Very interesting answer from Leica on 35mm 1.4

[tashley]

Evening all!

Some may remember my traumas with 2 copies of the 35mm F1.4 in a row which exhibited significant and increasing backfocus as the aperture was stopped down. Both have been sent back. Today I received an extremely clear and friendly email from Solms. Here's a clip from it:


....quote....
You described that the focus moves back when stopping down the aperture.
This effect occurs due to so called spherical aberration.

It is an image imperfection that occurs due to the increased refraction that occurs when rays strike the lens near its edge, in comparison with those that strike nearer the center.
It´s not possible to remove this aberration completely in the optical design, especially at high speed lenses . Different aberrations need different parameters to correct them.
Unfortunately it happens when you improve one aberration (e.G. distortion), that an other aberration increases. At the end it´s a compromise between all kinds of aberrations, size, weight etc.
The 35mm f1.4 lens is a well balanced lens regarding all competing parameters.
But unfortunately it´s not totally free of any aberrations.
It needs to be considered, that stopping down has an effect on the focus plane and probably it needs some exercises in order to get the right feeling for the lens.
...end quote...

The email goes on to explain that, since the first lens I sent back (they didn't receive the second one yet) was within the spec (and not at the bottom of it either) I might want to have it sent back to me.

Now I have to trust these good and helpful folk but I have replied that if the 35mm Cron does not exhibit this I might have that instead. I have also observed that the PDF technical spec for the 35 1.4 which I downloaded from their website appears to contradict the email in that the very detailed DOF charts show a one metre point of focus remaining within focus throughout the aperture range. I also observed that it might be helpful to let potential purchasers know in advance of purchase that this design constraint exists and that they will need to 'get the right feeling for the lens'.

So, any thoughts from you clever bunch? Am I just being too picky? Is there a good technical reason for the apparent contradiction between the PDF and the email response? And why don't reviews of lenses mention this more often, if it is a design characteristic that affects focus.

Hmmm!

Tim

[sergiolov]

Hi Tim,
just to let you know, I tested accurately for focus my 35 cron asph, and it shows this behavior:
-full open it front focuses a little, the focus area is 2/3 in front and 1/3 back.
-stopping down, the center of the focus area keeps going back, and at 5,6 you end with the classic 1/3 front and 2/3 back

This behaviour seems pretty well balanced, giving you more focus area in front fully open, wich is very good for close portraiture.

Adding to the problem that you noticed on your lux asph is the fact that (see the MTF) the lens at 2,8 is sharper offcenter,the mtf curve at 40 lp/mm is nearly sinusoidal.

Did you have a good time in Viareggio?
Ciao.
Sergio

[hankg]

It seems that the arrival of a digital M is revealing issues with lenses that where not apparent to most film users. Digital sensors are perfectly flat as opposed to roll film which was only an wavy, bowed approximation of flat. Leica may have to rethink it's priorities when balancing the various trade offs in optical design as focus would seem to be the one all users will be very sensitive to. When 5 minutes after receiving a lens you can be reviewing it's focusing accuracy at 5,000% - tolerance among customers for focusing error is going to be very small.

The new Zeiss 1.5 Sonnar seems to have lost much of it's anticipated appeal when descriptions of focusing issues of exactly the sort you describe appeared in multiple online forums. The new 85/2 Sonnar (which is was delayed around the time complaints began appearing about the 50 Sonnar) will have a progressive cam and floating elements. It will also be an order of magnitude more expensive then the 50/1.5, hopefully with the additional technology it will focus accuratley across its' aperture and distance range.

It seems to me it doesn't matter what else you get right if you can't get your subject in focus. All the extra money paid for a small edge in resolving power at fast apertures is out the window with the slightest focusing error. I would think it's back to the drawing board on some traditional design priorities. The law of unintended consequences.

