Notes and Rumors

Note that Ctein, in the post three down from this one, isn't saying that 400 MP cameras would be marketable. He's only talking about the general ballpark of where perfection lies, in technology extrapolated from today's.

One thing you might want to be on the lookout for in terms of the evolution of today's technology would be the first reviews of the new Fuji F200 EXR. Although this looks to be a bog-standard run-of-the-mill point-and-shoot, actually it's anything but. It's the first camera to feature Fuji's new EXR sensor technology [WARNING: link is a .PDF]. What's EXR? Basically, it's a sensor array optimized for pixel-binning, so it can alternate between fine detail capture in good light and pixel-binning for high sensitivity with less noise in low light. From the early samples, it looks pretty good up to ISO 800, which is a bit better than the typical. There's also happy talk of a combination mode, sort of an in-camera HDR mode implemented via software; we'll believe that when we see it, although one early sample is impressive enough. Word is that full production cameras are now abroad in the world, so we're in the crow's nest scanning the horizon with ye old spyglass, alert for the initial reviews. (Because it's been a while since I've run up to kick the pocket-digicam football only to have Lucy snatch it away at the very last second.)

The Canon 5D Mark II is continuing to meet a mixed, not to say rocky reception. Many buyers are "ecstatic" about the camera (why do I keep encountering that word from new 5D Mark II owners?), but Michael Reichmann reports a whopping 25% failure rate among the Canon 5D Mark II's on his just-completed Antarctica workshop trip (the loaner from CPS failed, even), and Canon-shooting fashion photographer Ron Purdy has thrown in the towel on his.

A little side-note to Ron's report of banding problems: I have a certain print that has a relatively small amount of shadow area, but pretty bad banding within those areas. When I show the print to people, most of them don't notice the banding until I point it out to them. And it's pretty flagrant. I'm just sayin'.

Rumors have been circulating that Panasonic is about to discontinue the popular LX3. 1001 Noisy Cameras is reporting that this is not true: the camera is in a "deep back order" situation (that's a quote of a paraphrase of an unnamed source, so git'cher caveats on), but is not discontinued. (Want my completely baseless guess? You'll see a refreshed LX4 at a significantly higher price by summer, as Panasonic strives to cash in on this camera's unexpectedly robust popularity.)

Meanwhile, back at the Waukeshavian Ranchhouse, the Canon G10 is now the all-time best-selling camera. More TOP readers have bought G10s through our links than any other camera. Note that we are not talkin' about huge numbers here.

Rumor has it that my #1 all-time favorite photo book is about to be reprinted again—I spoke to the distributor a few days ago. It's never been particularly rare or valuable, but it's been a while since you could buy it new. Word is that it'll be out in a couple of weeks. I'm on the case. Like the hawk.

What would a post about rumors be without fresh rumors of Leica's demise? This time, Andreas Kaufmann is said to be deep in talks to sell the company to Panasonic (why? Old joke, modified: want to know a surefire way to make a small fortune? Answer: start with a large one, then buy a struggling camera company). Seems to me I learned in Logic 101 that if you predict for long enough that something bad will happen, sooner or later you're bound to be right. Keep predicting rain, and sooner or later you'll be right. Repeatedly predict a future terrorist attack on U.S. soil, and sooner or later you'll look like Dick Nostradamus. Foretell a downturn in the market, and sooner or later...you get the point. I've been reading about the demise of Leica since the 'seventies. Nothing's forever, and at some point before the sun explodes Leica will probably cease trading. I suggest people talk about it then.

Okay, now back to the first topic. The point at which the march of DSLR resolution ceases will most likely be determined not by the technology reaching perfection, but by camera buyers by and large declining to spend actual money on further improvements. Consider just one perhaps chilling possible scenario. Say a huge conglomerate with a fledgling DSLR division came out with a really nice camera with ~25 megapixels, but it sold poorly. Then a seasoned mainstream camera company came out with a camera with the same number of megapixels, but priced it at, oh, some unrealistically stratospheric price, achieving predictably low sales as a result. Then say a nearly-as-high-resolution camera by the second company's main competitor, despite euphoric initial popularity, had some, er, nagging quality and operational irregularities, which perhaps causes the sales curve to tail off a bit too soon (this hasn't happened, I hasten to add. It just might). And say that while all this is happening, the world economy tanks alarmingly, casting a freezing pall on camera sales in general and high-end camera sales in particular. And say all these factors, in concert, effectively put the old kibosh on the development of further high-megapixel outliers; say nobody minds all that much, except a few fanatics; and say the standard settles down at, say, the 12–18MP range for the most part, which turns out to be plenty for most of the people actually willing to part with cash. When that happens, we'll reach the marketing analogue of the point Ctein describes at which further improvements are no longer detectable: the point at which further improvements no longer inspire additional sales.

I'm not saying it will happen this time, as in now, but I predict it will happen...sooner or later. I'm just sayin'.*

Mike

*Don't you wish I'd stop sayin' that?

