Very interesting answer from Leica on 35mm 1.4

[ho_co]

Advertisement (gone after registration)

William Smith--

If you can add to the conversation, please do.

 

If you can't, just read along.

 

And if you just like to be crude and show yourself ill informed, I'm sure you can find other forums where you can do that. But go away from here!

 

--HC

[Advertisement]

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

[bsmith]

William Smith--

If you can add to the conversation, please do.

 

If you can't, just read along.

 

And if you just like to be crude and show yourself ill informed, I'm sure you can find other forums where you can do that. But go away from here!

 

--HC

 

I enjoy the internet thread version of the Jerry Springer show !!!

[wlaidlaw]

Children, children! Remember it is still before the 8PM watershed in the USA and sensitive adults may be able to read this thread.

 

Wilson

[ho_co]

TOPIC:

Can the cause of the difference in behavior among (for example) Tim’s, Bill’s, Sergio’s and Jamie’s 35 Summilux ASPH lenses lie in the lenses’ 6-bit coding?

 

Let’s break the lens into three parts: a) the optical head, consisting of the glass and diaphragm; the focusing mount, consisting of the parts that hold the lens’s axis perpendicular to the center of the image plane and also move the optics forward and backward while communicating that movement to the camera’s rangefinder; and c) the bayonet mount, which attaches the lens to the camera and conveys lens information to the body.

 

Adding zebra coding to a lens means exchanging the bayonet. That could result in 1) moving the lens off-center; 2) cocking the lens so that its optical axis is not normal to the sensor/film plane; 3) placing the optical unit at a different distance from the image plane than previously; or 4) none of the above. As I see it, those are the only four possibilities when exchanging the lens bayonet.

 

The matter of focus shift when stopping down the lens is a strictly optical phenomenon. Whether or not the lens is realigned as in any of cases 1-3 above, the optics will still behave the same way as before. Cases 1 and 2 will result in loss of image quality; case 3 will result in loss of focusing accuracy. But the focus shift on stopping down will not be changed.

 

(Obviously, if case 3 applies, then a focus shift previously seen at 10 ft may now occur at 11 ft or at 9 ft, but the same shift occurs. Since the rangefinder is no longer getting accurate information about the focus distance, the problem may be more difficult to diagnose.)

 

That’s what I tried to say above (http://www.leica-camera-user.com/digital-forum/17699-very-interesting-answer-leica-35mm-1-a-16.html#post246519 Note 4): Coding a lens that is working fine will not cause it to begin to shift focus with changes of aperture. The optics will still behave exactly as they have been doing.

 

That is not to say that there isn’t a difference among different samples of the 35/1.4. Obviously there is a marked difference, but *logically* I don’t see how lens coding can have anything to do with it.

 

Granted, according to the current consensus, the uncoded lenses generally work fine and the coded ones don’t. But we could just as well say that the older samples work better than the newer ones. Or we could argue that the chrome versions work better, not because they are chrome or because they are uncoded, but because they are older.

 

That could imply a change in lens design between the older and newer lenses, but it wouldn’t be justification for not having a functioning lens officially coded.

 

Or perhaps the aluminum mount of the current anodized version puts different stresses on the optical unit, causing the lens to behave differently from the chromed brass version previously offered. (Call that an ‘unintended design change.’)

 

Just my opinion. Please feel free to shoot it down. I offer it simply to add another thought to the considerations here. I don’t know why some lenses work better than others, or why Solms says Tim’s lenses behave properly despite the fact that Jamie’s sample behaves better. In Jamie’s situation, I would also be reluctant to send my lens off for modification even though logic says that changing the bayonet mount could not change the characteristics of the lens with regard to focus shift.

 

And Tim—the image you posted again above of the yellow-masked model (http://www.leica-camera-user.com/digital-forum/17699-very-interesting-answer-leica-35mm-1-a-16.html#post246877) is one of my most favorite of the Venice series. It’s evocative and plastic and a perfect representative of what Leica’s lenses can do in the hands of a talented photographer.

 

I just wish it were as simple as you make it look!

 

And as I remarked before, you’re the guy with the consistent problems!

 

--HC

[ho_co]

… an aspherical lens surface only elimates spherical aberation completely for a specific pair of image/object distances.

