Guest post: Why I Am Leaving Leica for Micro Four Thirds


Today's guest post is by J Shin:

This is the morning of Christmas, 2012, and I am facing a moral crisis.

For years, I have not made a major photo equipment purchase. My partner and I have a deal where my income needs to meet a certain benchmark before I can purchase another one. I have been able to make “minor” purchases under $200, and have been playing with an Olympus E-PM1, bought refurbished at $170, and a $20 Photodiox adapter for my Leica R lenses. I just got a used Panasonic 14/2.5, for $160, with which I have been very happy. So happy, in fact, that, now that I did meet my benchmark (yes!), and am free to buy any one major equipment I want, I am contemplating getting the well-reviewed Olympus 75/1.8 lens, to fill a gap in my Leica lens line up, and to take advantage of this new-fangled thing they call “autofocus”, which I am just beginning to appreciate, thanks to my failing eye sight. It was either that or the NEX-VG900, a $3,300 full-frame madness.

Then, at 6 am Christmas morning, I wake up from a groggy dream, and innocently reach for my laptop to check the e-mails. There, like a misguided Christmas present from hell, is a notification from eBay.

A Summilux-R 1.4/35. For $2,600.

This is a lens I have coveted even before I bought my very first Leica, the R3MOT, with a Summicron-R 2/50. It is the very object I was preparing to buy when my partner put a clamp on my procurement process 4 years ago. It was going for about $1,200 then.

Now you are lucky to see it at $3,000. A “bargain” condition piece is occasionally available for about $2,500, with questionable optical flaws.

Here is a very MINT piece for $2,600.

I can afford it.

And then the soul-searching begins. This isn’t just about the one piece of equipment I am allowed to buy, but a question of what, exactly, I would rather spend my money on from here on, the entire future of my camera-buying career. Do I stay with the long-defunct, hyperinflated R system, or do I jump ship for a new system? Say, Micro Four Thirds?

I have also been looking at the Summilux-R 1.4/80, whose price has stayed at around $1,800 for years. If Leica is ever to produce a non-rangefinder mirrorless full-frame interchangeable-lens camera with an electronic viewfinder, the mythical beast I have been waiting for ever since they discontinued the R line; or if Sony or Fuji or someone is to produce the first non-camcorder full-frame mirrorless; or if I am to toss in the towel and get an EOS; or, better yet, get a Nikon or a Sony and drop a lot of money on Leitax... Under any of these circumstances, the Summilux-R 1.4/80 is surely preferable to the Olympus 75/1.8.

The moral crisis is this: I am no longer so sure! Me! The Leica apologist, who has held on to manual focus, manual metering, and sometimes manual film winding, no longer believe that either of the Summilux is a no-brainer!

$2,600! At that price, it is going to be sold before we are done opening the presents!

Why am I not clicking on the Buy It Now button before anyone else in the house wakes up?

Suddenly, I find myself believing that I am ready to leave the faith.

And here is the final offer from the ancestral spirits of the Leica faith. The Summilux-R 1.4/35, in mint condition, for $2,600.

I take it, and I stay a faithful. I turn it down, and I am excommunicated FOREVER. Well, until I could afford to come back, which amounts to the same thing.

[Expletive here.]

What happened?

It all began, of course, when Leica’s “R solution” turned out or was revised to consist of an EVF on an M. How much does an M cost, now? $7,000? There is no way I can afford that, and, even if I could, there is no way I would ever leave the house with it, or roam the streets of our town with it, never mind streets of any city or country where there might be anything interesting to photograph. Insurance I can afford. Bodily harm, I’m not so hot about. Leaving the house with $2,000 worth of battered-up R stuff makes me nervous enough that I can’t use the bathroom freely, unless there is a coathook that is strong enough to withstand the weight of a small car, and is located where a passer-by cannot reach it. I’m scared to leave my office at night with $100 to deposit, for earth’s sake.

$12,000 worth of M stuff? No way.

For that matter, could I go for a 6D or a D600? $3,800 for a new full-frame camera and a 85/1.4? $3,000 if I get a Sigma lens instead?

That’s going to be about the price of the full-frame NEX, if it ever comes out, isn’t it? That’s what they want for the NEX-VG900, after all.

But, there is a reason why I only carry the R3 when traveling. The R8 stays home, unless we are going to a “safe” environment, like a friend’s kid’s birthday; even then, I do not put it down anywhere.

I’m a sissy.

And there is this bit that I see a lot on the Internet: “The best camera is the one that you have with you.” If I cannot see myself toting it around, it’s not worth buying it.

