Focus off on lenses wider than 50mm?

[Bengerman]

Hello,

I'm having issues with my lenses that are wider than 50mm. I have a few lenses that range from 40-28mm that seem to focus approximately 1inch in front of the intended mark (when testing at minimum focus distance). When I test my 50's (CV f1.5 & ZM f2) and my 90mm Summarit, they are all spot on.
The M9 is still under warranty. I'm concerned if I have it sent in for calibration the 50/90 accuracy will be off. Does it sound like the wider lenses are off (Leitz 40mm, ZM 28 & CV 28 f1.9)? Odd that all three are off by a similar amount.
Is this normal? Any advice would be greatly appreciated.

Thanks, Ben

 

[Phil_F_NM]

You might be experiencing a bit of focus shift that may be inherent to the lens designs. A small difference like you're describing wouldn't be noticeable on film since the film has depth but since the digital sensor is perfectly flat it brings out these characteristics of lenses. Check to see if the point of sharp focus moves with shots at different apertures, then you'll know. Of course, this might be a big can of worms so proceed with caution.

As you say it could also be your camera and your longer focal length lenses may be coincidentally off calibration as well.

Good luck!

Phil Forrest

 

[SDK]

Shorter lenses are more susceptible to errors in back focus distance than lenses of 50mm or more. If the distance of sensor to lens flange is too long, this would cause close focusing and poor infinity focus, and the problem would be much more apparent on a wide angle lens than on a normal or telephoto lens. I had this issue on an early Hexar RF that had to be fixed because the film to lens mount distance was off on the body. It only showed up with my 28mm and 35mm but 90mm was fine. Afterward all lenses were fine with the body. You should get the body checked.

 

[Bengerman]

Thanks for the replies! I'll likely have the wide lenses sent out for inspection. Thanks again for the help guys. I'll follow up with any updates.

Regards, Ben

 

[SDK]

It could be the lenses or it could be the camera body. You probably should have both possibilities checked.

 

[Rayt]

Are you sure the lenses are at issue and not your eye? The M9 only has a .68x finder magnification and it is possible you are just not focusing it right. You should consider getting the 1.25x magnifier. For me I get 100% on focus shots with Noctilux on a high magnification set up versus 25% without.

 

[ferider]

Be happy that your longer lenses work. 1 inch is a non-issue for 40/2 and 28/1.9-2.8 (see http://dofmaster.com/dofjs.html).

 

[jaapv]

Are you sure the lenses are at issue and not your eye? The M9 only has a .68x finder magnification and it is possible you are just not focusing it right. You should consider getting the 1.25x magnifier. For me I get 100% on focus shots with Noctilux on a high magnification set up versus 25% without.

Umm.. a magnifier on wideangles will lose the framelines..

 

[Ben Z]

Be happy that your longer lenses work. 1 inch is a non-issue for 40/2 and 28/1.9-2.8 (see http://dofmaster.com/dofjs.html).

If he knows the focus is off by an inch, it must mean he can see it, so how is it a non-issue? I used to buy into the DOF myth, but that was when film grain limited crop+enlargement. For example DOF is supposedly huge at infinity even with telephoto lenses, but I could see obvious OOF (in the form of lowered contrast and increased fringing) in my 135mm Tele-Elmar at infinity, even on the LCD zoomed-in.

 

[ferider]

If he knows the focus is off by an inch, it must mean he can see it, so how is it a non-issue? I used to buy into the DOF myth, but that was when film grain limited crop+enlargement. For example DOF is supposedly huge at infinity even with telephoto lenses, but I could see obvious OOF (in the form of lowered contrast and increased fringing) in my 135mm Tele-Elmar at infinity, even on the LCD zoomed-in.

Then again, the OP didn't test at infinity, right ?

The 40/2 and the 28/2.8 have focus shift and field curvature. While the 28/1.9 is a stunning performer and quite flat field, it is LTM mount, and best performance will depend on the LTM adapter chosen. We don't know how the OP measured (in the center, and if yes, how relevant is this, etc), but not only is 1 inch covered in the DOF, but very much within the parameters and errors of the test.

From what the OP describes we have no idea if the lenses are "off".

I test and optimize my wides at infinity and optimum aperture (which would be f5.6-f8 for the 40/2 and 28/2.8, and f4 for the Ultron). Then I might play with the LTM adapter for the Ultron. Then, I would consider having the 40 and 28/2.8 calibrated if necessary.

Once the lenses are optimized for infinity, close up performance usually falls into place, is "good enough".

And regarding film grain being more limiting than the M9 sensor, I recommend to try a modern film such as Kodak Ektar or similar. Not just yet

Roland.

 

[jaapv]

It is not the film grain - it is the physical thickness of the emulsion and the fact that film is never 100% flat. (Unless you have a vacuum film pressure plate)

 

[elmer3.5]

Hi, i would suggest you to make a quick test in order to discard the rangefinder from being out of alignment.

focus your camera with each of the wides to infinity, choose some far away object that can be easily seen on the RF patch.

That way you´ll notice immediatelly if the Rf is missaligned.

It happened to me in a lot of cameras that alignment go off from time to time.

Many times Rf lenses are poorly set from factory and need some adjustment, also cameras have this issues.

Check for infinity it will help to discard some problems.

Bye and good luck!

 

[Ben Z]

When I shot film, grain became objectionable before I could reach the point of enlargement where objects could be seen OOF that were supposedly within the DOF of lens and aperture. I shot mostly 100-speed E6. This is just anecdotal from my own past experience.

I agree with Jaap it's the film thickness/flatness vs the absolute plane of the sensors that accounts for why so many lenses are out of tolerance with the M9.