Anyone got an 18mm F3.8 Super Elmar?

[Marty]

Advertisement (gone after registration)

I've taken this lens out again today but it was foggy, so no distance shots…

 

Another lens, my Cron 35 ASPH, has a similar "feature" at the near 0.7m end. You can rotate it a few degrees beyond that, and using a focus chart I could see that it does in fact focus closer than 0.7m. When I focus on an object exactly 0.7m away the scale is at 0.7.

[UliWer]

This is the first clearly confirmed report of a copy of this lens that performs as it should! You are one lucky dude!

 

Want to sell it?

:D:D

 

Thanks!

 

As I have looked at some pictures taken with the lens from objects far, far away and noticed how sharp they were and how good the resolution was I won't sell it (and you never can have too many Elmars....

 

And I am still confident that the red shift on the edges will be dealt with by Leica soon.

[tashley]

Tim, I've tried my Super-Elmar on my M9 and it's behaving as expected - you hit the end stop just as the two images on a distant object coincide. Still get red edges though...

 

Thanks for testing that Mark - it does seem that there's a fair mixture of them out there. My dealer has a new batch arriving Monday and I'll get to try them on Tuesday. The last three they had, and the one they had at Bruton Place, were all 'past infinity' copies..

 

Best

 

Tim

[ho_co]

... email from Leica saying that this is not a normal effect but has no influence on picture quality....

Sorry, OT insofar as I don't have the lens in question, but that's an absurd answer. Infinity stop, optical infinity and rangefinder must line up.

 

Optically, infinity is generally considered to be 1000 x focal length, or 18 m in this case. But your lens doesn't reach infinity either with the rangefinder or optically until focused to its infinity stop. At least the latter point is good: the lens' optical infinity and mechanical infinity match.

 

But since the rangefinder doesn't agree, the focusing cam is off and needs to be adjusted.

 

Suggestion: Try a through-focus test wide open at a distance of about 3 feet: From your description, the lens will likely be in focus on the sensor when focused (slightly?) past where the rangefinder says it's in focus. (In my experience in cases like this, the focus ring will have to be rotated just as far to get good focus close up as it has to be rotated from the rangefinder's indication of infinity to the lens's infinity stop.)

 

 

Tim, I'm glad you're around. Leica should eventually figure out that when you contact them, they should consider your remarks at the highest level. What did we end up with on your 35/1.4 thread, something like 300 posts? Even then, Leica didn't understand, and you had to take your (what, third?) sample to Solms to demonstrate the problem personally.

[ho_co]

And I am still confident that the red shift on the edges will be dealt with by Leica soon.

 

Uli--Hadn't you heard about the graduated cyan filters Leica is planning to offer?

[tashley]

Sorry, OT insofar as I don't have the lens in question, but that's an absurd answer. Infinity stop, optical infinity and rangefinder must line up.

 

Optically, infinity is generally considered to be 1000 x focal length, or 18 m in this case. But your lens doesn't reach infinity either with the rangefinder or optically until focused to its infinity stop. At least the latter point is good: the lens' optical infinity and mechanical infinity match.

 

But since the rangefinder doesn't agree, the focusing cam is off and needs to be adjusted.

 

Suggestion: Try a through-focus test wide open at a distance of about 3 feet: From your description, the lens will likely be in focus on the sensor when focused (slightly?) past where the rangefinder says it's in focus. (In my experience in cases like this, the focus ring will have to be rotated just as far to get good focus close up as it has to be rotated from the rangefinder's indication of infinity to the lens's infinity stop.)

 

 

Tim, I'm glad you're around. Leica should eventually figure out that when you contact them, they should consider your remarks at the highest level. What did we end up with on your 35/1.4 thread, something like 300 posts? Even then, Leica didn't understand, and you had to take your (what, third?) sample to Solms to demonstrate the problem personally.

 

Howard, I'm blushing!

 

I can't test the lens any more since I have declined to have it either returned to me or sent to Solms. It sounds churlish but my feeling is, if they don't check 'em at the factory gates, I ain't gonna pay for them! So I have another candidate arriving on Monday, though having tried four, so far, personally, and found they all have this issue on any body you try them on, I am not wildly hopeful!

 

I must say, and this said sadly, that the chances of my buying an S2, which were standing at around 90%, are now down to less than twenty. I observed in another thread that I am getting bored with ordering (and paying for) Porsches and taking delivery of Maseratis...

