Lies, Damn Lies, and Bias

Mike Johnston points out a Michael Reichmann review of the Panasonic L1.

Here is a one line summary of the review: Michael Reichmann says he should love Panasonic and the L1 design in particular and then tears the camera a new asshole.

I’m not surprised about the review. Reichmann’s “full frame” snobbery has been peeking through for a few years now and is now in full bloom. The fact that he qualifies his review by stating that he has a Panasonic bias is laughable.

Should I begin this review of the reviews and comments by stating that I have a Nikon bias? How is that relevant? If I wax glowingly of the the L1, does that make it more valid because I’m a Nikon guy? Typical fallacy.

The “4/3 Myth” is a myth

The reviewer’s first attack is on the 4/3 format. As regular readers know, I’ve done my share of denigrating the 4/3 format.

But Reichmann shows what a fucktard he is when he calls the size advantage “a myth.” He feels he blows a hole in my optical cost principle by comparing the Canon EF 24-105mm f/4L IS USM (on a 35mm body) to the Leica D Vario-Elmarit 14-50mm f/2.8-3.5 (on the L1), which even he admits is “a fair bit lighter.” See anything wrong here kids?

Hmm, let’s see.

1. He comparing an f/2.8-3.5 lens to an f/4 lens! How are these the same in any way? Why don’t we compare this to a 35mm kit lens that goes that wide? Oh right! That’s because that shit doesn’t exist because nobody would be stupid enough to make a kit lens that heavy!
2. The 14-50mm is 490g and the 24-105mm is 670g. That “fair bit” is almost half a pound!
3. Hmm, 24-105mm sounds familiar. Is Michael talking about this piece of shit? Why yes, I believe he is.
4. He neglected to mention that the flaming digital turd that is the 24-105mm costs you $1300. The 14-50mm is $1000 after the Leica markup! (As in, this is would be a $800 Zuiko, a $600 Canon/Nikon, or a $300 Tamron/Sigma/Tokina.)
5. The difference will get more pronounced when Leica 14-150mm f/3.5-4.6 OIS is introduced next year. For reference point you’ll be comparing that to the EF 28-300mm f/3.5-5.6 L IS USM which weighs 1600g and sets you back a mere $2300 ($3700 list). I wonder which will win in the price/weight category? Hint: the Nikon equivalent is my 18-200mm f/3.5-4.5G VR which cost me $650 and weighs 600g. Yes, I realize that the build quality of the 18-200mm isn’t as good as the 28-300mm. But I’ll stack up image quality any day of the week, plus when I’m around town or on a quick hike, I’m taking photos with mine; you left yours at home because it weighs too much (that’s assuming you had the $5k+ to buy it and a Canon 5D).

The take home point here is that the “4/3 Myth” is a myth. The disadvantage of 4/3 (2x multiplier) are more noise (or less megapixels), more depth-of-field (less background isolation), and diffraction effects at small apertures kick in sooner. The advantages of 4/3 (2x multiplier) are: smaller, cheaper (for a given quality), lighter and closer focusing. Oh, yeah, did I forget to mention that last one? Hmm, it appears Reichmann “overlooked” that one also. You can take these facts to the bank—literally, given how expensive medium format digital (16/3) relative to 35mm digital photography (8/3) relative to APS-C (~6/3) and 4/3 relative to digital point and shoot (1/3 or 2/3 sensor).

The “two points = a line” myth

I am guilty of this all the time. The idea is you throw two data points and draw a line, or you throw down three points draw an exponential curve.

For instance, if you are trying to convince a VC that your Web 2.0 company is worth investing you say something like, “Well when we launched we had 6 users, a month ago we had 30 users, and today we have 1000 users. Clearly we’ve reached a tipping point.” (Then a couple months later, you find yourself in the dead pool.)

Why?

It’s because you’re a fucking moron and believed your own bullshit you spewed.

Ever thought about why Moore’s law is so interesting? It’s because that shit almost never happens!

Take this one:

“In the early 2000’s, when the 4/3 format was being conceptualized and designed, sensors were expensive to make, because the larger the physical size the lower the yield. Constantly evolving fabrication technology has addressed this and we have recently seen the price of 8MP cameras with reduced-frame 35mm sensors drop to below $500, and full-frame 35mm sensors in camera that retail for under $3,000. Clearly sensor build cost is no longer as compelling a factor as it was in the early years of DSLR development.”

What Reichmann overlooks was the amazing development of CMOS image sensors good enough to compete with CCD. That isn’t an iterative development he implies, that’s a disruptive one!

But do you see the problem? The yield advantages of CMOS fabrication have already been reaped by the (very smart) move by Canon (and later Sony). You won’t be able to exploit Moore’s Law here because that talks about getting more circuits in the same package size. In photography your package size isn’t going to get smaller—it’s going to get bigger. Over time, serious photographers will demand larger sensors.

