Very interesting answer from Leica on 35mm 1.4

[Jamie Roberts]

Advertisement (gone after registration)

Rubén,

 

Yes, I think I'm very lucky on the 75; I had arranged to buy it back in January and the seller--a spectacularly nice guy--held it for me till now, which was very kind of him!

 

I know these shots are a bit soft, too--but not for any focus reason. At 1/16s at ISO2500, I wouldn't expect any different

 

I've shot this same thing with a dSLR and an IS lens, and it's nowhere near as sharp.

[wizard]

When a digi-slr focuses, it does so by contrast: so unless the lens is really out of whack the camera will compensate.

 

Tim, I tend to disagree. While it is true that a digi-slr focuses by contrast, it does so at full aperture. When a shot is taken using e.g. f8, the lens will stop down to that preselected aperture shortly before the shutter is actually operated, and if the lens exhibits a focus shift, that won't be compensated, as the camera has already decided on the focus position before the stopping down process and will not refocus to compensate for focus shift after stopping down has occured.

 

Cheers,

 

Andy

 

Edit: Lord, I did not notice how long this thread was. I have only now realized that Michael had already pointed out what I said before, but it's good to know my understanding seems to be correct

[Advertisement]

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

[Jamie Roberts]

@ Andreas--to say nothing of the fact most dSLRs will NOT autofocus in the light conditions I use 1.4 at

 

@ Tim--the legends of backfocussing <insert AF camera maker here> are legendary. PM me if you want a few thousand post examples , but as an example, all my AF lenses--and the camera bodies--went to the manufacturer to make precisely the compromises I'm talking about here.

 

At least with the M8 I can do it myself.

[tashley]

Steve,

it is a failing of mine, which I freely admit to, that I have short patience with fools. Also, my fuse is not getting any longer with my advancing years.

 

"But behind my back I hear

Time's wingéd chariot drawing near ..."

 

However, every vet knows the syndrome well. Dogs or cats without sufficient external stimulation do start scratching themselves obsessively, until they actually lacerate themselves. Horses bite the manger. One common solution, putting the poor beast away, is not applicable here. Thus my advise to go out taking some pictures. People who cannot find any interesting pictures should lay off photography and go into philately or the collecting of beer bottle tops.

 

A more worthwhile subject might be a discussion of the concept of depth of field. I think I saw somebody using the term 'circle of confusion'. A different thread?

 

The benign wise old man on the top of the mountain ;—)

 

 

Lars, we're not dolts and I for one at least can quote the poetry of John Donne correctly (I'd check yours if I were you) whilst simultaneously juggling lens caps and doing calculus in my head.

 

This circle of confusion stuff is hardly the unique province of the 'Wise Old Man' - more like DOF 101... and most of us here are fully cognisant of it when we place our posts in this thread.

 

Tim

[tashley]

Tim, I tend to disagree. While it is true that a digi-slr focuses by contrast, it does so at full aperture. When a shot is taken using e.g. f8, the lens will stop down to that preselected aperture shortly before the shutter is actually operated, and if the lens exhibits a focus shift, that won't be compensated, as the camera has already decided on the focus position before the stopping down process and will not refocus to compensate for focus shift after stopping down has occured.

 

Cheers,

 

Andy

 

Edit: Lord, I did not notice how long this thread was. I have only now realized that Michael had already pointed out what I said before, but it's good to know my understanding seems to be correct

 

 

And you were both right and I was being sloppy

[tashley]

@ Tim and Eric...

 

Eric first.

 

With the tweak, yes, you will get a slight amount of front focus. Actually, it's bang on, since the focus point is always in focus.

 

What you have right now is a small amount of actual back focus--since you can't focus on what you want with a 35mm lens, for heaven's sake!

 

Think about this. You're saying a lens which should be a no-brainer to focus is that way by design. Very odd.

 

FWIW, this is how my adjusment has affected other Leica lenses (I'll let you know about my CV 28 1.9 when it gets here next week):

 

24mm 2.8 ASPH--perfect focus

28mm 2.8 ASPH--perfect focus

35mm 1.4--perfect focus, with perceptible field shift to f5.6

50mm 1.4--perfect focus, with minimal, but perceptible, field shift to 2.8

75mm 1.4--1.5mm front focus at f 1.4 close up (pace Lars--we know); completley gone by f2.

