« DxO's Methods Explained | Main | Open Mike: Off to the Races »

Saturday, 21 April 2012

Comments

Perhaps a post for all the Canon fans who were disappointed about the 5dIII scores?

If Canon had been at the top of the list this time arounds, would you still have written this?

Or would you have reached the opposite conclusion (that scores do matter, because improving technology is what makes new photo opportunities possible)?

Something to ponder.

Very well said Mike. I totally agree with you on this. It explains why so many are playing around with Holga's and vintage cameras which can produce outstanding or unique photos in the right hands. Pixel peeping now seems to be the driving force behind the development of new sensors and camera sales.

While looking at somebody's work, "picture quality" has never been my top priority. Film and digital change and evolve constantly, it's the photographer's view what really matters.

Someone said to me "technology has made some great progress on the past few decades, thank God nobody takes pictures like those anymore".

Like those meant the Capa exhibition we were at.

I want to marry you, Mike. This is such a complete, careful, and well written essay encapsulating so much of how I feel about pixel-peeping, sharpness, and the overall obsession we all have (me included) with technical muckery.

The best I can manage, usually, is to grate out angrily "POINT THE CAMERA AT SOMETHING GOOD" which isn't quite it, because it's not the "subject" per se either, it's the whole image thing in all its dimensions. But I'm not very coherent, so that's mostly all I have on tap!

I think the idea that the history of photography is a trend toward more convenience is an interesting observation. Perhaps more to the point I think that the trend is toward more accessibility. Almost everyone now owns a camera of fair quality in some form. This accessibility has greatly changed photography. For one photographers are less iconic now. Until just a few years ago near the beginning of the digital photography age, one could fairly certain say who took a particular photograph. Most published photographs and the photographers who took the pictures were well known. Secondly the vast increase in accessibility to photographic equipment and the accessibility to photographic images for viewing have given us many more images to view resulting by the law of averages in many more quality images. I appreciate the Greeks' liberal appointments of gods. but the number of great images available for viewing is nearly sensory overload for me. I guess I still need a manageable number of icons.

The question is not bragging about a camera and its DxOMark score, it's knowing the tool you use with its strengths and weaknesses.
Nothing is perfect in this valley of tears : there are also image quality parameters ignored by DxOMark (pattern noise comes to my mind), and yes this whole scoring thing is not that helpful (except for bragging, of course).

But in the end, DxOMark helps to know these strengths and weaknesses, past marketing nonsense and common wisdom (or supposed so) : merci les gars!

While I agree wholeheartedly with your post, there is another side. It is much easier to remove quality* than it is to add it after the fact. Why did portrait photographers buy the sharpest lens' for their Hasselblad, only to put a soft focus filter in front of it? This is the same argument for using RAW. The more/better information you start with, the more lattitude you have for creating your vision.


*for some value of quality, be it sharpness, resolution, dynamic range or what have you

I'd guess that a lot of people are interested in which sensor is "best" for almost purely non-photographic reasons. Instead they're interested in what imaging technology is capable of - what sharpness/dynamic range/focus speed/etc. is possible, not what actually shows up in images.

I think this is akin to people caring about what a cars top speed is, and which car has the highest top speed, even when they'll never utilize those cars to their maximum capability. (Or highest G turn, or lap speed, etc.)

An excellent, sane post in the midst of a growing, thundering waterfall of camera insanity.

How about next discussing software?

"You don't get points for using the camera with the best sensor . . ." I agree.

As a fulltime freelance writer who shoots photos to illustrate my stories, I have had literally hundreds of pictures published in magazines that were taken with a 3-megapixel Olympus D-550. No editor has ever commented negatively on the quality of my photographs; they simply printed them.

When I first acquired the D-550, I ran an experiment. I submitted a story with half the prints generated from 35mm print film and half printed on paper from the D-550. The editor made no comment whatsoever. When the printed magazine arrived at home, I asked my wife to tell me which images came from the digital camera. She guessed wrong. So at this point, I figured the D-550 was "good enough."

Note well: I am simply a working writer, a yeoman photographer. I don't sit up nights waiting for the Pulitzer committee, and no one will confuse me with a "fine art" photographer. Do I get a charge out of big, clean images taken by master photographers with awesome lenses and sensors? Of course.

Great post, Mike. Perspective plays an obvious role in composition, but it is less often regarded when considering the equipment used.

Bravo. One of the reasons I'm a loyal reader of TOP. While I like to keep abreast of the latest advances and have the best gear I can afford, I'm much more interested in how I can take more interesting pictures that will compel people's attention. And, I think on balance, you cover more of the latter than the former.
Adam

Diego,
You're assuming that the purpose of today's post is to defend Canon? It really isn't.

