Digital Bridge Cameras
Why The Sony F828
Produces Consumer Angst
In the 1950’s Leon Festinger at Stanford University put forward the concept of cognitive dissonance. The theory suggests that when one belief system or pattern of behavior (cognition) that we have conflicts with another we change the weak cognition to conform with the stronger one. We see this in Aesop’s famous fable about the fox and the grapes. The grapes are out of reach and so the fox decides that they must be sour, thus relieving himself of anxiety over not being able to have any.
I pondered this topic on my flight to Europe on New Years Eve 2003, on my way to Tanzania where I was to lead a 10 day wildlife photography workshop. At issue was the first part of my review of the Sony F828, a just released 8 Megapixel digicam, that I had published a few days prior.
In my review I had offered a mixed report on the Sony, which is best summarized by the subtitle that I had given it, calling the camera — A Flawed Jewel. I praised the camera’s superb lens, excellent resolution, terrific build quality and very good colour characteristics. I criticized the poorly executed data buffering, lack of decent RAW software, and poor placement of certain operating function controls, which inhibit some aspects of shooting convenience.
I also compared the F828 to my Canon 10D, a 6 Megapixel DSLR, along with a couple of high-end Canon zoom lenses that would be needed to match the coverage of the Zeiss lens on the Sony. I pointed out that the Sony had higher chip resolution, and that its built-in zoom lens appeared to equal in performance the two Canon L series lenses that combined cost more than 2.5X times more than the entire F828.
I pointed out that the Sony was noisier than the Canon, by about 1.5 times, but that at ISO 64 and 100 — the Sony’s lowest sensitivities ratings — noise was not much of an issue, and that at higher speeds, though it got noisy quickly, current noise reduction software still allowed good prints to be made.
Chromatic aberration was the other issue that was addressed. I pointed out that the F828 displayed it in some situations — more so than the 10D, but that while visible in 100% enlargements on-screen, in prints I didn't regard it as that much of an issue.
Overall I thought that I had written a balanced review — pointing out the camera’s good features, its weak points, and also drawing a comparison not to other digicams (with which I don’t have much experience) but rather with a contemporary DSLR, with which I have about as much hands-on experience as any photographer / reviewer that I know.
The Sky is Falling
It seems that my review annoyed both Sony and Canon camera aficionados quite a bit, and in equal measure. The reason for this is, I believe, due to a cognitive dissonance.
On the one hand there is the widely held belief that to date digital SLRs produce higher quality images than do digicams (true). Then there is my review in which I make the claim that the Sony F828 has a lens that is comparable in capability to two expensive interchangeable Canon L series lenses, and that in prints up to Super A3/B size (13X19” paper) the noise and chromatic aberration issues are not a serious impediment to high quality prints. (Also true — at least that’s what my tests and field experience showed).
The problem was exacerbated by the fact that I published some 100% crops that showed the noise level of the Sony to be worse than that of the Canon. I did point out that it was worse, but many for whom the dissonance was too great used these as their hook, even though I pointed out that the noise and CA issues weren't a big deal in prints.
The dissonance thus lay in the fact that I had the audacity to claim that a mere fixed lens digicam could actually be a useable alternative to an interchangeable lens DSLR. The fact that this particular camera, the Sony, had some image characteristics that were inferior when judged at relatively large enlargement sizes was seized on as proof that I must be wrong.
Because of this the real point of my initial review was lost on many readers — that because of the F828’s high resolution chip, terrific lens and especially some aspects of its handling characteristics, it could actually be used to create very satisfying images suitable for professional use.
Noise / Grain and The Great Myth
Since the development of small format cameras in the early years of the 20th Century film grain was a something that photographers had to deal with. Low ISO (10-64) meant fine grain, while pretty much anything above ISO 100 meant coarse grain. Notice that I wrote fine grain, not no grain. This is because with small format cameras grain was simply a fact of life.
Then in 1999 along came the Canon D30, which was the first digital camera that most serious photographers could afford, and that essentially had no noise (the digital equivalent of grain) at low ISOs.
