Very interesting answer from Leica on 35mm 1.4

[ho_co]

Advertisement (gone after registration)

Um, isn't this why a lot of Leica's lenses are aspherical in design?

Jamie--

Aspherical surfaces just make one more tool in the arsenal of the lens designer. Aspheric design doesn't mean "no more spherical aberration" or any other one thing. You know that designing a lens is a balance of compromises, and this is just one more device in the toolset.

 

It isn't as simple as, for example, Nikon makes it sound: About the 17-55/2.8 they say, "3 Extra-Low Dispersion (ED) glass elements for minimized chromatic aberration and provide higher resolution and superior contrast." About the 80-200/2.8 on the other hand they say, "3 ED glass elements for high resolution and high contrast even at maximum apertures." They make it sound as if a lens designer said, "Hey, Mike, why don't you use an ED element here to eliminate the CA?" Same thing as, "Hey, why don't you put a capacitor in the circuit to improve the sound?" Or, "Hey, one of these pills will cure all your problems."

 

No one tool solves any one problem, because solving one problem causes other problems to grow, and lens designers need to balance the various effects in their designs.

 

I think by hanging ASPH on their lenses Leica is bending to marketing rules. They'll mention in passing that a lens uses 'anomalous partial dispersion' glass. Bandying about the abbreviation "ED" on a lens à la Nikon has always been beneath them; but designating lenses as ASPH comes close to the same thing IMHO.

 

A lens isn't better *because* it uses aspherical surfaces or glass with a different pattern of dispersion, but rather the *use* of those mechanisms makes it possible to make a better lens. And the use of a specific glass or lens shape counts in the entire design, not just in "increasing contrast" (Nikon) or any other thing.

 

I know I haven't given you anything to bite into; your question was succinct and my answer is nebulous. All I'm trying to say is that all tools used in a lens design are used to improve the lens; you don't eliminate any aberration by employing any tool or set of tools.

 

I'm sure someone can say it better, but aspherical surfaces aren't used to reduce spherical aberration. They are used to make the lens better, and that may or may not mean that they are put to use reducing spherical aberration.

 

--HC

[tashley]

That sounds very disappointing. I actually have the 35 lux, but I will try that. I sent my 28 back to DAG to see if there is any way to improve it. The interesting thing I find about the M8, which is totally uncomprehensible to me, is that it is easier to focus perfectly using longer lenses (50mm, 75mm) than with the 28, 35. The attached image (not for artistic purposes, thought), for example. I took it with the 75 lux @ 1.4. I just don't get anything close to that with my 35 lux @ 1.4. usm 70/1/0

 

 

You can get results like that with the 35 lux and cron: what you have to do is focus and recompose so the bit you want really sharp is on or outside one of the 'thirds'.

 

A surprising number of my really sharp shots from Venice (link below in my sig) were from the 35 lux: it was my first outing with the lens and I clocked on the first day that shots were blurry in the centre 1/3rd and adjusted accordingly. But it is a PITA, not to mention a pity!

 

Tim

[tashley]

Jamie--

Aspherical surfaces just make one more tool in the arsenal of the lens designer. Aspheric design doesn't mean "no more spherical aberration" or any other one thing. You know that designing a lens is a balance of compromises, and this is just one more device in the toolset.

 

It isn't as simple as, for example, Nikon makes it sound: About the 17-55/2.8 they say, "3 Extra-Low Dispersion (ED) glass elements for minimized chromatic aberration and provide higher resolution and superior contrast." About the 80-200/2.8 on the other hand they say, "3 ED glass elements for high resolution and high contrast even at maximum apertures." They make it sound as if a lens designer said, "Hey, Mike, why don't you use an ED element here to eliminate the CA?" Same thing as, "Hey, why don't you put a capacitor in the circuit to improve the sound?" Or, "Hey, one of these pills will cure all your problems."

 

No one tool solves any one problem, because solving one problem causes other problems to grow, and lens designers need to balance the various effects in their designs.

 

I think by hanging ASPH on their lenses Leica is bending to marketing rules. They'll mention in passing that a lens uses 'anomalous partial dispersion' glass. Bandying about the abbreviation "ED" on a lens à la Nikon has always been beneath them; but designating lenses as ASPH comes close to the same thing IMHO.