 

This to me is a perfect compromise; I guess I didn't make that very clear I only have one M8, and the front focus on the 75 1.4 is actually within the margin of focus error with the small rangefinder.

 

AT NORMAL DISTANCES, you don't even see this ff effect, because distance increase effective focus.

 

@ Tim, I actually think it is the design of the 75--a very old but optically interesting one--that makes for the ever-so-slight front focus.

 

My 85 R lux on the DMR does this too almost exactly, and it used to drive me a little batty till I realized it.

 

They're very similar lenses, as others have pointed out.

 

So--again. You have nothing to lose by attempting the adjustment. I really don't think your 90 cron is going to be affected at all, since the only effect I've ever seen is a neglible one at f1.4; nothing at f2.

 

Hope this helps

 

Enough already! I'll do it!

 

Remind me again, should I turn the allen key clockwise or anticlockwise? I don't want to get it wrong and then scratch until I lacerate myself...

Tim

[wlaidlaw]

Advertisement (gone after registration)

For those who want to play around with DOF, might I recommend a little programme called DOF-Master. I think it is available for various systems. I have mine on my Palm Treo 650 smart phone, so as to have with me "in the field". I used it a lot when I was having a go at large format a couple of years ago. It has all sorts of little goodies in it. You can alter COC as you see fit and it will show theoretical focus in front and behind various optimal focus distances. It also calculates hyper-focal distances at your required aperture, COC, optimal focus distance, etc. It is also available as a circular slide rule here:

Hyperfocal Distance and Depth of Field Calculator - DOFMaster

 

Wilson

[wlaidlaw]

Also a mac dashboard DOF widget here; DOFC - Depth of Field Calculator - Dashboard - Calculate & Convert

 

Wilson

[Jamie Roberts]

Enough already! I'll do it!

 

Remind me again, should I turn the allen key clockwise or anticlockwise? I don't want to get it wrong and then scratch until I lacerate myself...

Tim

 

Tim--Pascal even posted pictures (I love this forum):

 

http://www.leica-camera-user.com/digital-forum/13933-new-backfocus-thread-solution.html?highlight=backfocus#post146014

 

For backfocus, I believe it's clockwise. We're talking a very small tweak here; not even a 10th of a turn, more like a nudge.

 

Having said that, allow about an hour. I also took mine way out at one point just to bring it back to the right side of the focal field. No harm done

 

Read the thread; I did close up adjust, then infinity adjust, etc...

 

On the 35 with your test you are aiming to have point of focus just near the back of the field.

 

Once you have that, check infinity, it should be fine, but may need a nudge.

 

Then I'd check the 50 and 90 wide open. I think you'll be surprised.

 

Your ruler test setup is, I think, perfect for this.

 

BTW--since you're dealing with your own focussing ability too, I didn't have to use a tripod to do this. Just focus, adjust, focus, adjust until correct.

 

Remember too not to go right to the extreme edge of close focus range, because, of course, anything closer will be OOF:)

 

Gosh. It takes longer to describe than to do!

 

Let us know how it goes, Tim. If you don't want to do it, I'm sure there are third parties who would.

 

I swear if you were across the pond we could do this over coffee (and switch to beer once it worked )

[tashley]

Tim--Pascal even posted pictures (I love this forum):

 

http://www.leica-camera-user.com/digital-forum/13933-new-backfocus-thread-solution.html?highlight=backfocus#post146014

 

For backfocus, I believe it's clockwise. We're talking a very small tweak here; not even a 10th of a turn, more like a nudge.

 

Having said that, allow about an hour. I also took mine way out at one point just to bring it back to the right side of the focal field. No harm done

 

Read the thread; I did close up adjust, then infinity adjust, etc...

 

On the 35 with your test you are aiming to have point of focus just near the back of the field.

 

Once you have that, check infinity, it should be fine, but may need a nudge.

 

Then I'd check the 50 and 90 wide open. I think you'll be surprised.

 

Your ruler test setup is, I think, perfect for this.

