Rangefinderproblems on new M9s

[Julian Thompson]

Advertisement (gone after registration)

The pictures are sharp.

 

Things in the *distance* overlap fine in the RF patch.

 

Stars and very, very, (very) distant objects overlap-ish but do not fully coincide into one high contrast image - I am aware that if I were focussing the object normally I would twist the focus slightly more to get to where I wanted to be...

 

Does that make sense/sound right?

[SJP]

I think we are getting there. For me this sounds like the rangefinder is slightly overadjusted at infinity, as you do not report major back- or front-focus problems, but this is not a topic I am competent to comment on. Be as it may, If the lens is focussed at infinity, and produces sharp pictures, then objects at infinity should line up convincingly in the rangefinder, not slightly past the point of max. contrast.

[Julian Thompson]

They don't actually quite get to the point of max contrast, if you see what I mean - but oddly my M8 (did) and M9 both do the same things

[SJP]

OK, but in fact an error either way is not what should happen in the rangefinder, in the worst case I could imagine the rangefinder being slightly past infinity when you hit the maximum rotation of the lens barrel as some mechanical "play" is always a good thing to design in. This slight overrotation of the lens focus should be very close to infinity however and should not give noticably different results in the image itself.

[Julian Thompson]

Thanks Stephen

 

Therefore I guess if I adjust the rangefinder cam to set the infinity to perfectly line up I will then have incorrect adjustment across everything else.

 

This would explain - as Wilson said - why other people have less success with cam adjustment....

[wlaidlaw]

Julian,

 

For the infinity to be correct, you must have images overlapping or as you more correctly describe the situation, a noticeable increase in contrast AND the photograph you take of the same image must be perfectly sharp - difficult one to judge. Otherwise you can have two errors cancelling each other out. A lens maladjusted one way and the rangefinder the other way, would still show perfect coincidence on an near infinite object, but the image would not be in perfect focus on the imaging medium, film or sensor. Now we are getting into such fine adjustments, that it is possible that temperature variations come into play, with thermal expansion of the multitude of components in the train of the rangerfinder, throwing the whole thing out a tiny amount. I just don't know to what degree the RF is temperature compensated, like the bi-metallic strips on a mechanical watch balance wheel. en.wikipedia.org/wiki/File:Pocket_Watch_Balance_Wheel.jpg

 

Wilson

[ArtZ]

Advertisement (gone after registration)

Thanks Stephen

 

Therefore I guess if I adjust the rangefinder cam to set the infinity to perfectly line up I will then have incorrect adjustment across everything else.

 

This would explain - as Wilson said - why other people have less success with cam adjustment....

 

Julian (and the others),

 

 

Focus at infinite

 

 

This stuff about stars and planets is baloney.

 

To a first approximation (to keep the pedants at bay), camera lenses follow the simple equation:

 

1/u + 1/v = 1/f

 

where u and v are the subject and object distance and f is the focal length. For an object at infinity, 1/u = 0 and the object distance, v, is the same as the focal length, f, and the image comes to a focus behind the lens. For all possible values of u, v >= f.

 

The question is, if instead of using a celestial object several light years away, you revert to terra firma and instead use that Dunkin Donuts/<substitute your favourite junk food brand> sign over there, how big is the error in where the image comes to a focus? Taking Lars number of a junk food sign 200 metres (656 feet in old money), the error is 1/200,000 mm = 5 um which is the same tolerance within which Leica mounts the sensor in the camera. 200 metres is good enough.

 

If the infinity stop instead allows the lens to come closer to the focal plane that it should (v < f), it follows that u is negative so that nothing in front of the lens is in focus. The imperative on a Leica lens therefore is that the infinity stop should not be set beyond infinity but a fraction less than true infinity and rely on the depth of field to cover the difference.

[Julian Thompson]

Right - so - infact then Art my stars *not quite* lining up might be normal?

 

Someone else care to try it out + report back?

[SJP]

No the rangefinder should line up perfectly, that is what mine does. Or phrased differently I can focus the rangefinder on something at infinity and then the picture is sharp.

 

The point Mark is making is that there is little difference between infinity and 500 m or so (his 200 m may be too conservative, but OK for most of us), so you do not necessarily need a star or planet as feedback.

