"The 50mm Summilux ASPH was Designed to Front Focus" <-- Leica NJ Technician. Thoughts?

[eyeheartny]

Advertisement (gone after registration)

6 hours ago, oldwino said:

Sounds like they are admitting your original camera was a bit dodgy...You're at the point now where you need to get the "repair" back and try it out. Hopefully, the situation will be resolved in a positive manner.

The camera and lens will be back to me on Monday so I’ll hopefully know soon whether things were resolved satisfactorily. I’m very nervous about whether the body will be less tight/solid feeling after having been fully disassembled and put back together with parts from a different body. I’m prepared to engage American Express to help me fight this with Leica if they won’t get it working correctly. They have a responsibility to have working gear under the warranty or replace a faulty item. 

[01af]

Am 17.11.2022 um 00:16 schrieb eyeheartny:

"wide open, the Summilux-M 50 mm Asph was designed to front-focus, so it will never be fully accurate at f/1.4 because then it would not accurately focus at narrower apertures." Is that complete mumbo-jumbo as it seems to me?

Yes and no.

All lenses exhibit aperture-related focus shift—some more, some less—which basically is a consequence of inevitable residual spherical aberrations. At any given focus setting, the actual position of the image plane will shift with aperture, closer to the camera at wider apertures, away from the camera at smaller apertures. With a good lens, this effect will be too small to be a problem in real life.

So, any rangefinder lens (which will be focused for distance, not sharpness, unlike SLR lenses or lenses for use with electronic viewfinders) can be adjusted to be accurate at one aperture and will be slightly (slightly!) inaccurate at other apertures. Which aperture to adjust for is purely a matter of user's (or technician's) choice and is not inherent to the lens design.

For practical intents and purposes it makes sense to adjust a rangefinder lens for an aperture one f-stop down from full aperture—that is, f/2 for a Summilux. This minimizes the focus errors across the whole aperture range, and it usually also increases focus accuracy at full aperture because the (minuscule) front-focus will compensate for several factors which often lead to back-focus (such as re-framing after focusing). If you prefer, you can advise your service technician to adjust your Summilux lens to be perfectly accurate at f/1.4 rather than f/2 ... which however is not really a good idea.

If your lens is badly off (say, by several inches at portrait distance) then it simply is totally mis-adjusted and needs service. An error that bad has nothing to do with aperture-related focus shift.

[Advertisement]

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

[Harpomatic]

On 11/20/2022 at 6:20 AM, 01af said:

Yes and no.

All lenses exhibit aperture-related focus shift—some more, some less—which basically is a consequence of inevitable residual spherical aberrations. At any given focus setting, the actual position of the image plane will shift with aperture, closer to the camera at wider apertures, away from the camera at smaller apertures. With a good lens, this effect will be too small to be a problem in real life.

So, any rangefinder lens (which will be focused for distance, not sharpness, unlike SLR lenses or lenses for use with electronic viewfinders) can be adjusted to be accurate at one aperture and will be slightly (slightly!) inaccurate at other apertures. Which aperture to adjust for is purely a matter of user's (or technician's) choice and is not inherent to the lens design.

For practical intents and purposes it makes sense to adjust a rangefinder lens for an aperture one f-stop down from full aperture—that is, f/2 for a Summilux. This minimizes the focus errors across the whole aperture range, and it usually also increases focus accuracy at full aperture because the (minuscule) front-focus will compensate for several factors which often lead to back-focus (such as re-framing after focusing). If you prefer, you can advise your service technician to adjust your Summilux lens to be perfectly accurate at f/1.4 rather than f/2 ... which however is not really a good idea.

If your lens is badly off (say, by several inches at portrait distance) then it simply is totally mis-adjusted and needs service. An error that bad has nothing to do with aperture-related focus shift.

Having thoroughly tested the Summilux 50mm Asph for focus shift and accuracy the information you are sharing, although accurate, does not apply to the lens in question: focus accuracy wide open is absolutely spot on and at no aperture the focus shift lessens the sharpness of the plane of focus, which always falls well within the depth of field. This is thanks to the floating element design.

The OP is absolutely right in wanting the lens to focus very, very precisely wide open.

Having the Lux 50mm Asph adjusted for accuracy at F2 would be a mistake.

[eyeheartny]

For anyone interested in this issue, Leica returned the camera and lens to me. The problem is now reversed-- it focuses accurately when moving from the infinity stop back to the target, but going from the MFD stop to the target the lens back focuses. This is the reverse of the issue where it front-focused when moving from infinity. Clearly there is something wrong somewhere and I don't know where it is or whether Leica is capable of fixing this. It's deeply distressing that they cannot seem to rectify the issue despite me demonstrating clearly that something is very wrong with this system. The degree of back focus is approximately 0.5-1.5" each time, in repeatable tests indoors on a tripod. 