Send this post to a friend

Featured Comment by Paul: "Isn't the possibility of failure part of the price of admission for early adopters? Especially as the complexity increases? We accept/expect glitches in our computers, so we have auto-updaters; can it be that far away for camera manufacturers to have something comparable for their onboard software?

"Anyway, I think that on the high-end of the scale, reputations hinge more on how well manufacturers look after their customers, and how quickly they can fix problems.

"I'm just sayin'."

UPDATE from John A. Stovall: "Leica and Panasonic have dismissed as speculation rumours that Leica may sell part of its camera business to Panasonic."

Comments

[charlie d]

Antarctica is the death zone for cameras and other mechanical electronic gear.

I think a high failure rate is pretty normal. I'd bring an Olympus E-1 with a couple lenses and an Xpan. I don't do pano's but I certainly would there.

Panasonic buys Leica and sells $1000 rangefinders with free DVD players as incentive. I'm cool with that.

[Mike Johnston]

"Antarctica is the death zone for cameras and other mechanical electronic gear."

charlie d,
Quoting Michael R.: "As for the weather during which most of the failures happened, it was no worse than a drizzly day in winter in New York or Berlin. Nothing Antarctic about it at all."

Mike J.

[Barbu]

I've handled the G10; perfect control, nice images out of camera, but it's getting swamped with noise (or noise reduction) as soon as you move intro the shadows :)
I tested the LX3 even more thoroughly; perfect from a photographic point of view (at least on specifications), but the corners are soft and I suppose that happens because of the forced distortion control, even in the Lightroom RAW converter. Nope, I couldn't care less for the JPEGS out of camera; here I'm sold to the Nikon rendition for colours (in dSLRs; their efforts in compacts is laughable). And that's even if I'm mainly a Canon dSLR user.

I'm also considering the Fuji F200 EXR; can't wait to test it for a longer time. Meanwhile, a few subtleties about this, that are kinda lightly explained in your post:
1) Pixel binning: I don't know if Fuji has a particularly bad method for combining the charge of two pixels, but each and everytime I see that a simple software downsizing performs better than their "low noise" mode.
2) High pixel count: even in plentiful light the image lacks fine resolution/detail; maybe they should have bitten the same bullet :) as Panasonic did in LX3: less is more, and make the F200EXR a 6-8MP camera, with pixel binning for 3-4MP. Don't know if it's a problem with the overstretched capabilities of the tiny pixels or (more likely) the lens has abysmall quality, but the high-resolution part looks worse than G10 even if I downsample.
3) Finally, there is a better side to F200EXR: dynamic range. It does work wonders, and doesn't (seem to) rely on software interpretation. The results are as miraculous as those obtained with a Fuji dSLR (S2/S3/S5). In the SuperCCDs this is obtained by using two different sized pixels; in Fuji F200 is using its sensor as two separate halves, and amplifying their charge with the equivalent of two different ISOs. That's the feature of EXR that I'd take and run away with it :)
I'd be extremely curious to see a Fuji dSLR that uses an EXR sensor, even if only for the dynamic part. Here Fuji is basically one of the only two companies left that are still trying to innovate for the digital imaging world; the other one is Sigma, but they clearly lack the power to manufacture a decent sensor.
___________________

Now, a bit of wishful thinking rant :)
Now imagine what would happen if the Nikon engineers would be tweaking a Canon sensor; they seem to do wonders with the Sony manufactured ones. And what if Canon would adopt Fuji's idea of different pixels? Maybe even a combination of different sized pixels (à la SuperCCD) and different amplyfiers (EXR style). Now, even more, what if the Canon CMOS manufacturing prowess would be put to good use and they would make 3x sensors, per Sigma/Foveon model?
Only wishful thinking... And no, I won't say a word about the ongoing crysis; it might stiffle innovation, but I'm sick and tired of hearing about it.

[Barbu]

Quoting Michael R.: "As for the weather during which most of the failures happened, it was no worse than a drizzly day in winter in New York or Berlin. Nothing Antarctic about it at all."
I've read that post about a week ago, and mr. M.R. :) forgets something: even if the weather was kind, they were still on the ocean and even the air is full of salt. When you combine salt and humidity... Brr, I won't even think about what happens to those tiny little circuits :)
Still, it's weird to see so many 5DmkIIs failing. I guess it says something about the much-touted weatherproofing; I only wonder what that would be in the Canon 50D, which supposedly also features some form of weatherproofing, but even weaker than in the 5DmkII.

[charlie d]

It's called GLOBAL WARMING!!!

The average summer temperature is 20ºF. The average winter temperature is -30ºF.

http://library.thinkquest.org/CR0215022/climate.htm

[Pascal Scheffers]

I'm not sure if I'm disagreeing with you. You say one thing, and you say it not.

But we're not even close to maximum camera performance in DSLR. We lot have waaaaay too much money to spend. And it is not the way marketing and development work to part us from our money.

All silicon being equal, mega pixels will continue to go up for the foreseeable future in the DSLR market. They may be high-end tickets now, sure, but production prices will continue to fall. Even if the market isn't really asking in 2009, they will have the ability to do so at little extra cost. The current crisis may perhaps cause a small dip in the curve.

