Very interesting answer from Leica on 35mm 1.4

[tashley]

Advertisement (gone after registration)

Tim--

 

This is truly through the looking-glass...

 

So let's recap...

 

You've got an Elmarit that is sharper on one side than the other, and a Cron that is fuzzy in the middle but sharp at the edges for the same distance.

 

Logically, there's only one possible conclusion.

 

You, my friend, have a defective camera or the lenses need lens adjustments. I personally don't believe the sensor is be aligned with the mount properly--OR all the lenses need to be tweaked by someone who knows them.

 

The fact that the 90 and 50 focus is luck. We're talking about something being off by a minute amount (and I'd check the frame edges for the 50 and 90 wide open, too... you may find they're not as sharp as they should be...)

 

My elmarit is sharp all over the field. We already know about my 35.

 

And how, pray, can a 35 CV "focus better" than a cron or a lux? I think that actually nails it--something is very wrong here.

 

@ Eric--fixing the focus on the RF doesn't make it worse for other lenses, unless those lenses need adjustment to begin with. So don't give up on the 35s; many of us use them without problem!

 

As for my own tiny front focussing 75 Lux, I will eventually take it into Kindermann with the body and get it tweaked. But since I can live with it the way it is, it might be a long time before I do that!

 

Tim--I wish there was some old Leica know-it-alls close to you. There must be a UK repair depot for M, right?

 

Nope! Nice try! ;-) I have a zillion shots that are sharp as anything on a variety of lenses. The only lenses that don't behave are the 35 lux or cron, and they behave exactly as Solms tells me they will on M8's since they were designed for film Ms. The case against the elmarit is unproven, I noticed it in one shot and it may have been that the camera was not exactly parallel to the target. And the reason the CV focusses better is that it isn't aspheric. The reason the 50 lux works despite being aspheric is that it has a floating element and the 35 cron and lux don't...

 

So until I have tested that one I remain convinced of my case... but I admire your fortitude!

 

:-)

 

Tim

[Jamie Roberts]

Well Tim, I admire your fortitude too. There are a lot of people who have 35s in various states of focus, but none I've heard of focus at the edges but not in the middle.

 

I'll go to the wall here and ask for people with 35s to tell me if that's true or not.

 

The people from Leica were smoking something if they said their lens wouldn't focus in the middle because 'it's designed for film'...

 

Anyway, I'm talking with Leica US next week; I'll ask.

 

PS--test the Elamarit.

[tashley]

interesting to hear what they say but if you read through all the threads on the 35's you'll find plenty of people have said they are soft in the middle and sharp at the edges - it's just that we'dd all assumed it was a distance thing! The Solms guy did tell me that this was characteristic of the asph 35 behaviour on digital M...

 

:-(

 

Tim

[Jamie Roberts]

Tim, I hear you; only trying to help.

 

But softness is relative too--I've never seen any "current" "cron" Leica glass where the difference in softness is what you're showing with that birdhouse.

 

Bill Parsons--anyone else with a working 35--going to comment?

 

Or do I have the only magic 35 Lux in existence? If so, I'm upping the insurance

 

PS--very, very glad the CV 35 is working well It should be!

[marknorton]

Jamie, I've tested my 35/1.4 to death and I cannot reproduce what Tim is seeing. I've also tested a pre-ASPH 35/2 and it's fine too.

 

Tim, did you try moving the nesting box to the opposite corner of the frame, is it sharp there as well?

[egibaud]

There are a lot of people who have 35s in various states of focus, but none I've heard of focus at the edges but not in the middle. .

 

What about mine too? and what about all the posts in this thread? what about all the post on summilux.net ?

 

Beleive us, and me, it is the way Tim is saying.

 

Eric

[tom0511]

Advertisement (gone after registration)

From my experience if you have a M-body and a lens which work well together, then you can get very accurate focus.

I can focus my 50/1.4asph (which was calibrated) and my 75/1.4 very accurate, I have some other lenses which are fine but not totally perfect: The 35lux_asph and my 28cron asph; both had been sent to Leica for focus calibration as well.

I can just envy people to not give up on fast lenses.

[wparsonsgisnet]

Tim, I've read thru the thread and then reread the small print from the insurance -- sorry, it was from Solms....

 

I have 24 asph, 35 'lux asph, and 50 'lux asph. All focus properly with my M8.

 

I am reading your posts and feeling your pain. Before I sent another lens back to Leica, I'd put it on a different M camera and see what I got.

 

I agree with Jamie that there is something besides the lens at fault, here. Unless you got several 35's with poorly aligned mounts.

 

Since my 50 focuses as well as the 35 -- and not the other way around -- and then the 35 focuses as well as the 24, I think I'm in good shape. The longer lenses have the shorter DOF's after all.

