon 85 LTM 1.8 for parts with a good rear element, please PM me !


Jorge_Veras_test_Canon_85_LTM@1.8 by careca2013[/
Cheers
Jean-Marc

 

[menos]

Hey Jean-Marc, for what this beautiful lens is famous, your damaged copy looks as beautiful, as others!

I hope, you got the lens (and for a fair price).

I never saw any 85/1.8 "spares" though so far - the lens itself is not often to come by, while prices are on the high side.
Good luck with finding a nice rear element.

You want to make sure, that, when you have found one, you adjust your lens with BOTH shims.
First make sure, that the actual focus mount is shimmed properly, so it coincides correctly at infinity with your reference Leica M. This should be your starting point for adjusting the lens - don't make compromises and start to shim LTM adapters for "getting the lens right" - that is sloppy in my opinion.

Then shim the rear cell to modify the focal length until the lens coincides properly with the calibrated rangefinder (there is a brass shim between the body of the optical cell and the rear cell).

As the 85/1.8 also is a Sonnar design, very similar to the 100/2 LTM, you might need to make a compromise in best performance according to focus distance.

I adjusted mine, so it will perform best in portrait range - it will be soft wide open @ ∞.
Stopping down to ƒ2.8 or 4 it will get back sharp again @ ∞.

I really love the smooth look of this lens wide open for portraits and find it more pleasing in this regard, than the Leica 75 Summilux (it is also much easier to focus with it's loooong focus throw).

Good luck with the lens!

 

[jmanivelle]

quote /
Hey Jean-Marc, for what this beautiful lens is famous, your damaged copy looks as beautiful, as others!

Thanks Dirk!
I actually read your last year's posts about this only today and fortunately the lens I've tested is already calibrated at the near distances (didn't shoot it wide open at infinity) together with the adapter it is fitted with. It belonged to a pro photographer who had this done years back for his M bodies.
But unfortunately it flares a lot (more than it should IMHO) when there is a bright source in the frame and this is due to the rear damaged element ( I wonder what happened there).

I really love the smooth look of this lens wide open for portraits and find it more pleasing in this regard, than the Leica 75 Summilux (it is also much easier to focus with it's loooong focus throw).

I agree, I was stunned by that quality as well and the smooth gradation into the OOF areas .
It would not stop me to hope for a 75 Summilux but I know what you mean. Is there a smoothier 75 Summilux among the 3 or 4 types that exist?

Good luck with the lens!
Will keep you posted 🙂

Cheers
Jean-Marc.

 

[menos]

Hey Jean-Marc, you're lucky (that the lens is already properly prepared) and a bit unlucky then (about the condition of the rear element).

If there is no way, to find a spare 85/1.8, one way, to probably resurrect the lens to former glory might be a re-polishing at one of the shops, often mentioned here on RFF - I can't remember the names !@#$%

Basically the element in question would be examined, if the procedure is feasible and then polished and recoated.

Considering your purchase cost + cost of re-polishing it might be feasible.

Exactly, what you name as "the smooth gradation into the OOF areas" is, what blew me away, when I saw photographs, done with this lens the first time.
If you can combine this in a photograph with another strength of the lens - the beauty of beautiful, beautiful endless tones into the black without abruptly clipping the shadows, pictures really shine.

So prepare really dark backgrounds, light your subject's face and this lens really shows!

Regarding the Summilux 75, I have no facts, but only know, what is reiterated on the net over and over: all three versions are optically identical.

Upgrades have been made from v1 -> v2 in the form of loosing some weight + bulk and adding a slide out lens hood.
From v2 -> v3 (v2 being still made in Canada, while v3 is the only Summilux 75mm, that is made in Germany): the lens barrel has been further slightly reduced in weight, no changes to the optics have been made.

It looks like German made samples are priced more and more into collector territory, while I prefer the more compact size of the slide out hood version.
Therefore the v2 "made in Canada" is in my opinion the best user 75 Summilux.

I found, I am not much of a 75mm guy - framing is difficult for me, focussing the 75/1.4 feels much more difficult than the 50/1, which I use with ease.
You should really try one before buying.

 

[awilder]

One thing no one mentioned was confirming the M9's RF @ 1 meter is correct as that is calibrated by adjusting the length of the cam roller arm. If spot on for your other M lenses like the Noctilux, it's likely not an issue in this case. Assuming the camera's RF adjustment is correct, the fault may simply be that the lens' focal length doestn't exactly match the RF synch of the focusing mount. The cure would be to send it off to a good tech that can bench test and adjust the lens head optics or focus mount RF cam synch for an exact match. After all, maybe the lens head focal length doesn't exactly match the focusing mount RF synch or maybe someone swaped out the lens head with a different focusing mount given the 50 year age of the lens. Leica lenses with removable lens heads such as the early 90 Summicron always had a matching lens serial # clearly engraved or labeled inside the focusing mount. I always check for this as I've purchased used lenses where they din't match up indicating someone switched lens heads in the past resulting in focus errors.