[tashley]

Originally Posted by sergiolov

Hi Tim,
just to let you know, I tested accurately for focus my 35 cron asph, and it shows this behavior:
-full open it front focuses a little, the focus area is 2/3 in front and 1/3 back.
-stopping down, the center of the focus area keeps going back, and at 5,6 you end with the classic 1/3 front and 2/3 back

This behaviour seems pretty well balanced, giving you more focus area in front fully open, wich is very good for close portraiture.

Adding to the problem that you noticed on your lux asph is the fact that (see the MTF) the lens at 2,8 is sharper offcenter,the mtf curve at 40 lp/mm is nearly sinusoidal.

Did you have a good time in Viareggio?
Ciao.
Sergio

Hi Sergio!
Had a great time in Viareggio, thanks, and got a lot of unsatisfactory but fun shots of a very rainy carenvale (posting soon!)

I understand the idea that these lenses progressively back focus BUT the point for me is that the actual exact plane of focus I choose (very carefully, with a 1.25 mag and tripod) is in focus at 1.4 and then drifts slowly out of focus til F11 so it seems that the only way to get correct focus in the range f2-f8 is to guess. That really can't be either acceptable or right. Call me old-fashioned but a lens is there to focus light onto a plane!

Best

Tim

[tashley]

Originally Posted by hankg

It seems that the arrival of a digital M is revealing issues with lenses that where not apparent to most film users. Digital sensors are perfectly flat as opposed to roll film which was only an wavy, bowed approximation of flat. Leica may have to rethink it's priorities when balancing the various trade offs in optical design as focus would seem to be the one all users will be very sensitive to. When 5 minutes after receiving a lens you can be reviewing it's focusing accuracy at 5,000% - tolerance among customers for focusing error is going to be very small.

The new Zeiss 1.5 Sonnar seems to have lost much of it's anticipated appeal when descriptions of focusing issues of exactly the sort you describe appeared in multiple online forums. The new 85/2 Sonnar (which is was delayed around the time complaints began appearing about the 50 Sonnar) will have a progressive cam and floating elements. It will also be an order of magnitude more expensive then the 50/1.5, hopefully with the additional technology it will focus accuratley across its' aperture and distance range.

It seems to me it doesn't matter what else you get right if you can't get your subject in focus. All the extra money paid for a small edge in resolving power at fast apertures is out the window with the slightest focusing error. I would think it's back to the drawing board on some traditional design priorities. The law of unintended consequences.

You're dead right - the email said that peoples' ability to instantly zoom to 100% revealed what the lens actually does. BUT the PDF still shows the plane of focus as 'in focus' when the f stop is reduced - and on my two samples, that is not what happens. It may be within tolerence but it doesn't focus!

Best

Tim

[d_brown]

Hello Tim,

Sorry to hear about your 35mm problem. I'm also having trouble with my 35mm lenses (all three of them), as they backfocus terribly and are unusable. The solution to your dilemma might be to send the lens to a reputable third-party repair service and have them shim the lens so that correct focus occurs at f4 or so. That would put it nearer to the center of the DOF at all distances. Based on your information, I am going to call my repairman and ask him to do the above when he shims mine. The second 35mm Lux has been at Leica for a month now and I can't even get Solms to acknowledge having it, let alone fix it. The third lens is a Cron, brand new, and I haven't decided what to do with it. I finally bought a Zeiss Biogon and it focuses right on the money at all stops and distances.

Hope you get your problem resolved soon!

Dale

[mjh]

Originally Posted by tashley

So, any thoughts from you clever bunch? Am I just being too picky? Is there a good technical reason for the apparent contradiction between the PDF and the email response? And why don't reviews of lenses mention this more often, if it is a design characteristic that affects focus.

Unfortunately, the focus shift occurring with lenses under- or overcorrected for spherical aberration is something we have to live with. A good explanation of the phenomenon is given here: Spherical aberration.

[jonoslack]

Hi Tim
Glad you had a good time - sorry about the rain - looking forward to the shots.

I tried this with all my lenses when I got the camera back from Solms. the focus point seemed to be set on the widest aperture, and ALL of them exhibited this behaviour, not too much, but noticeably when testing. It did seem to be a great deal worse at close focusing distances . . . I don't know how close you want to get!