(Many photographers seem to be compensating for something: large lenses, comparing package sizes, and the like.)

I’m going to go out on a limb here and say that you will not see $500 dSLR bodies based on a 35mm CMOS sensor anytime in this decade. At best, we can expect them to get down to around $1500 (which might happen fairly quickly if you see Nikon, Pentax, or Sony enter this area). Today, we already see a $300 APS-C dSLR body (the Pentax *st DL).

The rest of the review

I actually like the rest of the review:

1. I actually wondered why Panasonic didn’t make the LCD is articulated like on the Olympus-330. (This is bothersome because the Leica lens loses one mode OIS when on the Olympus 330, besides the 330 is a very ugly camera.)
2. There are some compromises in handling when doing Live preview on an essentially SLR camera. OTOH, I don’t think the advantages of this (for street photography) are emphasized enough given the sort of comments I see on T.O.P.
3. I don’t know if I buy that the Venus III engine is so great as Michael makes it out to be. There are definitely some improvements in the technology in the sensor, etc. and some severe deficiencies in Olympus and Panasonic’s image processing. While they are really starting to “get it”, I’d be surprised if they reached Nikon and Canon.
4. Viewfinder limitations of the 4/3 system coupled with the use of mirrors in order to support the design (and also, the LiveMOS, which Michael neglects to mention) are an issue with this camera and many others. I’m happy to note the new Nikon D80 has the same eyepoint and pentaprism system as the Nikon D200, putting on par with the Pentax’s *st D series and way above Canon Rebel, Sony Alpha, Panasonic/Olympus 4-3, and early model Nikons (D200, D70(s), D50).
5. Dynamic buffering. Ouch! Remember when the Canon 300D performed this way? I wonder what Michael said about that? Oh that’s right. Nothing! In his defense, there wasn’t the Nikon D70 to put it to shame at the time. But it does serve to emphasize, once again, my old maxim: When Canon thinks nobody is looking, they screw you.

My conclusion

I probably would have approached it differently.

First, I wouldn’t have really harped on the cost relative to the Rebel XT + 24-105mm (not even the right zoom range when you account for the 1.6x APS-C sensor i the Rebel). I would have compared it to the Sony R1 and the Nikon D50 + 18-200mm f/3.5-5.6G VR which are more camera but both cost about half the Panasonic L1 kit.

But that shows you the entire review misses the point. It isn’t just about the Leica lens, doing so denigrates the work that the Panasonic engineers has done (who developed the Olympus E-330 sensor and co-developed the LiveMOS system they both use). Here is a challenge. Go to the store and pick up an Olympus E-330 and a Panasonic L1. Tell me which one you’d purchase if price is not an issue and you could buy the Leica lens with the E-330. I bet you’ll still feel the siren call of the L1 despite all it’s shortcomings.

Why?

Camera feel.

This camera screams “CAMERA!” It does so in a way different than Nikon D200 (Nikon F-series appeal) or the Canon 5D (engineering appeal). The manual aperture controls, the square body, even the silly shutter placement. The camera asks “Remember when rangefinders were cool?” And the camera whispers back to you, “I do.”

The fact that Panasonic got this right on the first try is nothing short of amazing. Companies like Sony (Alpha A100 descendent from Konica Minolta) and Canon (Digital Rebel series descended from the prestigious Canon EF system) still haven’t figured that one out. They both feel cheap in all the wrong places.

Live previews for street photography, 4/3 for instant macros… The L1 is a camera you might just carry with you before you go out today. Henri Cartier-Breeson reborn.

Final harshness

Actually, I respect Michael Reichmann a lot. His photos are nice and his understanding of digital photography is second to none—You won’t catch him with the sort of foot-in-mouth slip ups that Ken Rockwell regularly does.

The difference here is that while Ken has very emphatically declared his Nikon bias, Michael, like Bob Atkins, has an unstated Canon bias that comes from their “grumpy-old-man” love of 35mm and the reality that they have a lot more money to blow on photography than even me (and I have a pretty large disposable income). Unlike him, most of us can’t afford to purchase a Hasselblad H2D 39 with our own money, let alone 35mm dSLR bodies (and the expensive lenses that go along with them).

(Interesting story: When I was at Keeble and Schuchat a couple months ago, a guy there tried to sell me an H2D 39. I just about laughed in his face (no, correction, I did laugh in his face). He told me Bill Atkinson, who was mentioned in the article I linked, had bought his H2D 39 from there just a few months earlier.)

Michael forgets this sometimes and, in his zeal, he becomes so full of shit, it slips out when he opens his mouth.

The beginning of the Panasonic Lumix L1 review is one such example.