Apo-Summicron 50/2 Asph. Test!

[semilog]

Originally Posted by LCT

What's dude? Live view focusses behind the lens. Inferior system at real aperture when DoF is wide as i said above.

That makes no sense at all.

[semilog]

Originally Posted by jaapv

Huh?? The Monochrom is about 32 Mp equivalent.

No, it is not.

One may say that manufacturers of Beyer array sensors and Foveon sensors overstate their pixel count (an assessment with which I would agree), but an 18 Mpix monochrome sensor has 18 million photosites and is therefore 18 Mpix -- not 32 Mpix.

To say otherwise is -- and I choose my words here carefully -- blatantly dishonest.

18 million pixels is not equivalent, in any sense, to 32 million pixels. Except, perhaps, in the fevered imagination of marketing flacks.

In electron microscopy, where people are paying millions of dollars apiece for 4k x 4k CCD and CMOS sensors, you would be an object of ridicule if you tried to claim that such a sensor was equivalent to "32 megapixels" of resolution.

[rdeleskie]

Originally Posted by jaapv

Huh?? The Monochrom is about 32 Mp equivalent, the M9 is 18 Mp without AA and the M10 will probably be 24 Mp . Strange ideas.... In case you haven't noticed, Mp resolution is not a quality parameter any more for the last five years...

Whether the M9M is real-world 32MP is highly contentious outside of Leica marketing literature. Even if the M10 is 24MP it will be considerably smaller than the 36MP Nikon D800E which has no AA filter, is already available to purchase and will likely be supplanted by Sony or Canon by the time the M10 ships in quantity.

Admittedly, I know very little about sensor design, but since MP resolution is irrelevant, and other cameras are shipping without AA filters, by what magic is Leica going to produce a FF sensor in the near future that will require the resolution of the new APO Summicron 50mm? Particularly when this doesn't seem to be an issue for Sony, Canon, Nikon, etc.?

[LCT]

Originally Posted by semilog

That makes no sense at all.

Would you mind to elaborate?

[semilog]

Live-view allows direct assessment of focus at 100%. It is immune to misalignment of a rangefinder or an SLR focus screen, to parallax, and to focus shift.

For subjects not dead-center in the image, good live-view implementations allow direct focus, not focus-recompose. For lenses that have flat fields, this is important because focus-recompose results in front-focusing. For lenses that have significant field curvature (e.g. 35 Summilux ASPH FLE), this is important because it allows fine-tuning of which regions are critically sharp. For images where precise control over depth of field is important, it gives you a highly accurate DoF preview.

Now let's consider that 50mm 'con AA again. The MTFs provided by Leica will be for the plane of critical focus. MTFs fall off enormously, especially at higher spatial frequencies, as soon as focus is not critical. This decrease will be larger than the difference between (say) a Planar and a 'cron AA.

Thus, to see a meaningful difference, the 'cron AA must be used critically: tripod and/or short-duration strobe, and exceedingly careful focus.

For that, you want live view -- not a mechanically-coupled rangefinder.

Look at how Phase One does it: with a touch screen that allows you to tap to 100% live view (albeit with a slow refresh rate since they use a CCD) in any image region, to instantly check critical focus. Phase One knows exactly what they are doing here, and what their most critical users demand.

[BobYIL]

With Live-view and in critical situations one can overcome focus-shifting with some lenses by trying to focus at shooting aperture via hi-magnification.

[semilog]

Erwin Puts:

This is a stunning accomplishment, but as is the case with any excellent-quality French cognac, you need to have experience and taste to savor the true quality of the design. There are not many motives in real-world photography with such fine textures and details that the high definition of the lens is required and then the photographer has to have the qualities to record these details faithfully.

In other words, this lens really is all about the sharpness ("fine textures and details"). The lens must be used critically or you simply will not see what it can do versus the 10-times-less-expensive-and-very-nearly-as-good Planar. Puts clearly implies that even with fanatical, perfect technique, the improvements won't be visible except when photographing a minority of suitable subjects.

And that kind of (fanatical) critical use is not what the M body is really optimized for.

