Backfocusing lenses. Is it a Leica-only disease ?

[misha]

Advertisement (gone after registration)

Ed,

The backfocus problem on your 75 cron was similar to the problem I had when I bought a used 75 lux. Minimum backfocus close in but feet apart after a distance. I used the M8 rangefinder tweak to correct. Now it's spot on.

 

With backfocus, even if the factory starts calibrating the new M8s and new lenses to a closer tolerance, there are thousands of used lens stock out there that have to be matched to the bodies and to each other.

 

Jan

 

Leica NJ adjusted my backfocusing RF and 35 cron and 50 lux are now dead on. I just got a brand new 75 cron and it is clearly front focusing. Now, will re-adjusting M8's RF throw off the other lenses?

 

misha

[Guest guy_mancuso]

Sean

SW V1.102

The same thing happened as last time.

 

Switch on the camera, prepare to take a shot, the display shows the menu (?)and press the shutter, nothing happens, meaning display continues to show menu and shutter does not releaase.Yes, the battery was charged.

 

Turn camera off and menu menu display is still showing.(the camera is swithched off !)Take battery out , wait and replace, same thing.Try again same thing.

Take out 350d , switch on and take photographs.

Tomorrow off to Jerusalem, will probably buy a cheap WA not to miss too many opportunities...

 

regards

andy

 

 

That is not SDS , completely the opposite the menu stays off. Andy do you want to try reloading the firmware. I know your on the road but it maybe worth a try e-mail me at guymancuso at cox.net and i will send 1.102

[Advertisement]

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

[JWW]

Misha,

 

Unfortunately if some are backfocusing and some are front focusing, adjusting the M8 won't take care of it. When I adjusted the M8 for the 75 lux backfocusing, all my other lenses were still ok since none were frontfocusing. The ultimate answer is for Leica to come up with a focus test standard that all M bodies (film & digital) and all lenses can be calibrated to.

 

Jan

[sean_reid]

Sean

SW V1.102

The same thing happened as last time.

 

Switch on the camera, prepare to take a shot, the display shows the menu (?)and press the shutter, nothing happens, meaning display continues to show menu and shutter does not releaase.Yes, the battery was charged.

 

Turn camera off and menu menu display is still showing.(the camera is swithched off !)Take battery out , wait and replace, same thing.Try again same thing.

Take out 350d , switch on and take photographs.

Tomorrow off to Jerusalem, will probably buy a cheap WA not to miss too many opportunities...

 

regards

andy

 

Hi Andy,

 

That's similar to what many people have experienced, unfortunately. Did you happen to notice if there was much static in the area where this happened? Conditions? Dry?

 

Cheers,

 

Sean

[marknorton]

Misha, if your camera is focussing correctly with two lenses, adjusting the rangefinder to work with your 75/2 will inevitably upset the alignment for the other two lenses. As Jan says, all lenses and all cameras need to be calibrated independently of each other to defined standards to ensure any lens works with any camera.

[marknorton]

Actually, the CV 35/1.2 focuses spot on with virtually no focus shift as it is stopped down. Ditto for the 28/1.9, 35/1.7 and 50/1.5. See the 35 mm review but max aperture is probably a red herring here. Various people who have sometimes referred to CV lenses as being "off brand" and "second rate" must be scratching their heads on this one a bit.

 

There's some minor shift in the Zeiss 35/2.0 and CV 40/1.4.

 

The reality, which I hope gradually gains wider acceptance, is that various lenses have strengths and weaknesses and this is not always correlated with who makes them or what they cost. Status, cost and performance aren't always matched.

 

I certainly accept that different lenses have different attributes but the ability to focus accurately and remain focussed as you stop down is surely a pre-requisite before pretty much anything else. Much as Leica might like us to believe otherwise, the focus shift on a 35/1.4 is not that clever a party trick.

 

Since backfocus and focus shift are quite easy to find if you go looking for them with Leica's lenses, it seems clear they need to sharpen their act and redefine their calibration procedures.

 

Having played with the Nocti over the last couple of days, even though my copy focusses correctly, the focus shift is significant which renders it pretty much unusable in my book. I think with this lens you have to stop pixel-peeping and just ask the question, do I like the image the lens has created?

[misha]

Advertisement (gone after registration)

Misha, if your camera is focussing correctly with two lenses, adjusting the rangefinder to work with your 75/2 will inevitably upset the alignment for the other two lenses. As Jan says, all lenses and all cameras need to be calibrated independently of each other to defined standards to ensure any lens works with any camera.