The silliest thing in this is hinting at the end of progress because we don't want to spend any more money. The tools we use are essentially specialized computers, and pricey ones at that. Computers become cheaper when nothing else changes and time passes. This seems to be accelerating, if anything. Besides, I have never met a serious photographer or an amateur owning a DSLR*) who wasn't prepared to get more new stuff when equipment died, or just when it got 'old'. Or looked scruffy enough. Or just because. Money will flow. It may be somewhat less in 2009, so this crisis will put pressure on the manufacturers to produce the same products cheaper, but regression won't happen. It never has before, why would it now?

This may sound too optimistic in this doom and gloom financial crisis year, but Digital SLRs are just computers. Moore's law is, as others pointed out as well, highly applicable. We would not have had 25MP affordable cameras if it weren't. Elsewhere I asked which other component would need to be better to make a 400MP camera marketable, but a 100MP one**)? A few small things, probably. Today I would not even want one, not really, the files are just too big. I'd take that 30MP one, however. That is less of a problem. We take small steps.

Unless I am grossly mistaken (Ctein?) the Point-And-Shoot megapixel wars ended at about 12MP because cramming more pixels on the tiny sensors cannot yield more data. Basic physics, simply not much more there. Consumers sort-of-noticed, by-and-large driven by people like us (we are many too) pointing out that they could not see the difference between their 8 and 12MP P&S cams. So they started looking for cool other features.

Perhaps we should do the same. Weather sealing for starters. Built in Pocket Wizards. What more?

*) For a lack of a better word for all the monkeys consumers who make this hobby affordable for the rest of us.
**) Yes, diffraction limiting aperture is an issue. So what? We'll use those pixels for something else. The camera knows my F-Stop, maybe it'll switch to sRAW4 binning automatically. No noise at ISO128,000 Yay!

[Frank Petronio]

Panasonic is probably stalling the LX series so that enthusiasts buy G-1s instead. They have to make 4/3s work now before Canon and Nikon respond with a decent small camera.

I agree that DSLR development will probably simmer down. Except instead of innovation the manufacturers will just stick more "features" and complexity onto otherwise nice cameras. Video, Live View, HDR, Yawn Mode....

[Juha Haataja]

My LX3 has now about 30,000 images taken since September, and is showing some signs of problems (lens not retracting etc.).

I have been thinking of buying another, but LX4 for summer sounds great. I hope you are right!

[Alexander Thorp]

Where does the phrase "I'm just sayin'" comes from? As a non-US non-TV-watcher I just don't know. I imagine it being the catch phrase of a character out of a sitcom like The Fresh Prince Of Bel Air.

Maybe while you're about it you could clear up "your mileage may vary" too. It's totally obvious what it means, so no need for explanations there, but where the phrase originated would be interesting. It doesn't appear on our (UK) car adverts.

[Mike Johnston]

"Money will flow. It may be somewhat less in 2009, so this crisis will put pressure on the manufacturers to produce the same products cheaper, but regression won't happen. It never has before, why would it now?"

Pascal,
It never has before? Actually it always happens. One way of looking at the whole history of photographic technology is:

1. There is a defined need;
2. There is a race to sufficiency and beyond;
3. Sufficiency is reached and exceeded;
4. The point that is sufficient becomes the standard and further progress, although it often occurs, is not terribly significant.

It's happened more or less this way with all manner of things. There is a race to reduce the size of film; but then 35mm is small enough, and further miniaturization, although practical and implemented, doesn't catch on in the market. There is a race to improve the speed of lenses; f/1.4 is reached and exceeded but is largely deemed to be sufficient--faster lenses become niche products with very small markets, and nobody cares any more who has the fastest lens. SLRs are made smaller and lighter until they're small and light enough, and that's the end of that. Simplicity and ease of use for tyros is a holy grail until the "push here, dummy" point-and-shoots of the '80s, and then they're simple and easy enough and that stops being a driver of development. The speed of film, the search for the best fine-grain developer, the sharpness or range of zoom lenses, I could go on and on--it always happens this way.

Right now we're in the middle of, or nearing the end of, races for a) high-ISO/low-light performance with low noise, and b) more megapixels. But sufficiency in both these parameters will be reached and exceeded at some point, and people will stop caring so much, and then both those races will cool down and will no longer be driving the market or further development. And we'll move on to something else that everybody wants and "needs"--and will pay for. We don't know what features or capabilities future races will concern themselves with, but they'll happen too.

Mike J.

[Mike Johnston]

"Maybe while you're about it you could clear up 'your mileage may vary' too"

Alexander,
I'm not certain, but I think "your mileage may vary" comes from the 1970s when the Environmental Protection Agency first started testing and rating the mileage of cars. The early testing protocol was not very realistic and resulted in inflated numbers that actual buyers seldom were able to match. As I recall, initially "your mileage may vary" was used as both a straight advisory and also ironically....

Mike J.

[Mike Johnston]

P.S. I picked up "I'm just sayin'" from a funny bit Jon Stewart did on "The Daily Show." I can't find it online, but maybe somebody else can....

Mike J.