[marknorton]

Advertisement (gone after registration)

Not churlish at all. Leica have to shed this cottage industry approach and do it right every time. It's interesting that Leica lens quality issues are all to do with the mount alignment, seemingly never with the quality of the optical cell or lens barrel. Why can't they apply the same care to assembling the lens mount that they do to grinding, polishing, centring and coating the lens elements?

 

To a first approximation, the distance the lens barrel moves out from its infinity position to closest focus ("c", in mm) is ( f * f) / ( c - f ) so that for a 50mm lens and closest focus of 700mm, the barrel moves 3.85mm. It's no accident that this is similar to the distance the rangefinder roller moves in an out; you're nearly there if you just sense the position of the lens barrel and fine tuning can be accomplished by hand-grinding the focussing cam. That's how it was when the M rangefinder was first designed.

 

Fast forward to the SE and the lens barrel moves just 0.33mm over the same focussing range and you can't just sense the position of the lens barrel any more. There isn't enough movement to "drive" the rangefinder roller. Older lenses, like the 21mm Elmarit, use an expensive second helicoid - the first to move the lens barrel in and out (in the SE, about 1.3mm pitch over 90 degrees of rotation) and the second to generate the 4 - 5 mm movement for the focussing roller. Not so the SE - it uses a steeply profiled cam which rotates directly with the focussing ring to drive the rangefinder roller.

 

All well and good until you realise that with a steeply profiled cam, the point where the rangefinder roller makes contact is going to have a greater impact on what the rangefinder senses compared to a shallow cam. The fact that rangefinder adjustment changes the point at which the roller makes contact with the cam introduces some cross dependence between camera and lens where, ideally, there would be none. It's good engineering practice in any mechanical or electronic design to minimise, preferably eliminate, interactions between adjustments.

 

It may be the adjustment tolerances make this a wash, especially with the huge depth of field able to mask some nasties. However, if the lens barrel moves too close to the focal plane when set to infinity focus so that you are focussing beyond infinity, IQ will rapidly deterioriate; given the depth of field available, it's better with these lenses to err on the side of caution so that they approach, reach but never overshoot the infinity position.

 

Looking here at the 9 widest angle lenses (WATE through to 28mm Summicron and Elmarit), only two of them use this steeply angled focussing cam - the Super-Elmar and the 24mm Elmar.

 

If Tim still had his lens, I'd recommend focuss bracketing from 10m to infinity without using the rangefinder on a distance ojbject to find best focus and determine if the lens itself is focussing beyond infinity. If it isn't, and the rangefinder patch is over-shooting, I'm tempted to wonder whether the M9's newly aligned rangefinder is actually the source of the problem.

[tashley]

Thank you Mark. I don't know much about engineering so I can't dissect the rat but I know when I smell one and there is clearly something in common between all SEs that make them prone to this. I had found it from a big enough sample and you have now clearly explained what it is and I am very grateful!

 

Before the lens got sent back, I did sort of do what you suggested. Using a radio mast about a mile and a half away, I focussed on it using RF alignment and using barrel infinity. The shot using RF alignment was as sharp as anyone could wish for whereas the shot using barrel infinity was noticeably less sharp. I would estimate from memory that the RF patch, at barrel infinity, was about 15% of its own width overshot.

 

The loss of sharpness was clearly visibly, if subtly so, at a print size of 20 x 30 inches and had I chosen to upres in order to print larger, it would have been more noticeable still.

 

Thanks again Mark. Mystery solved!

 

Best

 

Tim

[ho_co]

@ Mark--Thanks for doing the math!

 

You've just answered a related question I had regarding the M8/M9:

 

Leica CL lenses could be used with 75% of previous M bodies, because on them the rangefinder roller was centered. But since the 'standard' Leica lenses all had flat cams, they would work whether the roller was centered or not. Before the CL, location of the tracking roller wasn't important (and wasn't a part of design tolerance) since all lenses had flat cams.

 

To the current point: It looks as if the steep focusing cams in the new lenses mean that the M8/M9 require that the focus tracking roller be dead center. Maybe there's a hair's width of variation from camera to camera in the location of the camera's tracking roller that Leica needs to consider? Or conversely: What if I put Tim's Super-Elmar on my M5, on which the tracking roller is noticeably off-center?

 

 

@ Tim--I messed up again, thinking I understood the situation of your 18/3.8, but equating the difficulty you were having with a different problem I had had.