Mike

Brilliantly true, Mike. I generally only print at 13x19 or smaller, and the decision for me to go from a bulky A900 to NEX cameras was simple. The IQ of these mirrorless cameras is plenty for anything that I'd personally do.

....."Many of you remember how much we were consumed in the first five years of the present millennium with the question of whether digital could, or had, matched the quality of film—not a question anybody cares about any more. "

I remember those times well, and am reminded frequently by countless silly forum debates....still. Evidently the issue hasn't gone away for many.

Better tools give us more options, more possibilities. They don't guarantee better results, but they do make new things possible to do and they give more choices for the photographer.

On a side note, it's also good that there's serious testing such as that of DxOMark available. I'd loathe to have photography to be dominated by the sort of mysticism that is prevalent in amateur audio.

I have seen a few award winning photos taken by photojournalists in the early 2000s which were tragically taken with low resolution DSLRs, possibly Nikon D1s. They look terrible when printed in large sizes. It is here that I start thinking "if only they had used film, or a better digital camera (which probably didn't exist at the time)".

"it's also good that there's serious testing such as that of DxOMark available"

Indeed it is, however, that doesn't mean that people won't argue the findings if they don't agree with their prejudices. If you already like Canon, it's fine if it scores high. If you're a Nikon fan there must be something wrong with the test, or they measured the wrong thing!

It all goes back to Magic Bullets: the illusion that buying a "better" camera will somehow make you a better photographer.

File under: "We hold these truths to be self evident..."

Mike, thanks for a sane comment in the storm. It's too easy to forget that what makes a picture interesting or powerful is very rarely absolute image quality. To be fair though, I think expectations for sharpness have increased as digital has matured.

The image linked below was not 'captured' using a state of the art dslr:)

http://www.rangefinderforum.com/photopost/data/500/Buttermere_1_of_2_.jpg

Best,

Mike

'While I agree wholeheartedly with your post, there is another side. It is much easier to remove quality* than it is to add it after the fact. Why did portrait photographers buy the sharpest lens' for their Hasselblad, only to put a soft focus filter in front of it? This is the same argument for using RAW. The more/better information you start with, the more lattitude you have for creating your vision.'

Whilst this is true, I think that this is a very different approach to working with a different medium. Although I fairly often convert digital images to monochrome, I find it much easier to work with black and white film if that's the intended end point.

In my experience there's a difference between a pre-dertmined outcome and making a series of choices using a computer after exposure. Adding a softar is a pre-exposure choice, not post.


'On a side note, it's also good that there's serious testing such as that of DxOMark available. I'd loathe to have photography to be dominated by the sort of mysticism that is prevalent in amateur audio.'

Actually, I htink that a lot of the adverse, unbelieving comment is very similar to audio mysticism. Some of the commentators seem to have a very similar approach - you can't measure it, but if you're not blind you can see it...

Regards

Mike

Extremely well put, Mike. I have a Nikon D800 and I can attest to the fact that this camera has not improved my photographic ability one iota. My buddy just recieved his 5DmkIII and, hot damn, that's what I wished the Nikon D700 would have been (only I didn't know it at the time :-))

Interesting essay, Mike. I have to agree with you. Cameras are just tools, and the DxO Mark tests allow us to compare some of their characteristics.

Improving technical capabilities of digital sensors have expanded the potential of what it is possible to capture with a photograph, but a good photograph is still the result of the photographer's vision more than the tool he or she is using.

I photograph fast-moving subjects containing a lot of fine highlight detail, and wide dynamic range, often in fading light. For this type of subject - breaking waves - the DxO Mark sensor comparisons provide a useful guide to which cameras offer the best capability to do what I want (if only I could afford to regularly switch systems!).

Having said that, some of my favourite wave photos were taken with a little Olympus XA and bw film.

Fantastic article, Mike-I have a 16x20 aerial pic I took of our place on my wall. Taken with a Canon 300D (6 megapixel) and Mk1 version of the 18-55. Virtually indistinguishable at first, (or even second) glance from the same pic taken last October with my 7D and 24-105. I sell cameras part time, and we're going to have a hard job convincing people that a Canon at 18 mp is going to be just as good as the new Nikon 3200 at 24 mp. Unfortunately, the megapixel race is on again.