The E-PM1, I don’t tote around everywhere, but I have been toting it around a lot. In the two months I have had it, I have shot five hundred frames, about a third of them usable, more than I have shot on the R8 over the last three years.

Because I had it with me.

There was a time when used Leica R lenses were sold at a price on par with a new Canon or Nikon. Granted, the R lenses were 20-odd years older, but they still performed, and compared well to the new lenses. This has changed. The Summilux-R 1.4/35, at absolute cheapest, goes for $2,400. This is a lens that can squeeze out 2000 lw/ih in the center when wide open, but only for a few millimeters in the center, and delivers corner resolution of only 500-1000 lw/ih, depending on where exactly you measure it. Contrast that to the Nikon and the Canon versions, which can, at $1,600, approach 1500 lw/ih in the center, and the new, fabulous Sigma version, which, at $900, approaches 2000 lw/ih in the center and 1500 lw/ih in the corner. Wide open. (These and other figures are based on’s MTF50 measurements; Leica R figures are derived from Leica’s MTF graphs. Other sites’ measurements tend to be higher, but paints the same relative picture.)


Even the much-maligned Olympus 17/1.8, at $500, gets about 1400 lw/ih in the center and 1100 lw/ih in the corner. Not too shabby. And it is absolutely tiny compared to the full-frame cousins. The Panasonic 20/1.7, at $350, gets 1650/1200 center/corner lw/ih, and the Olympus 75/1.8, at $900, gets 1750/1250 center/corner lw/ih.

(By the way, with Olympus in-body image stabilization, you use up to 2 stops slower shutter. Even with wide-angle lenses, which usually don’t come with optical image stabilization. So, with an f/1.8 lens, you get the exposure value of an f/0.9 lens and, thanks to the crop factor, the depth of field of an f/3.6 lens. To me, it’s the perfect combination. Depth of field at this aperture increases the perception of sharpness and contrast, while sacrificing little of the bo-keh; bo-keh, after all, is NOT determined by depth of field, contrary to a common misperception.)

I am only comparing measurements at maximum aperture because that’s what I need it for. If you intend to use it stopped-down, the Micro Four Thirds lenses fall further behind, not because the Micro Four Thirds lenses are inferior, but because their lower-resolution sensors have natural limit to the resolutions one can measure. (This is a subject for another day.) In fact, many of these cross-platform measurements are fraught with problems, because they are often done on very different bodies with very different characteristics. Furthermore, resolution isn’t everything; Leica lenses are unmatched for their bo-keh, color rendition, and overall contrast and clarity. These characteristics are functions of how the lens handles low-resolution and/or out-of-focus features, and are difficult to measure objectively; for some reason, German engineers are quite adept at things that cannot be measured objectively, something that has aggravated Japanese engineers for many decades.

I am also leaving out the current range of Leica M or S lenses. Their amazing performances are proof that unimaginable image quality, objectively or subjectively measured, can be achieved with current human technology, should one be willing to spare no expense.

Still, you get the point. One could argue that one gets what one pays for, but there comes a point of diminishing marginal return on expenditure. Leica lenses, current and old, have come to represent exponential expenditure for the limits of performance. Micro Four Thirds represents more than good enough performance for reasonable expenditure.

And, for my purposes, 1600 lw/ih is probably a reasonable center resolution anyway. I hardly ever print anything larger than 8”x12”, and 1600 lw/ih still gives me 200 dpi, more than necessary for a decent print. By comparison, Apple’s Retina displays are 220+ dpi on laptops and 320 dpi on phones, and “art books” are usually printed at 200 dpi. “Circle of confusion” used to determine depth of field on 35mm film is 0.03 mm, which translates to 800 lw/ih, or 100 dpi. Anything above that, I won’t notice, unless I take off my contacts and stand very close to the image. Our bedroom walls, in fact, are lined with macro shots with miniscule depth of field that nevertheless appear sharply in focus from near to the far. If I were to pixel-peep one of these shots, I will find that almost all of the image is, in fact, out of focus, but, even looking up close, that is not apparent in the print. (By the way, these were printed from slides, which could get 50 lp/mm, or 2400 lw/ih, which is slightly below the resolution of Micro Four Thirds sensors, although the Bayer filter does lower the perceived resolution of the sensor by about 1/3.)

But, wait, you might say, why are we comparing the top-of-the-line film system from the antiquities with the middle-of-the-line interchangeable digital system of today? Shouldn’t we be singing the praises of the current top-of-the-line?