Again, the shoddy tripod mount on a loosely-fitting bottom plate is our first clue.

The fact that the improvements are not readily detected by a resonably careful worker like Ming Thein underscores that this lens offers few or no advantages for practical photography in the classical Leica M modality.

Again, it's not at all clear what this "stunning accomplishment" was actually made to do, beyond being a "stunning accomplishment."

[jaapv]

Well, they have the Monochrome, which is a true 18 Mp denser, as opposed to the Bayer-reduced-by- 40% sensors of color cameras. But this lens is not really about resolution per se. For any print viewed at normal distance a 10 Mp Bayer sensor is ample. It is about micro contrast, frequency response, detail separation, color response, things like that. It will show its quality on any decent sensor, as the end result is the combination of lens and sensor and cannot be divided into either lens or sensor. the real world difference between 24 and 36 Mp is minimal btw. There is no real need for a higher resolution sensor for the lens to show its quality anyway. I can testify it is stunning on the M9 as well.

Originally Posted by rdeleskie

Whether the M9M is real-world 32MP is highly contentious outside of Leica marketing literature. Even if the M10 is 24MP it will be considerably smaller than the 36MP Nikon D800E which has no AA filter, is already available to purchase and will likely be supplanted by Sony or Canon by the time the M10 ships in quantity.
Admittedly, I know very little about sensor design, but since MP resolution is irrelevant, and other cameras are shipping without AA filters, by what magic is Leica going to produce a FF sensor in the near future that will require the resolution of the new APO Summicron 50mm? Particularly when this doesn't seem to be an issue for Sony, Canon, Nikon, etc.?

[semilog]

Originally Posted by jaapv

But this lens is not really about resolution per se. It is about micro contrast, frequency response, detail separation, color response, things like that.

1. How may we obtain "microcontrast," (high) "frequency response" or other "things like that" without high spatial resolution?

2. If spatial resolution does not matter, why should it matter whether there is a Beyer mask or antialiasing filter?

3. If the practical differences bewteen the new lens the Planar or the previous Summicron are so marked, they should be evident in (careful, fair) A/B comparisons.

4. Ming Thein has posted his (reasonably careful, apparently pretty fair) A/B comparisons (see link from the very first post in this thread). The differences are pretty minor.

The problem faced by Mr. Karbe is that Mr. Mandler's Summicron IV/V and the ZM Planar are already very, very good lenses. They are so good that superior technique is already required to extract what they can deliver.

[semilog]

Erwin Puts, again:

Some future M-cameras will incorporate a CMOS sensor and by definition may offer live view functionality. This will amount to a huge change in the rangefinder style of photography. For the first time in 80 years the Leica M photographer will not have to rely on the opto-mechanical linkage between finder mechanism and lens mount, but can view the image in the real optical image plane.

It doesn't take a weatherman to know which way the wind blows.

[MCTuomey]

Originally Posted by semilog

Again, it's not at all clear what this "stunning accomplishment" was actually made to do, beyond being a "stunning accomplishment."

You want to convince people that the accomplishment isn't practically worthwhile for M-camera users because they aren't sufficiently technically competent to make use of the lens?

And do you persist in this line of reasoning when it comes to refining bicycle frames and components? Sports car drive trains? Sports optics? That marginal improvements are of no real benefit because riders/drivers/users aren't skilled enough to make use of or need them?

Why would you presume anything about skills others bring to their avocation? Maybe I have a very fine tripod, ballhead, camera bracket, lots of experience and can avail myself of the incremental benefit. Maybe not in my case. But some folks will. And there is something in me that appreciates that Karbe designed and led this lens into production for those who can afford to use it.

[MikeL]

I'm constantly amazed at how much people care about a company (like or dislike) with all of their posts on a company's products. It's like an obsession.