I'd say that today's cameras are still sometimes insufficient for a rookie like me. I'll screw something up and lose a picture that I would like to have had, maybe by hopelessly blowing the highlights, or accidentally leaving the camera on ISO 1600 after using it last night, or missing focus, or whatever. But as the cameras get ever better, they can save me from myself more often. For the truly proficient, I imagine that the "almost always sufficient" point may have been passed a while ago.

Mike,
This was most thought provoking and well written. My first "good" digital was also an Agfa. It was around 3 mp I think. It had the rotating main body, a la the Nikon 900 series. I loved it. I remember my first photos giving me the same kind of thrill that my first darkroom experience did years earlier.

Your point is perhaps best illustrated by the rise of iPhonography (sorry Kirk).

The picture is always about the picture.

Thanks for stirring up some useful thoughts about digital photography for me.

Cheers JD

Similarly, phonographic discs were inferior in sound reproduction to wax cylinders- but considerably more convenient (as are CDs to LPs).

Excellent essay, Mike, and a good tonic for much of the insanity now gripping "photography" (i.e. gear) forums on the interwebs.

My midrange DSLR produces files that make near-perfect 20x30 prints, far better than I ever made in a darkroom even with medium format. For me, the resolution wars are over. Now it's time to find more vision...

This article is one reason I'm excited by the Canon G1X: good enough for practically any published work, lens covering all my usual photographic needs, even professionally. My only concern is that it makes a sharp and correctly colored 11X14, past that, who cares...

My goal is to make my end of the professional photographic process as easy as it was when I was taking film...now if everyone (instead of just Nikon), would make a .tiff file part of the selection (instead of just jpeg and raw), it'd be all over for me but the crying...

What about the evolution of lenses? With no empirical evidence for this statement, I bet the qulaity of the lens, on any quality of camera, has always and will always make more of a difference than the pixel count (or DXO score) of the camera. Any thoughts?

cfw

I'd still like to see how film does on a DxO score. Frederic Guichard says that digital sensors have, in practice, a wider dynamic range than film. How does film compare to digital color rendering? I'd really like to know, but of course who prints a rigorous test like that? Do digital sensors really have, in real-life practice, a wider dynamic range than film? According to what was published on Twin Lens Life a while back, they don't. So yeah, I'd like to see digital and film rigorously compared.

Quality only matters when it really does matter. That's why, for over 100 years, large format cameras have been hanging around. No small camera or sensor beats these behemoths. Are they easy to use? Of course not. But do they deliver quality? Oh, yes. Are they popular? Only when you want top quality. A while ago there was a comparison of an IQ180 back to an 8x10 camera by Tim Parkin. Good comparison, with a good attempt to control the variables between them. Film more than held its own.

Yeah, the DxO mark matters. With film, if you don't like one emulsion, you can immediately swap in another. With digital, you are absolutely stuck with that sensor. You want something different? You have to buy a brand new camera, like it or not. So yeah, the DxO tests matter.

Hi Mike,
Just wondering then why you are upgrading from the GF1 to E-M5?
I'm considering it myself as a present GH1 owner, but if image quality (dynamic range) turns out to be the same I don't think I could justify the purchase.. even with the other differences (improved handling, sensor stabilization etc).

Yes ! Imaging-resource reviews include a subjective evaluation of prints that basically tells me any state of the art m43 or APS-C camera meets my needs. dxomark is good for sanity checking, maybe good for those trying to determine whether an upgrade is worth it, and good for those of us who are just endlessly fascinated by the details, but the practical differences between state of the art 43 and APS-C cameras is really insignificant compared to practical concerns (usability, cost, system accessories). So my last upgrade was driven by a desire for a camera with a quiet shutter, not by a desire for better image quality. My 4-year old Sony A700 did fine, thanks ! (Not that I don't appreciate the cleaner & more detailed results from my new D7000).

"...the entire history of photographic techniques, with a couple of important sidesteps, has been a steady march toward greater convenience, not greater quality. (The one really big sidestep was the advent of practical color, which was indeed a qualitative evolution. But we'll leave that aside for now."

Never one to leave aside something with which I disagree -- please consider that color, practical or otherwise, can also be seen as a distraction that lowered the quality of photography. I certainly view it that way.

"...the question of whether digital could, or had, matched the quality of film—not a question anybody cares about any more."

On the contrary, those of us who shoot with 5x7, 6.5x8.5, 8x10 and larger view cameras do indeed still care about that question. For us, the answer is "not yet."

"...the swift shift away from printmaking and toward online viewing of pictures..."

Monitor screens are totally unacceptable as image viewing devices. That the masses accept them for this application earns no more respect from me than did Disc Camera machine prints. Some of us aren't even satisfied with enlargements from our view camera negatives. :-)

Spot-on, Mike.
Your insight about convenience rather than quality as the driving force is pitch-perfect. A similar evolution can be seen in recent audio technology.