[jaapv]

Hi Stephen,

 

Thanks for coming on board!

 

What we're looking at here is that I've adjusted my roller cam again (seems like i have to do it on every M camera I have!) and I'm getting great results, but when I look at a star through my old M8 (back last year) and my M9 now I can't get the image and the rangefinder patch to fully conincide. I 'run out' of focus and the rangefinder patch hits the stops just before the 'two' stars become one.

 

So, when Wilson talked of 'infinity focus' being right I wanted to know, does this mean that my 'two' stars should indeed become one - or just that the image that results from that should be in focus?

I think that if you are working on your roller cam on all your M cameras, you are maladjusting 67% of them, seeing that the focus adjustment has three points to do so....

[Guest mc_k]

Right - so - infact then Art my stars *not quite* lining up might be normal?

 

Someone else care to try it out + report back?

 

The two images of the star should align.

[lct]

The two images of the star should align.

Agreed. Now a slight misalignment at infinity will be compensated by DoF so no worry if the RF's accurate at usual focusing distances.

[Julian Thompson]

Well I thought about this some more, and then some more :-)

 

When you adjust the rangefinder wheel you are not, of course, adjusting anything physical in the lens.

 

If you set your lens to infinity then it's at infinity, and whether your rangefinder is adjusted or not does not have any bearing on the image.

 

So we are left, as I see it, with infinity FOCUS adjustment only being possible by shimming the sensor or adjusting the lens.

 

But, as we move the lens away from infinity the rangefinder also moves obviously, and then you need to focus USING the rangefinder, and so the infinity REFERENCE becomes more important since it is from this that the base point for the other adjustments is set.

 

In practice I have done some far distance checking today and just as lct says it's a none issue because at the kind of distances we're talking about the depth of field means you just can't tell if any slight focussing changes have been made.

 

So in this case my M9 - post roller adjustment - is performing very well and I have solved the focus issue.

 

But - I know that to be 'technically perfect' I would first need to set the infinity roller then find (I don't even know where it is!) the other adjustment wheel to set the closer focus.

 

Is this behind the red dot with the vertical adjustment?

[Guest mc_k]

...So we are left, as I see it, with infinity FOCUS adjustment only being possible by shimming the sensor or adjusting the lens...

 

I didn't follow your reasoning. You have two focusing systems, rangefinder and lens scale, and either can be off. Whichever one you are using, if it's off, you can misfocus something at infinity (a.k.a. a long ways away). The stop at infinity can also be off.

[howard_cummer]

About focusing the 90 Summicron APO - I had the same problem with this lens on my M8.

I had it adjusted by the Leica Rep in HK and then adjusted again by a local repairman and could not consistently "nail" accurate focus. None of my other lenses had this problem and finally I sold the 90 'cron and replaced it with the 90 Elmarit which focuses perfectly. My 2 cents YMMV

Cheers

Howard

[Guest mc_k]

...there is clearly something 'going on' that I haven't quite understood yet as to why some people are successful with the cam adjustment and some are not.

 

I wil try to think about this, and I'll also double check all my lenses at a range of apertures and shooting distances and report back.

 

I think this is key, that your adjustment technique will only be as good as your testing technique. At a middle distance, I think very few could do an accurate job of it.

 

I actually have the same question you do, but would phrase it a different way: is there something like a service manual that gives the details and mechanics of the adjustment procedure? I have heard a couple of different versions, and I am not even sure you have the number of adjustment points correct.

[Steve Ash]

Beside the the point that all my dhree lenses work perfectly with my M bodys, I do not get the point of adjusting the rangefinder for infinity. Where is infinity btw? Wouldn't it be logical to adjust the rangefinder for the hyperfocal distance instead of infinity? Just for technical interest.

 

 

Regards

Steve

[wlaidlaw]

Beside the the point that all my dhree lenses work perfectly with my M bodys, I do not get the point of adjusting the rangefinder for infinity. Where is infinity btw? Wouldn't it be logical to adjust the rangefinder for the hyperfocal distance instead of infinity? Just for technical interest.

 

 

Regards

Steve

 

No! Infinity is the point from which all other adjustments/settings flow. Hyperfocal is different for each lens and close focus point changes as well. the only thing they all have in common is infinity.