P—

Your statement raises a couple questions for me:

 

1) I would have thought that perhaps only a single image distance would be corrected, rather than a pair of distances. Or do I misread you and did you intend to say that an aspherical surface could only eliminate SA for a single specific image/object distance? (In other words, are you taking into account the possibility of a macro correction as well, where the lens might be only millimeters from the object but meters from the camera?)

 

2) Am I correct to infer then that designing a lens with 14 lens surfaces, seven of them aspherical, could correct for SA at seven focus distances? And could I go a step further and infer that a compound lens could be designed with no spherical aberration at all if it contained only aspherical surfaces? (That is, I know that astronomical telescopes use a single aspherical collector to get near-perfect performance at a single aperture, but I would not have thought the idea could be extended to multiple surfaces.)

 

Thanks!

 

--HC

[Jamie Roberts]

TOPIC:

{snipped} case 3 will result in loss of focusing accuracy. But the focus shift on stopping down will not be changed.

{snipped}

 

Howard, here's my problem with this whole line of reasoning.

 

What we're talking about *is* a lack of focusing accuracy on Tim's (and others) cameras. We're obviously not talking about an optical design issue--unless the design of the 35 1.4 lux ASPH has actually been changed between when mine was made and the present.

 

Mine is tack sharp at the focal point. Tim's isn't, but it can be if you compensate for the "shift..."

 

I'll say this again, though. On my system, by f4, there are *several feet of DOF* on the M8 at, oh, say 10 foot distances.

 

By 100 feet, the DOF is very large indeed, as I would expect. Everything is in focus for 20 feet or more around the focal point, or thereabouts.

 

So again, even though Tim's could focus--it would be off *by several feet* at f4 using the rangefinder patch in the middle.

 

That's astonishing to me. Really. For any modern lens ever made except maybe a Lensbaby!

 

I mean, do you guys shoot Canon or Nikon at all? Or a DMR? If I had a 35mm lens on a cropped digital body off by "several feet" at the center at f4 at 20 or 30 feet, well, that's a completely broken system.

 

Period.

 

No-one has designed lenses to be off with those kind of design parameters since, well, the mid-20th century maybe? Maybe even earlier!

 

So--if this is the lens--just the optical system--and not the lens / body interface--then there are only two significant differences between mine and Tim's.

 

One is the chrome vs black build. I refuse to believe that's the issue until someone could prove that.

 

The other is the updated mount / new mount. One doesn't change any tolerance or design at all; the other at least potentially does. And the other, at least, anecdotally fits the symptoms.

 

OR--the whole problem is the interface between perfectly good lenses and perfectly good M8s--both of which just happen to be mis-calibrated to give the worst possible effect when combined.

 

But that's it. Those are the choices. ASPH or not, focal field shift or not, I really don't have a magic lens, guys.

 

So the mount is suspect. The lens run is suspect. Heck--the particular lenses and the particular M8s are suspect, IMO.

 

But I shot it again all day yesterday and today, between 2.0 and 5.6... and I purposely aimed for the center every time. Virtually every shot is in focus, and ridiculously sharp--and at f4 there's a ton of leeway too.

 

So how many angels can dance on the head of the aspherical optical pin, it seems to me, will not solve the problem.

 

These 35s that won't focus properly(like Tim's)--or even nearly properly--at f4 at distance are as far away from the DMR, Canon, Nikon, Oly and other maker's wide camera combos as they are from my 35 Lux or Sergios.

 

Something big is wrong here--not something small. IOW, IMO, we're looking for a gross differentiator here, not a fine or subtle one (or at least if a subtle cause, it still must account for the gross effect!)

[JWW]

Advertisement (gone after registration)

After all this discussion, I'm tempted to test out my chrome 35mm ASPH Summilux (non-coded) on a tape measure. I've been happy with its performance on the M8 and was not really going to try to fix what ain't broke.

 

Jan

[rvaubel]

Jamie, et al

 

I just started to follow this thread so I don't know anything. I must say that although I have seen focus shift it has always been of mostly theoritical interest and ALWAYS covered by the increased depth of field of the smaller aperature. At most it would reveal itself in the receding domino shot. The center of best focus might shift(slightly) but the subject domino would still be in focus.

