M8 Framing Accuracy?

[hestoft]

Advertisement (gone after registration)

I remember from my M6 film days that the frame lines were designed with a little "slop" built in so that one always got more on film than one saw through the viewfinder. I never really tested this, but, since I shot mounted slides, I felt that it always worked out in my favor, with the "slop" getting cut off my the slide mounts. All in all, I felt that I could frame my shots pretty accurately.

 

With the M8, we now have the advantage of seeing the actual framing results on the screen, in real time as it were and I am a bit surprised with the amount of "slop" actually present. In fact, I had to dig out the manual to check and see if I was using the proper frame lines!

 

In the enclosed example, shot with a 90mm f2.8, moved backward and forward until the outer edges of the frame lines lined up with the edges of the planks on the bench and alligned the top frame line with the top plank of the bench and you can see the result. To me, it seems like the actual frame is WAY too loose compared to what I saw and lined up in the finder.

 

I am wondering what others think based on their experiences with the M8. (By the way, I know you can crop, but since I am used to my 22 megapixel Leaf Aptus and 16.7 megapixel Canon EOS 1Ds Mark II, I really don't want to waste any pixels if I don't have to)

 

Thanks in advance for your input.

 

Ralf-Finn Hestoft

HESTOFT | flash detection

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!

[marknorton]

Interesting test, the 90mm frame is tiny of course, so you might expect there to be more safety margin with this frame. Were you framing to the inside or outside of the lines? The other thing to bear in mind is that the focal length of the lens changes as you close focus and the frame lines are set to be most accurate at closest focus.

 

Right now though, it looks like there's 20% more in the image than you expect which I agree is a little too slack.

 

Care to repeat the test with a 35 or shorter lens?

[ho_co]

If I'm not in error, S Reid says he was among many photographers requesting that Leica improve their framing accuracy, and he says they followed his recommendation.

 

After reading Reid's review, where he said the M8 framelines are more accurate than previously; and Puts's review, where he said the M8 framelines are less accurate than previously, I asked Leica Technical in New Jersey and got the following response:

 

 

"The framelines are absolute exact at 0,7m. So they will show more when focused at infinity.

 

"Leica left the old way of beeing [sic] exact at 2m and cutting at 0,7m and adding something at infinity.

 

"Aim was not to cut anything. Better is to add something, because this can be easily erased digital if it annoys the photographer."

 

 

--HC

[mike prevette]

It's possible that not all leica 90's render the same field of view. Lenses of the same mm often vary slightly in FOV. For example canon 90's arent the same FOV as Leica's or nikons.

 

Even the exact same lens can change FOV depending on focus. There are WAY to many variables for leica to take into account in order to make the frame lines accurate.

[ho_co]

It's possible that not all leica 90's render the same field of view. Lenses of the same mm often vary slightly in FOV. For example canon 90's arent the same FOV as Leica's or nikons.

Mike, you're right that there are a lot of variables, but remember that with Leica lenses 50mm and longer, the actual focal length is engraved down at the end of the ft/m scale, so you can be sure exactly what the focal length is of the lens in your hand when focused to infinity. The engraving shows the last full millimeter plus the first decimal.

 

For example, a 50mm might be engraved "19," indicating 51.9 mm actual f.l.

 

Or a 90mm might be engraved "95," which would mean 89.5 mm.

 

Actual focal length is far more important on a rangefinder than on a reflex since you're not focusing through the lens: The focusing cam of the lens must be matched to its actual focal length.

 

--HC

[adan]

You may already know this, but I'll state it for completeness. When a regular lens is focused closer, it moves further from the sensor or film (not true for internal or rear-focus lenses, but no M lenses fit those categories). The effect of this movement is like pulling a slide projector further away from the screen - the projected image grows larger.

 

Therefore, lenses crop more when focused on near objects than on far objects, and longer lenses (e.g. your 90) move more and change cropping more than wide-angles.

 

The Leica M viewfinder compensates for parallax changes with subject distance, but not for this slight "zooming" effect. To be sure that nothing is accidentally cropped out at ANY distance, the framelines are biased towards correct framing at close distances and thus loose at longer distances. And the longer-lens frames are looser in this regard than the wider ones.

 

The 90 frames on a film M, at 15 meters/50 feet (across a city street) actually show an area about like a 105mm lens (I checked this with an M 90 and a 105 on a 100% Nikon F body, once).

 

Mark isn't far off - that's about 15%-20% of slop - and it is normal and necessary. Just one of the joys of non-ttl viewing with an RF.

