LFI on Backfocus - Left Hanging!

[gtownby]

Advertisement (gone after registration)

Nice article in the latest edition of LFI on M lenses on the M8. Odd though, because in several places, the article hints that backfocus shift "can be compensated for easily." Then, the article ends without ever explaining how to compensate!

 

(BTW, I've noticed no backfocus problem with my 28c or 50c, nor, according to the article, should I expect it from the 75c on order.)

 

--Gordon Ownby

Los Angeles

http://www.gordonownby.net

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[guy]

Maybe by sending the lenses back to Solms?

[johnbuckley]

I generally love LFI's coverage, and think in the early days of the M8, they were very good about trying to identify solutions to problems this forum had picked up on. But this article on backfocus was strangely defensive and obscure...

[lars_bergquist]

I got the impression that they suggested that you should lean back a little after focusing ... which is crap of course. I think the only 'cure' is simply not to use superspeed lenses at close distances, say inside 2 meters. A good plan in fact, if you are taking pictures of people, no matter which lens you are using.

 

The old man from the Age of f:4.5

[ho_co]

As I read it, LFI's recommendation was that once you bring the rangefinder into focus, you should nudge the focus ring a smidgen toward closer focus.

 

I found the article to be exceptionally complete, describing the problem in much more detail than we had worked out on the forum and even naming some lenses as likely offenders that we hadn't seen much complaint about.

 

As for the article being defensive, I felt that apparently Leica doesn't want to get a slew of lenses back for adjustment, and there may be good reason. LFI says (and I think someone on the forum discovered the same after sending a lens in for adjustment) that the adjustment simply amounts to setting the lens to frontfocus slightly, so that the focus shift is evened out between fore- and back-focus.

 

--HC

[erl]

I got the impression that they suggested that you should lean back a little after focusing ... which is crap of course. I think the only 'cure' is simply not to use superspeed lenses at close distances, say inside 2 meters. A good plan in fact, if you are taking pictures of people, no matter which lens you are using.

 

The old man from the Age of f:4.5

 

My copy of LFI has not arrived yet, so I won't comment on the article. Lars, I have just gone through a "Very heavy" session with backfocus involving several lenses, principally my Noct, coupled with the M8. All problems, in my case, were exacerbated for longer focus distances. Close focus distances were not too bad. Currently, my M8 + Noctilux are back in Solms to pin down the problem(s).

 

Howard, your comment about LFI suggesting to compensate for backfocus by "backing off" the focus ring a little for each shot is a PITA. I have, in the last two weeks, shot 2 - 3000 images on assignment doing just that. There is no way you can be sure of correct focus doing this. I have a loan camera while mine is returned for checking. The loaner, and several others I tried, all appear to have the same problem, to a greater or lessor degree. I fear there is a much deeper seated problem lurking somewhere.

 

I expect my units back in November. Iwill report anything of interest then.

[ho_co]

Advertisement (gone after registration)

... Howard, your comment about LFI suggesting to compensate for backfocus by "backing off" the focus ring a little for each shot is a PITA. ...

Erl--

1) It's not 'my comment.' That's their recommendation on page 51, where they say making the adjustment is quickly learned.

 

2) I didn't say the 'solution' wasn't a pain. I merely answered Gordon's question.

 

3) On the same page, and in fact in the same sentence, there's a point I like: After pointing out how film's thickness tends to mask focus shift as compared to the completely flat sensor of the M8, the magazine goes on to say that "The M8 is ... an excellent instrument for the analysis of previously almost imperceptible niceties of the M lenses."

 

Isn't that a nice, positive spin on the matter? "Oh, your lens exhibits focus shift? And you hadn't noticed it before? Well, just think how lucky you are to have an instrument like the M8 for such finely tuned analysis! You can use the M8 to explore the subtleties of your lenses!"

 

I guess one could say, a trifle cynically, use your quirky camera to study your quirky lenses.

 

--HC

[erl]

Howard, not for a moment was I attributing the comments to you. I fully understood that you were commenting on the comment(s). All this M8 stuff is a very good parallel to training a racehorse or driving a Maserati! Highly tuned .... highly temperamental, but great when it works to plan:D.

