Leica M: 3 or 4?

[Richard G]

KTM, I've used Leica M for decades and never personally bonded with the M3. My eye favours the 35mm lens. At the same time, i feel the advantage of the M2/M4 when using the 50mm lens is the context you see in the viewfinder as you can see the action outside the 50mm frameline. Although i used them, i found the multiple viewfinders of the M4-P,
M6,MP to be overly busy. My first Leica was the M2, although these days i'm using a '68 black paint M4. I'm sure DAG could retrofit the M2R loading system into an M2 for you.

I agree with all of this. And don’t forget (as you used one before) the M3 frame liens for the 50 have curved corners. I was shocked when I first looked at those and have never given an M3 a moment’s serious thought as an acquisition since.

 

[Rob-F]

I agree with all of this. And don’t forget (as you used one before) the M3 frame liens for the 50 have curved corners. I was shocked when I first looked at those and have never given an M3 a moment’s serious thought as an acquisition since.

The rounded corners were to echo the same shape as some slide mounts of the time. Not a deal breaker for me; I just don't bother with the M3 owning to its lack of 35mm frame line.

 

[Huss]

Maybe try an M4 with the framelines that you don't want, removed. It can be done during a CLA.
Phil Forrest

Yeah I had the 75mm frame lines removed from my M7, M-A and M5. I'd never use a 75mm lens as I have 50s and 90, and the 75 - well I just don'r get it. Then again the last time I used a 90 on an M camera was a long time ago.
Anyway, remove the 75mm frames and the view is much nicer.

The M2 and M3 are not built to the same level. The M2 has a cheapened RF mechanism compared to the M3 and has been susceptible to light leaks through it. Wonder why there have been quite a few threads about M2s and light leaks on RFF? That's why.

Apparently the rf parts r not the same - stuff is cheaper on the M2 (obvious example is the film counter). Obviously light leaks only affects a few so perhaps a fit issue with those in conjunction w the diff parts.

https://en.m.wikipedia.org/wiki/Leica_M2

“The rangefinder system was also simplified from that of the M3 and this made it potentially more prone to rangefinder flare. The M2 has a rangefinder with a 0.72 magnification and framelines for 35, 50 and 90mm lenses instead of the 0.91 magnification and 50, 90 and 135mm framelines of the M3. This made it better suited for photojournalists who favour shorter lenses or for spectacle-wearers using a 50mm lens who sometimes find it difficult to see the framelines on the M3. The ground glass frameline illumination window of the M3 was replaced with a fresnel-type plastic lens. Finally, the ornate beveling around the various windows on the front of the M3 were flattened on the body of the M2. Unlike the M3, the widest framelines were not always visible so only one set of framelines were ever displayed at one time. “

 

[Erik van Straten]

Y The ground glass frameline illumination window of the M3 was replaced with a fresnel-type plastic lens.

Not so on the early M2, it is frosted glass with a fine linear structure on it.

Y Finally, the ornate beveling around the various windows on the front of the M3 were flattened on the body of the M2. Unlike the M3, the widest framelines were not always visible so only one set of framelines were ever displayed at one time. “

The M2 top plate as we know it, was the original design for the new M Leica in the early 1950's, made by Heinrich JANKE, the chief of the design department at Leitz in those years. However, the bosses at Leitz did not like it, it wasn't looking "technical" enough. So Janke made another design, the one that we know now as the M3 top plate.

Only later, in 1958, the original design was used for the then new M2. In the end it became the standard design for the top plate of many M-cameras, in fact all but the M5.

True story.

Erik.

 

[Beemermark]

Why is the M3 better than the IIIG? Except for the mount the load system, rewind, VF, etc are basically same. Lots of great screw mount lenses available.

 

[splitimageview]

You might ask DAG if he still offers the tulip-style takeup spool for M2. It’s a removable piece, and basically is M4-style loading. It will also work in the M3 but the frame counter won’t reset unless it’s pulled out and then replaced, which kinda defeats the purpose.

All this talk about M2 vs M4 vs M3 build quality makes me laugh. Just buy a well-cared for camera with the features you want, make sure it’s recently serviced by a quality tech, or if not, get the service done. This makes more of a difference than any quality differences that happened during manufacturing at Wetzlar 50+ years ago...

Can also modernize with a Ti shutter, multi coated viewfinder windows, new frameline mask, and other customizations. 🙂

 

[Huss]

Why is the M3 better than the IIIG? Except for the mount the load system, rewind, VF, etc are basically same. Lots of great screw mount lenses available.