 

Put this 35 on another body and see if it works well.

 

Is this the asph lens? I don't see it specified.

[Jamie Roberts]

Bill--thanks!

 

And IIRC Mark Norton's works as well... and Sergios, and....so if this is *endemic* it's not consistent, either.

 

Eric--I've never heard of Summilux.net. I'll go look.

 

But--by analogy here, please don't clobber me--at least 3/4 of the people I knew complaining that the Canon 24-70 wouldn't focus (backfocus) never had the thing looked at by Canon with their camera.

 

I suspect something similar is happening here--regardless of what Leica has implied.

 

BTW--I'm as skeptical as Jono--I'd find someone who has personally serviced M stuff through the years, and perhaps not just send everything to Leica. Dunno why, except their best folks are perhaps otherwise employed right now Just a hunch--I have no knowledge about this.

[ho_co]

I do think however that Leica should have revised that 1/30th of a millimeter standard way back in 1935 or so – but sitting on an industry standard does of course give you a nice, cosy feeling.

Lars--

And remember, Leica invented that standard. Compare the Contax lenses up into the 50's and 60's; their DoF scales were engraved for much smaller enlargements. (In other words, they showed much greater DoF than their Leica counterparts.)

 

And think of all those fine lenses out there with those Neolithic DoF scales!

I did a bit of study on DoF and rangefinder accuracy when the M8 came out with its reduced baselength, and found some interesting stuff, which led to a thread on this forum which some found confusing (http://www.leica-camera-user.com/digital-forum/7446-rangefinder-accuracy-charts.html).

 

E Puts is very vague about the formulas he uses, and uses different formulas at different stages of his calculations. The DoF and Hyperfocal Distance formulas are well known, but when you get to calculating rangefinder accuracy, it becomes clear that he makes a number of 'unjustified' assumptions--i.e., 'unjustified' in the sense that he does not clarify why he chooses them, and does not even formulate them clearly

 

The upshot is that by the time you get around to adjusting the CoC for calculating enlargement sizes, you're changing the assumptions that the rangefinder calculations are based on; and I have found no one who shows how to adjust the rangefinder formulas to account for those changes.

 

In other words, the origins of the rangefinder calculations seem to be lost in the mists. In the LFI article relating focusing accuracy of various VF magnifications to focal length and speed, the author points out that the graphs should actually be redone with a smaller CoC. But when you do that using the formulas provided, you 'prove' that we can't focus lenses like the 75/1.4.

 

That is, the assumptions are out of date, and apparently the original assumptions are lost, so we seem to be locked into these out-of-date assumptions. [see--finally back on topic! ]

 

As you said above in other words to the annoyance of a couple other posters, "Der Ball ist rund. Das Spiel dauert neunzig Minuten. Alles Andere ist Theorie."

 

--HC

[tashley]

Tim, I've read thru the thread and then reread the small print from the insurance -- sorry, it was from Solms....

 

I have 24 asph, 35 'lux asph, and 50 'lux asph. All focus properly with my M8.

 

I am reading your posts and feeling your pain. Before I sent another lens back to Leica, I'd put it on a different M camera and see what I got.

 

I agree with Jamie that there is something besides the lens at fault, here. Unless you got several 35's with poorly aligned mounts.

 

Since my 50 focuses as well as the 35 -- and not the other way around -- and then the 35 focuses as well as the 24, I think I'm in good shape. The longer lenses have the shorter DOF's after all.

 

Put this 35 on another body and see if it works well.

 

Is this the asph lens? I don't see it specified.

 

 

Thanks for the empathy!

 

I am 100% convinced by this - Iv'e tried three different 35mm Leica asperics, latest versions, on three different M8s (not every lens on every body but nearly) and get exactly the same results with all.

 

I do think that despite what Solms has said, there may be a tiny sweet spot where the RF is calibrated to the edge of its tolerance in one direction and the lens in the other, where it might be made to work but I can only tinker with my RF and not my lens. I might do as some have suggested but it looks like my new 35,, CV pancake lens is an easier option since on first showing at least it seems very sharp.

 

We live and learn (and spend!)

 

Tim

[tashley]

Jamie, I've tested my 35/1.4 to death and I cannot reproduce what Tim is seeing. I've also tested a pre-ASPH 35/2 and it's fine too.

 

Tim, did you try moving the nesting box to the opposite corner of the frame, is it sharp there as well?

 

 

I'll try in the morning but I am sure it will be. I have loads of shots of buildings where the edges are sharp and the centre silghtly oof, and the edges seem equally sharp!