 

OTOH, in the studio at 1 meter I had to be very careful with the 90, because the frames really did show the exact cropping, with no room for error.

 

Leica's old accesory-shoe finders for 90 and 135 lenses actually had double lines or internal tick marks to show framing at close and far distances. Unfortunately, there was never a "120mm" accessory finder made that could work with a 90 on the M8.

 

There are various rules of thumb (or eye) for accomodating this - use the outside of the frame lines for medium distances, the inside for close distances, and the outside plus one frameline thickness for infinity. I haven't yet tested whether they still apply to the M8.

 

- or just do what you've done in this shot. Chimp the results until you develop a feel for how to place the line for a variety of subject distances.

[mike prevette]

Advertisement (gone after registration)

HC- I never new what those numbers meant. Heh you learn something new every day.

[carstenw]

HC, I never *noticed* those numbers before. Thanks for the insight.

[osera]

The engraving shows the last full millimeter plus the first decimal.

 

--HC

 

Interesting, thanks. On my 75 summicron, there's a little "50." Does that mean, then, 75.0?

 

Eye position relative to the finder also makes a difference to my perception of the FOV, especially with the WA. I get this when I have the viewfinder magnifier on for a longer lens, then switch for a, say, 24mm and leave the magnifier on. (And I wear glasses ) I have to move my eye around a little to really see the frame edges, and it makes for certain inaccuracies. It also makes it difficult to very accurately gauge the frame edges with anything wider like the 21, so I've always (with film) preferred external viewfinders - but that's just me.

[harmsr]

Personally, I like the old philosophy of Leica relative to framelines best. (example my M5)

 

The M8 is much looser than the M5 for the reasons given above.

 

I rarely take photos for my style at .7m where the frame lines are accurate. Basically when focused in mid range I use the outside of the frame lines plus a line. When closer to infinity, it becomes a two line "guess"

 

I also liked the older frame line construction of my M5 much more, as the lines are thicker and bolder (covering more area).

 

Maybe Leica will offer a la carte options / replacements in the future. (My hope.)

 

Ray

[sean_reid]

I discussed this in Part One of my review but the frame lines are are as accurate as they can be, given the variables. What's really happening with DRFs (and this also happened when people started using the R-D1) is that immediate LCD feedback is showing people a discrepancy that they may have not fully realized with their film RF cameras. It's been there forever with RFs, in fact.

 

The trick is to use a familiar set of lenses and learn where the edges will tend to fall with those lenses on the M8 relative to the frame lines (at various distances). This is also something I've discussed in more detail in various articles.

 

Cheers,

 

Sean

[ho_co]

Interesting, thanks. On my 75 summicron, there's a little "50." Does that mean, then, 75.0?

Allen--Yes, exactly.

 

It's actually one of the features that used to be necessary and is becoming less so, but Leica keeps up the tradition. Thanks for noticing, Allen, Carsten, Mike!

 

In the old days, lenses were more likely to vary in focal length; Leica kept the quality high on individual samples, but allowed a bit of focal length tolerance. These were the days when a lot of lenses had removable heads and could be used on the Visoflex, so a finished lens head could easily be matched to a particular focusing helicoid.

 

You'll notice, for example, that all Dual Range Summicrons have exactly the same focal length (51.9mm as I recall), but other Summicrons produced at the same time had several different actual focal lengths. That's because the double helicoid of the Dual Range was quite expensive and Leica produced only one version. Other Summicron 50 lens heads would go into other mounts, suited to their own focal lengths (and of course, some of those would also be 51.9). This 51.9mm was actually the design focal length for all the 50 Summicrons at the time.

 

With modern production practices, Leica tends to hold the actual focal length with less or no variation.

 

Now, go out and amaze people with your new knowledge!

 

--HC

[nrj]

Hi HC

 

I think there were two versions of the Dual Range Summicron - I seem to remember some information on this on the Cameraquest site. I've owned both versions. The one I have now doesn't have any little numbers on it, so thanks for giving me the exact focal length.

 

Nik

[nrj]

If I'm not in error, S Reid says he was among many photographers requesting that Leica improve their framing accuracy, and he says they followed his recommendation.

 

After reading Reid's review, where he said the M8 framelines are more accurate than previously; and Puts's review, where he said the M8 framelines are less accurate than previously, I asked Leica Technical in New Jersey and got the following response:

 

 

"The framelines are absolute exact at 0,7m. So they will show more when focused at infinity.

 

"Leica left the old way of beeing [sic] exact at 2m and cutting at 0,7m and adding something at infinity.