 

A lens isn't better *because* it uses aspherical surfaces or glass with a different pattern of dispersion, but rather the *use* of those mechanisms makes it possible to make a better lens. And the use of a specific glass or lens shape counts in the entire design, not just in "increasing contrast" (Nikon) or any other thing.

 

I know I haven't given you anything to bite into; your question was succinct and my answer is nebulous. All I'm trying to say is that all tools used in a lens design are used to improve the lens; you don't eliminate any aberration by employing any tool or set of tools.

 

I'm sure someone can say it better, but aspherical surfaces aren't used to reduce spherical aberration. They are used to make the lens better, and that may or may not mean that they are put to use reducing spherical aberration.

 

--HC

 

In my VERY layman's speak, if you use ASPH design to correct for field aberration, you risk the introduction of a spherical focus aberration.

 

???

[wattsy]

The interesting thing I find about the M8, which is totally uncomprehensible to me, is that it is easier to focus perfectly using longer lenses (50mm, 75mm) than with the 28, 35.

 

I made the point in an earlier thread that if the sensor plane to lens mount distance is slightly out, it will have a more significant effect with shorter focal length lenses than it will with longer ones. Is it possible that your M8 sensor has been sloppily shimmed/adjusted during manufacture?

[carstenw]

Howard, high-precision aspherical lens manufacture is a Leica core competency, so I can understand that it makes it onto their lenses. This is not easy to do, and many other companies which tag ASPH on their lenses don't see the same kind of quality or gains as Leica does.

[tashley]

I made the point in an earlier thread that if the sensor plane to lens mount distance is slightly out, it will have a more significant effect with shorter focal length lenses than it will with longer ones. Is it possible that your M8 sensor has been sloppily shimmed/adjusted during manufacture?

 

 

Hi Wattsy,

 

I guess that could be the case for some of us but isn't it interesting that the OP had no problem until he'd had his glass coded, and that Jamie's Wonder Lens is uncoded?

 

Best

 

Tim

[wattsy]

Advertisement (gone after registration)

But it is a PITA, not to mention a pity!

 

I think it is also unacceptable with lenses at this price point. If there is a genuine inherent problem with focus shift involving some lenses used in conjunction with the M8 (and its perfectly flat sensor plane) then Leica need to come clean about it and not recommend the use of such lenses on the M8. Nobody should be expected to pay £2k for a lens that provides very indifferent performance between F2 and F5.6 when used on an M8. At this kind of price, a lens should be usable without compromise.

[tashley]

I think it is also unacceptable with lenses at this price point. If there is a genuine inherent problem with focus shift involving some lenses used in conjunction with the M8 (and its perfectly flat sensor plane) then Leica need to come clean about it and not recommend the use of such lenses on the M8. Nobody should be expected to pay £2k for a lens that provides very indifferent performance between F2 and F5.6 when used on an M8. At this kind of price, a lens should be usable without compromise.

 

 

I kind of agree BUT.. when I was talking to the chappie from Solms about it, I said 'what other wide lenses do you have that don't have this behaviour?' and he said 'the new 28 elmarit but you won't get the pop you have in some of your Venice shots from it.' Or words to that effect. So if there's no other easily code-able glass that can do it, I'd rather have the option. I have kept my 35 cron so far, after all.

 

IMHO Leica should allow menu selection of lenses to allow us to use other manufacturers' glass where their own has issues, and they should include a technical not in their marketing literature pointing the issue out in advance. But I bet people would still buy the glass because when you learn how to work with it, the results are stellar. But would I use it for mission critical work (like photographing the Dalai Lama?) - No!

 

Tim

[wattsy]

But would I use it for mission critical work (like photographing the Dalai Lama?) - No!

 

But that is really my point. The fast wide lenses are now around £2k each and should be fully usable in a professional capacity. If there are insurmountable issues with some of these lenses used in conjunction with the M8 then Leica should be very clear about this.

[Jamie Roberts]

Tim--interesting what Leica's been telling you!

 

First, that new 28 Elmarit, in addition to being sharp as tacks and easy to focus, is, if anything, a bit *too* snappy for people. It's a high-contrast sort of lens, and is easily a great bargain!