 

Wilson

[Julian Thompson]

Quote:

Originally Posted by Julian Thompson

...So we are left, as I see it, with infinity FOCUS adjustment only being possible by shimming the sensor or adjusting the lens...

 

I didn't follow your reasoning. You have two focusing systems, rangefinder and lens scale, and either can be off. Whichever one you are using, if it's off, you can misfocus something at infinity (a.k.a. a long ways away). The stop at infinity can also be off.

 

To explain further - what I am trying to say is that it is only possible to adjust the rangefinder patch. You can't adjust the actual lens (well we can't - solms can I believe) - so that means that when the lens is at infinity that's it - no amount of rangefinder twiddling will actually affect the real-world focus the sensor sees. - Does that make more sense? (you could yank the rangefinder out of the camera and the focus will still be the same with the lens set to infinity - so when we refer to 'setting the infinity focus' all we mean is setting the rangefinder to a reference point.

 

 

 

But - Wilson - I think Steve may just have a point here - and it's funny because I was thinking similar last night but I didn't use hyperfocal distance in my reasoning.

 

What I was thinking is that most of our lenses are marked up to 10m and then just a quick rotation round that's it - the lens is at 'infinity'. So the distance from 10m to infinity is just a few degrees round the lens barrel as you turn it.

 

So what I'm getting at (and I guess so is Steve) - is that once you're focussing on something say 20m or more out from the camera you are not needing the rangefinder of the camery anyway - it's superfluous to requirements because the lens will be on infinity at that point anyway.

 

- that an image taken at 25m from the camera requires the same 'focus' as taking a photo of the moon that's '000000' s of miles away.

 

I wonder, (and I'm at 'the point' at last you'll be thrilled to hear!) - if in fact when Leica refer to the 'infinity' adjustment they really mean the point at which you can't focus 'any more' - ie 20m (say)

 

Because if this is the case it would explain why, when you set the rangefinder to 'align' two stars (or say, masts 500m away) that as you then wind on some focus to come closer in you are not able to get it to be right.

 

(This would make sense because, as is obvious, once the distance to an object goes beyond about the 20m point you are no longer going to be using the rangefinder anyway)

 

Sorry if this is a ramble - hope I am making my point in there somewhere (well it's more of a question than a point!)

 

:-)

[Advertisement]

Hallo,

Look here: Leica M

[jaapv]

Well I thought about this some more, and then some more :-)

 

When you adjust the rangefinder wheel you are not, of course, adjusting anything physical in the lens.

 

If you set your lens to infinity then it's at infinity, and whether your rangefinder is adjusted or not does not have any bearing on the image.

 

So we are left, as I see it, with infinity FOCUS adjustment only being possible by shimming the sensor or adjusting the lens.

 

But, as we move the lens away from infinity the rangefinder also moves obviously, and then you need to focus USING the rangefinder, and so the infinity REFERENCE becomes more important since it is from this that the base point for the other adjustments is set.

 

In practice I have done some far distance checking today and just as lct says it's a none issue because at the kind of distances we're talking about the depth of field means you just can't tell if any slight focussing changes have been made.

 

So in this case my M9 - post roller adjustment - is performing very well and I have solved the focus issue.

 

But - I know that to be 'technically perfect' I would first need to set the infinity roller then find (I don't even know where it is!) the other adjustment wheel to set the closer focus.

 

Is this behind the red dot with the vertical adjustment?

Yes, it is, and it needs a special tool. To adjust the body, you need:

1. A collimator

2. Targets of 1, 3 and 10 m length and something outside far away.

3. A reference lens that is known to be right in the middle of its tolerance span.

4. A fixed testing bench.

5. Special tools

6. Knowledge and experience.

 

For adjusing a lens you need the same, except you obviously need a reference body, plus tools to disassemble the lens and a stock of varying shims.

 

Each and every competent Leica technician or workshop has all of this and can do the job quickly and reliably for you. It is really not smart to start playing aroung yourself, unless you want to go the whole way and teach yourself to be a qualified technician.

 

The only thing that works sometimes and is very practical as an emergency procedure when you are out in the field is to give the roller cam a twist. You have a chance of improving things, there is a chance of making it worse, but it is reversible, and you will need to have your camera checked when you get home.