[Gingerbaker]

Great.

Five people out of what, ~ half million sold (?), have problems with their 5DII's and you have dismissed the entire model as problematic?

That "whopping 25% failure rate" was what - four cameras? Add Ron Purdy to make a grand total of five people. That's quite a sample size.

You have a well-read blog here, Mike, and with that comes a responsibility to have some perspective and care about the products that you talk about.

Are there some folks with complaints about sub par performance of their 5DII's? Of course. But the overall percentage of owners with complaints is *tiny*. Even the Ron Purdy link you provided had comments that spoke glowingly of their particular iteration of the cam.

Using terms like "rocky reception" is professionally irresponsible. But, hey, it's your blog, you can call 'em like you see 'em, right?

[Mike Johnston]

Gingerbaker,
How's yours doin'?

Mike J.

[Don Jagoe]

Also, on the 5DMk2 issue (and mine has worked splendidly in rain and fog), I believe most or all of the Antarctica failures had the non-sealed, attached multiple battery grip.

[mike b]

Due to a not very successful attempt to actually try and make money from photography last year, I have a bit of a tax refund windfall on the way.

I have no desire to upgrade my 5D and find myself instead obsessing over av receivers and loudspeakers and buttery leather chairs.

I've hit some kind of tipping point, I guess.

[phule]

This is not really meant as a counter-point to the 5D Mark II failures in the field. It's more a "wow, these guys are having fun with new technology" point.

Anyway, 5D Mark II + Video + 15mm fisheye ==

http://vimeo.com/3155182

[improbable]

I'm eagerly awaiting more F200 EXR samples, especially the extended dynamic range type. The sample you lined to doesn't seem all that impressive to me, because it's hard to tell (without having been there) how contrasty the light was. If the clouds are just right, this could be a normal camera.

Unlike the Fuji S5 etc, this seems like a much simpler modification of the usual chip design: if I understand right, they can set different bias voltages for even and odd pixels, which probably doesn't carry a huge penalty in terms of how fabrication difficulty or pixel area. And if you set them all to the same sensitivity, you're back to a normal bayer sensor. So here's hoping we'll see this idea becoming more common.

There should be no advantage, other than file size, to pixel-binning in the camera (the high-iso mode). But if for some reason your down-sampling algorithm on the computer is poor, then maybe there is some advantage. And of course if you're looking at the result at "100% pixels" as people are so fond of doing, then the downsampled one will look less noisy!

[Eric]

I can't remember the last time I was "ecstatic" about anything that didn't involve a member of the fairer sex. For my part, 12mp is plenty, and the only advance that might tempt me at this point is post-bayer developments or camera body progress, like modular bodies, etc.

[Gingerbaker]

"Gingerbaker,
How's yours doin'?

Mike J."

I don't own one, Mike.

I DO own a 5D, whose resale value might be higher if the 5DII was universally judged to have "nagging quality and operational irregularities, which perhaps causes the sales curve to tail off a bit too soon."

Alas, that is not happening as the 5DII is selling so quickly that it is back ordered or in very short supply just about everywhere.

Fess up, Mike. You are not particularly enamored of Canon DSLR's, and you felt like taking a little shot.

[Mike Johnston]

Gingerbaker,
If you must know, I've been dealing with this problem personally and professionally for at least fifteen years. The fact is that there isn't any clear way for a reviewer to tell in the early going how reliable or problematic any particular piece of equipment is. The manufacturers would sooner commit hari-kiri than release DOA and return statistics for new models; quite the opposite--they'll try to sweep any early problems under the rug insofar as they can. In the era of the internet they can actually get bitten the other way--early teething problems and not-quite-ready-for-prime-time releases can get blown out of proportion and cause excessive bad publicity.

Still, we're left trying to figure out reliability by some combination of intuition, Kentucky windage, and a finely-tuned ear to the ground. And I'm very familiar with that game. Eventually the truth gets sorted out, and everybody knows what the issues were and their relative significance, but that's after a fair amount of time has passed--usually far after the early reviews are history.

You might not like it, but an expedition of a couple of weeks' duration where 26 cameras are used daily and six of them fail is far better data than we almost ever get about how a new model is performing in terms of reliability in the field. It's a significant sample and the failure rate in this case was abysmal. If that's not worthy of a link, then I don't know what is.

Sorry if you disapprove.

Mike J.

[Dan K.]

Gingerbaker,

Sounds like you're a bit of a fanboy eh?

As for Mike's description in regards to the 5DII ("rocky reception"), I would say that is a pretty accurate account. Considering the store I work at already had a used 5DII and the fact that most of our pre-orders for the camera asked to be removed from the list (from 15 to 5), I do believe the camera has had a rocky reception. (remember, I'm not discussing about the great IQ of the camera, but rather its reception)

Anyways, thanks Mike for the wonderful insights.

[Malcolm]

I seriously think that there are either two different factories turning out 5DMKII bodies or that there was an early production change. I haven't seen any of the problems anyone has reported, especially shadow banding at any ISO below 25600. This is just what DPReview found in their in-depth review:
http://www.dpreview.com/news/0902/09021302canon5dmarkiireview.asp
How can I have 4000 shots where the only bad ones (misfocus, exposure, etc.) were my fault and not the camera whilst others have "thrown in the towel"? It just doesn't make sense to me.