 

 

[andybarton]

I have hidden several posts in this thread, for unnecessary fighting between two members.

This really isn't called for and it would be appreciated by all if it would stop. Thanks.

[jaapv]

Actually this lens is notorious for problems like these. The tolerances on the floating element mechanics are on the limit of what is technically possible and quite a few examples need adjusting by Leica, sometimes twice, before they are perfect. Don't instruct Leica to adjust at f 2.0 when sending it in. When it is finally sorted it will not exhibit focus shift, that is the reason it has a floating element. 

I agree with LCT that it is self-defeating to start fiddling with a lens when any error it might have does not affect your photography. 

[jaapv]

Advertisement (gone after registration)

14 hours ago, eyeheartny said:

For anyone interested in this issue, Leica returned the camera and lens to me. The problem is now reversed-- it focuses accurately when moving from the infinity stop back to the target, but going from the MFD stop to the target the lens back focuses. This is the reverse of the issue where it front-focused when moving from infinity. Clearly there is something wrong somewhere and I don't know where it is or whether Leica is capable of fixing this. It's deeply distressing that they cannot seem to rectify the issue despite me demonstrating clearly that something is very wrong with this system. The degree of back focus is approximately 0.5-1.5" each time, in repeatable tests indoors on a tripod. 

 

 

As far as Leica M memory reaches, it has always been the technique for optimal focus to go down from infinity. The other way will be slightly inaccurate due to unavoidable mechanical tolerances in the mechanism. Your lens is calibrated perfectly if focus is spot-on from infinity. Always return it to infinity for the next shot. Hunting focus will result in focus errors on a rangefinder. 
Thank you for highlighting this half-forgotten fact. 

[eyeheartny]

6 hours ago, jaapv said:

As far as Leica M memory reaches, it has always been the technique for optimal focus to go down from infinity. The other way will be slightly inaccurate due to unavoidable mechanical tolerances in the mechanism. Your lens is calibrated perfectly if focus is spot-on from infinity. Always return it to infinity for the next shot. Hunting focus will result in focus errors on a rangefinder. 
Thank you for highlighting this half-forgotten fact. 

 

8 hours ago, jaapv said:

Actually this lens is notorious for problems like these. The tolerances on the floating element mechanics are on the limit of what is technically possible and quite a few examples need adjusting by Leica, sometimes twice, before they are perfect. Don't instruct Leica to adjust at f 2.0 when sending it in. When it is finally sorted it will not exhibit focus shift, that is the reason it has a floating element. 

I agree with LCT that it is self-defeating to start fiddling with a lens when any error it might have does not affect your photography. 

 

To be clear, as I have stated multiple times, the errors DID AFFECT my photography, to the point where numerous images were ruined from an anniversary trip due to the small focus error-- out of focus eyes, etc. Anyway. 

It's disappointing to learn of how finicky and messy this lens is to get working properly. For the price, I expected way, way more from both the body and lens. 

[jaapv]

As I explained - it is working properly now - as long as you focus from infinity down, and not from minimum focus distance up, which is the way it should work. Yes, rangefinder photography is an acquired skill, especially with demanding lenses, like the Noctiluxes, some Summiluxes, the Apo-Telyt 135, etc.
I agree that a Leica M, any Leica M, is an idiosyncratic camera which appeals to many, but not to everybody. But if you master it, you'll love it.

To repeat: any mechanical linkage has an unavoidable (minimal) amount of play, so it matters for the accuracy which way you turn it.

[Jeff S]

13 hours ago, jaapv said:

 When it is finally sorted it will not exhibit focus shift, that is the reason it has a floating element. 

 

Well, technically a secondary benefit; the floating element is primarily designed to improve performance at close focusing distances (reducing spherical aberrations). 
 

We’ve been repeating this exchange since 2014.  In Karbe’s words (see also comments from Rick and 01af)….

Jeff

[archive_all]

On 11/21/2022 at 5:02 PM, eyeheartny said:

For anyone interested in this issue, Leica returned the camera and lens to me. The problem is now reversed-- it focuses accurately when moving from the infinity stop back to the target, but going from the MFD stop to the target the lens back focuses. This is the reverse of the issue where it front-focused when moving from infinity. Clearly there is something wrong somewhere and I don't know where it is or whether Leica is capable of fixing this. It's deeply distressing that they cannot seem to rectify the issue despite me demonstrating clearly that something is very wrong with this system. The degree of back focus is approximately 0.5-1.5" each time, in repeatable tests indoors on a tripod. 

 

 

I was interested to find out what happened. I just did a test and even though I thought my lens was perfect I got a similar result. In normal use I've never noticed an issue. So doing an infinity to in focus test and then mfd to in focus test, I get accurate focus from infinity and then maybe about .25" back focus from mfd. It's a small enough amount where I never noticed it but like @jaapvsaid, I always start from the far end.

