Focussing Issues

[phancj]

Advertisement (gone after registration)

Perhaps you have the impression that the image is sharper on the camera screen because you more often have a correct focus now

 

Saw your photos, they are very very impressive.

 

BTW can I ask why your choice of the 35, then 75...do you consider the 50mm on the M9 not as important? I am still contemplating the 35 cron or the 50 lux, hence my question.

[thighslapper]

nope

something else is definitely different

[Guest mc_k]

...

 

Let's go through the possible outcomes and what it means

 

[A] image is in focus, distance scale is at infinity. Most likely everything is fine.

 

image is not in focus, distance scale is not at infinity. Most likely the rangefinder is off.

 

[C] image is in focus, distance scale is not at infinity. Most likely both rangefinder and lens are off.

 

[D] image is not in focus, distance scale is at infinity. Most likely lens is off

 

...

 

[a] image is in focus, rangefinder images do overlap. Most likely everything is fine.

 

image is not in focus, rangefinder images do not overlap. Most likely both rangefinder and lens are off.

 

[c] image is in focus, rangefinder images do not overlap. Most likely rangefinder is off.

 

[d] image is not in focus, rangefinder images do overlap. Most likely lens is off

...

 

It would be great to have some kind of a chart to diagnose all of the camera and lens pathologies from their symptoms, to answer the age-old question: do I send in the lens, or the camera, or everything. I really doubt this is possible.

 

First of all, any test needs a rangefinder check at both a near distance and infinity (and you can bracket around infinity like suggested, note when you hit the stop, etc.) If you enumerate all of the different outcomes and what could cause them, you will see there is no case in which you can narrow things down to just a lens or just a rangefinder at fault. Taping off a distance to check the scale will still not provide enough information. The best you can do is in some cases say **at least** the RF is off. And if you check two lenses on the same camera and they differ widely, you can say **at least** one lens is off. If you are lucky enough to have a large number of cameras and lenses you can make a good guess about what might be off, but until you compare to a reference you will never know for certain.

 

Anyway that's my opinion, but your setup is really interesting.

[Lindolfi]

You could be very practical and say: I don't care whether the rangefinder is off or the lens is off, as long as I have correct focus at infinity and at 1 meter. And that would be fine, because you can work (as long as you can actually reach infinity in terms of true focus in the image plane).

 

But that attitude is not tenable as soon as you work with 2 bodies and/or 2 or more lenses. Since although the 1 lens - 1 body combination works, the other combinations may fail in correct focus.

 

So you do need to know what the culprit is.

 

Therefore, I used the wording "Most likely..."

 

Now let's think about the mechanics of the camera and the lens and the places where things can go wrong

 

 

[1] The contact between lens flange and camera mount ring

[2] The engagement of the rangefinder arm by its roller through a pushing action of the rangefinder cam in the lens

[3] The eccentric setting of the rangefinder arm roller (in the M9 to be set with 2mm hex key)

[4] The effective length of the rangefinder arm (in the M9 and earlier to be set with a screw at the base of the arm)

[5] The position of the rangefinder cam in the lens relative to true infinity of the optics

[6] The speed of travel of the rangefinder cam relative to the speed of image distance (setting of distance scale)

 

 

Now let's go through possible causes of all these items which get them away from what it should be.

 

[1]

[a] flange on lens may be loose

mount on camera body may be loose

[c] there may be dirt on the lens flange

[d] there may be dirt on the mount of the camera body

[e] the lens may be mounted incompletely

 

[2]

[a] the cam in the lens may be dirty, damaged or worn

the roller of the rangefinder arm may be dirty, damaged or worn

 

[3]

[a] the roller may have changed its correct position by build up of friction at its bearing (so that the hex screw got turned)

the roller may have changed its correct position by bending of the rangefinder arm because of collisions with the lens cam or flange during mounting

 

[4]

[a] the small screw at the base of the lens may have changed its setting

the rangefinder arm may have been bent

 

[5]

[a] the cam or cam mechanism (both are used in Leica lenses) is loose

the cam or cam mechanism is damaged or worn

[c] the helical groove of the optics in the barrel is loose or turned

[d] one or more glass lens elements are loose

 

[6]

[a] the shape of the cam is changed (its slope) from wearing or damage

the coupling between cam and optics has developed play

 

 

If we assume everything was fine with the body and lens(es) when coming from the factory and we look at this list what would be the most likely places and causes where things go wrong?