[ricx1]

"I'm just sayin..."

It is interesting to hear word or phrases come in and out of the lexicon. I think the "I'm just sayin'..." phrase came originally from a 1970's TV sitcom "All In The Family". It was said by the character Archie Bunker (Carroll O'Connor) or his son-in-law Meathead (Rob Reiner).

All of which probably goes back to some Boston or New York City street-speak.

I'm just sayin'....

Rick

[Jay Moynihan]

I had read the Luminous Landsscape tech post.
As an "old timer", i was surprised how well the other digital cameras did in those conditions. That is pretty encouraging for the breed, generally.
Re the new Fuji, i wish it had an optical finder, or peep whole or something. That is lack of feature i still do not like about my Fuji F20.

[Juha Haataja]

"Panasonic is probably stalling the LX series so that enthusiasts buy G-1s instead."

Well, the LX3 has only been on the market for half a year, so I wouldn't say that they are stalling. But if there won't appear a new one in the series this time next year, then they would be stalling.

That said, I sure hope for an improved version of the LX3. It would be great to have the same lens, but better dynamic range and 1-2 stops better sensitivity. Would that be possible?

[Mark Sirota]

That shot of the Fuji F200 makes it look like it has pincushion distortion built right into the case.

[aizan]

i always said canon should take the vertical grip off of a 1-series body and be done with it. just because it's smaller doesn't mean it can't be heavy duty.

[Tim Medley]

From MR's 2007 Antarctica Trip Report:

"About 85% of the trip's members were shooting with Canon gear, mostly 1Ds MKII. There were a couple of 1D MKII and 5D's, with several people having Rebel XT and XTi bodies as backups.

There were quite a few camera failures, most of them occasioned by our shooting in rain on the first day ashore in The Falklands. The rain was not heavy – more a persistent drizzle than a heavy rain, but it took its toll. In all, 90 minutes of wet shooting produced six 1Ds MKII cameras which stopped working for one reason or another. Three of them recovered after a night of drying out. Three remained hors de combat for the rest of the trip.

Three Canon 5D's died that day, with one subsequent recovery. Two Rebel XTi's lost their rear LCD's, though otherwise continued to work (which is a real hassle, because though one can keep shooting, there's no way to change any settings, or at least to know what the changes are).

We also lost two video cameras, one with a dead sound board and the other a total cardiac arrest. Several lenses bit the dust during the trip, including two Canon 70-210mm f/2.8L IS, and a Canon 28-135mm."

http://www.luminous-landscape.com/essays/aa-07-worked.shtml

(I'm a Canon owner.)

[Kurt K.]

From what I have read, the Fuji does not have RAW support. That makes it a non choice for me...

[paul in kirkland]

The LX3 thing is really bugging me. I've been trying to get my hands on one for a while now, but no retail store carries it. And now it's back-ordered on B&H should I choose to buy it sight unseen.

I'm not sure what to think, but I need to get a pocket P&S soon. Maybe I'll just get a different Panasonic

[Ken Tanaka]

"Ron Purdy"? Who?! Sorry to learn of his 5DII "troubles". My 5DII has been just fine, although I've not yet taken it swimming.

Having spent the past week largely immersed in the "real" art and photography worlds I admit to returning with a thinned tolerance for the Internet-based maelstroms and celebrities.

[MBS]

Taking a hard left...

Eric said, "I can't remember the last time I was 'ecstatic' about anything that didn't involve a member of the fairer sex." Made me laugh, but also got me to thinking.

I've had a couple of religious experiences that I believe the word "ecstatic" would apply to; but I also have had some experiences that were practically religious in nature, certainly entailed a rapturous joy, when viewing particular works of art and when listening to specific passages of music.

As a for instance, and not to belabor the point (though perhaps I’m too late), I still remember coming around a corner my first time in the Getty—when it still housed its art in the villa in Malibu—and seeing Hendrick Ter Brugghen’s "Bacchante with Ape." I was stunned. I stood there for several minutes, then noticed that someone had kindly placed a bench just in front of the painting. I don’t know how long I sat there, forty minutes at least. Unusual for me, as I generally am a pretty quick viewer. I was, as an old jazz musician might’ve said, solid gone.

[Dave Richardson]

My brand new Canon 5D Mark II showed up on Tuesday with a dysfunctional metering system and had to be sent back for exchange. But then, maybe it's just me. My 20D had to go back to Canon 3 times in the first two months that I had it.

-dave-

[hugh crawford]

If you are indulging rumors, there is Nikon's "Big" announcement on Monday night at the MGM Grand Arena in Las Vegas.