 

But I actually bought the CV 1.9 to work with people--faster, and less contrasty! Oh--and I would have bought the 'cron but the Noctilux must come first.

 

I truly think we're onto something here with the coded mount. That's the real variable we've all been looking for, and certainly if I was a betting man, that's the one most likely to be at fault, no?

 

Well, maybe not. Sigh. Maybe it's chrome

 

Dunno, but this thing is sharp sharp sharp. Here's one right out of C1, absolute center of frame (you know, it's hard to find real world examples where I haven't "shot and recomposed" for rule of thirds with this thing, but I *know* I focused dead center and fired here--not that it would matter, given the DOF anyway!)

 

35mm ASPH @ f4--Expodisc WB-- ISO 320--no filters--C1 only (no PS)

 

First the frame for reference:

 

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!

 

Now the 100% crop from a C1 low-res JPEG.

 

I don't think I could reasonably ask a system to be sharper here, could I?

 

[Jamie Roberts]

I made the point in an earlier thread that if the sensor plane to lens mount distance is slightly out, it will have a more significant effect with shorter focal length lenses than it will with longer ones. Is it possible that your M8 sensor has been sloppily shimmed/adjusted during manufacture?

 

Wattsy--right on.

 

@Rami--I find the M8 incredibly easy to focus, period, but the wides are much easier to focus for me than the 75.

 

I think Tim will agree that the 50 Lux, for instance is ridiculously sharp. But so is the 24 Elmarit for its focal length (it's easier, maybe, to make a sharp 50 than a sharp 28. The amazing thing about all Leica 50s I've used (R and M)--excepting the very early R Lux, which is a wonderful "soft focus" lens overall--is just how d*** good they are edge to edge--and wide open)

 

The really cool part is that with the extended crop you get more effective DOF than you would with a FF sensor (or film), so I've actually shot the Noctilux at f1 from the hip and had a pretty good focus hit rate, and the 75 Lux at f2-ish as well.

 

In the dark!

 

Try doing that with an AF dSLR

[castelletta]

Tim--interesting what Leica's been telling you!

 

I don't think I could reasonably ask a system to be sharper here, could I?

 

 

No, you don't.

In fact I never got that sharpness at f/4 with my coded 35 cron

[magd0328]

Jamie--

Aspheric design doesn't mean "no more spherical aberration" or any other one thing.

Exactly. My 35mm f2 asph, for example, has 7 elements, which is 14 optical surfaces. Only one of those is aspherical. And besides, an aspherical lens surface only elimates spherical aberation completely for a specific pair of image/object distances.

 

 

I'm sure someone can say it better, but aspherical surfaces aren't used to reduce spherical aberration.

--HC

Yes, they are, but see above!

 

Regards

[AlanG]

Tim,

 

At one time aspherical elements had to be hand made and were very expensive. That has changed but "perfect" lenses will never be made. And a perfect camera won't ever exist either.

 

Maybe what we are seeing here is that the M8 and the lenses are really pushing the maximum of what can be expected in a mechanical focusing system that is trying to remain compatible with the legacy lenses. As an example, you and others say some of the newer lenses have less focus shift when stopping down than some of the older lens designs. Along with that comes the thought that a lens such as the 35 1.4 may need to be re-designed now that photographers are examining their images more closely and see its shortcomings. And I don't see how Leica can't be considering making some new more telecentric wide angles for a future body that won't require IR cutting filters.

 

So why not think of an entirely new way to go and elegantly solve most of these problems? The M8 or a future model can stay in existence for those who need backward compatibility and a new camera and lens system can be developed that will get the most out of future sensors, lenses, and digital technology. This could enable Leica to overcome most of the limitations that are causing various problems and will give them a wider path in the future.

 

I think at minimum, the lenses need a way to electronically communicate their f stop and focusing distance to the body. Perhaps an electronic or hybrid electronic/mechanical servo controlled range finder could be developed that eliminates or supplements the cams. You manually focus the lens and it sends a signal for the servo to move the rangefinder prism. Then it would be an easy matter to program in focus compensation for all f stop and distance settings. The potential for digital focus confirmation lights (in the regular and auxiliary viewfinder) and even the possibility for auto focus would be there too. This system would require less technology, speed, and precision than is found in the head actuator mechanism of a hard drive. But this is just one possiblity.