But a bit of preoccupation with pixel-peeping or MTF curves can, within reason, sustain a quest for creative freedom. Just as, in film days, we used to seek exposure latitude, or slide films that would combine decent colour rendering and sharpness at slightly more than ASA 25. It's good to know that an affordable sensor combines good dynamic range and comparatively low noise at higher ISO: such margins allow for a more easygoing approach. An audio analogy: In the reel-to-reel Jurassic, we used to have heated debates about higher pre-magnetisation of the magnetic tape, which yielded a staggering 1.5-2.5 dB more signal-to-noise headroom. Nowadays, even my cheap pocket recorder, really a glorified digital dictaphone capable of PCM recording, provides way more dynamic range and better linearity than the Nagra IV-S of my radio days — the Leica M of tape recorders — ever had. I'm getting acoustic takes under conditions which would have overwhelmed my previous analogue equipment, and I'm getting them because I'm care-free about getting them. It is often said that technical limitations of the equipment can be a factor for creativity. I find the converse equally true: not having to worry about too flagrant limitations of one's equipment can liberate one's energies and help concentrate the mind on the image rather than on ancillary servitudes.

Tonight, I was startled to read Lloyd Chambers' speculations regarding the 24 MP sensor in new Nikon D3200. Lloyd has not been hitherto known for reviewing cheap cropped-frame cameras, but even he wonders whether Nikon was able to let some of the D800 magic trickle upon their new entry-level DSLR. Quote Diglloyd: "…if the sensor is good, the Zeiss 25/2.8 Distagon or Zeiss 28/2 Distagon or similar might prove very interesting, which is why I might review the D3200…"
Grab a really good lens, if you have one, put it on a cheap no-frills camera with a good sensor — the modern equivalent of a film holder — and off you go, catching the light and painting with it. It's back to basics, in a sense, and if DxOMark helps us identify the best "sensor holder"* for a given lens and a given budget, more freedom to us, and kudos to them.

(* Yes, I am aware that the processor and associated electronics are also essential; I'm just focussing on the film-sensor analogy.)


Quote: "Does DxOMark Matter?"

Answer: NO

Great post. Other than an inherited Fujica st705w* which I took maybe a dozen rolls with, I "grew up" shooting digital. With my DSLR, I find myself thoughtlessly shooting 200 photos a day (many poorly exposed because I can't use the camera's Program mode and be a serious photographer right? RIGHT?) and then bemoaning my less than stellar results and trying to post process the pain away.

As a sort of cathartic therapy, I've started shooting with a Holga and trying to squeeze decent images out of it. The cost of film and the fact that DxO would score it (-)47 keeps me aware of what I'm doing and focusing on technique.

*Unfortunately, the Fujica was stolen when our house was burglarized in January, along with my dSLR containing photos of my son's first birthday. Rot in hell, burglars.

There are dozens of photos that would prove your point. I've made something of a jerk out of myself on a few forums (including this one? I can't remember) by suggesting that no great photo -- ever -- has depended on resolution for its impact.

But for those that think that resolution does have some critical importance, I suggest a study of the Ruth Snyder execution photo. What possible technological advance could have increased power?

This could, at least in part, explain why so many of us have had to wait so long for what we consider to be cameras made for photographers. It seems to only be in the last few years that companies have started to spend less time pushing the boundaries of what is scientifically possible in favor of refining and enhancing what is required to perform the act of making pictures. After all, how many of us are buying cameras based on how well a speck in the distance looks blown up by 500% ?

If you use full frame and m43 cameras at the same time, you will notice the difference reflected by sensor score. Among many, one example is DR, blown high light is always blown, no matter how small you print or you only view online, shadow noise is always higher on 43 as well. Do a long exposure and if you try to post process you will see. So yes, the score does matter

"...the notion that "image quality" is an essential building-block of a successful photograph. Actually, it isn't. Actually, it isn't at all."

AMEN.

Right now I am working on 40x60" prints done with stitched photos from a Panasonic GH2 and an old Kern Macro-Switar. The (technical) quality is simply breathtaking. I would not have been able to produce so much detail and subtle rendering with my 5x8" Sinar.

But that is only one side. The other side is, that with digital photography I somehow have totally lost my love for taking pictures. Now I feel more like an operator of a very complex picture taking machinerie.