[Peter Cameron]

When determining whether the Canon 5dMkII failure rate is good or bad one also must consider the target market. It's not a camera aimed at the "charge up the camera battery Martha, it's little Johnny's birthday tomorrow" crowd. It's a camera aimed at pros (as a backup body) or advanced amateurs/enthusiasts. These people expect it to be nearly as rugged as a full-pro model and are likely to use the camera under less-than-ideal conditions.
Thus, a 6 out of 26 failure rate is abysmal.
However, if it was 26 Rebel XTi or Nikon D40x bodies we were talking about it would be a case of "Wow, 20 of them survived."

[Tom Passin]

On the pixel-binning of the Fuji F200 - I don't know if they do it this way, but it seems to me that pixels should be binned only in the low-light areas. The binned pixels could then be split again to fit into the original, higher-resolution array. In this way, dark shadow areas would have lower noise with lower resolution, and the rest of the picture would have the full resolution. The best of both worlds.

If Fuji isn't doing it this way, but rather binning over the whole image for dark scenes, then it doesn't seem that it would be too hard for someone to write a (e.g., Photoshop) filter to do the same thing.

[robert e]

If the D3x had an interchangeable back, $8,000 would be considered a screaming bargain.

[Dave Brown]

Mike,

I think you're right about where the megapixel race ends up. After all, on the list of things that makes a photograph "good", well, resolution isn't even on my list.

I can think of two things that will spur camera sales in the future. First, as cameras these days are more software than hardware, manufacturers can simply abandon older models, for example, making new raw converters incompatible with older cameras. I'm sure some people are already seeing their older cameras become obsolete as they upgrade their computer operating system. Perhaps a cynical view, but I think time will prove me right.

Second, (and this possibility is the only thing about digital imaging that truly excites me) advances in high iso performance will eventually allow the "auto iso" function to work in a meaningful way. Imagine - instead of choosing your aperture, and compromising on shutter speed (or vise versa) you will be able to choose both the precise aperture and shutter speed you desire, and pick the iso you need to get the exposure right. This will truly revolutionize the way we photograph.

[Mike Johnston]

"Re the new Fuji, i wish it had an optical finder, or peep whole or something. That is lack of feature i still do not like about my Fuji F20."

Jay,
Don't get me wrong, I'm not saying that the F200 is likely to be some sort of paragon. It's mainly interesting to me because it's the first proof of concept of the EXR sensor to make it to market. Fuji's sensor experiments are not always world-beaters as real-world products, but they're very interesting, I think.

Mike J.

[Ken]

It's called GLOBAL WARMING!!!

The average summer temperature is 20ºF. The average winter temperature is -30ºF.

http://library.thinkquest.org/CR0215022/climate.htm

I can't tell you how much this post made my day. For many years I was a judge for thinkquest. I think the students would be thrilled to see that research by a group of fourth graders was used as a supporting document.

[Mike Johnston]

Gingerbaker,
I added a new line to the second-to-last paragraph of the post for you. Just to be clear that it was only a scenario....

Cheers,

Mike J.

[Dave Kee]

Just a few comments on the current economic slowdown. Unlike other recent recessions this one will not soon turn around. We have allowed income and wealth inequality to grow to the point that the bottom 90% of our population now control only 20% of our wealth. This means that they can no longer maintain the spending needed to drive our consumer economy, nor save enough to insure their own economic security. No conceivable amount of government stimulus spending can correct the resulting downward spiral. The only solution is to use the tax system to redistribute some amount of the accumulated wealth of the super rich to the middle and lower classes, but that acton is and will be politically unacceptable until long after the problem has reached critical proportions. Hate to say it, but the future megapixel cap on our DSLR's is kind of a minor part of what we are facing.

[Matthew Robertson]

What I don't understand is the complaints about the pricing on the D3/D3x. Yes, they're expensive, and there's a big price gap between them. But the situation is no different with the Canon 1D/1Ds - essentially identical bodies with different sensors and electronics. Yet I don't recall seeing the same disgruntlement about the pro Canon pricing. It's as if people have different standards for what a Nikon is worth.

I'm just sayin...

[erlik]

"Imagine - instead of choosing your aperture, and compromising on shutter speed (or vise versa) you will be able to choose both the precise aperture and shutter speed you desire, and pick the iso you need to get the exposure right. This will truly revolutionize the way we photograph."

Er, don't understand this. I can put my E-3 into manual mode, set shutter and aperture, and if the camera is in Auto ISO setting it will pick the sensitivity it thinks appropriate. Is that what you meant?

[Eolake Stobblehouse]

"(Because it's been a while since I've run up to kick the pocket-digicam football only to have Lucy snatch it away at the very last second.)"

Hehe. I really wish Schulz had made Charlie actually kick the dang thing. He even said in an interview that he had realized with sadness that Charlie never would.

[Rick Schreiber]

Do not take a non rated weather resistant camera to weather prone locations and be critical that they do not work. I would have taken my Canon D mark III or DS mark III. Sorry, my Canon 5D mark II works great in non weather shoots. I do wish it was weather proof.

[Eolake Stobblehouse]

"... And say that while all this is happening, the world economy tanks alarmingly"

Stop it, you're scaring me. Surely all those things couldn't happen at once or at all.

[Mark Muse]

I think the image quality of the 5D2 (yes with L lenses) sucks (a technical term) in relation to the price tag. I will have trouble making 36 x 24 exhibition prints with it.