 

As Leica has already programmed various image corrections into its software, there is no reason why they can't take into consideration firmware compensation as part of the overall design consideration when making new lenses. DxO improves my images by mapping lens corrections and applying them in its raw conversion engine. Once the camera communicates focus and f stop info., I don't see why Leica can't be as thorough.

 

Well maybe Leica can squeeze out a little more performance by sticking with the current designs and just making some new lenses. But I think they are just about at the limit for this design and as you and others have demonstrated with all of the focusing tests, the current lenses are not always capable of getting the most out of the M8. And sometimes the M8 is not always capable of getting the most out of the lenses. While compatibility with the legacy lenses satisfies the needs of those who have Leica lenses, that need has been met with the M8 but it is of no consequence to first time Leica buyers.

[ho_co]

Howard, high-precision aspherical lens manufacture is a Leica core competency, so I can understand that it makes it onto their lenses. This is not easy to do, and many other companies which tag ASPH on their lenses don't see the same kind of quality or gains as Leica does.

Fair enough, Carsten. But just to take another example: In the 60's Pentax made a big advertising move by "Super-Multi-Coating" their lenses. Leica had already been multicoating those elements that needed it, and not multicoating those that didn't. Pentax made headlines as if they had discovered a new technology, and Leica just kept building optics, doing what needed to be done.

 

That's what I expect from Leica; I don't want them to start the string of designations that (for example) Nikon has, none of which says anything about image quality. I think Leica will be careful about that, but I want to raise the alarm.

 

--HC

[ho_co]

Tim--weren't your lenses purchased coded? I believe that someone has introduced a red herring into the discussion: You report in your initial post that Germany considers this a trait of the design; it has nothing to do with after-the-fact coding.

 

...when I was talking to the chappie from Solms about it, I said 'what other wide lenses do you have that don't have this behaviour?' and he said 'the new 28 elmarit but you won't get the pop you have in some of your Venice shots from it.' Or words to that effect.

Aha! News that fits. I don't think you had remarked on the 'pop' part of the equation before, Tim, or I had overlooked it.

 

SPECULATION:

In some designs, Leica uses aspherics to create 'anomalous' SA (my term).

 

EXPLANATION:

In the traditional explanation of spherical aberration, masking the edge rays serves to shift the actual focus plane (in object space) in one direction only.

 

However, Jamie, Tim and others have described situations in which the focus plane moves away from the camera as one stops down from wide open, but then after a particular point, the focus plane begins to shift toward the camera with further stopping down.

 

In other words, the focus shift changes direction as one stops down--something completely impossible with traditional optics, and achievable only with aspherical designs. That is, Leica is using aspherics to *stretch* the effects of SA in such a way as to improve image qualities.

 

NOTE 1:

This is theory, but it fits the observed facts. (Definition of a good theory, as I understand.)

 

NOTE 2:

This would be an addition to the earlier post in which I inferred that the reason Leica optics are so good and so highly respected is the same reason we currently see some of these effects (http://www.leica-camera-user.com/digital-forum/21425-focus-question.html#post226600).

 

NOTE 3:

Don't shoot me for this, please. Maybe the upshot is: You're right, the lenses exhibit this behavior. They're designed to do so. They do so on film or on a sensor, but we can enlarge the performance proof immediately when it is presented in digital form. CONCLUSION: Shoot the lens, don't condemn it. If it doesn't work for you, get a different model.

 

NOTE 4:

Do you really believe that the process of changing a lens mount for a coded one would cause the lens to lose its current ability to perform? If so, arent't you saying that either Leica designers or Leica technicians are incapable of precision work?

 

--HC

[Jamie Roberts]

{snipped}

NOTE 4:

Do you really believe that the process of changing a lens mount for a coded one would cause the lens to lose its current ability to perform? If so, arent't you saying that either Leica designers or Leica technicians are incapable of precision work?

--HC

 

Well, design mistakes have been known to happen, yes. Even to Leica.

 

So it's not that Leica is not capable of precision work. But they're human like everyone else. Perhaps the mount on a certain run of coded lenses isn't quite right. It won't take much to put the focal field just out of whack, especially since it *does* move around as you stop down.