 

 

thats what i thought. is calibrating a brand new lens more time consuming than adjusting an M8 RF?. i just got the 75 cron and it would be a pity to leave it behind (i am going on a one month trip). in any case, lets see what people in NJ have to say tomorrow

[Peter Branch]

........ The ultimate answer is for Leica to come up with a focus test standard that all M bodies (film & digital) and all lenses can be calibrated to.

 

As I have posted several times the problem being discussed, (setting aside calibration errors which should simply be corrected), is one more unintended consequence of the M8's technology. The sensor is absolutely flat. Film was never flat.

 

When lenses were set up for best performance on film they were set a small fraction of a millimetre further into the body than the theoretical engineering distance. This is not the optimum setting for a digital sensor. It is thus simply not possible to have a single standard which gives the best possible results on film and on the M8.

 

I don’t know if my lenses were adjusted when they were coded and I have not used them on film since, but in the everyday world of taking pictures, hundreds of them, these “defects”, which I have easily been able to replicate, have had no noticeable impact whatsoever.

 

The M8 is a much more precise and sensitive tool than the film M cameras. Artefacts that never troubled users of film are now much more easily seen. We just have to learn new skills.

[wlaidlaw]

Zeiss has for years had a bit of a fetish about flat film paths. On their top of the range SLR's, the RTSIII and I think the N1, they actually had a tiny vacuum pump to suck the film flat onto a porous precision ground ceramic plate (a Kyocera speciality). They said there was no point making ultra high quality lenses, if the film was not held flat at the correct distance. In that case they would make all their lenses to precisely the same flange to imaging medium distance. I would guess this philosophy has continued in their new Cosina made lenses.

 

I still talk occasionally to the ex-Zeiss/Kyocera UK guys. They have said that in contrast to Kyocera, where there were constant rows about quality control and the high rejection rate of the Zeiss inspectors, Cosina has gone from strength to strength. Orginally Zeiss were going to make a number of the more difficult lenses in Germany but now I think only the 15 and 85mm ZM lenses are not made by Cosina.

 

I think that my two ZM lenses actually feel better made than my Noctilux and Elmar-M 50. Only my slightly older Elmarit-M 90 (about 6 years old) feels superior. The Noctilux's wobbly plastic sun hood is a piece of junk and the floppy, loose aperture ring on my new Elmar feels like something made by Lomo.

 

Wilson

[marknorton]

Misha, the 75/2 has a double helicoid - set the lens to minimum focus and look into the back of the lens and you can see them. As the focusssing ring is turned, one helicoid adjusts the position of the lens barrel relative to the lens mount; the second adjusts the position of the focussing cam relative to the lens mount but at a different rate. In addition, the focussing cam is slightly profiled to achieve the required characteristic.

 

Adjusting this lens will likely mean rotating one helicoid relative to the other.

[tashley]

The two terms "focus shift" and "back focus" are used interchangeably but to my mind they are actually two distinct phenomena which will spoil your day.

 

 

Day! LOL Mark... howzabout since February...?

 

;-)

 

t

[sean_reid]

I certainly accept that different lenses have different attributes but the ability to focus accurately and remain focussed as you stop down is surely a pre-requisite before pretty much anything else. Much as Leica might like us to believe otherwise, the focus shift on a 35/1.4 is not that clever a party trick.

 

I think with this lens you have to stop pixel-peeping and just ask the question, do I like the image the lens has created?

 

It's partly a question of degree and, also, of what one finds unacceptable. Focus shift from aperture change is not new to photography although knowing that fact doesn't necessarily make it palatable to experience now. The 35/1.4 Aspherical lenses I tested showed some focus shift as they were stopped down but the effect was not dramatic. Others have reported seeing no focus shift from that model while still others have reported significant focus shift. The degree to which one should attribute that difference to sample variations, vs. variations in personal perspectives, is hard to know.

 

And, lest we be picking on the Lux, there's also some focus shift in the 35 Cron and 35 Biogon as well.

 

I wouldn't hesitate, and have not hesitated, to shoot an assignment with the 35/1.4 Aspherical but the relentlessness of the M8 (with the 100% monitor view) may indeed cause some to expect no focus shift from these very expensive first-rate lenses. What might we be trading off in order to gain that lack of focus shift? I don't know.