I don't know the answer either - but it's educational to look at splendid photographs from the 60's and 70's and to notice that most 35mm shots leave something to be desired from the focusing point of view!

As Sergio and others have noted, we look at things much closer, and we're very much more aware of issues - one can only hope that it makes us better photographers!

[tashley]

Originally Posted by mjh

Unfortunately, the focus shift occurring with lenses under- or overcorrected for spherical aberration is something we have to live with. A good explanation of the phenomenon is given here: Spherical aberration.

That is a very interesting link and I thank you for it... very clear and detailed.

What it leaves unexplained is the Leica PDF for the 35 lux, which clearly shows the selected point of focus as remaining in focus as the F stop is reduced.

I can therefore conclude that either:

1) The marketing department has some say in the production of the DOF charts in the pre-sale information or...

2) There is something I have not understood - this is perfectly likely since I do not design lenses for a living, but I do test them, by using them or...

3) I have been sent two miscalibrated examples.

I may never know the answer but the questions interest me nonetheless!

Tim

[tashley]

Originally Posted by d_brown

Hello Tim,

Sorry to hear about your 35mm problem. I'm also having trouble with my 35mm lenses (all three of them), as they backfocus terribly and are unusable. The solution to your dilemma might be to send the lens to a reputable third-party repair service and have them shim the lens so that correct focus occurs at f4 or so. That would put it nearer to the center of the DOF at all distances. Based on your information, I am going to call my repairman and ask him to do the above when he shims mine. The second 35mm Lux has been at Leica for a month now and I can't even get Solms to acknowledge having it, let alone fix it. The third lens is a Cron, brand new, and I haven't decided what to do with it. I finally bought a Zeiss Biogon and it focuses right on the money at all stops and distances.

Hope you get your problem resolved soon!

Dale

Hi Dale,

I would describe my 35 lux as unusable also, in that the point on which a well calibrated rangefinder mechanism selects as focus is consistently far enough in front of the actual point of focus as to make most images significantly OOF between F2 and F8, at most distances from the plane of focus. I have been much luckier than you in that Solms has been quick and considerate, but I cannot help but feel that I have yet to get to the bottom of this issue. The job of a lens is to focus light. Mine have failed to do this and I would like to understand why.

Tim

[woodyspedden]

Originally Posted by tashley

You're dead right - the email said that peoples' ability to instantly zoom to 100% revealed what the lens actually does. BUT the PDF still shows the plane of focus as 'in focus' when the f stop is reduced - and on my two samples, that is not what happens. It may be within tolerence but it doesn't focus!

Best

Tim

Interesting though that to keep costs lower, Zeiss chose to use no aspherical elements in the new series of M lenses. The elements are all spherical and yet, at the same time, Zeiss advertises that they have gone to lengths to minimize spherical aberrations and particularly focus shifts with aperture.

Interesting

Woody

By the way, although not immune from the problem by any means I have not found focus shifts to render the lenses (I have both the 35 cron and Lux) "unusable" as several forum members have described. I also don't understand the relationship between the MTF curves and this phenomenon, at least not at this magnitude. Worth more study to be sure.

[tashley]

Originally Posted by jonoslack

Hi Tim
Glad you had a good time - sorry about the rain - looking forward to the shots.

I tried this with all my lenses when I got the camera back from Solms. the focus point seemed to be set on the widest aperture, and ALL of them exhibited this behaviour, not too much, but noticeably when testing. It did seem to be a great deal worse at close focusing distances . . . I don't know how close you want to get!

I don't know the answer either - but it's educational to look at splendid photographs from the 60's and 70's and to notice that most 35mm shots leave something to be desired from the focusing point of view!

As Sergio and others have noted, we look at things much closer, and we're very much more aware of issues - one can only hope that it makes us better photographers!

Hi Jono!

Rain is irritating but presents certain opportunities!

My lenses all do this a bit, but the 35 lux is off the scale. The 50 lux is close to perfect (Lord I love that lens!) and the 24 2.8 is good if not perfect. The 90 f2 has a bit of back focus too.