I hate this soulless, ugly machines with 30 knobs or dials. I only need a good finder—which do not exist anymore—and a trigger. MLU combined with a self-timer always helped and still helps. And I want to control the aperture. Autofocus is useless, it most of the time degrades the nominal resolution of a given camera to one third.

Since there is no more such a thing, which was once called finder, liveview is a very welcomed new invention. But it per se detaches you from the camera. And that melting process—the eye seeing through a camera—was once the core of picture taking. Just to see and to learn, how three dimensions translate into two dimensions.

The fulfillment I once had in composing a picture is totally gone now. That is, because I really feel a disgust in putting a computer between my eye and any given scene. This changed my way of photography, totally and forever.

And yes, the thing with the numbers: Everyone who has any insight, would tell you, that there is such a big margin of error and unpredictability (if this is an english word),
that you simply can forget about it all. It does not matter. My best photos would be marked by DxO with 4 to 8 out of 100 in sheer technical picture quality. And I now make my living out of these.

MJ has to be thanked a lot for bringing this up.

Lucid, cogent, and thoughtfully argued. In six short paragraphs you've provided as much insight on the current state of image quality as anything I've read.

I typically enjoy your writing, Mike, but you've really hit it out of the park with this one.

Despite how quantifiable and objective DxO measurements are, I doubt that any serious photographer buys into a DSLR system because of DxO scores. Real world ergonomics and subjective "look" of the out-of-the-box files delivered by the combination of the camera paired with one's favorite (most-often used) lens is the ultimate decisive factor.

The trouble with most reviews is that they all appear to be written for the landscape or architecture photographers who have the patience to use tripods and choose the best lens for the occasion - a scenario that isn't realistic for street/candid/travel shooting where what you have on you at any given instant is all that matters, regardless of any scores. I suspect the latter is the scenario that applies to the vast majority of us, including most professionals who often end up renting Medium Format systems on most commercial client assignments anyway.

What I find amusing is the audacity of many so-called reviewers to declare one camera "objectively better" than another based on controlled tests that pay no heed to the diverse needs of professionals working in different kinds of photography. The only thing more amusing is the fact that so many novices buy into these "reviews" and seem to get caught up in these "scores" to rage against anything that allegedly denigrates their chosen brand "religion".

How many technically "epic" images does one care to see of someone else's pets, kids, and friends on FaceBook or Flickr that might be capable of being printed tack sharp 24"x36" but in all likelihood will never be printed beyond 8.5"x11", if even that?

It seems to me that for the first couple of decades of digital photography, much of the effort was devoted to making digital at least as good as film. Now that we're there (and have been for a while), things are getting interesting because effort will be spent on making digital do things that film never could. Any of the readers of this site can think of examples: HDR, stitched panoramas, focus stacking, light field (Lytro). My personal favorite is Nokia PureView. I'd like to have a pocket camera made with that technology. http://conversations.nokia.com/2012/03/07/the-story-behind-the-nokia-808-pureview/

Excellent. I wrote a similar take on this very issue about a month ago on DPReview as well: http://www.dpreview.com/articles/1747066263/my-little-photo-essay-how-i-learned-to-stop-worrying-and-love-photography

This is not a "defense of the Mark III" but the Mark III hullabaloo is a case in point that forces the discussion of this issue for sure.

Drew

I should add that in regard to my previous post that it was my observations about the Reiman magazines that caused me to start pondering this a long time ago!

Drew

There has always been heated discussions in photography, but it used to be a privileged discussion.

For me, the "thing" with the digital evolution is about countering the old art challenge of elitism. Basically anyone can create wonders in their computer now-days, and anyone can take part in the discussion.

The question is not to find the universal answer to pixel-peeping, the question is to find out what's right for you.

I'm a great fan of Shorpy', of historical photography in general and I think the technical developement is goes further than just increased convenience. There are things posible now, which were not possible a 100 years ago, which to me makes it qualitative changes, not just color. For instance in this shot from 1908 (http://www.shorpy.com/node/12778) there are several figures blurred or even reduced to ghosts because they walked too fast. And this was in full daylight.

From another field: in classcal music, there is a movement to use 'historical' instruments, that is use the gear the composer had available. But as a conductor once said: Of course, Mozart would have loved a modern piano instead of his quirky pianoforte.

I agree with the main points--"image quality" has little to do with picture quality--but of course it is understandable that pros who are in competition to produce images with an edge over their peers, and shmoes (me included) who simply don't want to feel foolish for committing a lot of money to something when they could have gotten something else that worked better, do want to evaluate their cameras/potential purchases.