Now, just a few short years later, we expect (and usually get) images from digital cameras that are routinely noise free at the lowest sensitivity settings. So, when we do see some noise in a camera’s output it because a focal point for comparison. “Well, Camera X clearly shows more noise in 100% enlargements than does Camera Y, therefore it’s clearly not as good”. The fact that in typical sized prints the difference is vanishingly small is of no account. Y is better than X, and that’s that.
Of course the factors that really do count much more, such as control layout and other usability factors, simply aren’t as quantifiable, and most reviewers don’t explore them in depth. Can the typical reader then be blamed for focusing on those aspects that can be more easily compared, and which are the main topics that reviewers write about?
A final thought about noise and grain. When we look at photograph by Cartier Bresson, for example, do we care one whit about whether it’s grainy or not? Of course not, and of course it’s grainy. The film emulsions that were available for 35mm cameras in the 1930s though the 1950s were very grainy, even by today’s film standards let alone by contemporary digital standards.
So let’s develop a sense of proportion when we consider such small differences.
Suitability for The Task
I don’t know how to say this politely, so I’ll just put it on the table. Make of it what you will. When the better automotive magazines review a new model, sure — they write about horsepower and the technology of the latest Sat-Nav system. But for the most part they address the issue of how well the car performs its intended role. If it’s a sports model, how quick is it on the track, and how’s the handling? If it’s a new SUV, how does it do in the mud and snow. And, minivans are rated on how many kids and shopping bags can be carried in comfort.
Unfortunately the same paradigm is not usually applied to camera reviews, at least not by some of the reviewers currently writing on the Net. They focus their attention on the things that can be readily observed and compared with other models; things like resolution, noise characteristics and the regurgitation of a laundry list of the manufacturer’s specs.
But those are not the core issues, are they? What should really matter in a review is the suitability of the product for its intended task based on the reviewer’s experience as a photographer. And this is where the interests of what have been called the pixel peepers and these reviewers go hand-in-glove, but I believe, to the determent of the broader community of photographers.
Here’s the thing: of much greater importance to photographers than whether one there’s a few lines per millimeter difference in resolution between one lens and other, or if there’s a bit more chromatic aberration on a particular digicam than in another model, is how well the product performs its overall job, and why and why not. To not do so is to render the reader a great disservice, because it focuses his or her attention on to what are likely relatively trivial differences.
Let me be the first to add that I am often as guilty of this as are other professional reviewers. For this I apologize, because I now believe that to do so perpetuates the current misapprehension about what’s important, and what isn’t.
In my product reviews I do try and address design and usability issues in addition to the more usual feature and spec comparisons. But as regular readers of this site, my magazine articles and reviews, and viewers of The Video Journal know, I am usually more concerned with how well the product meets the real needs of the photographer, and especially with cameras, how well it performs in the field. The small image differences seen between systems is of much less importance than more subtle but often more detrimental design flaws.
By The Numbers
I see a very disturbing trend developing on Net discussion boards. There has been a virtual hijacking — for lack of a better word — by the pixel peepers and the equipment junkies at the expense of those for whom cameras are tools to an end. That end is the production of photographs — ones that inform, ones that record a time and place, ones that express an emotion, and also of course, ones that satisfy technically.
To place almost the entire emphasis of a review on the device's technical performance (or lack thereof) is to miss the very important consideration of the camera as tool, rather than just as a set of specifications. As but one example there are sites that rate cameras, comparing them with numerical values such as 8.1, vs another that might have a rating of 8.0. These numbers are an amalgam of various numerical sub-ratings of various features. The unsophisticated reader naturally skips to the end of the review, sees that Camera X has a higher rating than Camera Y, and then becomes a defender of one over the other, often debating his favourite's merits online without actually having ever taken a photograph with that particular device. This is nonsense, of course, but others read these commentaries, and since there's no way of knowing the experience or credentials of whoever wrote the diatribe such comments are often assumed to have validity.
A Call to Reviewers & Readers
There's more to cameras than specifications and numbers. There's more to photography than knobs and switches. There's more to our craft than MTF charts and optical aberration measurements. Photography is an art. Photography can be a way of life. Cameras are more than just tecky devices to be fondled sitting on the couch in ones den, or debated ad-nausium on web forums.
