RF accuracy problem with 90mm lens (It's OK with a 50mm lens)

[steve_taylor3]

Can anyone please help?!

 

I have just bought a rather old 9cm f/4 Elmar lens in M mount. It

looks nice, but in tests, I have noticed that it appears to have

focusing errors. If I focus on something, the film image does not

have that item in focus, but something further away.

 

I have done some tests with the camera (an M6 0.58 bought new in Feb

02) on a tripod. To do this check, I opened up the back door, and

opened the shutter with a cable release locked on B. I got a focusing

screen and a loupe and stuck them together so that the loupe was

focused on the fresnel side of the screen. I put the fresnel side of

the screen on the film running bars (the inner set of the two sets)

of the M6. I could then see a magnified image of what the lens was

showing at the film plane.

 

With a 50 f/2 (bought new same time as the camera), the image

produced by the lens was perfectly in focus when the rangefinder

patches were coincident. This was at about 2m and at f/2. When I put

the 9cm lens on, and focused so that the rangefinder patch was

coincident, The image thrown by the lens was clearly out of focus.

When I focused at the film plane, the rangefinder patches were out by

quite a bit. (The gap was probably about a sixth of the RF patch

size - a significant error.) When the focus was correct at the film

plane, the lens scale indicated 7 feet. When the RF was corrected to

make the same image coincide, the lens scale showed 6.3 feet about. I

don't think that it is that I have the low viewfinder magnification -

the error is blatently obvious!

 

 

 

My questions are:

 

Can I expect better accuracy than I am getting? I would imagine so.

 

Is the camera or the 90mm lens out? I suspect the lens since the 50

lens focuses fine. Even though a 50mm lens at f/2 has about a half

more depth of field than a 90 at f/4, I would say that the

combination of the camera and the 90 lens is worng.

 

Can you get the lens adjusted? I do not suspect the camera as much as

the 90 lens since the camera is pretty much brand new and the lens I

would imagine is about 40 years old. Also, the camera and the 50 lens

is fine. The camera is also fine with a 35mm lens, but this has not

been tested using the setup above, just that the pictures look fine

and what I wanted in focus was. I appreciate that the wider the lens,

the less critical the RF accuracy is. The RF is coincident on clouds

and far objects when focused on infinity.

 

I have never had a lens longer than 50mm on an RF camera before, so I

do not know what to expect. I just need to know what can be fixed

before I go back to the lens vendor and ask them to get it adjsuted

or replaced.

 

Sorry this has been long - I wanted to get everything in that I could

think of. If anyone can help I would be grateful.

 

Thanks in advance

 

Steve Taylor

[albert knapp md]

Steve: there is an error in the lens and you should send it in for

collimation and CLA. DAG or Leica NJ will handle it for you and

make you a happy camper.

[charles barcellona www.bl]

Steve, the focusing screen on the inner rails will result in some slight amount of inaccuracy. The reason is, when the pressure plate is in place, it rests on the outer rails, but the distance to the inner rails is greater than the thickness of film. What happens in there, is that the film makes a sort of doube curve when looked at lengthwise. It starts out on the inner rails, then bends toward the pressure plate, until it contacts it someplace between edge and center. Then in the very center it lifts from the plate a bit, and bows back toward the lens/rails. This is reversed from center to the opposite rail. In other words, film is not completely flat, and manufactureres take that in to consideration.

 

Here's a test that will be conclusive.

 

1. Test BOTH lenses at infinity focus, using the moon as a test target. They both should align at infinity perfectly.

 

2. At 2 meters or so, set a ruler on a table top, at about a 45 degree angle from the film plane. Angle is not too important, just as long as its not parallel or perpendicular to the film. Mark one of the divisions on the ruler, and this will be our focusing index point. With camera on tripod, focus on the marked division. Using medium speed black and white film, take a photograph using electronic flash with the lens wide open. Flash here will eliminate camera shake, while a wide open lens will reduce depth of field to a minimum. De-focus the lens and repeat this two more times, focusing carefully each time. Do this with both lenses.

 

3. Process the film. If the prints are very high quality, you ought to be able to see any error in them, but you might want to critically examine the negatives with a strong loup. The results ought to show the area that the camera is actually focusing on even though you focused on the marked division.

 

That will tell you if the lens and camera are in coincidence.

 

If both lenses are off, the body can be adjusted. The 90 lens may be able to be adjusted a bit if its off overall, but the "pitch" of the mount is not adjustable. Unless the 90 has a removable head, and someplace its been swapped for an unmatched one, there should be no problem with the pitch of the focusing helical.

[Rob F.]

Steve: Yes, the lens can be off, and very likely is. There are some threads in the archives on this. I haven't had to do it, but I understand that one can adjust the lens. I think your Elmar must have a kind of a pushrod that transfers the focus from the focusing part, back to the camera's cam follower. If this thing is sticking, it might be the problem. Let's see. The "pushrod" has a spring to keep it against the cam, which moves forward as you focus farther away. So if it doesn't stay tight on the cam, the rangefinder will think you are focused closer than the lens is really set for. Yep, I think that's it. Check the archives for more on this.

