Let's Play Another Game
The Digital Middle-Ages are Now Over
Following the publication of my recent review of the Canon 1D Mark II the issue of image quality came up in a large number of e-mails, and also on this site's discussion forum. Everyone wanted to know why I hadn't commented in the review on the camera's image quality. I addressed this in an addendum to the review that same day, which read in part...
Why no mention of sharpness?
I didn't mention "sharpness" because there isn't much to say. The camera is from the same image quality mold as the 10D and 1Ds, and Rebel for that matter. The imaging chip may have slightly different pixel pitch and size, but otherwise they are of the same design, though as I stated each generation gets a bit better.
People shouldn't look for some dramatic change in this area. The real story with this camera is the shooting performance, not the image quality — which is terrific.
At 8MP it will produce bigger prints than those from a 10D and smaller ones that those from a 1Ds. There will be small differences in noise at any particular ISO and small differences in colour balance. Otherwise Canon's current cameras produce pretty similar image quality. The rest is pixel peeping.
The Internet has produced a most curious phenomena — individuals who are fixated on image quality, often to the exclusion of anything else. I have recently written two essays that explore different aspects of the subject; The Case of the Nit Picking Pixel Peepers and Digital Bridge Cameras and Cognitive Dissonance. But I feel that a closer look at the topic deserves more than a few paragraphs in an addendum to a review of a single camera, and so here are my further (and likely final) thoughts.
Before the Pixel
I believe that debates over image quality are becoming largely irrelevant. In each category, 5MP digicams, 8MP digicams, 6MP DSLRs, etc, while there are modest differences between cameras, these are actually now quite minor when real-world prints are viewed. Usually what people obsess about at 100% on-screen disappears when seem in a print, especially when every other aspect of image processing and production is less than exceptionally well done.
Prior to the appearance of semi-affordable DSLRs in the late 1990's photographers always enjoyed discussing how different lenses affected image quality. Zeiss vs. Leica was a popular topic, with Zeiss partisans rooting for contrast and Leitz aficionados on the side of resolution. Then in the 50's and 60's the debate was — were Japanese (read Nikon) lenses as good as German lenses? In later years it became — is a $250 Sigma/ Tamron / Tokina as good as a $1,500 Canon L lens?
No one debated whether a particular camera body produced better images than another because people realized that unless it was out of alignment or had a faulty shutter mechanism the camera itself doesn't affect image quality. It was simply a light-tight box for holding film. Of course one could (and did) debate ergonomics, features, handling, build quality and materials, but to do so meant that you actually had to go out and use the equipment to take photographs to be qualified to comment on such issues.
Canon 1Ds with 500mm f/4L lens at ISO 250
Into the 21st century
But in the late 90's as silicon began to replace film the game changed. In addition to being a light-tight box cameras become the recording medium as well. The quality of the imaging chip, its resolution and its noise characteristics were important factors. Some were better than others (some were downright poor). This, combined with the ability to post large files on the Internet, along with essentially unmoderated discussion forums, allowed anyone to critique a particular camera. You didn't have to be a photographer. You never even had to have seen let alone handled the camera that you were critiquing. You simply had to download a file and tell the world what you thought you saw.
Experience? Credentials? Who cares! One opinion is as good as another, right? Now everyone could be a critic.
This was fun for a while, but it's really started to pale, and here's why.
The Differences are Shrinking
The truth of the matter is, that there is now (Q2 — 2004) precious little real-world difference between comparably speced cameras when it comes to image quality. During the past year or so I have tested digital cameras and digital backs from Nikon, Canon, Kodak, Pentax, Olympus, Leaf, Minolta, Phase One, Imacon, Sony and many others. These have not been casual tests. In almost every case I have shot with these cameras and backs on locations around the world, in each case from hundreds to thousands of frames. I have dissected image quality, looking at differences at various ISO settings, long exposure setting and so on. I have put these cameras through tests on the DxO Analyzer system (a new $20,000 optical testing system designed for digital cameras and lenses).
Do I see differences? Of course I do. In almost every area with every camera. But — and here's the main point, so don't doze off — the image quality differences between competitive cameras are becoming very small indeed. In fact when I'm doing comparisons unless I am very careful in naming files and labeling prints it is sometimes almost impossible to tell which is which.
In the early days of digital cameras such differences were more noticeable. But with the current generation of digicams and DSLRs the differences are becoming increasingly small. Sure, sometimes one camera's images are slightly better than another in a specific performance area, such as noise or colour reproduction. But it's extremely rare to find a real dog.
Read carefully. I'm not saying that there are no difference. I'm saying though that the differences are small, and while sometimes noticeable when doing a direct comparison, in prints up to A3 they are rarely an issue. The differences that I prefer to report on, but which don't lend themselves to Sunday morning quarterbacking, are the ones which I believe make a much bigger difference to serious photographers — features, ease of use, build quality, user interface and the like.