[rdeleskie]

At the risk of fanning the flames, here's a quote from Ken Rockwell from some time ago that, ahem, seems to say it all:

"Because every other LEICA SUMMICRON f/2 lens has been updated to ASPH (28mm in 2000, 35mm in 1996, 75mm in 2005 and 90mm in 1998), a LEICA SUMMICRON-M 50mm f/2 ASPH seems like a logical update, however LEICA tells me this is unlikely simply because the performance can't be much improved, but the price would become about the same as the f/1.4, so why bother? Hey, why bother to send men to the Moon? We bother precisely because we are men and continuously must explore and improve."

Blind pig, acorn, and all that.

[Nando]

I don't care about digitalia. I want to see this lens truly tested with Adox CMS-20 or SPUR Orthopan UR.

[semilog]

Originally Posted by MCTuomey

You want to convince people that the accomplishment isn't practically worthwhile for M-camera users because they aren't sufficiently technically competent to make use of the lens?

I didn't say that.

Erwin Puts came a lot closer to saying that than I did.

What I said is that extracting that kind of performance from an M-body is antithetical to the Barnack ideal, and that (current) M bodies are far from ideally designed for taking the kind of picture that would highlight off the unusual qualities of the 'cron AA.

Since you gave the appearance of caring what I thought.

[LCT]

Originally Posted by semilog

Live-view allows direct assessment of focus at 100%. It is immune to misalignment of a rangefinder or an SLR focus screen, to parallax, and to focus shift...

Well i won't repeat myself at nauseam. The problem is DoF. The wider the DoF the more difficullt focusing proves to be with anything else than rangefinders.

[LCT]

Originally Posted by semilog

...It doesn't take a weatherman to know which way the wind blows.

Indeed. True rangefinder + live view and accessory EVF. Best of both worlds.

[LCT]

Originally Posted by semilog

... extracting that kind of performance from an M-body is antithetical to the Barnack ideal, and that (current) M bodies are far from ideally designed for taking the kind of picture that would highlight off the unusual qualities of the 'cron AA...

Branack's ideal was to take the best possible photos in the smallest package. Exactly what Leica are trying to achieve with the smallest FF bodies and the best 50 ever made. According to photogs having used it intensively, the Apo Summicron delivers superb photos with both M9 and MM. I have no experience with this lens though. I have therefore no reason to rain on others' parade so far.

[jaapv]

It is amazing how expert people who haven’t even seen the lens, haven’t spoken to Mr. Karbe and other Leica developers and haven’t seen a single printed image are....
Once again: Leica has set out to build a prestige lens, to show off what the company can do and they have succeeded. It was their intention from the start to make this a connoisseur lens. If you want an excellent lens for Joe Average, there is the classic Summicron. That is all there is to it. As for Barnack’s ideal, have you noticed lately that Ford cars are available in other colours than black? Leica of 2012 is not the same company as Leica of 1920, fortunately. if that were the case they would be embalmed and stuck in time.

Originally Posted by semilog

1. How may we obtain "microcontrast," (high) "frequency response" or other "things like that" without igh spatital resolution?

2. If spatial resolution does not matter, why should it matter whether there is a Beyer mask or antialiasing filter?

3. If the practical differences bewteen the new lens the Planar or the previous Summicron are so marked, they should be evident in (careful, fair) A/B comparisons.

4. Ming Thein has posted his (reasonably careful, apparently pretty fair) A/B comparisons (see link from the very first post in this thread). The differences are pretty minor.

The problem faced by Mr. Karbe is that Mr. Mandler's Summicron IV/V and the ZM Planar are already very, very good lenses. They are so good that superior technique is already required to extract what they can deliver.

[semilog]

Originally Posted by jaapv

It is amazing how expert people who haven’t even seen the lens, haven’t spoken to Mr. Karbe and other Leica developers and haven’t seen a single printed image by the lens are....

I don't claim expertise, although a major portion of my professional work involves quantitative applications of high-performance scientific imaging, and I have had to do comparisons of microscope objectives and CCD sensors that cost a good deal more than the 'cron AA... or for that matter, the S2.

At any rate, if you've done careful side-by-side comparisons of the Planar and the 'cron AA under real-world shooting conditions, you're not sharing the data. You also seem to take a certain amount of pride in throwing around quasi-technical language ("microcontrast," "frequency response") without defining what the terms that you are using actually mean.