For quite a while, while cameras were light-tight boxes into which we all put mostly the same film, there really was not much difference. Autofocus changed that (for some genres); now you actually could credit or blame the camera for pictures working or not. And digital completely changed the game, because different cameras were (and are) making new kinds of photos possible with every iteration: pictures made in ever increasing darkness, or across dazzling dynamic range scenes, or photos which work because of the juxtaposition of more and more discernable details or textures. I for one went from favoring a 90mm lens on my om3, to a 50mm on my 5d, to a 35mm on my m9, in large part because the increasing resolution made it practical to compose increasingly complex scenes, which needed to be printed larger to "work" as intended. (for other reasons too, of course.)

I guess my point is just to say that even if realizing the ultimate capabilities of our present camera is rarely the road to exceptionally good photographs, the changing potential capabilities of our cameras affect the sorts of photos we attempt to create (which you already pointed out under the heading of convenience). So, i guess, we shouldn't underestimate the significance of convenience in the generation of great photographs, at any level.

IQ is never as importent as AQ (Artistic Quality) but IQ can be measured, can be fussed about, can be compared, and can become a goal in itself.

But the trend towards ever bigger pictures by professional, Lartique printed 30 x 40 (cm not inch) nowadays you see 6 meter prints by Gursky and companions. That drives IQ up the wall since these still have to maintain image detail in order to "work".

But for you and me who print at 13 x 19 12Mp is okay. I wonder what his gourouness Ctein will be able to wrench out of his Oly. I guess that will be the ultimate test (personally I think 26 x 40 can be achieved with good glass a high contrast subject and some nifty upscaling).

Greetings, Ed

I'd like to make a couple of points...

1.Part of the greater convenience of digital cameras is their greater versatility. One camera can be used to good effect for stills & video, still lifes to action, bright light, low light, hand-held, fill-flash, B&W, colour, etc etc. Amazing, when you think about it.

2. I have a different take on the cause / effect indicated in this sentence:

"the early years of digital were consumed with an anxious search for adequate, affordable quality — resulting in habits of mind which have definitely continued to now"

I've read that something like 25% of the population are analytical / details orientated. Digital photography didn't create "pixel peeping" ... it just gave analytical people something else to obsess about. Compounding this, analytical people (I'm talking about nerds, here) have little understanding of aesthetics & personal expression. When they look at a photo they only see the pixels, the noise, the slightly crooked horizon, the incidental person cropped at the edge of the frame.

Hmm. I am not so sure about this.

Photographers have been obsessed about quality as long as I can remember, it's just that achieving it was usually out of the price and skills bracket for "normal" people. However within that price bracket, there was a lot of discussion and quality obsessed nonsense. I remember. I was there.

Thing is only "professionals" could typically dedicate the time and resources to LF or MF photography and learn the required skills. We "consumers" had to accept whatever manufacturers could give them and whatever shoddy film processing we could afford.

35mm changed the world. It was easily good enough for reportage (newspaper prints were a low quality standard) but it could get you places that a LF camera could not. However improvements in 35mm film meant that this small, portable format became the amateur standard. It was relatively affordable, adequately portable and produced adequate quality. However it was hamstrung by the technology itself. Film improvements were glacial.

But what would Capa have done with a D3? What digital has done is achieve both convenience AND quality at an affordable price. I can shoot in less light, I can print 24" prints from a small SLR or MFT, I can capture fast action, all with cameras that are far smaller and more convenient than their 1980 equivalents. In what way is this a bad thing?

Sorry I don't see the downside, and as for dumb amateurs buying stuff they don't need, well what about all those Leica owning dentists in the '80s? It's just become more democratic is all because it's now a lot cheaper.

There is far too much sour grapes these days from the community that once had the field to themselves. "What other people do" is not very interesting. "What can I do" is perhaps a lot more interesting. And now I can do a lot more than I could in 1980.

As I get into printing, I'm seeing a lot of pixel-peeping behavior from printmakers too. I wonder if it's the same thing - if papers have just gotten so good in the last five years that people are used to making distinctions that mattered then but really don't now.

If you substitute "target rifle" for "camera" I think the argument becomes clearer. A perfect gun will not make you an Olympic competitor. You will need to learn and train to shoot straight. However, an inferior weapon, one that puts a bullet in a different location every time you pull the trigger, can certainly limit your ability to hit the bulls eye. Will a camera that has astonishing resolution, dynamic range and color accuracy make you a better photographer? No. But a good photographer has a much better chance of creating the image that he/she envisions if the camera does not get in the way. I would much rather have to grow into a camera's capabilities than to be limited by them.

Both Steinway and Kawai make pianos with 88 keys. I will readily admit that all too many students can make better music on the latter than I can on the former. Numbers all too often are just numbers.