[rwfreund]

Several of these focus issues will be with us for as long as we use rangefinders.

 

This focusing system is inherently open loop. That means that as the adjustment to the focus ring is being made, there is no immediate visibility of the effects on the sensor itself. Instead of viewing the effect on a lcd or a ground glass, for example, we observe a proxy of that focus plane which is the half silvered focusing patch.

 

For all of this to work perfectly all of the time requires besides an accurately calibrated rangefinder roller and pivot, but also that the ramp match the focusing helix, or perhaps not in the case of a floating element, and that lenses at all distances and apertures have absolutely no spherical aberration.

 

It is enough to give the inner engineer the heebie-jeebies.

 

All of this requires absolute precision, not just well-damped feedback mechanisms. I think this reflection is giving me more appreciation for the Leica optical precision fixation. It is not merely enough to be the best, bu it is necessary to be better than that because of the inherent design constraints of the M system.

 

I still love the camera, heck I own two of them, but I am thinking that some sort of auto focus detection aid that provides closed-loop feedback is necessary in future models.

 

Digital sensors, all theology aside, are capable of resolving more than most film, and we as users feel justified in instant zooming to 1:1 or greater to determine the quality of the shot. There may never be sensors dramatically denser in pixels per square mm than what we currently are using due to optical limitations and prohibitions in manufacturing cost unless some dramatic departures are implemented.

 

-bob

[Peter Branch]

The thing I don't understand is why the focus shift on the Noctilux was not more obvious to more people with film. Even with objects at a considerable distance, ~ 50m, if you focused critically at f/1 and then, without changing anything other that the aperture and shutter speed, stopped down progresively by f/2 the image has markedly deteriorated. If you continue to stop down the image improves again - but - this is due to the rapidly increasing depth of focus not because the focus shift has gone away, in fact it is still getting slightly worse.

 

I fully understand that the M8 is a far more analytical tool but the effect was well known to those who used the lens with film. There is no real answer other than to shift the focus a bit.

 

I sold my Noctilux simply because it is not good enough for use with an M8 and transferred to a Summilux. I believe that Leica will have to pay a great deal of attention to several aspects of lens performance such as focus shift and field curvature in any new lenses it introduces.

[Pedro]

I bought an M8 a couple of weeks ago and I like it.

 

I was having some difficulty focusing at close distances but sought it was part of the learning curve (my other camera is a Canon 5D which focus very well automatically).

 

Now that I read this thread I understand what is going on...

 

Made some tests with a ruler and its clear my M8 & 35mm f2 are focusing to a longer distance than my selection in the range finder .

 

This is not acceptable in such an expensive camera.

 

Are all of you guys just guessing the focus by "backing off" the focus ring a little for each shot???

[hankg]

Made some tests with a ruler and its clear my M8 & 35mm f2 are focusing to a longer distance than my selection in the range finder .

 

This is not acceptable in such an expensive camera.

 

Are all of you guys just guessing the focus by "backing off" the focus ring a little for each shot???

 

Either your lens or your camera's rangefinder is miscalibrated and which ever it is should be adjusted to a standard. A few high speed lenses like the Noctilux, the 35mm Summilux and the Zeiss 50/1.5 Sonnar-C shift focus as they are stopped down which is a different problem to a degree that it is a noticable problem. The 35/2 once adjusted should be fine at all distances and apertures.

[jaapv]

As I read the article the conclusion was that a design change was needed on the Noctilux and Summilux 35asph and they suggested some stopgap measures to compensate with the current lenses. Whether one wants to live with this is another matter.

[Pedro]

Either your lens or your camera's rangefinder is miscalibrated and which ever it is should be adjusted to a standard. A few high speed lenses like the Noctilux, the 35mm Summilux and the Zeiss 50/1.5 Sonnar-C shift focus as they are stopped down which is a different problem to a degree that it is a noticable problem. The 35/2 once adjusted should be fine at all distances and apertures.

 

http://idisk.mac.com/pedro.h-Public?view=web this is the address of my Mac public folder where you may see the results of my test. I was focusing on the 31.

 

So must I send a brand new lenses and camera body to Leica for calibration? And if I buy a new lens must I send it again?