 

Tim

 

ps Mark, can you try the birdbox test with yours? Choose something about that size, about twenty five feet away, with a big structure further back - say another thirty feet - and focus on the near object. Expose with that object at exactly centre frame and at F4, then move that subject to about 1/3rd in from the edge and see if it is sharper.

[ho_co]

In other words the degree of bacfocuss of the 35 cron on an M8 depends not just on aperture but also on where the subject is in the frame.

Tim--

Relax, my friend!

 

This is the field curvature the guy at Solms implied was present in the lens.

 

There is nothing wrong with your lens. This is the way it was designed. We see it now because unlike film, the sensor is completely flat and unforgiving.

 

BUT if you have a lens (24) with one side sharper than the other, there is something wrong with the lens: The optical head is crooked in its mount. Puts says this lens is a stellar performer. You say yours isn't. Send it in for adjustment/replacement: It is defective.

 

AND as others have suggested, there's nothing wrong with packing up the lens (35) and body and having them calibrated together. Leica specifically recommended that practice with the Noctilux 1.2 in fact.

 

OR as others have suggested, though it's 'impossible,' you may have a misaligned sensor. Best bet is to try the lenses on another body--just to keep you busy, of course!

 

--HC

[tashley]

Tim, I hear you; only trying to help.

 

But softness is relative too--I've never seen any "current" "cron" Leica glass where the difference in softness is what you're showing with that birdhouse.

 

Bill Parsons--anyone else with a working 35--going to comment?

 

Or do I have the only magic 35 Lux in existence? If so, I'm upping the insurance

 

PS--very, very glad the CV 35 is working well It should be!

 

I know - and you know how much I appreciate it! I've learned more about lenses and RF calibration in the past few days than I ever thought I would, most of it from you!

 

Tim

[tashley]

Tim--

Relax, my friend!

 

This is the spherical aberration the guy at Solms told you was present in the lens. (The fact that it has an aspherical surface doesn't mean it can't still produce spherical aberration. Aspherical surfaces are merely one more tool for the lens designer, but he still has all the same aberrations to take care of.)

 

There is nothing wrong with your lens. This is the way it was designed. We see it now because unlike film, the sensor is completely flat and unforgiving.

 

BUT if you have a lens with one side sharper than the other, there is something wrong with the lens: The optical head is crooked in its mount. Puts says this lens is a stellar performer. You say yours isn't. Send it in for adjustment/replacement: It is defective.

 

--HC

 

Thanks for the kind words!

 

I sent the first two (both luxes) back and solms didn't even blink: they just said, as you do above, 'this is what they all do'. For quick and easy, the CV 35 pancake is very nearly as good, just a bit more viggy, uncodable (there's a screw in the way) and slightly less sharp at the edges. But it's a fab lens!

 

Do you have any idea why Jamie and a few others are not seeing this effect?

 

I unscientifically tested the 24 elmarit and it does seem softer on the left - someone else here reported this too - but it's marginal and in general it is very sharp.

 

Thanks for your help!

 

Tim

[ho_co]

Do you have any idea why Jamie and a few others are not seeing this effect?

Tim--You might as well ask why _you_ *are* seeing the effect.

 

Seriously, I think you understand Solms's explanations correctly. The problem is inherent in the 35/1.4; you discovered that and they verified it for the lenses you sent for checking.

 

The problem is less for the 35/2, but still present. That's the lens that I thought you might want to have calibrated to your body.

 

See E Puts's comments on the Biogon 35/2 at http://www.imx.nl/photosite/comments/c016.html: He feels the Summicron is better overall, but the Zeiss lens is better in regard to field curvature, which is the problem your birdhouse pics show.

 

Not to diminish the sincerity, ability and proficiency of Jamie or Mark or Eric or you or anyone else, I think you answered the question of why others don't see the problem earlier when you said these tests must be performed rigorously. You are very much interested in the performance of your lenses, both technically and artistically. Others may not notice the same problems simply because they shoot differently or because their technical interest is less.

 

In other words, the problem is probably present to some degree in all the 35/1.4's, but because of their shooting practices and habits, and possibly because of manufacturing tolerances, some other folks simply haven't noticed it.

 

That's where Lars's recommendation to just go take pictures comes from. (I mean after all, this thread is already over 170 posts long! ) Of course, Lars's advice doesn't work for everyone because we don't all want the same pictures.

 

My guess is that your body is fine and that the 35/2 could be tweaked modestly to give slightly better performance with your M8 if you sent both for adjustment. But the design really doesn't seem quite up to the flatness of the digital sensor, so you probably would still not be fully satisfied.

 

It may even be (I'm no lens designer) that designing a flat field lens would require a diminution of performance on other counts, as is the case with the Biogon.