 

Cheers,

 

Sean

[Guest guy_mancuso]

My new 35 cron is dead to nuts stopping down, my 35 lux was not. If your going to adjust the cam yourself leica mentioned using a 75 lux or 90 cron maybe the best lens to adjust the cam too because of the tight tolerence. which is exactly what i did yesterday and everything else fell into place except my 50 lux and 135 apo needed a shim. There in NJ now getting calibrated along with one M8 with a vertical adjustment issue , which went out when in Germany. i have had this stuff done before with Canon, Nikon , Hassy and others . this is really nothing that new about it , we just did not test for this stuff in the day with film. You looked at a loupe that had 8x power and not a 40 x60 print equivelent on screen. All my Hassy gear went in almost every 6 months for a CLA when i shot MF and every time it needed it. Folks are buying 20 year old lenses that maybe never had a CLA and than blame leica for it being out of focus. You have to simply take your lenses in and get CLA more often than 5 to 20 years. I understand some of these issues that are very real, and with new gear a issue but CLA are normal and should be done yearly on all lenses

[misha]

My new 35 cron is dead to nuts stopping down, my 35 lux was not. If your going to adjust the cam yourself leica mentioned using a 75 lux or 90 cron maybe the best lens to adjust the cam too because of the tight tolerence. which is exactly what i did yesterday and everything else fell into place except my 50 lux and 135 apo needed a shim. There in NJ now getting calibrated along with one M8 with a vertical adjustment issue , which went out when in Germany. i have had this stuff done before with Canon, Nikon , Hassy and others . this is really nothing that new about it , we just did not test for this stuff in the day with film. You looked at a loupe that had 8x power and not a 40 x60 print equivelent on screen. All my Hassy gear went in almost every 6 months for a CLA when i shot MF and every time it needed it. Folks are buying 20 year old lenses that maybe never had a CLA and than blame leica for it being out of focus. You have to simply take your lenses in and get CLA more often than 5 to 20 years. I understand some of these issues that are very real, and with new gear a issue but CLA are normal and should be done yearly on all lenses

 

sure, except my 75 cron is brand new and it is somewhat annoying. that said, 35 cron has been perfect thus far.

[Guest guy_mancuso]

I know Micheal the brand new ones are the ones that folks are having focusing issues with are more the concern. I don't disagree here. i have seen them being built in the factory last month and it is amazing how it is done and we did see the calibration room for every lens that goes out the door. So i know the frustration , a couple of mine are being adjusted also even a pretty new 50 lux

[ho_co]

That's why (and I know HC doesn't agree with me) I think lenses like the 35/1.4 need to be redesigned for digital to take account of these issues which are assuming greater significance in the era of the M8.

Mark, if it turns out that the 35/1.4 problems are as widespread as the forum makes them seem, you may be right. (Notice, I said "may..." )

I really think that Leica has gone down hill from the old days.

Ed, even in the old days there were problems. The Internet makes us think there are more now, but there weren't. Remember the Noctilux 50/1.2? Leica even recommended that anyone buying the lens should send the lens and the body in to have the two matched.

The sensor is absolutely flat. Film was never flat. ... The M8 is a much more precise and sensitive tool than the film M cameras. Artefacts that never troubled users of film are now much more easily seen. We just have to learn new skills.

Peter, you're right. As Guy said, we seldom used more than an 8x loupe with film, but with immediate chimping or the ability to see errors on the computer screen at 100% or larger, we are becoming immediately aware of difficulties that were always there.

Zeiss has for years had a bit of a fetish about flat film paths.

Wilson, so has Leica. Remember the long film path of the screw-mount cameras? So long you can't load them today without increasing the leader length. Remember the first M3s' glass pressure plates with the need for doublestroking to avoid static discharge in dry weather? Remember the Leicaflexes' tiny pressure plates because of the new geometry of the film tunnel?

 

E Puts pointed out that Zeiss has designed its current lenses with the assumption that film will bow backward, while Leica's assumption is that film will bow forward. And the bayonets on Leica bodies were 1/100 mm further forward than the film dictated, in order that they could 'wear in' instead of 'wearing out.'

 

Here's the point: Leica has since the 60's or so designed at the cutting edge of the possible. In bodies that was always the case, but Leica really put the move on lens improvement after the M3's appearance. Zeiss has been a bit less demanding. That is, according to Puts the Zeiss designs are simpler and have more room for assembly error. That's part of why they can be so much less expensive than Leica lenses.