 

'Bye for now . . .

[al_kaplan1]

I'd suspect that the lens head is matched up with the wrong mount. Most of these lenses had the serial number, or partial serial number on the inside of the mount where it's easily visible. It's usually engraved (scratched) through the black finish.

[Rob F.]

Or did I get it backwards? The cam moves forward as you focus closer, not farther away. So if the spring-loaded piece sticks, the camera will think you are focused farther away, when you are really focused closer. That seems like the opposite of what you said you get: The camera thinks you are focused closer than you really are. So it may be something else. What about what Charles said? Is your lens head revovable from the mount? If so, does the serial number on the lens head match that on the mount? What is the serial number? I can look the lens up by its #.

[Rob F.]

"revovable" should be "removable" of course.

[kelly_flanigan1]

The Leica cam is evolved around about the approx 52mm focal length normal lens. The lens block movement of this first lens was the same as the lenses cam; ie they move together. This 2.5mm movement focusing from infinity to 1 meter on LTM and a 52mm lens defines the leica rangerfinder mechanism for all Leicas.

[Rob F.]

I searched over half the archives until my eyeballs crossed, and didn't find anything about adjusting the lens itself, only the camera. I thought I remembered one about adjusting the lens, but I may have been wrong.

 

Unless this a very early Elmar, it will have a detachable head, and it's important to know if the head and focusing mount belong with each other.

[doug_hagerman]

Another possibility is that the lens head might not be screwed all the way into the mount. When you take the head off to check the serial number, check whether there's any gunk on the threads that might be keeping it from going all the way on. Another possibility is that the lens itself might have been disassembled and then reassembled incorrectly, but that would almost certainly show up as bad images in general.

 

The only way to "adjust" the focus that I know of is to modify the cam profile by filing it (which is how they do it at the factory), but once the lens has been assembled there's no reason why that should ever have to be done again.

 

The quality standard you should expect in this area is pretty much perfection: A properly adjusted camera and lens should focus on infinity as exactly as you can detect in the rangefinder. Age of lens or camera should not be an issue. (Although I must admit that my old M2 doesn't meet this standard...but my new M6 does.)

[mark-j]

Steve:

Another thought is that the lens may have had a shim installed between the lens head and body originally. Removal of a required shim would shift the focus in the direction you mentioned. I would test the lens as Charles suggested and this will show you what the lens is doing. I checked one of my lenses this way, only I checked it at multiple distances. Near focus (1 meter), 2 meters, 3 meters and infinity and then shimmed the head to the body to correct the lens as needed.

 

The lens has more depth of field at infinity, hence the checks at multiple points in the nearer range.

[charles barcellona www.bl]

I believe there are several type of adjustments, depending on the mechanics of the lens. They all involve setting infinity on the lens, so that it matches standard infinity of an M rangefinder.

 

On some (maybe all) lenses, the relative point of infinity can be adjusted. That is, the infinity stop is variable.

 

On some lenses, the front (focusing) helical can be adjusted in reference to the rear (RF) helical, and also in reference to the infinity stop.

 

The important thing to remember is: However the lens is adjusted, its the INFINITY point that is adjusted. Keep in mind that this type of adjustment sets the position of the glass to the RF cam. In other words, the RF cam says we're in focus, but what is the glass doing? This is normally set at infinity.

 

The other adjustment, if you want to call it that, is done at the manufacturing level, by choosing the pitch of the helical threads to closely match the actual focal length of the optics. Someplace I thing I remember reading that Leica had 5 different helicals for 50mm lenses, each with a different pitch. I'm sure something similar exists for 90mm lenses, and thats why the optical heads were matched to the helical mounts by serial number on lenses that had removable optics (for viso use).

[steve_taylor3]

Thanks to all that responded. A few answers & questions:

 

Interesting that the focus takes into account that the film is not perfectly flat! I never knew that. However, the results of real tests with film prompted me to do focusing screen test. With the film, I was focusing on a point using the RF and something slightly farther away was in focus. I could clearly see this on a light box with a 6x loupe.

 

 

The serial number of the lens is 1162261. Does this have a removable head? Actually, I do not know which bit should you be able to remove? Is it the front part before the aperture ring?

 

I will check tonight to see if there is another serial number and if it matches the one on the front of the lens.

 

Thanks again all

 

Steve Taylor

[Rob F.]

Okay, the serial # places it at 1954. It should be the very slender chrome model with a milled focusing ring, and the vulcanite band at the base. This is the ELANG (ELANG-M for the bayonet mount). Hove/Laney and Lager both indicate that the lens head can be used on the Visoflex, which means it's removable.