 

BTW--since you're dealing with your own focussing ability too, I didn't have to use a tripod to do this. Just focus, adjust, focus, adjust until correct.

 

Remember too not to go right to the extreme edge of close focus range, because, of course, anything closer will be OOF:)

 

Gosh. It takes longer to describe than to do!

 

Let us know how it goes, Tim. If you don't want to do it, I'm sure there are third parties who would.

 

I swear if you were across the pond we could do this over coffee (and switch to beer once it worked )

 

Thanks Jamie - you can now ignore the PM I just sent you asking for exactly this info! Talk about a pre-emptive strike...

 

I think I'll need a Scotch before, a coffee during and a beer after. There, I'm actually looking forward to it now!

 

:-)

 

Tim

[carstenw]

Even if it doesn't end up working, and there really was something different about your copies, I do think that this is a worthwhile and educational exercise.

[egibaud]

@ Tim and Eric...

 

Eric first.

 

With the tweak, yes, you will get a slight amount of front focus. Actually, it's bang on, since the focus point is always in focus.

 

What you have right now is a small amount of actual back focus--since you can't focus on what you want with a 35mm lens, for heaven's sake!

 

Think about this. You're saying a lens which should be a no-brainer to focus is that way by design. Very odd.

 

FWIW, this is how my adjusment has affected other Leica lenses (I'll let you know about my CV 28 1.9 when it gets here next week):

 

24mm 2.8 ASPH--perfect focus

28mm 2.8 ASPH--perfect focus

35mm 1.4--perfect focus, with perceptible field shift to f5.6

50mm 1.4--perfect focus, with minimal, but perceptible, field shift to 2.8

75mm 1.4--1.5mm front focus at f 1.4 close up (pace Lars--we know); completley gone by f2.

 

This to me is a perfect compromise; I guess I didn't make that very clear I only have one M8, and the front focus on the 75 1.4 is actually within the margin of focus error with the small rangefinder.

 

AT NORMAL DISTANCES, you don't even see this ff effect, because distance increase effective focus.

 

I am very happy to read this. My intention is to use the Lux35 about 70% of the time and the rest of the pics would be a 75cron and a CV15mm (but I understand this one does no need focusing the way we mean it).

 

It sounds quite adventurous to do it myself, but I am not sure that sending it to Solms and wait for weeks will get the exact tweaking I need.

 

Last piece of advice to me?

 

Tim let me know what result you got. I'll get an alen key on Monday.

 

Thank you again.

 

Eric

[tashley]

Firstly: it is easy, and easy to undo. Tips:

 

1) with a fine black marker pen, make a mark at the top of the wheel within which you find that allen nut so that if you want to regain your starting point you need only rotate the assembly back to the 12 o'clock mark you have made for yourself. Be aware that this might alert later professional personnel to your activities and have some impact on your warranty!

2) Shoot finest JPEG not RAW: most RAW developers make it less than instantaneous switching images for comparisons. It's much faster in RAW and you will likely shoot a lot of files

3) Take a lot of 'before' shots with a ruler at all f stops and your main lenses. Annotate them carefully.

4) As you peer into the body of the camera, with it's top plate facing down so that you can access the allen adjuster, an ANTI-clockwise turn brings focus nearer ( do this first if you think your RF may be backfocussing) and vice versa.

5) Turn by TINY amounts.

 

My results were mixed:

 

I was able to adjust the RF so that my 'Cron behaved as I would like. Namely, at F1.4 my chosen point of focus was right at the back of the very narrow DOF, and as I stopped down that chosen point remained *just about* in focus as the lens backfocussed. Even at f4, the previous worst, it was acceptable. Then as the F stop tightened further, DOF increased and everything sharpened up. This was a radical improvement.

 

BUT it ruined focus on my other lenses and threw infinity out by a lot. Even a tower 6 miles away had twins in the RF when the lens was set to infinity.

 

So I then decided on the following strategy:

 

I put the 50 lux (my best and main lens) on the camera. I then dialled back (i.e. turned clockwise) the compensation I had so far achieved until at f 1.4 and approx 80cm distance to subject, my chosen subject was in focus but at the back of the zone of focus. This is tough. At 1.4 this lens has about a 1cm DOF. But after fiddling I achieved it.