 

From your most recent post, it sounds to me as if Leica may be right that the discrepancy will indeed make no photographic difference. But if so, it means that everything Leica has trained us about the rangefinder camera has been tossed out.

 

After reading Mark's post, it seems they'll either have to learn to get it right the first time, or to pay later for "fixing it" to the standards they've taught us to expect.

 

One of the ways they've used to reduce the price of the new lenses seems to have turned around and bitten them.

[ho_co]

Hmmm. Maybe Leica needs to learn from Panasonic's manuals.

 

When you take pictures at close range, the resolution of the periphery of the picture may decrease slightly. This is not a malfunction.

--Panasonic LX3 instructions, p 170

Extremely high precision technology is employed to produce the LCD monitor screen. However, there may be some dark or bright spots (red, blue or green) on the screen. This is not a malfunction.

--Panasonic LX3 instructions, p 180

 

 

Leica could say:

In certain cases, rangefinder infinity may not line up with a lens' infinity stop. This is normally not a malfunction. Just use the rangefinder to focus the lens as usual.

Or:

In certain cases when using discreet and continuous modes together, the camera may hang. This is not a malfunction. Simply remove and replace the battery to resume normal usage.

 

[tashley]

From your most recent post, it sounds to me as if Leica may be right that the discrepancy will indeed make no photographic difference. But if so, it means that everything Leica has trained us about the rangefinder camera has been tossed out.

 

 

Thanks for all the good giggles in these posts Howard! Just thought I'd correct one thing though: my posts are intended to show that the discrepancy does in fact result in a real difference, and my reading of Mark's post is that this is what would be expected where a super elmar focusses past infinity!

 

Best

 

t

[ho_co]

First, Tim, I’m always happy to know that someone sees the humor in what I intend to be humorous.

 

... my posts are intended to show that the discrepancy does in fact result in a real difference, and my reading of Mark's post is that this is what would be expected where a super elmar focusses past infinity!

 

Second, in my attempts at both wit and concision, I failed to make clear in my post 29 that I agree with both of those contentions.

 

 

More directly put:

1) There is certainly something wrong with the lenses that behave as you describe.

 

2) However, I think that the problem is mechanical only, and not optical; if so, the Leica CS rep is correct that it will not manifest itself photographically under normal usage. (A reversal of my contention in post 24, as first recanted in post 29.)

 

3) I accept Mark’s calculations, but see below.

 

 

Summary:

I think the focusing cam of the 18 Super Elmars that behave like yours is properly set and that through-focus tests such as Mark (post 27) and I (post 24) have recommended would show no focus problem.

 

Clearly, the infinity stop of the 18 Super-Elmars that behave like yours is improperly set, should never have passed final Quality Control and must be adjusted.

 

 

Reasoning:

A) Your lens is sharp and agrees with your rangefinder when the rangefinder shows proper alignment at infinity. That is, “rangefinder infinity” agrees with “optical infinity.”

 

The location of your lens mount’s infinity stop doesn’t match “optical infinity.” You can turn your lens beyond the point at which the rangefinder indicates infinity focus, and when you do so, the image quality deteriorates because you have “focused past infinity” to reach the mount’s infinity stop.

 

B1) The situation described in B should not obtain; however,

 

B2) The situation described in B will not influence your images so long as you either focus with the rangefinder or zone focus;

 

B3) The situation described in B becomes photographically apparent only if you crank the lens to its infinity stop, and there is never a reason to set a lens to infinity since that discards at a minimum half its depth of field.

 

 

It's a minor mechanical adjustment, but a major oversight for Leica QC.

 

You in Solms and Carsten in Berlin (and possibly others as well) had to demonstrate the focus problems of the 35/1.4 directly to the people responsible, and those problems were then corrected to the degree that they are no longer showing up in the forum.

 

The same will happen here I'm sure.

[markowich]

First, Tim, I’m always happy to know that someone sees the humor in what I intend to be humorous.

 

 

 

Second, in my attempts at both wit and concision, I failed to make clear in my post 29 that I agree with both of those contentions.

 

 

More directly put:

1) There is certainly something wrong with the lenses that behave as you describe.

 

2) However, I think that the problem is mechanical only, and not optical; if so, the Leica CS rep is correct that it will not manifest itself photographically under normal usage. (A reversal of my contention in post 24, as first recanted in post 29.)

 

3) I accept Mark’s calculations, but see below.

 

 

Summary:

I think the focusing cam of the 18 Super Elmars that behave like yours is properly set and that through-focus tests such as Mark (post 27) and I (post 24) have recommended would show no focus problem.