I dunno, the "image quality" people have been around forever IMHO. Pre-digital there were always any number of mailing list or forum posts telling you that there was no way you could ever make a decent photograph with those puny 35mm pieces of film. And if you wanted real detail you had to haul our your 'Blad or 4x5.

I don't think the obsession over sensor performance is really any different.

Something I learned a long time ago, if someone or some business is selling something the data that they present is ALWAYS suspect.

Not that they mislead on purpose but again they are selling something and they tend to try to make their product "more attractive". When digging into the DxO site you will find it almost impossible to get a simple/understandable definition of the measurement terms. "Overall" score is a classic example, it is mostly emotion and should make one cautious about using the other values as anything but a relative comparison. BUT if they are just relative how important are the differences, good luck finding any information on this on the site.

(not sure you want/should publish the following)
I used the DxO file processing product for about 2 years, version 4 0r 5 (forget which) came off the drawing board way to early, it was a disaster. Was emailing/talking to one of the tech reps who was getting me through some of my problems, one day I sent him an email concerning the raw to jpeg conversion, a 20+ meg file was converted to a 2.5 meg jpeg file as the max file size. The answer came from a marketing type, his explanation was DxO has a special process to do this, it is better than any other vendor. After this I was never able to reach any tech rep, all communication was done thru marketing/damage control.

MORAL;
They are selling a product, they are suspect.

Yes, yes, yes . . . but in a small way, no. There are photo books that I look at, that I'm continually drawn back to, that of course have good content, but they also have something else - the photos were made with medium or larger formats.

The tonal richness, the micro contrast, the tiny details, either in landscape photos or in environmental portraiture, just yields another experience, an intangible something that is compelling.

Does that mean that there are not thousands of superb, fascinating, compelling images from 35mm, no, of course not. The medium is even dominated by superb small format images, but in the end, when an image calls for the qualities of larger format that I named above, then 35mm cannot fulfill that.

Its nice to be fully committed to analog photography and not to have to worry or even pay attention to all the digital nonsense. For me the question of digital vs film has never been worth debating. Film is simply better to me, and always will be, for inummerable reasons, most of which are not objectively quantifiable. To an artist, debating the medium they use with other artists is pointless.

This is why I recently ordered a Chamonix 45N-2, which is replacing my Toyo 45A ... and I will be making speed/dev tests on slower films, experimenting with staining developers and investigating older opitcs like Velostigmats, Tessars, etc. They are the tools that can fulfill the vision I have, which is the real quality. If the vision changes, I will change the tools.

The only nod given to ease and accessibility is the Chamonix; it's simply much lighter than the Toyo, and will allow me to use a much lighter tripod to boot. (Well, the Chamonix is much prettier than the Toyo, too ... that's important to me!)

I don't care about being on the leading edge of any technical curve. I'm too old for that crap.

I would politely but firmly disagree with the assertion that image quality doesn't matter at all. It's contingent upon on what you're trying to 'say' photographically. It's certainly true that Capa's blurred, grainy, poorly exposed D-Day image wouldn't be improved (in fact would be substantially weakened) by sharp, grainless perfection. But technical image quality was essential to what Ansel Adams was trying to say about the world we inhabit. As John Szarkowski put it, Adams' technique was rigorous, but only as good as it needed to be for the message he was trying to convey.

The sharpest lens backed by the highest resolution sensor will not miraculously produce great photographs. Far from it; that's where the whole 'art thing' comes in. But the technical means must be sufficient for one's vision or goal for the image. And for some ends, like a large scale print including a forest with lots of detail, the means have to reach a fairly high level (of resolution among other features) to be adequate to the task.

My somewhat longer take on this is that it makes sense to worry about gear when you have a specific goal in mind. I, for instance, wanted to take available light photographs of my daughter, which drove my purchase of a low-end DSLR and a fast prime. (holy cow, this rig can practically see in the dark)

If you want to make and sell prints up to 30x40 inches as a business, this will also drive a set of requirements. And so on.

What I rail against is the acquisition of gear in the hopes that it will in some way simply "make you better". If you're not happy with your photographs, in a non-specific way, you should spend more time working on your ideas, and less money on your gear. Eventually, your ideas will firm up to the point that you're not happy with your work in very specific ways, and if those specifics are technical, then you will know what gear to buy.

Then, go nuts!

DxOMark certainly matters commercially. With all other things equal, someone will naturally prefer a higher-rated camera to a lower one. I'm sure Nikon optimized the design of the D800 to achieve their excellent score. What better marketing collateral could they wish for than an independent tester saying it's the best sensor they've ever seen, including cameras costing 10 times more?