[eyeheartny]

14 minutes ago, archive_all said:

I was interested to find out what happened. I just did a test and even though I thought my lens was perfect I got a similar result. In normal use I've never noticed an issue. So doing an infinity to in focus test and then mfd to in focus test, I get accurate focus from infinity and then maybe about .25" back focus from mfd. It's a small enough amount where I never noticed it but like @jaapvsaid, I always start from the far end.

How far from the target were you shooting? 

[archive_all]

4 minutes ago, eyeheartny said:

How far from the target were you shooting? 

That was around 1 meter more or less. By 1.5 - 2 meters the depth of field increases enough where it looks identical to me in either direction.

[eyeheartny]

1 minute ago, archive_all said:

That was around 1 meter more or less. By 1.5 - 2 meters the depth of field increases enough where it looks identical to me in either direction.

Yeah I’m seeing a variance of around 1-.5” at about 1.3 meters. Much more than you are seeing it seems. 

[archive_all]

33 minutes ago, eyeheartny said:

Yeah I’m seeing a variance of around 1-.5” at about 1.3 meters. Much more than you are seeing it seems. 

Are you outside of the return period? If you have a local Leica dealer maybe see if you can also test a lens side by side. 

[eyeheartny]

1 hour ago, archive_all said:

Are you outside of the return period? If you have a local Leica dealer maybe see if you can also test a lens side by side. 

I am outside of the return period. Purchase was September 8. I thought for a while that the issues were user error as I’m new to rangefinders but not to photography. I’m kicking myself for not getting a focus target and testing rigorously right away as I could have returned earlier. Some of why I believed that was this forum and others that push the idea that one just has to learn to use the system. It was only after the initial front focus issue ruined a bunch of anniversary trip photos that I realized something was wrong and it wasn’t user error. 
 

There is not a Leica store near me. 

[hmzimelka]

3 hours ago, archive_all said:

I was interested to find out what happened. I just did a test and even though I thought my lens was perfect I got a similar result. In normal use I've never noticed an issue. So doing an infinity to in focus test and then mfd to in focus test, I get accurate focus from infinity and then maybe about .25" back focus from mfd. It's a small enough amount where I never noticed it but like @jaapvsaid, I always start from the far end.

What camera are you using? My M11 does the same as Eyeheartyny, but recently compared it to another new M11 which gave the same degree error with my Nokton 75/1.5 which is HUGE. About 8cm back focus when pulling focus from close focus at a subject distance around 1.5m.

My Summicron 50mm  also behaves like this, but I haven't measured the difference yet. My Ultron 28mm ASPH II does it too. 

I haven't got access to another M camera such as a M10-R or M10M to compare it to. 

My M11 has a lag/hesitation when switching focus directions that my Leica M3 simply doesn't have. The issue is very very frustrating as I can't pin point if its solely lens or both lens and camera related... and if it's the camera that it's the same with all M11 cameras. I understand that this issue might be intrinsic to the M system, but the degree of error we're experiencing (Eyeheartyny and myself)  just doesn't sit right with me.

Edited November 23, 2022 by hmzimelka

[jaapv]

3 hours ago, eyeheartny said:

and it wasn’t user error. 
 

Well, technically it is unless it is more than specified, as this is simply the way a rangefinder system. behaves. All rangefinder cameras must be focused this way.
As for the variation between cameras, mechanical linkage play has a tolerance span and I have no idea what it is. Only Leica can tell whether a camera/lens combo is within tolerance. They can only test this with Leica lenses.

[eyeheartny]

15 minutes ago, jaapv said:

Well, technically it is unless it is more than specified, as this is simply the way a rangefinder system. behaves. All rangefinder cameras must be focused this way.
As for the variation between cameras, mechanical linkage play has a tolerance span and I have no idea what it is. Only. Leica can tell whether a camera/lens combo is within tolerance. They can only test this with Leica lenses.

My 50 Summilux was initially front focusing by 2-3” when moving from infinity and was accurate when moving from MFD. This was a repeatable and consistent issue. That’s not user error if it’s the same every time: accurate from one direction and 2-3” off in the other. I know you guys like to blame everything on user error but you’re simply wrong here if that’s what you are saying. 
 

Leica themselves acknowledged that that behavior was problematic. It was so bad it led to them replacing the entire internals of the camera and also servicing the lens. So again, you are wrong: it was not user error. 

Edited November 23, 2022 by eyeheartny

[Advertisement]

Hallo,

Look here: Leica M

[jaapv]

Being adjusted the wrong way around  is not user error, but being accurate in the direction from infinity only is, as Leica puts it so nicely "System-bedingt" 
Your camera is fine as it is just now, unless the play in the system is out of tolerance. It just being serviced it is unlikely, but by all means have it checked.