Look, I'm a scientist. If you want to change my mind, show me the results of well-controlled experiments, and tell me how you did the work. Ming Thein has posted A/B comparisons. You don't seem to like them, but you have not actually raised any good criticisms of the methodology or presented your own or anyone else's comparisons that address the presumed shortcomings of that work.

If you want to persuade others there is a difference that makes a difference, don't merely assert. Show it.

[Nigel Meaby]

Originally Posted by jaapv

It is amazing how expert people who haven’t even seen the lens, haven’t spoken to Mr. Karbe and other Leica developers and haven’t seen a single printed image are....
Once again: Leica has set out to build a prestige lens, to show off what the company can do and they have succeeded. It was their intention from the start to make this a connoisseur lens. If you want an excellent lens for Joe Average, there is the classic Summicron. That is all there is to it. As for Barnack’s ideal, have you noticed lately that Ford cars are available in other colours than black? Leica of 2012 is not the same company as Leica of 1920, fortunately. if that were the case they would be embalmed and stuck in time.

Extraordinary statement! You've just insulted everyone who has used any version Summicron up until the "Apo" was introduced! Whether you intend it to or not. Remarks like that come across as a little Elitist and Snobbish. You don't have to be a "Connoisseur" to think nothing of spending that amount of money on the "Apo" You can be a "Connoisseur" and choose to own a Zeiss Planar because you like the results you get with it irrespective of how much it costs. Choosing one over the other doesn't make you "Joe Average" You can have the most expensive equipment available. If you don't have the talent your pictures will be perhaps "average" at best.
Looking at the objective comparison work done by Ming Thein, a very accomplished Professional Photographer, I have to say his views hold a lot more credibility than someone who appears to defend Leica's every move.

[semilog]

Originally Posted by jaapv

have you noticed lately that Ford cars are available in other colours than black? Leica of 2012 is not the same company as Leica of 1920, fortunately. if that were the case they would be embalmed and stuck in time.

Actually, that's precisely why I purchased an X-Pro1 instead of an M9: it has imaging performance as high and likely higher than an M9 with a 50/1.4 Summilux ASPH (that is Erwin Puts's assessment, among many others), in a more modern package and at about half the weight. And it absolutely crushes the M9 in low light performance. It has live view, and its lenses focus far closer than M lenses.... all at about one fourth the price. That will provide some solace if the camera is stolen, or stomped by hippos.

I will keep my M6 around for film. It's a lovely camera and I have enjoyed using it for almost 15 years. It reminds me of the first camera that was ever placed in my hands, my father's M3 with a DR Summicron mounted. I do understand the appeal of the Leica M system, and I well understand both the system's significant technical strengths and its considerable limitations.

[sandermarijn]

I would like to see this new Leica AA 50 compared to the Contax G 45mm. The current 50 Summicron could not impress me after the Contax 45 had set a standard. The Contax is that good a lens (to my eyes).

BTW, I rate the Contax 45 much higher also than the ZM 50 (shadow detail, micro contrast).

[Steve Ash]

Semilog,

I looked at your portfolio but could not find one image that is technically difficult to shoot with any M as they are right now.

The only situation where I loose a shot with the RF once in a while is if someone approaches me at high speed with no time to prefocus. But I doubt liveview would be helpful to me in that kind of situation.

Regards,
Steve

PS:

BTW, I like your signature:
There are two kinds of photographers:
those who are interested in what a particular camera can't do,
and those who are interested in what it can do.

[semilog]

Originally Posted by Steve Ash

Semilog,

I looked at your portfolio but could not find one image that is technically difficult to shoot with any M as they are right now.

I agree. The M is a good tool for most of my photography, which is mainly handheld and at moderate distances. But that work is also to some extent shaped by the difficulty of using an RF for certain types of work -- particularly close-up stuff. So I tend to avoid those sorts of photos. The X-Pro is already changing that.

Kirk Tuck has a really nice piece on the benefits of EVFs and live view.