 

This focus thing shifting by feet at moderate distances seems crazy. I don't doubt that there is a problem but I doubt it has anything to do with the focus shifting with aperture change. Of course its easy for me to dogmatic about my disbelieve because I don't have renegade lens in my collection. Obviously something is going on and thats why this is a interesting thread.

 

Keep hammering away... there is an answer

 

Rex

[ho_co]

After all this discussion, I'm tempted to test out my chrome 35mm ASPH Summilux (non-coded) on a tape measure. I've been happy with its performance on the M8 and was not really going to try to fix what ain't broke.

Jan--

That's the best course of action.

 

The viewpoint of everyone with the problem is, "If you've got it, you know it."

 

As Rex says, the matter of focus shift has always been and should always be of merely theoretical interest.

 

--HC

 

Or even better--sell it to me!

[ho_co]

Jamie—

I hope you understand I’m actually on your side. I’m just trying any logical route I can find. But that’s not solving the problem, is it?

 

What we're talking about *is* a lack of focusing accuracy on Tim's (and others) cameras. We're obviously not talking about an optical design issue--unless the design of the 35 1.4 lux ASPH has actually been changed between when mine was made and the present.

I agree. And since the problem seems to appear with some but not all of the lenses, old, new, coded, not, green, black, pink, what have you, I agree that design change is not a likely explanation.

 

… So again, even though Tim's could focus--it would be off *by several feet* at f4 using the rangefinder patch in the middle.

I’m not doubting you, but is that so? Is Tim’s lens off that much? Although I’ve read the entire thread, different parts stick with me than do with you.

 

… If I had a 35mm lens on a cropped digital body off by "several feet" at the center at f4 at 20 or 30 feet, well, that's a completely broken system.

Again, I agree completely. But when I look at Tim’s Venice pictures, I think, ‘gee, the guy’s good, and so is his equipment.’ In those pictures, the problem isn’t evident; but nonetheless the problem did bring him to switch to a different 35mm lens.

 

… So--if this is the lens--just the optical system--and not the lens / body interface--then there are only two significant differences between mine and Tim's. One is the chrome vs black build. … The other is the updated mount / new mount. ...

OR--the whole problem is the interface between perfectly good lenses and perfectly good M8s--both of which just happen to be mis-calibrated to give the worst possible effect when combined.

Again, possible. But why is it so inconsistent? And why did Solms tell Tim that both his Summiluxes were to spec? Why didn’t they suggest to him that he ought to send in the body to have it checked? Why did Guy Mancuso say he ‘fixed’ his 35/1.4 (I think that was the lens) by tweaking the camera rangefinder so that it was just a little off in one direction at close distance and a little off in the other at infinity (IIRC)?

 

… So the mount is suspect. The lens run is suspect. Heck--the particular lenses and the particular M8s are suspect, IMO. …

Something big is wrong here--not something small. IOW, IMO, we're looking for a gross differentiator here, not a fine or subtle one (or at least if a subtle cause, it still must account for the gross effect!)

Jamie—

Your analysis is a perfect counter-argument to mine. Thanks very much for taking the time to respond in such detail.

 

Whatever the problem may be, awareness of it is also growing as more people see this thread.

 

Possible solution: You and Tim (and Sergio and Bill and whoever else and maybe a couple ham-fisted wrestlers) get together and go to Leica Germany with your lenses and hold the lens techs to the mat till you get an answer.

 

Possible solution: Someone who understands the issue and can reject technospeak presents the question to Leica for an explanation. (That person may also need wrestlers, preferably ones who don't care about technospeak but respond quickly to verbal commands.)

 

It’s fine for someone to say, “I’m sorry, sir, but that’s the way the cookie crumbles.” But it’s a lot harder for him to say that if he is face to face with both that person AND someone else whose cookie didn’t crumble that way.

 

I’ll add one more thing, and I hope it’s irrelevant. I used to work for Leitz/Leica as a sales representative in the US. At one time the company replaced a highly respected lens with another of supposedly improved performance. When I saw images from the new lens, I commented to a very tech-conscious and honest person that I didn’t think the new lens was as good as the previous one. He answered, “Oh, yes; optically, I don't care for the new one either. But we had so many problems with the mount of the old one that we needed a new design.”