And as for lenses, those vaunted and very pricey L lenses are quite a long way from perfect. Both of mine yield aberrations on virtually every shot. And the coatings are not the best I've used either. The 17-40 is so unsharp it couldn't cut mustard—and that is in the center of the frame. The corners are simply a smear.

Then there is the user interface: Consistency? Forget it. Properly prioritized? Hardly. Obtuse? Plenty of that. It is as much a a gizmo crammed with all the latest gotta have adolescent wiz-bangs as it is a professional tool.

A disgruntled... no, make that pissed... 5D2 owner. Rant off.

[Paul Amyes]

Up and till recently I have taken cameras to -35 to -45 degrees Celsius. Admittedly they have been film cameras, but out of all the ones I've taken (OM1n, 2n, 4, Mamiya c330, and various 'blad bodies) I've never ever had a failure. So to have 6 out of 26 cameras of one particular model fail during a relatively warm Antarctic summer is quite an eye opener.

When the original Canon 5D came out it was touted at the landscape photographers and the travel light photojournalists. I see Canon is still pitching at that market with the Mk ii so I would say something is seriously amiss. Firmware failures on the camera are one thing as they can be rectified easily, but when it comes to the hardware, particularly body shell integrity then that's a sign of bad design. Canon have dropped the ball badly recently with the Id Mkiii autofocus debacle, and if the 5D Mk11 turns out to be a problem child then Nikon and Sony will be rubbing their corporate hands together with glee.

To put the record strait I have a considerable investment in Canon equipment which is used professionally day in day out. I was hoping to get a 5D Mkii at the end of our financial year in June, but I will hold off and see what happens. I don't want to buy a camera that has issues, the whole 1D Mkii debacle was enough for me.

[Dennis]

Of course, it'll be a cold day in you-know-where (or a hot day in Antarctica) before an expedition of photographers includes more than a couple Sony users, but if 25% of Sony cameras failed on such an expedition and that was reported online, Sony would be written off with no further investigation.

Just sayin' ...

[Mike]

This Czech site has some DR examples. The most interesting one is about 3/4 of the page down - an archway leading to bright light. Mouse over the right image to see the DR effect. (The other samples are much less dramatic.)

http://www.fotografovani.cz/art/forec_amater/Prvni-zkusenosti-Fuji-F200EXR.html

[charlie d]

"I can't tell you how much this post made my day. For many years I was a judge for thinkquest. I think the students would be thrilled to see that research by a group of fourth graders was used as a supporting document."

I pulled that with only a cursory glance to support my claim! I like it even better that it's Highland Park Il 4th graders.

I am such a fraud

[k4kafka]

Professional shooters expect better from a company like Canon.
The recent production problems with both the EOS 1D Mark III and the 5D Mk 2 show a troubling trend.
Many years ago I switched from Nikon to Canon because I felt Nikon was making too many point & shoot consumer-level cameras while ignoring the needs of the professional shooters who built Nikon's reputation. Canon seems to be marching steadily toward a similar marketing strategy.

[Mike Johnston]

"I am such a fraud"

We like you, though.

[g]

Mike

[Jerred Zegelis]

To say that there might be a relative freeze in technology leaps sounds a bit wrong to me. Maybe you can say that if you only look at the mad rush for more megapixels... but there are always innovators out there ready to push the boundaries of what we think is possible. People will pay money for truly outstanding innovative tech, even in tough financial times.

I remember thinking to myself, 20D in hand, "It can't get better than this. This is the perfect camera. I will never need another." HA! Although I have a special place in my heart for the 20, as it's gotten me some of my all-time best photos in my portfolio, I would never go back to it after all the time I've spent with my 5D and Mark III.

There are plenty of innovations coming, but I don't think the most important will be tied to megapixels, and instead will be innovative ways that the cameras actual capture the light, regardless of megapixels... kind of like the Fuji pictured above, but with our beloved DSLRs.

[jay moynihan]

"Jay,
Don't get me wrong, I'm not saying that the F200 is likely to be some sort of paragon. It's mainly interesting to me because it's the first proof of concept of the EXR sensor to make it to market. Fuji's sensor experiments are not always world-beaters as real-world products, but they're very interesting, I think."

Oh, i understand why you posted about it. My Fuji F20 has the super-ccd sensor (6.3 mp) as the fabled F31. While i complained about its lack of somekind of optical view finder, it is the only camera i have that can literally fit in a shirt pocket.

And, i can get a printable (i only print 8x10ish or under) iso 400 image, and in B&W, iso 800. Therefore, i too am very curious about their new little machine :).

[Gingerbaker]

"Gingerbaker,
I added a new line to the second-to-last paragraph of the post for you. Just to be clear that it was only a scenario....

Cheers,

Mike J."
.
.

Thanks, Mike, that is appreciated. :)

Re sample size:

The point has been made here that

(1) the Reichmann expedition results are significant and based on a proper sample size, and

2) the Arctic conditions were very mild, which is even more a reason for consternation

Folks, do you not see how both points can not be true simultaneously?