 

To remove it, grasp the head end at the very end, where the filter mounting threads are. Don't twist it by using the aperture ring for a handle! Grasp the base end at the vulcanite. Turn the lens head to the left like you are removing a thermos bottle stopper. After a few turns (maybe more turns than you expected) it will come out of the focusing mount.

 

Now check for a serial number, probably scratched into the inside front of the mount. Make sure it matches the number on the lens front retaining ring. If not, that's the problem. If it matches, then we need to dig up a thread from the archives (if any) that tells how to do the adjustment Charles mentioned. I thought I remembered that there was one.

 

With the lens rotated to the infinity stop, does it line up with the infinity mark? Also, could you try focusing on a distant TV tower (instead of a cloud!) and see whether the lens reaches infinity by the scale, and runs out of travel or not.

[steve_taylor3]

Thanks so much for this. I had no idea as to its age. I will give try and get the front off tonight and also do the focusing test as you indicated, and will report tomorrow.

 

Again, thanks - this is above the call of duty.

[charles barcellona www.bl]

Steve, with the groundglass on the inner rails, the image you see on the groundglass would be at some distance farther out than what the rangefinder reports.

 

The inner rails are just slightly closer to the lens than the film, and the shorter distance to the lens equates to focusing at a more distant object.

[Rob F.]

Charles, he said he was getting this result with the image on the film, not just with the groundglass. I'm having trouble understanding why you say the inner rails are the wrong place to test from, because they are what the emulsion side rests on. The bottom rail is only an alignment guide.

[steve_taylor3]

More reports...

 

I got the lens head off the mount last night as you instructed. I did find that the part in front of the aperture ring unscrews as well! I found a metal baffle inside the mount with the number 261 on it, which corresponds to the lens serial number. I did a few experiments with testing focus and found that by putting an elastic band between the mount and the head I could adjust the head until things were in focus on the ground glass, and the adjustments would stay where they were.

 

The point about adjusting infinity is well taken, as you are moving the whole lens in and out. I have not been able to do infinity tests as it is dark outside when I get a chance to do the tests, i.e. in the evening. I did do a quick test and can confirm that the lens does stop rotating at infinity. I think the RF patch lines up but I would like to wait until daylight to do a proper test on this, so I can see things clearly. I will report back more then.

 

Thanks again - this is most helpful

[steve_taylor3]

Further to my post last week, I thought I had better finish it off with some results and findings. Thanks to the advice posted in this thread, I had some good pointers. I can report that I have got the lens focusing fine now. What I did is posted below.

 

Thanks to the advice in this thread that the head came off, and my initial focusing error, I concluded that the lens head was too close to the camera to agree with the rangefinder. Hence I needed to pad it in relation to the focusing thread. What I used for this padding was bits of HP5 film! I put it between the flange of the lens head and the barrel so that the lens head screwed down on it.

 

I put the camera on a tripod and put some HP5 in it. I ran tests at 3.5', 10' and 20' focus. (I couldn't do any more as this was last week at night and that was the max distance I could get away from something in the room I had.) For each focusing distance I took pictures at f/4. f/5.6 and f/8. For each focusing distance and aperture I put my padding between the lens head flange and the barrel. I did this for 0, 1, 2, and 3 thicknesses (0 thickness being the head screwed tight onto the barrel). Hence I had 36 exposures. 3 per focus distance, each with 3 per aperture and each in turn with 4 padding thicknesses. Each picture was exposed using an appropriate shutter speed to give Z5 exposure and a cable release was used.

 

I deved the film in XTOL 1+1 at 75F for 6 mins, and looked at the results on my lightbox with a 6x loupe. The consistent winner was 2 thicknesses of (good old!) HP5. Hence I used 2 bits of HP5 to pad the lens head.

 

Now I needed to wait until the weekend to test the longer focusing distances and if it would still give infinity focus. (This was in question since I had extended the lens a little.) I did these tests yesterday in the light by focusing on things 100' and 300' away - no problems. Infinity was fine as well - a flag pole 2 miles away was in focus with the lens set at infinity and using f/4. All these daylight tests were on a tripod and a cable release as well, to ensure consistency.

 

So the upshot of all this is that I conclude that I have it sorted. The only problem is that I cannot take the head of the lens without upsetting my padding. This is not much of a problem since I don't plan on taking the head off much.

 

I concur with Mr Barcellona's findings that my test with a focusing screen on the film guide rails would cause a small error with that recorded on the film. I have seen it - when the focus was slightly off on the screen, it was perfect on the developed film.

 

In conclusion, thanks to all that helped with this and I think I have got it sorted. Hope all this is clear!

 

One more thing - even though the lens is 45 years old, I am very pleased with its rendering. Very nice.

 

Cheers

 

Steve Taylor

[Rob F.]

All's well that ends well. But I still think that if you put the frosted surface of the groundglass against the same rails that the emulsion rides on, then the glass will see what the emulsion sees.