 

This was the basis of the use of my wording "Most likely..."

 

But you are absolutely right: there is always the possibility that 2 or more factors compensate each other. That is why Leica uses better methods than the one I have outlined to find the culprit(s). But because it is not nice to be without a camera and/or lenses for while, it is sometimes nice to be able to make an educated guess what may be wrong. That is why I offered the method I have used up till now with great success with my 2 bodies and 5 lenses.

 

And by the way: with digital M, checking of focus has become much easier. Within half an hour you can check all lenses at infinity and 1 meter.

[Lindolfi]

Saw your photos, they are very very impressive.

 

BTW can I ask why your choice of the 35, then 75...do you consider the 50mm on the M9 not as important? I am still contemplating the 35 cron or the 50 lux, hence my question.

 

Thanks phancj.

 

A bit off topic here, but given my former posting, I may have the liberty to answer your question here

 

Well the choice comes from my post experience in which I did have the equivalent of 50 mm in the bag next to other focal lengths. It appeared that I had a preference for the 35 and 75 mm (at both sides of the "normal" 50 mm) as the most often used lengths, with wider (21 mm) and longer (135 mm) as a second preference, but never really in the "true" middle.

 

I think that is very personal and has to do with how you approach a subject. 35 mm is really so wonderful because it allows you to be part of the scene while you can leave out too much from around the scene (as is hard in 21 mm or wider). 75 mm puts you slightly outside the scene but offers the intimacy of correct perspective and that wonderful shallow depth of field.

 

So the choice between 50 mm and 35 mm is hard and personal. For me it is clearly 35 mm, with the added advantage of the size of the lens

[Guest mc_k]

...The best you can do is in some cases say **at least** the RF is off. And if you check two lenses on the same camera and they differ widely, you can say **at least** one lens is off...

 

had this one wrong, apparently

[Guest mc_k]

Advertisement (gone after registration)

...

Let's go through the possible outcomes and what it means

 

[A] image is in focus, distance scale is at infinity. Most likely everything is fine.

 

image is not in focus, distance scale is not at infinity. Most likely the rangefinder is off.

 

[C] image is in focus, distance scale is not at infinity. Most likely both rangefinder and lens are off.

 

[D] image is not in focus, distance scale is at infinity. Most likely lens is off

 

...

Let's again go through the possible outcomes and what it means

 

[a] image is in focus, rangefinder images do overlap. Most likely everything is fine.

 

image is not in focus, rangefinder images do not overlap. Most likely both rangefinder and lens are off.

 

[c] image is in focus, rangefinder images do not overlap. Most likely rangefinder is off.

 

[d] image is not in focus, rangefinder images do overlap. Most likely lens is off

 

...

 

Not so sure about many of the conclusions in A-D and a-d. It's puzzling you allow for a lens and body to fully cancel, i.e. case C (capital C), but don't allow they can partially cancel (more likely). Some of these conclusions take for granted what is on or off at distances besides infinity, but you don't mention this. I really wish you would show more of the justification. I don't think you need the specifics of what's off mechanically--I think it's purely a formal problem.

 

I do think this is very clever, and much more of an effort to sort all this out than I have seen before.

 

Using the rf and the scale at infinity and another distance, I think you can break things down into the cases: all o.k,, rangefinder bad, both bad, and lens-or-both bad. In other words, you can just about sort everything out, but not quite.

[Lindolfi]

I do think possible causes are important, because some of them are more likely than others. For instance changing the slope of a lens cam takes quite some damage or wear, but the change of the shape of the rangefinder arm can happen much more easily.