 

So I dunno--is that better or worse than saying the lenses themselves are out of alignment?

 

All I do know is that mine works; Tim's doesn't.

 

And we're talking a gross difference; Tim's Lux was out *by several feet* at f4.0 in the center. Mine is razor sharp at f4 in the center (and at the edges too, given that with the crop I've got the sweetest spot of the lens)..

 

One of the main differences in the lenses that seem to work (mine, see above) and those that don't (Tim's and others) does seem to be the coding, which is a function here of age and of a new mount (not just a coding on the old mount.

 

When I heard they were actually replacing the mount, well... it's just another potential variable here.

 

FWIW I will hand code my 35

 

Mine was purchased used. I don't know how old Ser

[ho_co]

Jamie--

I understand your reluctance to possibly mess up your "magic lens."

 

But what we seem to be seeing is that the current lenses shipped new with zebra stripes are misbehaving. That would imply that something has changed in the lens--design, mounting, assembly, something.

 

And that means that those of us who don't already have the 35/1.4 might want to think twice before purchase.

 

The Solms tech apparently told Tim that the 35/1.4 is 'supposed' to behave as his lens does, but still has more 'snap' than the 28/2.8, which doesn't exhibit the focus shift.

 

There's no doubt that Leica can mess up. Apparently a batch of bodies went out with improperly set rangefinders, but that was quickly corrected. And with the 35/1.4, we are seeing case after case of peculiar behavior. I think it's more likely an anomaly in assembly than in the coded mount.

 

My NOTE 4 was more afterthought than anything else. I hope Tim and Bill can compare their lenses and report on their findings.

 

No offense intended to you or anyone else. I'm as interested as anyone else in why some lenses don't tend to exhibit the peculiar behavior, and as I said, until we know what the reason for that is, I don't blame you at all for holding tight to your copy.

 

--HC

[tashley]

Tim--interesting what Leica's been telling you!

 

First, that new 28 Elmarit, in addition to being sharp as tacks and easy to focus, is, if anything, a bit *too* snappy for people. It's a high-contrast sort of lens, and is easily a great bargain!

 

 

 

As usual in the interests of brevity I got sloppy: when I paraphrase him as saying 'pop' what I meant was the particular effect seen in images such as

 

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!

 

from a 50 lux

 

I think this is less to do with outright contrast but more a sort of 3D effect created by a combo of micro-contrast, acuity, and the way that transitions to OOF and bokeh are handled - and I do think that the 28 elmarit is unlikely to match the luxes in this regard

 

Regarding many of the other more recent posts here, I can only hypothesise further:

 

* I am wrong, as are those others with 'the problem' (don't believe it)

* Jamie is wrong (don't believe it)

* Leica and I are right in general but Jamie's lens is an older design and something relating to a change in optics or mount is indicated

* There's a bad batch of new coded mounts around (and yes I do think it possible: to say that Leica, like anyone, can screw up sometimes is not to say that they are incapable of precision manufacture).

 

The jury is still out! But Juryman Reid is about to pronounce...

 

Best to all

 

Tim

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[sean_reid]

As usual in the interests of brevity I got sloppy: when I paraphrase him as saying 'pop' what I meant was the particular effect seen in images such as

 

[ATTACH]36266[/ATTACH]

 

from a 50 lux

 

I think this is less to do with outright contrast but more a sort of 3D effect created by a combo of micro-contrast, acuity, and the way that transitions to OOF and bokeh are handled - and I do think that the 28 elmarit is unlikely to match the luxes in this regard

 

Regarding many of the other more recent posts here, I can only hypothesise further:

 

* I am wrong, as are those others with 'the problem' (don't believe it)

* Jamie is wrong (don't believe it)

* Leica and I are right in general but Jamie's lens is an older design and something relating to a change in optics or mount is indicated

* There's a bad batch of new coded mounts around (and yes I do think it possible: to say that Leica, like anyone, can screw up sometimes is not to say that they are incapable of precision manufacture).

 

The jury is still out! But Juryman Reid is about to pronounce...

 

Best to all

 

Tim

 

Indeed, I am but one member of the jury looking at just one example of each 35 mm lens in the test. They're all here however and I'm testing them now.

 

Cheers,

 

Sean