I hate to say it but I like my shots to be extremely sharp at the point I choose. I also like good bokeh where I want it. But if the only way to get sharp focus on a 35 lux is to bracket focus then it doesn't work for me, not at that price!

When a digi-slr focuses, it does so by contrast: so unless the lens is really out of whack the camera will compensate. On a RF you have to trust the lens design, calibration and the camera RF calibration to be exact or you get shoddy results.

I might be able to live with this since the results are so good when it all comes together, but I do not like having to guess at that which should not require guess work!

Hope all is well with you!

Tim

[jonoslack]

Originally Posted by tashley

What it leaves unexplained is the Leica PDF for the 35 lux, which clearly shows the selected point of focus as remaining in focus as the F stop is reduced.

You're quite right - on page 5 it clearly shows the closest focus point moving towards the camera at smaller apertures.

BUT

I think we've all accepted that the 'acceptable' depth of field for digital cameras is much smaller than for film - I haven't tried the calculation, but I wonder whether if you reduced the acceptable dof as a percentage, then it might change the figures so that it did move away.

[tashley]

Originally Posted by jonoslack

You're quite right - on page 5 it clearly shows the closest focus point moving towards the camera at smaller apertures.

BUT

I think we've all accepted that the 'acceptable' depth of field for digital cameras is much smaller than for film - I haven't tried the calculation, but I wonder whether if you reduced the acceptable dof as a percentage, then it might change the figures so that it did move away.

But that PDF is their pre-sales bumpf! It makes no inference that things might be different on the M8 than on a film M, though we all know this to be true. And in any event if the DOF moves both forwards and back as aperture tightens then there's no way that on any M it should randomly move only forwards. There's more to this than meets the eye. Trust me. And I'm not even a doctor.

t

[russell]

No problems with the 50mm pre-asp Lux. I also noticed this on my R-D1. The only lens I could rely on to focus consistently. A friend of mine also has this lens and he noticed the same things. It's a keeper...

[jonoslack]

Originally Posted by tashley

But that PDF is their pre-sales bumpf! It makes no inference that things might be different on the M8 than on a film M, though we all know this to be true. And in any event if the DOF moves both forwards and back as aperture tightens then there's no way that on any M it should randomly move only forwards. There's more to this than meets the eye. Trust me. And I'm not even a doctor.

t

Trust me (even though I have a red AGA). The depth of field isn't moving back very much in those figures - you don't have to reduce the acceptable dof much as a percentage for it to move forwards instead.
If you do a calculation at each aperture for each distance, you will find the middle of the dof range moving away.

[eronald]

I believe the reason why the 35 lux has this nice look is precisely the uncorrected spherical aberration ...

Edmund

BTW, I have a new blog entry up at Edmund Ronald's Color Management News

[charlesphoto99]

This is an interesting thread as I'm looking to possibly unload a lens in order to finance the purchase of an M8. The 35 summ asph was the one I was thinking about though it's taken some real nice pics with M7 esp wide open (I also have 21 pre asph, 28 summ, 24, and 50 pre asph summ, 90 apo, and 135 elmar) esp at night.

I've had real back focus issues with my Nikon d200 - even sent it back for adjustment though not sure if it made that much of a difference. I use older manual focus lenses, and the problem seems to get worse the smaller the aperture. Was shooting some portraits with a ring flash and the 28 f2 (which I believe has aspherical element and is considered one of Nikon's best) and couldn't get the flash power down below f11-16. Well, I thought, no problem with focus there because of depth of field. Wrong! All backfocused. Fortunately was able to clean up with sharpen brush and they were only getting printed 1/4page anyway. Yeah, it's a real problem with digital and I'm sorry to hear M8 suffering as well.

Anyone have issues on other wide asph lenses? Who knows, maybe pre asph lenses will suddenly become sought after...

One saving grace of course is M8's 1/8000 shutter speed. At least easier to keep things down below f4.