 

And a number of possible causes can easily be excluded by checking they are not present (like loose flange or dirt)

 

Using the rf and the scale at infinity and another distance, I think you can break things down into the cases: all o.k,, rangefinder bad, both bad, and lens-or-both bad. In other words, you can just about sort everything out, but not quite.

 

Fully agreed, especially the "...but not quite" bit.

 

Lastly: it is nice to understand more of how the camera works and how the interaction with the lens works within its tight tolerance. The fact that it is so critical (0.1 mm of rangefinder arm motion is easily visible in the rangefinder patch) and still can work very well, increases my feeling of love for this remarkable line of camera's.

[Guest mc_k]

Now let's assume you use the infinite end of the distance scale of the lens as the adjustment value (for instance when you can not find the rangefinder to overlap)

 

Let's again go through the possible outcomes and what it means

 

[a] image is in focus, rangefinder images do overlap. Most likely everything is fine.

 

image is not in focus, rangefinder images do not overlap. Most likely both rangefinder and lens are off.

 

[c] image is in focus, rangefinder images do not overlap. Most likely rangefinder is off.

 

[d] image is not in focus, rangefinder images do overlap. Most likely lens is off

 

 

To get back to the original poster: thighslapper found outcome [c] of this second list.

 

I don't know, I'm back to square one and not sure there is a rationale for this.

[sidmuer]

would there happen to be detailed instructions anywhere on how to adjust the swing arm (roller) on the m9?

 

I don't want to send it in, and I've concluded that is the culprit.

(a little back focus issue)

 

thanks.

[Lindolfi]

Three links that may be of use:

 

http://www.l-camera-forum.com/leica-forum/leica-m8-forum/13933-new-backfocus-thread-solution.html#post146014

 

Dante Stella

 

http://nemeng.com/leica/binz/rf_adjust.txt

 

Finally, here an image I made yesterday, focussing a 135mm on an M9 without a magnifier in a very short time. Correct working rangefinder coupling is critical here! Click

[bocaburger]

would there happen to be detailed instructions anywhere on how to adjust the swing arm (roller) on the m9?

 

I don't want to send it in, and I've concluded that is the culprit.

(a little back focus issue)

 

thanks.

 

Midway down the page HERE

 

In your case (back focus) you'll want to turn the ring CLOCKWISE. (Front-focus would of course be the opposite, counterclockwise). The adjustment is not as sensitive as the infinity adjust (i.e., it takes more rotation to effect a noticeable adjustment). When you're done, re-tighten the lock-down screw. Be cautious what driver you use, because if it has a straight shaft you're going to be coming at the slot at an angle and can easily burr the slot.

[Guest mc_k]

...

 

thanks for the reference--it describes the two adjustments as a gain and an offset. I'm still hoping to find this in a book or repair manual.

[Guest mc_k]

Can anyone straighten me out about what kinds of adjustments are made to lenses for front focus / back focus? Regrinding the cam profile seems like an infrequent adjustment. For the rest of the adjustments, do these all have an effect the same as a shim between the body and lens? Then, if a lens and a "reference" body is off at one distance, in theory it is off at all distances?

[bocaburger]

thanks for the reference--it describes the two adjustments as a gain and an offset. I'm still hoping to find this in a book or repair manual.

 

The "offset" refers to the infinity adjustment, and the "gain" refers to the arc of travel. It is a combination of both that calibrate the rangefinder. Step 1 is to adjust the roller eccentric (offset) for coincidence at infinity. Step 2 is to adjust the arm-length (gain) for coincidence at 1m. Step 3 is to re-check (and if needed, readjust) the offset for infinity (typically only needed if the gain adjustment was severe).

 

I don't know if it appears in a book or manual, but the procedure can be confirmed by any Leica repairperson. This is how I've adjusted my M9 and it works perfectly. The only difference is that the "official" procedure uses a reference standard (either a lens or a jig). If you DIY then you have to be certain the lens you use for the adjustment is perfectly in order. In my case with the M9, I had 11 of 12 lenses all front-focusing by exactly the same amount, and not being a strong believer in coincidence, I made the judgment that those 11 were collimated correctly, and adjusted my M9 so it focuses all of them perfectly. Lens #12 is currently off being re-collimated. If I had found among my 12 lenses an assortment of mis-focusing behaviors, then I would have had to send the lot of them along with the camera for a pro to figure it all out.