Of course. D800, 5D, a99, the new M, these are all fantastic cameras that can achieve image quality far surpassing anything from the film camera era. If money is no object, then there is no doubt what is better and what is worse, and the Micro Four Thirds system is definitely not on the top of the heap there. But, again, where is the point of diminishing marginal return on expenditure?

That point is when the image starts being “good enough.”

I will bluntly say, the lenses for the classic SLR and rangefinder cameras offered by Canon, Nikon, Olympus, Minolta, Rollei, Pentax, Contax, and the like rarely came close to Leica M or R quality. There was a time when you could pick up a Yashica and a Leica, shoot a few rolls of film, and place them side-by-side, and the winner was readily apparent. Yes, readily.

No more. All the cameras and lenses now simply exceed the capacity of a semi-trained human eye to tell the difference. Have you stared and stared at the comparison shots between the X1 and the X100 to see if you could really decide which one was better, only to realize that the real difference was in the point of focus, or the JPEG algorithm, or exposure, or some such? Sure, the Leica images have more contrast, but who’s going to be able to tell that you ratcheted up the contrast in Lightroom? Sharpened it and de-noised it? Corrected the distortion? If you have to run the images through a computer to tell the difference, there is no real difference that matters.

Even if we consider the one black mark against most Micro Four Thirds bodies: dark noise. You don’t have to magnify the image that much to see it. It’s there. It is annoying. It gets worse the moment you try to lighten up the image. Surely, 10 bits of dynamic range cannot possibly be good enough?!

Actually, though, I have almost always shot slides, and, with slides, you get used to the fact that either the image is right or it isn’t. Unlike with negatives, there is not much latitude to play with if the shadow is too blue or the face is too dark. It is right, or it isn’t. From this point of view, 12MP, 10-bit images are either good enough when they come out of the camera, as they often are, or they aren’t. Lighten it up a stop. Correct the color balance. That’s about it. Also, with film, you expect to see a little grain at ISO 400; it is part of the look. Because of this, for me, the point of diminishing return still starts right above the $170 I paid for the E-PM1.

Also consider the fact that these digital bodies will last 10 years if one is lucky. My R3 is almost 35 years old. I paid about $250 for it in 1991, and you can pick up a very good copy for about $150 now. Apart from having lost 1/1000 and M sync, which I never use, it’s still clicking away without problem. It is the only body I trust sub-freezing, and $200 will bring it back to perfection. What you are paying for when buying a new digital body, then, is what you would have spent on film. Back when everyone shot slides, a roll cost about $10 per roll to purchase and process, or about 25 cents per frame. With digital cameras, people seem to shoot about four times what they would have shot on film, so a $2,000 body is worth about 32,000 frames. If you shoot less than that over 10 years, or about 9 shots a day, you wasted your money. For an $8,000 body, that is 12,800 frames over 10 years, or 35 shots every day for 10 years. Not impossible, but not very feasible for an amateur with a day job and a family, unless you are the kind of person who only takes 50-shot panoramas, or 20-shot sequences at sports events, etc., which, of course, people tend not to do with an M. Or maybe you just don’t know how to set things up to get that one right shot, like someone trained with view cameras quickly learn to do.

Pros, of course, have very different needs, and very different calculus of expense vs. cost.

Not being a pro, my needs are simple, I realize, and my needs can be met with a cute little bargain camera with teeny lenses that deliver surprisingly good results.

Surprising, that is, only because I am used to lugging around a small cannon to get a decent shot. My Apo-Telyt-R 2.8/280, which I love dearly, is so heavy that I have never been able to hand-hold it steady enough to take advantage of its legendary 100 lp/mm resolution. Eat your heart out, that’s 4800 lw/ih. Even cropped on a Micro Four Thirds body, that is 2700 lw/ih. That’s what we call “pushing the limits of performance”, 1984-style.

But, if I can’t hold it still, I can’t take good pictures with it, diffraction-limited or not. I can barely handle the tripod I need to use it with, and tripods are not very useful when taking shots of skittish animals. I have an elaborate body frame to keep it steady, but, some days, I cannot even get the lens cap off of it with one hand.

So, for the first time in my life, I’m quite seriously considering a zoom lens. Yes, the unmentionable Panasonic 100-300/4-5.6.

Whopping 1350 lw/ih for only $500. Who could ask for anything more? Not me.

This entry was posted in LR Guest posts. Bookmark the permalink. Both comments and trackbacks are currently closed.
  • Thinkinginpictures

    I’d love to see this article updated. I had read it about two years ago and it is just as good now as it was then. I’d like to see this logic expanded in the face of…yet again…another push for ultra high res cameras such as the A7RII, RX1R II and 5DS. Great article.

  • Back to top