 

I was unaware of any problems with the older version, and I still don’t know what he meant by ‘problems with the mount.’

 

As I said, I hope the example is irrelevant here. Whatever the problems were, that was pre-Internet.

 

Michael Hußmann summed it up well (though it could be clearer that the problem obtains with both Summicron and Summilux):

We all know about the undercorrected spherical aberration of the Summicron. If that was all there is to it, then we would need to get used to it and adjust the focus accordingly. Only there are those copies of this lens that deliver sharp images at all apertures, which leaves open the tantalizing possibility that there is some magical calibration of the lens – an ever so slight front-focus that preserves the sharpness at f1.4 and at the same time shifts the focus back into the DOF at smaller apertures.

--HC

[tashley]

Howard, here's my problem with this whole line of reasoning.

 

What we're talking about *is* a lack of focusing accuracy on Tim's (and others) cameras. We're obviously not talking about an optical design issue--unless the design of the 35 1.4 lux ASPH has actually been changed between when mine was made and the present.

 

Mine is tack sharp at the focal point. Tim's isn't, but it can be if you compensate for the "shift..."

 

I'll say this again, though. On my system, by f4, there are *several feet of DOF* on the M8 at, oh, say 10 foot distances.

 

By 100 feet, the DOF is very large indeed, as I would expect. Everything is in focus for 20 feet or more around the focal point, or thereabouts.

 

So again, even though Tim's could focus--it would be off *by several feet* at f4 using the rangefinder patch in the middle.

 

That's astonishing to me. Really. For any modern lens ever made except maybe a Lensbaby!

 

I mean, do you guys shoot Canon or Nikon at all? Or a DMR? If I had a 35mm lens on a cropped digital body off by "several feet" at the center at f4 at 20 or 30 feet, well, that's a completely broken system.

 

Period.

 

No-one has designed lenses to be off with those kind of design parameters since, well, the mid-20th century maybe? Maybe even earlier!

 

So--if this is the lens--just the optical system--and not the lens / body interface--then there are only two significant differences between mine and Tim's.

 

One is the chrome vs black build. I refuse to believe that's the issue until someone could prove that.

 

The other is the updated mount / new mount. One doesn't change any tolerance or design at all; the other at least potentially does. And the other, at least, anecdotally fits the symptoms.

 

OR--the whole problem is the interface between perfectly good lenses and perfectly good M8s--both of which just happen to be mis-calibrated to give the worst possible effect when combined.

 

But that's it. Those are the choices. ASPH or not, focal field shift or not, I really don't have a magic lens, guys.

 

So the mount is suspect. The lens run is suspect. Heck--the particular lenses and the particular M8s are suspect, IMO.

 

But I shot it again all day yesterday and today, between 2.0 and 5.6... and I purposely aimed for the center every time. Virtually every shot is in focus, and ridiculously sharp--and at f4 there's a ton of leeway too.

 

So how many angels can dance on the head of the aspherical optical pin, it seems to me, will not solve the problem.

 

These 35s that won't focus properly(like Tim's)--or even nearly properly--at f4 at distance are as far away from the DMR, Canon, Nikon, Oly and other maker's wide camera combos as they are from my 35 Lux or Sergios.

 

Something big is wrong here--not something small. IOW, IMO, we're looking for a gross differentiator here, not a fine or subtle one (or at least if a subtle cause, it still must account for the gross effect!)

 

Hi Jamie,

 

Just to add to the mix, I have a new 35 2.5 Skopar (the screw mount not the M mount) and whilst it is pleasantly on the soft side wide open, at F4 and in centre it beats the cron into a corner and makes it whimper, providing the sort of sharpness you get and I want! Further weight to the idea that my RF isn't off, on either body, if not yet proof!

 

I follow Howard's logic but can't conclude anything new from it. I would just question one of your statements though: to say that the focus is off by several feet at longer distances to subject is not strictly accurate: as we know, DOF proceeds on average about 1/3 in front and 2/3 behind: but we also know that even the magic ASPH lenses on the M8 often push this more back than forward as they focus shift. So in the birdbox test at 20 or 30 feet and F4, though the box is clearly OOF and the far background in focus, I know that the box is very near the front of the focus field - and of course, the edges, where the subject is flat, are in focus. So a small forward tug on the focus tab will generally move the centre into focus without pushing the edges out.