[Guest mc_k]

Can anyone straighten me out about what kinds of adjustments are made to lenses for front focus / back focus? Regrinding the cam profile seems like an infrequent adjustment. For the rest of the adjustments, do these all have an effect the same as a shim between body and lens? Then, if a lens and a "reference" body is off at one distance, in theory it is off at all distances?

 

anyone?

[giordano]

It all boils down to one of these three:

 

1) Adjusting the mount so that when the focusing ring is at infinity, the lens is focused at infinity.

 

2) Adjusting the rangefinder cam so that when the lens is focused at infinity the rangefinder cam pushes the roller in the camera body to the position at which the rangefinder is focused on infinity too.

 

3) Adjusting the shape of the rangefinder cam.

 

The practicalities of each vary with the design of the lens mount.

 

Then, if a lens and a "reference" body is off at one distance, in theory it is off at all distances?

 

This gets complicated. The rangefinder coupling is based on the focusing travel of a "50mm" lens. This means (for 50mm lenses):

 

• If the relationship between the optical cell and the rangefinder cam is correct the rangefinder will focus the lens focus correctly even if the shim between the bayonet mount and the fixed part of the focusing mount is wrong. If the shim is too thick, the lens will be focused closer than infinity when the focusing ring is at infinity - but the rangefinder will correctly show the distance focused on. If the shim is too thin, the lens will be "focused beyond infinity" when the focusing ring is at infinity - and so will the rangefinder. So incorrect shimming between the bayonet and the mount proper means the lens will only be "off" if you rely on the infinity position (and in theory on the focusing scale, though on modern lenses this isn't accurate enough to be significant).
  
• If the relationship between the optical cell and the rangefinder cam is wrong, the lens will be "off" at all distances.
  

When focusing lenses of other focal lengths, the optical cell moves at a different rate from the focusing cam (faster for long lenses, slower for wide ones). So it's possible for a non-50mm lens focus correctly at one distance but to gradually get further "off" at progressively longer or shorter distances.

[ho_co]

_, obviously, a lens for a rangefinder camera is much more labor-intensive to produce and much more complicated to adjust than a lens for an SLR, since with the SLR all you have to do is get infinity right.

 

You've seen the pictures of the rangefinder adjustment station for the M9, where the adjustments are fixed for four different distances.

 

 

 

To oversimplify: in the lens is a lens cell, which holds the optics.

 

That sits in a focusing helicoid, which focuses the optics when you turn the focusing ring.

 

There's a third part, which is the rangefinder cam. On some older 135's, it was simply a metal tab coupled to the focusing mount. On some lenses, it is part of the focusing helicoid. On others, it's an additional separate helicoid.

 

"What do they do?"

They adjust the lens head, the focusing helicoid, the focusing ring and the rangefinder cam till all match.

 

No, they don't file the cam.

 

Some people say that once everything is properly aligned, it can be adjusted by shimming the lens tube as it sits in the mount.

 

It's because of this mechanical complexity as well as manufacturing tolerances that most(?) older lenses 50mm and longer have the actual focal length of the lens unit engraved on the lens barrel: A different focusing helicoid is used for the specific focal length of the lens. If you put a lens with a 51.9 mm focal length into a mount for a lens of 51.4 mm focal length, obviously the lens can focus correctly only at one distance.

 

You can see how all these different helicoids work together in the Leica technical drawings. Here's the one for the Tri-Elmar 28-35-50, one of the most mechanically complex lenses ever produced for the general public.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[Guest mc_k]

...

No, they don't file the cam.

...

 

by accounts maybe sometimes

 

http://www.l-camera-forum.com/leica-forum/leica-m8-forum/36701-focusing-madness-noctilux.html

 

but I don't know, that's why I asked.

[Advertisement]

Hallo,

Look here: Leica M

[Guest mc_k]

thanks to H and J for the detailed and helpful replies.