 

Clearly, the infinity stop of the 18 Super-Elmars that behave like yours is improperly set, should never have passed final Quality Control and must be adjusted.

 

 

Reasoning:

A) Your lens is sharp and agrees with your rangefinder when the rangefinder shows proper alignment at infinity. That is, “rangefinder infinity” agrees with “optical infinity.”

 

The location of your lens mount’s infinity stop doesn’t match “optical infinity.” You can turn your lens beyond the point at which the rangefinder indicates infinity focus, and when you do so, the image quality deteriorates because you have “focused past infinity” to reach the mount’s infinity stop.

 

B1) The situation described in B should not obtain; however,

 

B2) The situation described in B will not influence your images so long as you either focus with the rangefinder or zone focus;

 

B3) The situation described in B becomes photographically apparent only if you crank the lens to its infinity stop, and there is never a reason to set a lens to infinity since that discards at a minimum half its depth of field.

 

 

It's a minor mechanical adjustment, but a major oversight for Leica QC.

 

You in Solms and Carsten in Berlin (and possibly others as well) had to demonstrate the focus problems of the 35/1.4 directly to the people responsible, and those problems were then corrected to the degree that they are no longer showing up in the forum.

 

The same will happen here I'm sure.

 

hmmm, my 18mm super elmar neither shows red corner on the m9 nor it has the 'past infinity' issue. instead i have got a bad 24mm lux (finally decided it was bad). CA all over, up to f11. back to maker.

peter

[jonoslack]

hmmm, my 18mm super elmar neither shows red corner on the m9 nor it has the 'past infinity' issue. instead i have got a bad 24mm lux (finally decided it was bad). CA all over, up to f11. back to maker.

peter

 

Should have got the WATE Peter

[markowich]

Should have got the WATE Peter

 

or the 24mm elmar....

p

[tashley]

I tested another two 18mm super-elamrs today form a fresh batch. All the same, focussing waaay past infinity. Grrrr. What to do???

[carstenw]

You decide But do it quickly, before I try to buy one. IMO, if it is sharp everywhere, and you can focus it accurately even at infinity, there is no issue. Just always focus it. It would be nicer if the infinity bump was closer to the intended spot, agreed, but we can only deal with what is in front of us. Perhaps a roundtrip to DAG is the way to go.

[ho_co]

Tim--"focuses waaay past infinity." Understood.

 

Specific, different question:

 

A) When you focus with the rangefinder on an object at infinity, does the infinity mark of the lens line up the focus mark of the lens?

 

Or, instead, does the lens barrel's infinity mark match up with the lens' infinity stop?

 

That is: When you pick up the lens in the shop, without putting it on the camera, do you have to rotate the focusing ring past the engraved infinity mark to reach the lens' infinity stop?

 

 

Those are two different possibilities and represent two separate but easy problems for Leica to fix.

 

The lenses shouldn't have left Solms that way, but if you use the rangefinder, the discrepancy will make no difference in either case; and in only one of the two cases will accurate focusing by scale be precluded.

[Guest mc_k]

...there is never a reason to set a lens to infinity since that discards at a minimum half its depth of field...

 

Sorry, am not quite following this...how should I focus, for example, the moon?

Thanks

[Advertisement]

Hallo,

Look here: Leica M

[ho_co]

Thanks, k, you're right. Never trust a guy on a soapbox.

 

For astronomical shots, depth of field is meaningless, and focusing closer than infinity will cause a loss of sharpness where you want it.

 

But I think you're generally not going to shoot the moon with a wide-angle.

 

If you're shooting a landscape with moon, you may want to focus for the moon and let the depth-of-field work for you in the landscape. Or you may want to put the moon inside the far depth of field but focus nearer than infinity to get a sharper foreground.

 

You're right to call me on my statement, because it was a gross exaggeration.

 

I should have said that for most work--unless one has a specific reason not to worry about near depth of field, in other words--one won't be focusing at infinity.

 

I think that generally with an 18 mm one will be more concerned with composing in depth than in focusing for infinity--but that's just my view, and I'm glad to be corrected.

 

 

On the topic of depth-of-field, if you haven't seen them, Gary Ferguson has two brief but interesting articles at LuLa on the fact that with digital, we may need to rethink what we thought we knew about depth of field, starting at Digital Focusing Part One. (There are some folks on this forum who disagree with his findings, by the way.)