 

That's why I keep the cron: if there's time to focus bracket I know it will give amazing results in at least one frame! But if I need to be sure and fast, it's the CV or the Cron wide open...

 

Best

 

T

[mjh]

to say that the focus is off by several feet at longer distances to subject is not strictly accurate: as we know, DOF proceeds on average about 1/3 in front and 2/3 behind: but we also know that even the magic ASPH lenses on the M8 often push this more back than forward as they focus shift. So in the birdbox test at 20 or 30 feet and F4, though the box is clearly OOF and the far background in focus, I know that the box is very near the front of the focus field - and of course, the edges, where the subject is flat, are in focus. So a small forward tug on the focus tab will generally move the centre into focus without pushing the edges out.

The distribution of the depth of field in front of vs. behind the focused distance isn't really dependent on the lens, and it isn’t 1/3 in front and 2/3 behind either. Actually the distribution is more like 1/2 in front and 1/2 behind at short distances; it gradually changes with greater distances until at the hyperfocal distance the distribution is 1/2 the hyperfocal distance in front whereas the DOF behind extends all the way to the end of the universe. There is just one distance where the 1/3 to 2/3 rule holds, so it's really more an exception than the rule. Having said that, I agree with your main point that at greater distances, a healthy amount of front-focusing won't hurt background sharpness, but would suffice to push your subject safely back into the DOF.

[wlaidlaw]

Looking at my Fuerstentum.net DOF calculator widget, it shows 2/3 in front and 1/3 behind, not the other way round.

 

Wilson

[castelletta]

A simple question to all the readers. Just to better understand...

 

Is there anybody that owns a new black 35mm asph (lux or cron) CODED by Leica that focuses perfectly (and for sure) at all apertures?

 

Thank you

[tashley]

A simple question to all the readers. Just to better understand...

 

Is there anybody that owns a new black 35mm asph (lux or cron) CODED by Leica that focuses perfectly (and for sure) on all apertures?

 

Thank you

 

 

I do!

 

OK, that was a joke...

 

;-)

 

tim

[castelletta]

I do!

 

OK, that was a joke...

 

;-)

 

tim

 

I suspect that the list will end here.......but for now let's wait

[ChrisC]

....Is there anybody that owns a new black 35mm asph (lux or cron) CODED by Leica that focuses perfectly....

 

I was thinking the same question. In the mountain of agonising on this subject I have lost track of who owns what and when they purchased it. I have an uneasy feeling [no evidence] that the coincidence of the newly introduced M8 and some M lens focus issues is more than just being able to quickly see problems on monitor screens. I also cannot recollect a similar reported issue with a Zeiss, or a CV lens on an M8.

 

Gross cock-ups happen of course, but lens engineering has been Leica's proud forte and it seems highly unlikely that the new lens mount would be engineered to give focus mis-registration on an otherwise [previously] acclaimed lens design. I can't help suspecting that to coincide with the M8 era there has been a lens design tweak, or a lens construction change, or even both.

 

My reasoning is mostly ignorance based, but nevertheless could still be correct.

 

............Chris

[castelletta]

I can't help suspecting that to coincide with the M8 era there has been a lens design tweak, or a lens construction change, or even both.

 

............Chris

 

Let's make the situation clear and after that, if the problem will end up to be a construction change or fault (or both), I'll pretend a lens swap from Leica.

 

It's not conceivable (for me) that I spend 2.150€ for the cron or 3.000€ for the lux and have to compensate manually the focus for each aperture stop.

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[tashley]

I have just PM's Sergio to ask him which his lens is, since his was the one that passed the test most convincingly...

 

We should probably start a list of who has what, either way so in the next post (Title "Leica 35mm Lens Survey Response" I give my answer.

[tashley]

Mine is a Black 35mm Cron bought direct and new from Leica this year and already coded. It has focus shift that places the point of focus OOF between F2.8 and 5.6 or 8.

 

I had previously two 35 luxes which were identical in every respect to the above description except the OOF behaviour began at F2.

 

Tim