 

One of the things that makes so many of the Leica lenses so good wide open is the way Leica makes use of the intentional focus shift. Don't ask them to drop it; they just need to get it under control for the M8.

 

On the other hand, it looks as if they're beginning to re-think things along the same line as Zeiss: In his review of the new 28/2.8, Puts said its assembly has some features that make it less expensive to produce. He didn't say what, but he said it's a change for Leica.

 

Thread topic asks specifically about backfocus. No, it's not Leica-only. Just for example: Canon has had a great deal of trouble with some lenses on some models. On the 1D III, they have given the control to the user: You can make a setting on the body that says "When lens XXX is mounted, take into consideration that it backfocuses and correct for it by ZZZ micrometers" or some such.

 

Final point: We all love to understand these things, and we all learn from others here. But lens design is not so simple. Everyone--Canon, Nikon, Panasonic, Leica--uses the same software to design lenses. (That wasn't so until computers became ubiquitous. Till that time, Leica used Curta calculators and when school was out lens design progressed much more rapidly because of the extra bodies Leica could get to calculate ray traces.) So if Leica's lenses are going to outperform, Leica must push the design with special glasses or unique designs or extreme optimizations. Sometimes (e.g. Noctilux 1.2) they get a little ahead of themselves. I think that's where they've landed now with the M8. They've always solved the problems before, and they will now again.

 

That's my belief.

 

--HC

[nscali]

HC,

I think you are spot on.

My Noctilux and 35 lux (both new coded samples) are spot on wide open and one stop down. There is some focus shift to about f5.6, but only really in the absolute centre of the lens. I can't help but think this is an intentional design thing. You can 'cheat' by adjusting your RF to ever so slightly front focus wide open (a little softer when wide open) and then it is spot on in the centre as you stop down. I don't however, see the point.

Inveresly, my canon 50mm L (1.2) on a 5d is terrible wide open (very soft) but get's better as you stop down.

I tend to think that if you have fast lenses (1.0 and 1.4) you want them to perform their best wide open. That's what makes these Leica lenses so great.

Nicky

[jaapv]

HC,

I think you are spot on.

My Noctilux and 35 lux (both new coded samples) are spot on wide open and one stop down. There is some focus shift to about f5.6, but only really in the absolute centre of the lens. I can't help but think this is an intentional design thing. You can 'cheat' by adjusting your RF to ever so slightly front focus wide open (a little softer when wide open) and then it is spot on in the centre as you stop down. I don't however, see the point.

Inveresly, my canon 50mm L (1.2) on a 5d is terrible wide open (very soft) but get's better as you stop down.

I tend to think that if you have fast lenses (1.0 and 1.4) you want them to perform their best wide open. That's what makes these Leica lenses so great.

Nicky

It is hardly intentional - It would be better to call it unavoidable....

[nscali]

Jaap,

Very true.

The bottom line is that it is just a fact of life as we know it.

It is not a 'defect' or poor QC.

The problem is we see how incredibly sharp and how much dynamic range there can be in images produced from this system. We then expect to to see this stopped all the way down.

I was so preoccupied with focus shift/back focus issues until I picked up the 5d again and realised how good these fast lecia lenses are.

[Advertisement]

Hallo,

Look here: Leica M

[wlaidlaw]

Wilson, so has Leica. Remember the long film path of the screw-mount cameras? So long you can't load them today without increasing the leader length.

 

--HC

 

Howard,

 

A bit OT but I was shown some years ago by the man behind the counter in R.G. Lewis in Holborn, how to load a normal cassette film in an LTM Leica, if you do not have the leader cutting template with you. You slide a small card down the back side of the sprocket nearest the cassette side of the camera. The best sort of card is one of those shiny floppy membership cards that comes stuck to a letter. The card MUST have rounded corners or you risk damaging the shutter blind. You then slide the film and cassette in behind the card i.e away from the sprockets and shutter. Once the film is fully down, you gently slide out the card and the double row of sprocket holes on the film now settles nicely over the sprocket teeth. I have been doing this for about 10 years with no problems. In fact if you asked me where my film leader cutting template was, I could not tell you.

 

The doorbell has just rung and it's my Voigtlander 35/1.2 finally arrived. Now lets give it a back-focus test ;-}}

 

Wilson