 

Sorry for all these questions but I am new to the range finder world and i must learn... to keep enjoying it.

 

By the way, browsed your web pages and liked what i found. Thanks a lot.

[jaapv]

 

So must I send a brand new lenses and camera body to Leica for calibration? And if I buy a new lens must I send it again?

 

Sorry for all these questions but I am new to the range finder world and i must learn... to keep enjoying it.

It happens not too often, but too often if you get my meaning. And is a pita. But chances are it is a one-time thing, and once the body is adjusted 99.99% of new lenses are fine.

[hankg]

.Mac - iDisk this is the address of my Mac public folder where you may see the results of my test. I was focusing on the 31.

 

So must I send a brand new lenses and camera body to Leica for calibration? And if I buy a new lens must I send it again?

 

Sorry for all these questions but I am new to the range finder world and i must learn... to keep enjoying it.

 

By the way, browsed your web pages and liked what i found. Thanks a lot.

 

Thanks for the compliment. The RF of my M8 was dead on, but I did get a new 90/4 that was off and returned it for one that was calibrated properly. It is unfortunate but some cameras and lenses are going out out of spec. I have the retailer check focus before sending me any new lens now (I always figure an adjustment at DAG into the price of a used lens).

 

You need to find out if it's the cameras or the lens and only service can tell you that. Hopefully you won't have this problem with future purchases.

[Dan States]

Several of these focus issues will be with us for as long as we use rangefinders.

 

This focusing system is inherently open loop. That means that as the adjustment to the focus ring is being made, there is no immediate visibility of the effects on the sensor itself. Instead of viewing the effect on a lcd or a ground glass, for example, we observe a proxy of that focus plane which is the half silvered focusing patch.

 

For all of this to work perfectly all of the time requires besides an accurately calibrated rangefinder roller and pivot, but also that the ramp match the focusing helix, or perhaps not in the case of a floating element, and that lenses at all distances and apertures have absolutely no spherical aberration.

 

It is enough to give the inner engineer the heebie-jeebies.

 

All of this requires absolute precision, not just well-damped feedback mechanisms. I think this reflection is giving me more appreciation for the Leica optical precision fixation. It is not merely enough to be the best, bu it is necessary to be better than that because of the inherent design constraints of the M system.

 

I still love the camera, heck I own two of them, but I am thinking that some sort of auto focus detection aid that provides closed-loop feedback is necessary in future models.

 

Digital sensors, all theology aside, are capable of resolving more than most film, and we as users feel justified in instant zooming to 1:1 or greater to determine the quality of the shot. There may never be sensors dramatically denser in pixels per square mm than what we currently are using due to optical limitations and prohibitions in manufacturing cost unless some dramatic departures are implemented.

 

-bob

 

 

Bob, there are two seperate issues at work...Range finders or lenses that are miscalibrated vs. focus shift.

 

The first can be adjusted for, the second is a natural tendancy for many high speed lenses regardless of focus system.

 

Focus shift occurs in many SLR lenses, just as it does in some of the high speed M lenses. In the past we never enlarged our films enough for it to be an issue. 100% crop on screen plus the 1.33 crop factor in the digital sensor are making previously invisible faults stand out.

 

On a camera like the D200, the AA filter often masks the effects of focus shift. On the M8, the files are naturally of very high resolution so we see it all the more.

 

Leica will just have to take this more into consideration in their future designs. It should be remembered that the kinds of shift we are discussing are generally invisible at normal print sizes and viewing distances. (Not counting the Zeiss Sonnar 1.5 which is off the charts)

 

Best wishes

Dan

[bradreiman]

weird-i think i got a defective m8. mine focusses with all of my lenses including the noctilux at closest focus, infinity and at different apertures. really it does. im dumfounded by these threads. i use my noctilux at f1 alot at all distances and accidently alll my shots are in focus?

[Advertisement]

Hallo,

Look here: Leica M

[carstenw]

Try the Noctilux at 2m and f/4. If that works too, never sell it. Some Noctilux copies have noticeable focus shift. Some 35 Lux Asph copies too. Most other lenses are much more in line.