I don't want to adopt a holier-than-thou attitude on this, because I have at times been as much at fault as other writers. This site and my various magazine articles and other publications do concern themselves with product tests, comparisons, ratings and the like. But I would like to think that foremost in my writing's about products and the reviews that I produce I try and focus primarily on a product's suitability for its intended task. This includes hands-on evaluations on location, shooting many hundreds and even thousands of frames before drawing conclusions. It means giving emphasis to ergonomic issues (handling), and also the sometimes indefinable things that make a given product (especially cameras) "connect" with the photographer.
The call to my colleagues is to make an effort to stop feeding the tech-junkies that have hijacked the Net's photography discussion forums. Let's take the time to educate readers about some of the less measurable aspects of a product's feature set. Camera's are not refrigerators. The Consumer Reports approach to camera reviewing doesn't truly serve would-be purchasers, because cameras are about much more than just measurements and specs.
The call to readers is — read the various reviews available on major sites on the Net. Gather together the impressions that reviewers provide and then go and handle a camera for yourself. You'll learn a lot more about whether or not a particular camera is the right one for you this way than listening to nattering's of discussion forum regulars who gauge a device's goodness by its measurements and not by its results.
So, What about the Sony F828?
Which brings us back to the initial topic — how the Sony F828 seems to have become a lightning rod for so much on-line angst and confusion. The first part of my on-line review ( the first by a major site of a full production camera) was controversial because I didn't downgrade the camera as much as some thought I should have because of its higher than usual Chromatic Aberration and higher than typical noise levels at ISO ratings above 100.
Indeed I took Sony to task for a number of other design flaws, some of them quite serious. But, in the end I angered and confused a great many people because I said that I believed that this was a landmark camera, possibly the first of the new "bridge" or "cross-over" digicams that can challenge DSLRs at their game, and which can produce images suitable for some professional applications.
This is where the cognitive dissonance comes in. Digicams are for snapshots, beginners and hobbyists. DSLRs are for serious photographers. Well, maybe that's been the case till now, but the game is changing. It could be that there are other models currently available (early 2004) that might fall into the same camp as the Sony F828. I don't know, as I haven't tested them all. But that's not the point. The real issue is that we are about to enter a new era where the line between interchangeable lens DSLRs and fixed lens digicams starts to blur, and where measurable parameters like pixel count and sensor size are becoming increasingly of lesser importance than how suitable the device is for its intended task. And the things that determine suitability can't always be quantified.
A Final Comparison
In the early to mid-90's Contax brought out a series of film cameras, the G1 and G2, which excited many photographers. They were beautifully built and featured superb Zeiss lenses. These were small and elegant "rangefinder" cameras. If one read the glossy brochures (this was before the Internet) or read the reviews in magazines like Popular Photography, they seemed like a dream come true. A possible competitor for M series Leicas.
But many photographers found that while the lenses were great, the shutter speeds were highly accurate, and other measurable parameters seemed top-drawer, in the end these were not very good photographic tools. They were neither rangefinder cameras nor point-and-shoots, but rather something in between. The viewfinders were too small, and they contained several other flaws which we have no place to explore here and now. I owned both models and ultimately abandoned them, as did many photographers who I know who thought that these cameras would be alternatives to an M6. They weren't. (If you're a Contax G1 or G2 aficionado— great. Enjoy. Please don't write telling me I'm wrong. I simply have a different opinion than you, and this isn't the real point of my example in any event.)
Here's the thing — the Sony F828 may be noisier that some competitors, may have more CA than other digicams in its price range, and it certainly does have a number of other flaws that I have enumerated in my review. But, it also has some features that lift it above run-of-the-mill digicams, and which allow it to perform as a truly useful photographic tool. These include..
— a superb 28-200mm f/2-f/2.8 (equivalent) Zeiss lens that combined with the camera itself can literally fit in the palm of ones hand. This speed and focal length combination meet the needs of the majority of photographic situations, and would require two large, heavy and expensive 35mm lenses to provide equivalent coverage and speed.
— a responsive manual zoom. Digicams with electronic zooms simply don't offer the responsiveness necessary to compete head to head with interchangeable lens SLRs.