 

The result, after much testing, is as follows:

 

At this distance to subject (about 80cm) the 50 lux has very very slightly sharper focus at f1.4 than it did before, and the chosen point of focus remains in focus as I stop down (it always did before). DOF proceeds 1/3 before and 2/3rds behind the subject as I stop down.

 

Setting infinity on the lens means exactly coincident images on very distant objects in the RF as it did before I started with the allen key - but on my 90mm macro, resolution at infinity is now noticeably better.

 

The 35 Cron still doesn't focus properly! It is better, a little at close range, somewhat at about 1.5 -2 metres and quite a lot better at longer distances. Yesterday I focussed on a bird box at twenty five feet way with a garage a futher twenty five feet behind it. The box was OOF and the garage crisp (at F1.4). Today, the box is less (but still somewhat) OOF and the garage is also OOF. This is progress.

 

Should Eric (or anyone else with an apparently accurate RF but a backfocussing 35 cron or lux) try this?

 

1) It's a useful experience and it appears to have very slightly improved my Lux and my 90 macro, though I have yet to test them at all Fstops and distances. This implies that my RF was very close but set ever so slightly wrong.

2) It should improve the 35mm lens, reducing the degree of error. But it will not fix it (not without throwing other lenses out) and it will still be backfocussing noticeably between F2.8 and 5.6, worst at F4. You will need to compensate less, but given that the degree of compensation has to be learned in either case, if you already have it nailed then I wouldn't bother changing a thing.

 

Upshot. Whilst I am very grateful to those here who have given me guidance and support on this, I remain convinced that what Solms told me was correct: the 35 lux and cron were designed for film Ms and do not work happily on the M8, exhibiting a characteristic of backfocus which is exacerbated by the crop factor and greater demands of digital.

 

The remaining mystery is, why do some people with these lenses not have a problem?

 

Theory:

 

1) There may be a tiny sweet spot where the RF is set to the very back of its tolerance and some individual lenses to the front of theirs. I may get my lens looked at but it has done EXACTLY the same as the two 35 luxes I tested.

2) Jamie has a silver lens, mine is black, there may be a difference.

3) No more theories.

 

I will update this when I've given more lenses more of a workout on my newly calibrated machine but for now, good evening!

 

Best

 

Tim

[rami G]

Ok--a demonstration here--just to let you see this...

 

This is the 75 lux in an incredibly darkened room, wide open, and my 3 year old son is watching "The Polar Express" (his favourite movie--he's train mad!).

 

This is an *incredible* lens and system. If anyone's wondering why I love the M8, this is a perfect example.

 

Specs:

• M8 RAW converted to JPEG in C1. No PS.
  
• 75 Lux @ f1.4 @ 9ft; focussed on my son's eye, then re-composed for the shot
  
• ISO 2500 @ 1/16 s, handheld, no filter.

What do you think? Not bad for a handheld 75mm shot in almost total darkness, eh? BTW--I took 5 snaps, they're all in focus.

 

[ATTACH]29981[/ATTACH]

 

Oh--and what the heck--here's the 100% crop of the above. No manipulation other than C1.

 

Please allow for the 1/16s here, the worst possible light, and for the percentage of the overall shot this represents.

 

You're also seeing the signature softness of this lens wide open--it's sharp as tacks stopped down.

 

[ATTACH]29982[/ATTACH]

 

 

Jamie, I have had much sharper results with my 75lux at 1.4, both on my M8, on my R-D1 and on my M film bodies. The lens is not that soft at 1.4. Unfortunately, my lens is at DAG for repair, but I will post some shots if I will find, or wait for it to return. I don't think it is Camera shake due to the 1/16 but a slight miss of the focus. I get, under similar conditions a very low presentage of perfectly focused shots.

 

BTW, I think the lens has beautiful characteristics even when its focus is less than perfect. It is only on really large prints that the difference comes out and I think your photo is perfectly usable. Just that the 75, even at 1.4 can reach rather scientific results.

[Jamie Roberts]

{Snipped}.