 

So at the moment there probably is no solution: You can choose another lens, or learn to live with the quirks of this one. I think you've done a very great job in making us aware of the limitations of these lenses.

 

This kind of exactitude is what makes this forum what it is.

 

One dealer I spoke with told me what his Leica rep told him about the 35s. I won't repeat it here. You tell me if you think the 35s don't have a problem.

Aw, Dale! Come on! Tell us! Pretty please!

 

Seriously, it would be only hearsay, and that's all you're getting from most of the rest of us! Remember, the rep may or may not know (I know because I was a rep); and his information is being filtered through a dealer who had an alternative lens he wanted to sell you.

 

So IMHO there would be nothing wrong with adding that tidbit to this already long-enough thread.

 

--HC

[tashley]

Thanks HC, my sanity is restored and I agree with all of the above.

 

The funny thing is, I was never picky about this sort of thing until I bought my M8 and lenses and started hanging out here. But has it been a learning curve!

 

I took a whole load of shots for fun this afternoon and my new CV35 is a doll. Bit smeary on the far left, not so as you'd notice, but otherwise very much sharper on centre than the cron. I like it a lot but you rarely hear much about it. And the new WATE is great too.

 

Thanks for your clarifications. I'm off to Barcelona for the weekend and I will get back to the real job in hand - taking pictures!

 

Best

 

Tim

[Jamie Roberts]

HC--we will agree to disagree I performed my tests just as rigorously as Tim, and I still don't see

 

1) an OOF point of focus (IOW at the center)

2) anything like sharper corners than edges

 

--though field curvature certainly takes place with all lenses.

 

This usually means, however, that things at the edge stay in focus when you might not think they would due to them being at a slightly different distance.

 

Normally, for example, I would put a row of people I was shooting in an arc to keep them in relatively the same distance from the lens.

 

However, due to field curvature this may in fact be off a bit (and stopping down helps).

 

But I've never seen field curvature to explain why a center point would be OFF and the edges SHARP--unless the lens is mis-aligned to begin with.

 

Sorry. This is still a mystery, and making Tim feel better, while a worthwhile goal, doesn't explain all the people (critical ones, too, for whom the 35s work just fine at the center at the focus point (sheesh--how limiting!) through the aperture range.

 

Finally, HC, the tests might have to be performed rigorously, I guess, but Tim is seeing an error of several feet at f4 at distance with a wide angle lens.

 

Sorry--that does not compute. Maybe if we're talking a lens baby or holga; certainly NOT a summicron.

[wparsonsgisnet]

Tim, I got curious and took the attached pix:

 

35 'lux asph, closest focus, f1.4, 1/125, iso 160, nothing else; then, f5.6, rest the same.

This was focused on the edge of the paper, at the center of the pic. The 31 and the scribble are in the same plane of focus.

 

The 1st is at f1.4; the second, at f5.6.

Both are in v1.092, with Edmund LoSatTuned2 profile, and WB. No other processing.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[sergiolov]

Hi Tim,

a few points. First the mistery of the bird box.

The focus and reframe procedure is usually misunderstood. A lens focuses a plane on another plane, not a sphere. So the focus distance is valid only for objects that are perpendicular to the sensor. Remember Pytagora theorem? If your focus distance is 10 meters perpendicular to the focus plane, an object displaced 10 meters on that same plane will be on focus, but its distance from the focus plane will be 14,14..meters.

If you put this on a drawing, you'll see the consequences..

 

Also, did you take a look to the MTF table for the lux asph, as I suggested in my preceding post? You'll notice that the 40 lp/mm line at 2,8 is strongly sinusoidal, with a variation of more than 2 stops. This accounts for a sharpness variation across the frame.

 

The question of the lens and the opinions of the guy in Solms. What he declares is only generically correct, but has no relation with the case we are discussing here.

It is completely true that the lens backfocuses when stopped down. The question is if it is possible to maintain the focus plane inside an acceptably sharp area while focusing and stopping down. If this is possible even with one lens, than all others are practically misadjusted.

Now we know that my lens and a few others focus with reasonable accuracy. I showed on my test that this is due to the fact that, with the lens full open the focus plane is positioned to the extreme rear position of the sharp area, while stopping down this plane progressively comes nearer to the front position of the sharp area. This is a good compromise, but is connected to the positioning of some shims that millesimally change the focus distance of the lens, as pointed by Brown.

And we also know that it is possible that, ignoring the opinions of that smart person from Solms, the operator and his Quality Control are sometimes a little distract.

Your "sharpness with focus guessing" is just like changing the thickness of a little shim.

I know that this is of no help, and that you want a working lens.

Ciao.

Sergio