IIIG - slow speeds separate from the faster speeds.
Adjusting speeds is more awkward and slower as have to lift and rotate shutter speed dial.
Film advance knob instead of lever
Separate vf to focus from the framing vf
Need to adjust the diopter on the focus vf depending on subject distance
Shorter RF baselegth
Harder to load as only the base comes off and so need to trim film leader
Can only use LTM lenses.

M3 - none of the quirks from the above, and can use LTM lenses as well as modern state of the art M mount lenses. Also has a light meter which syncs with the shutter speed dial to make for semi AE readings.

Yeah, big difference as to why the M3 replaced and is much better than the IIIG

 

[splitimageview]

IIIG - slow speeds separate from the faster speeds.
Adjusting speeds is more awkward and slower as have to lift and rotate shutter speed dial.
Film advance knob instead of lever
Separate vf to focus from the framing vf
Need to adjust the diopter on the focus vf depending on subject distance
Shorter RF baselegth
Harder to load as only the base comes off and so need to trim film leader
Can only use LTM lenses.

M3 - none of the quirks from the above, and can use LTM lenses as well as modern state of the art M mount lenses. Also has a light meter which syncs with the shutter speed dial to make for semi AE readings.

Yeah, big difference as to why the M3 replaced and is much better than the IIIG

Interestingly, the IIIg (and other models of II and III) has a shorter RF baselength but greater magnification, so the rangefinder is actually more accurate than an M3, by roughly 10%. It's a bigger gap of course to the 0.72x viewfinders.

 

[santino]

Is magnification really a subsitution for baselength ? I mean the wider the baselength the more accurate is the rangefinder, the greater the magnification the easier it is to focus.
In FSU world I don‘t beliebe that a Zorki 1 rf is more accurate than a Kiev II‘s rf - but the Kiev‘s magnification is clearly inferior to the Zorki‘s.

 

[splitimageview]

The mathematical equation for accuracy uses both the mag and the baselength...

 

[Huss]

Interestingly, the IIIg (and other models of II and III) has a shorter RF baselength but greater magnification, so the rangefinder is actually more accurate than an M3, by roughly 10%. It's a bigger gap of course to the 0.72x viewfinders.

That simplifies it. The effect on accuracy is determined more by the rf base length than by the magnification. Otherwise you can take a low magnification vf like a .58, stick a 1.4 X magnifier on the VF and then claim it is a more accurate vf than a .72 vf.
It isn't. You just made the image in the VF larger.

 

[Bill Blackwell]

So, I bought my first film Leica in 1974, a CL. I acquired an M3 around 1977 and kept it a few years. I have owned an M2 (four or five bodies), a couple original M4's plus the M4P and M4/2. I also owned a couple M6's and M6TTL when they first appeared. My latest film camera is an M5 from Sherry and I truly believe it may be the best shooter if you want a metered M. I recently traded it for an M2.

.... The M2 is my favorite viewfinder while I like the more modern film loading and rewind of the M4. A meter is nice but I really don't like the frame lines in a modern Leica as I find it just clutters everything up. And the reason I still use Leica's is the viewfinder and rangefinder focusing. ...

I presently own a M2 and an MP240 and several M mount lenses from Leica, Zeiss and Voightlander. ... Luckily I can own more then one camera.

... My question is which film body?

... what am I not seeing about the M3? ...

With your wealth of experience, why would you need to ask such a question?

Seriously!

 

[ktmrider]

I ask the question since the last time I owned an M3 was in 1977. I was simply wondering if there is something I missed that my memory (now 43 years later and not quite as good at 67 as it used to be) did not remember.

Nothing is perfect out in the world, even Leica. Each model is incredibly familiar to one another but with subtle differences which causes endless debate. As I stated early on, my favorite is the M2 primarily for its viewfinder. Guess I will explore having a rapid loading system installed or finding one in Hong Kong or Tokyo in a few months.

 

[splitimageview]

That simplifies it. The effect on accuracy is determined more by the rf base length than by the magnification. Otherwise you can take a low magnification vf like a .58, stick a 1.4 X magnifier on the VF and then claim it is a more accurate vf than a .72 vf.
It isn't. You just made the image in the VF larger.

That’s exactly right, except the claim is correct; RFs are more accurate with higher magnification, even by simply adding an accessory magnifier.

The formula is:

b' = (e * f^2) / (k * z) where b' is the effective base length of the rangefinder, e the visual acuity (0.0003 at approx. 1 arcmin), f the focal length, k the aperture and z the circle of confusion (0.020mm for APS-C, 0.030mm for 35mm film)

The effective baselength is the actual baselength times the vf magnification.