If the weather conditions were no worse than anywhere else in the world, and the sample WAS significant, then we would be seeing 25% of all the 5DII's in the world failing spectacularly due to mere misting rain.

But, we do NOT see 25% of the world's 5DII's failing. So either point (1) or point (2) or both are wrong. Yes?

That said, Canon needs to get its QC and camera reliability up to industry standard. They have taken their eye off the ball in recent years.

If all your competitors offer true or superior weather-sealing, you must keep pace. Canon has not. The 1DIII focus debacle was just that - a debacle. A complete redesign of a focusing mechanism goes out into retail without proper vetting - what were they thinking? Years of 1D excellence reputation gone overnight.

That Nikon can surpass Canon in full-frame noise performance in their first get-go shows what a really top notch engineering department they have developed. IIRC, new younger managers are to be applauded. Canon needs to keep pace here as well.

[Mike Johnston]

Gingerbaker,
The difference is that photographers on an Antarctic photo expedition are taking pictures outside all day long every day for however long the expedition lasts, whereas camera owners in the rest of the world, I would imagine, like most of us, are going about their work and lives and only occasionally taking their cameras out to shoot--and when they do, it's not unreasonable to assume that on the occasions they choose it's not showering, drizzling, or misty...for the most part.

The Antarctic Expedition's failure rate is a data point, not probative but potentially indicative nevertheless. I'm sure many people will not care simply because they never get their cameras damp or wet. For those who do, it might be a useful caution.

Mike

[David Bostedo]

"That Nikon can surpass Canon in full-frame noise performance in their first get-go..."

I assume you're referring to the D3. I would submit that this was a terrific marketing decision, rather than a feat of engineering. Based on their other efforts, it's likely that Canon could have released a 12 MP sensor with noise performance at least as good as the D3. Where Nikon scored is by actually choosing to make a relatively low-res full frame sensor; That's what allows it to be so low noise. Most companies, Canon included, didn't see the market for this apparently, yet it was a hit.

[Ctein]

Dear David B,

I'm a bit late to the party, but you've made me think of something.

Camera makers have a problem similar to car makers. Even at today's breakneck pace of product release, the camera designers are always working a few years ahead of market (car makers it's more than 5). So you have to guess what people are going to care about.

Right now, the vox populi is obsessed with noise (and make no mistake, it is an irrational obsession). Five years ago, it was pixel count. Five years from now it could be color accuracy (I doubt that, personally) or skin tone/neutral color balance (which is entirely possible-- it's TOUGH getting this right under all conditions).

So, to some degree, the poor camera designers are stuck with trying to guess whether next year's irrationality will be like last year's.

Glad it is not my job.

pax / Ctein

[Gingerbaker]

David B said:

"I assume you're referring to the D3. I would submit that this was a terrific marketing decision, rather than a feat of engineering. Based on their other efforts, it's likely that Canon could have released a 12 MP sensor with noise performance at least as good as the D3."

Canon had years of full-frame and lower noise sensor manufacturing and design experience under their belt. Yet, Nikon, in one fell swoop, surpassed their noise performance with their first entry and did it using a sensor developed out of house.

Their success, I opine, was because they had really good noise processing engineers on staff. They needed them! For years they were refining their algorithms to deal with the godawfully noisy sensors Nikon kept putting into their DSLRs. I think their marketing genius was to scrub chrominance noise on-chip which really wowwed folks. That they are able to do this and still maintain good color saturation and balance on a chip made, after all, by Sony (!) is very impressive.

So, I see it as great engineering that, as you say, was effectively marketed. They make a very good product.

[Gingerbaker]

Mike said:

"Gingerbaker,
The difference is that photographers on an Antarctic photo expedition are taking pictures outside all day long every day for however long the expedition lasts, whereas camera owners in the rest of the world, I would imagine, like most of us, are going about their work and lives and only occasionally taking their cameras out to shoot--and when they do, it's not unreasonable to assume that on the occasions they choose it's not showering, drizzling, or misty...for the most part."

Good assumptions, Mike, but they are only assumptions. It is not only folks on expeditions who use their cameras in the rain, and for every 5DII on that expedition , there are probably a million 5DII's in use by ordinary photographers in all kinds of conditions, and for months now, not weeks.

We cannot take a small sample size like the expedition, and inductively apply it to a large sample size. If we want to make *valid* assumptions about the large sample size, we must use an appropriately large sample size. If we had done this, we could then make inductive assumptions about the smaller expedition size.

Mike said:

"The Antarctic Expedition's failure rate is a data point, not probative but potentially indicative nevertheless. I'm sure many people will not care simply because they never get their cameras damp or wet. For those who do, it might be a useful caution."

I agree with you here, Mike. And, to combine two statistically small populations, this is not the first time that Canon cams failed under those Antarctic conditions in numbers that did not compare well with their competitors.

[macsuperfly]

There is a website with links to multiple sample photo pages for the f200exr. It's at http://f200exr.com . There are a few "real world" pictures (taken by amateurs) at 800 iso that don't look promising (noisy) the f20/f30 series would have done a much better job.