Most pros understand this. They make commitments to gear based on their shooting style, previous lens commitments, personal preferences, local dealer support, and so forth. And once they've bought one of these tools they stick with it until something considerably better comes along, and better usually means more productive, not just something that occasionally shows a slight image quality difference.
Where's the Beef?
Here's the crux of the matter. Photographers now need to turn their attention back to issues that are much more important when it comes to judging cameras. Things like user interface and ergonomics play a far bigger role than the minor differences that can be seen between images from one camera and another.
Case in point. In March 2004 I tested all five new 8 Megapixel digicams from Sony, Canon, Nikon, Minolta and Olympus. All five use the same Sony-sourced imaging chip. I've done extensive tests and comparisons between them and have had all five cameras in my shooting bag on location at the same time.
Frankly, there isn't much difference between them when it comes to image quality. I could take shots done with each, make A3-sized prints, lay them out on a table, and no one would be able to say that one was demonstrably better than the other. Yes, there are small differences visible, but small is the operative word. (Anyone who has seen my exhibition prints or attended one of my seminars knows what a stickler I am for image quality — so it isn't as if my standards are particularly low either.)
On the other hand there are huge differences in the way that these cameras handle. A couple of them are so badly designed that I could hardly bear using them. They frustrated the hell out of me. Others worked fine but had some annoying operational flaws. One stood head and shoulders above the rest when it came to usability, but it appears to have some quality control issues.
These are the things that matter — Usability. Build quality. Suitability for the task. Design ethos.
But, these aren't things that the pixel junkies can load up on their screen at home and critique. To evaluate these you need to actually go out with a camera and take photographs, and you need to have some experience as a photographer, as well as a frame of reference from which to draw conclusions.
Sony DSC-T1 @ ISO 100
Frankly I despair that this will change. The marriage of Internet discussion forums along with the ability of anyone to analyze to death those things that lend themselves to on-screen scrutiny is a tide that can't be turned. While it gives me some solace to air my concern I know that few will find these comments of more than passing interest, and many will deride them.
There are now so many newcomers getting into photography because of digital that they naturally want to know what's what. What should they buy? What's good? What isn't? How much should I spend? Which camera takes better pictures?
Tell someone asking these questions that there isn't a huge difference — that most comparably priced and featured cameras will produce comparable images, and that issues like handling and user interface are much more relevant, and he'll ignore you. People simply don't want to hear this. I've seen it in action. Just stand at the sales counter of any large camera store and listen to the dialog between salesman and customers. The customer wants to know "which camera takes better pictures", not which one has a CF card slot so badly designed that the card can't be extracted without a pair of tweezers.
Therein lies the problem. If a camera is so poorly designed (as many are) so as to get in the way of straightforward usage, nit-picking differences in image quality are largely irrelevant.
In The End
I know from countless e-mails that I receive weekly that there are a large number of photographers out there who understand these issues. These are primarily pros who use their photographic equipment on a daily basis to make a living, and who understand the importance of the issues of ergonomics, design, build quality and user interface. They know that when it comes to a camera like the Canon 1D Mark II, for example, the fact that the image quality is going to be somewhere between very good and excellent is a given. What they care to read about about are the things that I care to write about — how it handles in the field; how well its features and functions perform their intended tasks, and whether as a tool (not a fetish object) it can get the job done.
Because I know from the correspondence that I receive, and from personal discussions at my seminars and workshops, that there are large numbers of serious photographers out there that understand these issues, I will continue to focus in my digital camera reviews on the topics that I consider to be of importance. I will also continue to rail against manufacturers who sell us cameras that can produce very fine images but that are badly flawed in design and execution.
If you have an interest in the same things that I do, I look forward to sharing my impressions and results with you in future tests. But I'm no longer going to count angels on the head of a pin — or pixels. I've been guilty of feeding the obsessions of such folks in the past, but I no longer am willing to play this game. The digital Middle-Ages are over.
I enjoy reviewing cameras as well as lenses, software, scanners, monitors and the countless other photographic products that interest me and which cross my desk each month. But I'm about to change the rules of the game, not just in what I preach — as discussed above — but also with regard to what I do.
There are numerous very good web sites which review digital cameras. These include DPReview, Steve's Digicams, Imaging Resource and Megapixel.net. If you are the kind of person that wants to read in-depth reports on each spec and knob, as well as comprehensive image quality analysis, I recommend these sites to you. I now intend for the most part on leaving these subjects to those that continue to have an interest in them — both the people that write them and the people that read them.
I will continue to review cameras on this site along with the wide range of other photographic tools that I'm interested in. But I will focus my attention on those aspects which I regard as being of greatest importance — as outlined in the above essay. If a camera's image quality stands out in any particular area I'll comment on it and even analyze it, but otherwise I plan on focusing my attention on what I consider to be of greater significance — design, features, ergonomics, build quality, operational ease, and suitability for the camera's intended market and task.
Michael Reichmann — April 1, 2004