Here is an RF accuracy chart with various models of RF, some with accessory magnifiers, listing focal length and max aperture:

https://docs.google.com/spreadsheets/d/1BWB8ZaryAV8Q6UlBGFs6I5wwj-MdxeaSkLP_kiZVQgY

I put together this spreadsheet back in 2008 and have updated it periodically with more bodies and lenses; the original thread is here.

 

[leo mm]

One point that I haven’t seen anyone raise: to my understanding older Leica Ms can sometimes eventually suffer from viewfinder separation, which is an expensive repair. I have read this has something to do with an organic resin used in the viewfinder assembly, which was replaced with a different cement in the M4 onwards. This is something that has always made me a little wary of buying an M3 or M2, as appealing as they are. I have a few M4s. Obviously many people don’t worry about this and enjoy their older Ms.

Also, Tokyo and Hong Kong are very nice places to buy cameras in person and find unusual things, but these days the prices there are normally higher than in the US and Europe. Not to discourage you from enjoying the shopping!

 

[splitimageview]

Prism separation can happen but there are now several techs who can repair the prism for reasonable prices.

 

[Shac]

Isn't the potential limitation of the M3, the availability of parts - especially those for the rangefinder. I don't believe that this is the case for the M2/4/6 etc.

 

[Rob-F]

That simplifies it. The effect on accuracy is determined more by the rf base length than by the magnification. Otherwise you can take a low magnification vf like a .58, stick a 1.4 X magnifier on the VF and then claim it is a more accurate vf than a .72 vf.
It isn't. You just made the image in the VF larger.

But magnification still counts enough so that a magnifier on a .58 Leica can improve focusing accuracy. Here's why:

The focusing optical system consists of the Base length, the magnification, and the eye and nervous system doing the focusing. A fresh, rested eye and nervous system can focus the camera better than a tired one. The fresh and rested eye might even be able to focus a .58 finder as well as a tired eye can do with a .72 finder. The tired eye will probably do better with a .72 finder; or if using a .58, then augmenting it with the magnifier can make up for the eye's limitations. And, if the eye in question is not the equal of the perfect young eye with 20/20 vision, then the higher magnification assumes a greater role.

So the magnification becomes an integral part of an optical system that includes the user's vision. Having adequate magnification then becomes as important as base length.

 

[Sarcophilus Harrisii]

The theory is important, obviously—but the reality of manufacturing a rangefinder camera system can't be overlooked, either. In an ideal world, if production tolerances were perfect magnification versus physical base length probably wouldn't matter. But as it stands, when two rangefinder systems have identical EBLs, but differing quantities of optical magnification—assuming both are built to identical standards of mechanical precision—the one with the longest actual base length will be the most accurate. If it was as simple as magnifying the patch image up to a usable size, camera makers could have designed a half inch base length and simply amped the system up optically. The best did the exact opposite and implemented the longest possible physical base length, knowing that it was a superior approach which would not unduly exaggerate the effects of inaccuracy in the beam deflection, inherent in any mechanical and optical system almost regardless of cost. If you greatly magnify your patch to improve usability you are greatly magnifying whatever tolerances you've decided you have to live with in the mirror or lens deflector, it's pivot, plane alignment, etc, etc.

It seems nobody (well perhaps other than myself) ever mentions the ratio of deflection, either. Put simply the further the patch image travels across its usable arc from infinity to minimum distance, the more finely graduated the adjustment will be. In simpler terms you could think of it as being akin to gear ratio of the system. Less arc = quicker to use and more responsive—more arc = slower to respond but, potentially, far more precise.

According to no less a source than Lipinski. the swinging wedge system used by another first quality maker of rangefinder cameras *in itself* yields an increase in RF precision of an astonishing four times greater than the screw mount Leica. That is even before the gigantic physical base length involved, and the EBL, is factored in. As superlative as it undoubtedly is—by comparison—even the M3 is merely the best of a bad lot. But it's a wonderful viewfinder with integrated frame lines and lots of good features, so that's OK...

EBL is helpful for broad brush comparisons between different cameras. But it does not tell the whole story behind what makes a RF accurate. Not by a darned sight.

 

[splitimageview]

Ultimately the practical focus accuracy of these cameras is more dependent on the individual photographer and the calibration of the RF and the cam in the specific camera body.

Assuming infallibility of the shooter and perfect calibration, the formula above is useful especially when considering fast, longer lenses. Those cameras with short EBLs also generally aren’t good with such lenses anyway as the viewfinder and/or RF window itself can be partially blocked...