 

"Aim was not to cut anything. Better is to add something, because this can be easily erased digital if it annoys the photographer."

 

 

--HC

 

I'd be interested in the reasoning behind this choice - did Leica sound out users or was it an engineering choice I wonder. I've shot with M film cameras for a long time and never bothered about frameline coverage, probably because I don't use them up close much. So lines accurate for 2m was a good choice since the field of view doesn't change as much from 2m to infinity as it changes from 2m to closest focussing distance. But when I use my Mamiya 7 where I think the framelines are set for their closest focusing distance I usually end up with more in the frame than I bargained for - which can get annoying at times. When I had a Plaubel 670 the framelines changed according to the distance focussed -great design, but I guess very expensive.

 

Nik

[thrid]

Well, I was afraid that Leica would go this route.

 

Its been an issue with every metered M body since the M6, when Leica decided to shrink the framlines. Of course they are now accurate at .7 meters, but how often do you shoot that close up? The older M cameras showed coverage at 1 meter and are more accurate at real world working distances. The new 50mm markings in cameras like t he MP/M7 are so far off the mark that I simply gave up shooting that focal length on those cameras.

 

Obviously I do not expect the framing accuracy of an SLR from a rangefinder, but there is a big difference between an acceptable margin of error (as we had with the old markings) and being wildly off the mark, as the new markings are.

[carstenw]

It's actually one of the features that used to be necessary and is becoming less so, but Leica keeps up the tradition. Thanks for noticing, Allen, Carsten, Mike!

 

Interestingly, I only have these numbers on my Wetzlar R lenses, not the Midland ones...

[ho_co]

Interestingly, I only have these numbers on my Wetzlar R lenses, not the Midland ones...

Carsten--Remember, the actual focal length is engraved only for lenses 50mm and longer. You may be looking at lenses shorter than that. My Midland 75 has the engraving "50."

 

Or of course, I may be wrong...

 

--HC

[ho_co]

Nik, P.G.--To my knowledge the relationship of frame coverage to photographed area has been the same with the M series from the beginning till now. There was a minor adjustment to the frames when the extra framesets were added (75mm and 28mm) in the M6, but I think the coverage stayed the same.

 

As for why they made the choice, I think they were open to new ideas: Before the M8, at close focus distance you would get a little less than what was in the frame lines--in other words, you had to remember to add a little room to be sure you got what you wanted. So the compromise was to frame for a common working distance, 2m, lose a little on the close end (as you say, how often do you work at closest focus?), and although you gain a little at greater distances, the slide mount will cover most of that.

 

But when did you last shoot slides? Since the M8 is digital, it's practical to get more around the edges, which you can trim if you wish, and guarantee that you always get at least what you think you'll get.

 

In practice, I think the M8's design is better for today's way of working, just as the design of the earlier M's was a better design for shooting transparencies.

 

As S Reid said: It's not like an SLR, but you quickly get the hang of it.

 

--HC

 

By the way--welcome to the forum, P.G.! I haven't tried the MP or M7, so I can't argue with you on using a 50 with them. It works just fine on my M6, and I'm still waiting to try it on an M8.

[carstenw]

Carsten--Remember, the actual focal length is engraved only for lenses 50mm and longer. You may be looking at lenses shorter than that. My Midland 75 has the engraving "50."

 

Or of course, I may be wrong...

 

My Midland 50 Cron-R and 135 Elmarit-R lenses don't have the numbers. My Wetzlar 80 Lux-R (95=79.5) and 180 Elmarit-R (90=179.0) do. "My" 35 Cron is no longer mine and I can't check any more

[Advertisement]

Hallo,

Look here: Leica M

[ho_co]

Hi HC

 

I think there were two versions of the Dual Range Summicron - I seem to remember some information on this on the Cameraquest site. I've owned both versions. The one I have now doesn't have any little numbers on it, so thanks for giving me the exact focal length.

Nik--Very interesting! This is news to me!

 

As I recall, you can unscrew the lens heads of 50 Summicrons of that vintage. If you do that, I think you'll find the actual focal length lightly hand-engraved on the normally-hidden part of the lens head mount. And then on the focus mount itself you'll find engraved the serial number of the lens to which it was originally mated.

 

Actually, since all Dual-Range Summicrons had to have the same focal length because there was only one focus mount assigned to it, maybe they didn't bother to engrave the actual focal length.

 

And by the way, don't bet the farm just yet! I *think* the design focal length of the Summicrons at that time was 51.9 mm. If anyone finds a different value engraved on his DR mount, then I'm wrong.

 

--HC