 

BTW, I think the lens has beautiful characteristics even when its focus is less than perfect. It is only on really large prints that the difference comes out and I think your photo is perfectly usable. Just that the 75, even at 1.4 can reach rather scientific results.

 

Hi Rami

 

We agree to disagree, though you're completely right about how beautiful the less than perfectly focussed areas of the 75 Lux are. I've also had much sharper shots than this from the 75 Lux already, but not at 1/16s.

 

But sharpness is all relative; the 75 1.4 is, as I said before, not exactly soft but very low contrast wide open. The shot I posted was not mis-focussed, because nothing farther back or further in front was focussed more sharply.

 

Camera shake is, without a tripod, inevitable at 1/16 handheld with such a large lens. My surprise came at how *little* shake there was, but it's still evident at 100%. But that won't affect normal sized prints much.

 

Anyway, I stand by my feeling that the look of the 75 1.4 is downright soft at 1.4 compared with the newer 50 1.4 Lux or even the 35 1.4 Lux, where the ASPH design, I believe, has made things much sharper / contrastier wide open. The 50 R Lux is also much sharper at 1.4

 

BTW, both those M lenses, wide open, are also much sharper edge to edge than the 75.

 

By f 2 or f2.8--on my copy, at any rate--the 75 becomes much "sharper" though it is a matter, again, of contrast more than fine detail.

 

Erwin Puts, by the way, agrees with me, too, for whatever that's worth You should read his lens analysis available at Leica. Because he explains this so well, I've excerpted the relevant passages, but the whole article is much longer and definitely worth a read or three!

 

ANYWAY--and this is a big anyway, hence the caps--

 

I love the lens, and the look of the lens wide open. It's fabulous, and absolute sharpness and contrast aren't always pleasant in, say, portraits, which is where I'm using mine.

 

I also *love* the look of OOF areas--they are superb, just like the 80 R lux, which must be this lens's spiritual twin!

 

This will be one of my most "go to" lenses for people... I can tell.

 

From E. Puts on the 75 Lux:

At full aperture the lens exhibits a medium to high overall contrast, with extremely fine details quite visibly recorded. Very fine details are clearly resolved with some softness at the edges. Some astigmatism is visible in the outer zones, which softens the finest possible textural details.

 

This performance holds over most of the image field, with a detectable reduction in the outermost zone. The corners, although much softer, still record very fine details with good visibility. Stopping down to f/2.0 achieves the high overall contrast needed to record extremely fine details with clarity and crispness. Higher contrast generally gives the fine details more clarity and sharper edges. The outer zones now also improve and only the extreme corners lag a bit beyond this performance.

 

At f/2.8 the contrast is slightly higher yet and now the micro contrast is at its top, allowing the clear and crisp rendition of exceedingly small details.

[d_brown]

Tim, very nice post about your adventure! My findings roughly parallel yours. This is not a matter of getting eyelashes in focus instead of the nose, it is a problem of getting the whole head out of focus when you focus on the nose. Or your garage instead of the bird box. I have two 35 Luxes and you have one that exhibit the same problem. My 35 Cron and my 50 Lux do the same. I also have 15 other M lenses which focus correctly. Others may have 35 Luxes that either don't have such an extreme problem, have only a slight problem which they haven't noticed, or no problem, in which case I am happy for them.

 

The 35 Lux is an incredible lens if you can get it into focus. It's ability to capture minute detail and render overall clarity and brilliance of color are amazing. Yes, you can easily mis-adjust the rangefinder to compensate for an out-of-tolerance lens but he result, of course, is making the camera mis-focus all other lenses which are within tolerance. This is fine if one lens is the only lens you use on a camera. But this violates the idea of interchangeable parts and components, practiced since early in the last century, and is the whole idea behind the Leica system.