— fast turn-on time, low shutter lag and very quick autofocus. I have tested digicams that can take five seconds or more to turn on, and which can take many seconds to autofocus and release the shutter. The F828 is ready to shoot in about 2 seconds, comparable to many DSLRs, and it can autofocus as quickly.
Without such characteristics no camera, even one that was absolutely noiseless and totally free of CA would be an adequate photographic tool. A camera that can produce technically perfect images, but which isn't able to properly capture them is useless, and no better than having no camera at all. On the other hand, a camera that has some image flaws, but which doesn't get in the way of capturing the moment, is a joy to use, and small image quality flaws can be forgiven when the image comes before the technology.
Maybe the grapes do taste good after all.
In the weeks following the publication of the above essay I have received literally dozens of e-mails from readers supportive of the thoughts and sentiments expressed above. Interestingly, not one has been negative, though there have been a few critical naysayers on some Net forums. In any event, I think that the letter by Mike Sims below presents another perspective on the issue, and it is therefore reproduced with permission.
Ps: I certainly don't see myself as another Ivor Tiefenbrun, though I too consider myself an audiophile, and enjoy a high-end tube-based stereo system. Measures terribly, sounds great.
I found your article fascinating regarding a number of issues. In particular the "hijacking" issue.
This is not the first time that technophiles have hijacked discussion of the merits of equipment in the way that you describe - looking only at technical data and thus loosing track of what the equipment is actually used for.
This happened many years ago in the area of High Fidelity (Hi Fi) music reproduction, and I agree with you that it is happening now in the area of High Fidelity photographic reproduction.
I draw the parallel because I believe that examination of what occurred in the world of music reproduction may throw some light on what is happening now in the world of photographic reproduction.
I got interested in Hi Fi in the seventies. Transistor powered hi-fi had recently become affordable to a huge number of people. There was a lot of hi fi stuff around at all price levels. There was a huge interest in the technical capabilities of electronics, and of course in the abilities of loudspeakers to reproduce (in particular) low and high frequencies.
Previously, many people had marvelled at the fact that using a steel needle, a revolving table and a horn you could hear a scratchy and very constrained recording of music or song. It was not much like the original, AND DID NOT PRETEND TO BE, but still the reproduction conveyed something of the original and most were delighted.
With the advent of Hi Fi the claim was made that this equipment could convey Faithfully the original sound. But it did NOT! Those of us that listened to live music knew that it didn't! But it should!
So we started to listen to the equipment and not to the music!!!!!! We examined every detail of the equipment, distortion, signal to noise ratio, wow & flutter etc etc, believing that if we could only improve the specification that little bit more we would (at last) have a system that sounded perfect.
Amplifier manufacturers advertised the amazingly low distortion of their amplifiers, loudspeaker manufacturers emphasised the flat frequency response, record deck manufacturers the vanishing small amounts of wow & flutter etc. They never ever mentioned what the equipment sounded like.
Then, out of the blue in 1972 Ivor Tiefenbrun, owner of a Hi Fi manufacturing company named Linn Sondek shook the Hi Fi world by suggesting that he had built a record deck that could, irrespective of the way that it measured, convey the musicality of the performance in a way that no other record deck could.
He was laughed at by many in the industry, and by the pundits - until people started comparative auditioning his record deck against others. People found that they were beginning to compare the musical reproduction in terms of the satisfaction that the performance gave them - truly a revolution!
It is now accepted that we must judge Hi Fi equipment by its ability to move us, entertain us, satisfy our musical needs.
I suggest that digital cameras have been an equipment revolution in the photographic world in the way that transistors were in the music world. The first digital cameras were poor, and since technology provided the digital camera in the first place we have expected technology to improve them.
So we improve whatever we can measure, more pixels, lower noise, less distortion etc, forgetting that what we are after is a device that can deliver the highest percentage of keepers (aesthetically pleasing pictures) whenever or wherever we use it.
Perhaps it is up to you to become the Ivor Tiefenbrun of the photography world by pointing out that no matter how the damned thing measures, what counts is what percentage of keepers the guy (or gal) who paid for it can get by using it.
I wish you luck
You may also enjoy reading a related essay titled:
The Case of the
Nit Picking Pixel Peepers