 

What is the problem? The lens is composed of several components, among them are the mount, the helicoid, which controls actuation of the rangefinder as the lens if focused, and the lens element assembly. The problem with my lenses, and probably yours, is that the lens element assemblies are not positioned correctly in relation to the helicoids. How can that be fixed? The position of the lens element assembly, in relation to the helicoid, is controlled by interchangeable shims. Simply, change the shims to the proper value! The lens then will focus properly, the rangefinder will still provide co-incident images at infinity, and all your other lenses will focus properly. Unfortunately, for me, at least, this is not as simple as it sounds. I have never disassembled a Leica lens, don't know how to, and never want to. But I could learn how to do it if I have to. My 35 Lux has been in Solms for two months. They don't acknowledge they even have it, although Leica NJ says they do, and I have three more unusable lenses that desperately need attention. I am not-so-patiently waiting for my Lux to be returned. If they fix it, wonderful. I'll send them three more. If they don't and I get the old tap dance, I'll have to resort to plan B.

 

I saw a post by Robert Stevens a few days ago regarding lens focus calibration which caught my eye. He kindly responded to my email about it with a suggestion about a Leica repair manual, which I have ordered. I don't want to fix the lenses myself (unless I have to), but rather determine the exact amount by which the elements are mounted out of tolerance. The manual may help me refine the method I have in mind for determining this distance. I have the tools to accurately measure this once I get the methodology finalized. Preliminary testing shows that the distances they are out-of-spec are between .002 and .006". When I feel I can confidently specify the exact amount for the necessary correction, I'll let someone else change the shims. Unfortunately, this is all a very slow, frustrating process, but necessary to recover the $12,500 I've invested in lenses I can't use and would not feel comfortable selling to someone else in their present condition. It is too bad that after all the research, design, precision grinding and mounting of the elements, and making such nice mounts, the Leica would slip up on the last, simplest, part. And then forget to test it before it was shipped!

 

All this takes time which should be devoted to making some nice photos. The M8, with all its faults and all the frustrations it brings, is a wonderful camera. If it were not, I would have unloaded it long ago and wouldn't have to bother with things like this!

 

Tim, I hope you get everything going the way you want it to very soon.

 

Dale

[sergiolov]

Dale, I think you are perfectly right, and this is what I am planning to do with my trielmar 28-35-70 that has strong front focus.

Can you point me to the manual?

Thanks.

Sergio

[Jamie Roberts]

Dale--what a story! I hope Leica fixes your problem soon.

 

For those of us in Canada, if I thought something like that was out with my lenses, I'd call Kindermann Canada--they've been fixing M stuff for ever and are very, very good.

 

I'm waiting nervously to see how Tim did. I sure hope his 35 isn't suffering from the same thing as yours, Dale!

[egibaud]

I am lost !!!!

 

Tim detailed tweaking experience proves what I said since I wrote my first post in this thread. Tweaking will affect all lenses!! So the decision is basically which lense do I use most, tweak for this one and do guess focusing on the rest.... no quite Leica's image I had.

 

Other option... bye to my Lux35 and get a new Elmarit 28 designed for the M8. As Leica has stopped the 75lux I would not be surprised to see a new 35 soon (that maybe to years in Leica time scale)

 

Could anyone tell me if they have focus problems with the following lenses as I do need to use some kind of lens on my M8 and opinions would be welcome before buying lenses.

 

Elmarit 28

Cron 75

Tri-Elmar 28-35-50

Cron 50

 

Thanks to all, and Tim

 

Eric

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[marknorton]

I am uneasy about adjusting the camera to fix a problem with a lens. As a number of people have pointed out, adjusting the camera will affect its operation with all lenses to varying degrees. It would be far better to adjust the camera to an absolute standard and then all lenses separately to the same standard.

 

One of the issues they faced with the M8 design was that the lens is mounted forwards of the body compared to an M7 to make space for the sensor/LCD package at the back. They therefore had to move the pivot point for the roller forwards by the same amount and adjust the geometry of the rangefinder so that the relationship between the position of the roller relative to the lens mount and the rangefinder focussing distance exactly matches that of the film cameras and the basis then for all M lenses made to date.

 

Question is, how well did they do? I am wondering just how closely the relationship matches that on a film body even in a camera which is supposedly adjusted correctly .

 

Lenses exploit that relationship through the profile of the focussing cam so that if they are no longer working to the same standard, focussing will be out.

 

I think there's a real need to be able to check the range finder independently of any lenses - a lens mount which sets the roller to infinity one way and 0.7m when turned through 180 degrees would help users decide if their camera or lenses are at fault.