I use the individual measurements to compare a camera I know to one I haven't seen. It's a way to guess whether I will like or regret a purchase.

Speaking of individual measurements, there's been much interest in the past year or so in dynamic range, in which the D800 excels. If I remember this right, our gracious TOP editor, writer, and host purchased a Pentax K-5 partially based on 14 stops of dynamic range. I may have missed articles, but I don't recall discussion of experience with the K-5 and the practical advantage of the additional DR. Did the increased dynamic range turn out to useful?

There's another quite different way to look at this, and that is that amateurs have always been concerned with things that can be measured by numbers, and real photographers have consistently benefited from that. The manager of a camera store I worked with over 40 years ago put it well: I could thank that rich dentist who just left with the latest Canon toy (some motor drive he would never use, in his case--he simply bought every new thing they announced) for the fact that Canon made enough cameras [for amateurs with too much money] to bring the price down far enough that I could afford to buy one.

Tests are certainly not new--the major photo magazines were doing lens tests over a half-century ago. Those tests and the attitude they fostered are the direct reason that we don't have to worry about the quality of what we buy--that it will all meet a reasonable standard--and we should be happy for that. It's one more thing not to have to worry about.

"Many of you remember how much we were consumed in the first five years of the present millennium with the question of whether digital could, or had, matched the quality of film—not a question anybody cares about any more."

Except, of course, for the occasional volcanic outburst of wailing and nashing of teeth over said topic. Or color versus b&w..or format size... Or

Wax versus vinyl ?!?!? Really?

Lol, folks.

Visual image quality was always a part of photography. The other day I had a look in Cartier-Bresson's India book: The tonality from the Leica + excellent darkroom work (post-processing) are essential to convey the message. This would not have worked if he had used a Kodak Instamatic and machine copying!
It is your type of photography that dictates what equipment you need. Most great photographers has spent much time and energy on equipment and the best way to utilize it for intended purposes.
DXOmark is ONE way to compare cameras and lenses to each other. Myself I have found out, now having more time for my photography, that my current m43 Pansonic G3 with Olympus 12mm, 45mm and Panasoning 25mm does not deliver to my needs and ambitions. It is about dynamic range, exposure accuracy and to some extent resolution. Neither is usability, ergonomy and viewfinder good enough.
Different tests, including DXO mark, and sample images are excellent to single out equipments that could suit my photography. The next step is to get a feel of physical handling of selected photo machines. - And check my wallet.
As consumers we are under pressure to buy the newest. Before digital I did not change camera very often. Now cameras are improving all the time: Just an example: Thom Hogan did not expect much improvements from the upcoming D4 compared to D3s. Even an expert as he is was suprised by imprivements!
My ambition is to buy into a system with a camera that hold up for five years! Anyone thinks that is possible?

As if this topic needs one more damn opinion... but here's mine anyway.

Re: DxOMark: Does it "matter"? With no object to the question I don't know. But I do appreciate having an independent source of consistent testing in the world of digital imaging. My camera buying days are mostly past but, yes, DxOMark is to digital cameras as Consumer Reports is to vacuum cleaners and washing machines.

Re: Image Quality: I largely agree with your position in its spirit and essence, Mike. But I would present my own position by saying that technical quality participates in the overall impression of an image but that it is neither the foundation nor the last word in most such impressions.

As you know, I am fortunate to see a remarkable number of famous, and not-so-famous, prints of works representing photography's complete history. Some, such as many of Irving Penn's studio works, amalgamate technical excellence with creative talent to create timeless wonders.

But there are many, many examples of other photographers' works that favor creativity and instincts over technical values. Davidson, Arbus, Levitt, DeCarava, et.al. knew their cameras, knew their medium, but achieved timeless value largely through their instincts and persistent creative juices. Many of their best photos could have been taken with -any- quality camera or lens.

But what about all those gaze-freezing monster prints from technical image geeks like Burtynsky or the Dusseldorf School gang (the Bechers, Hofer, Gursky, Wunderlich, et.al.)? These are images heavily reliant on technical quality to achieve their impact at their usual scale. But it's still the scenes and arrangements in front of their creators' lenses that makes the impression.

Moral of the story: The value of technical image quality is relative to the objective and results, rather like the value of a seasoning is to a food dish. It's an ingredient, no less, no more.

center their hobby around exhaustive comparisons and evaluations ... what you have is adequate for what you're trying to do.

Those two phrases sum it up. I aspire to live in the arty side of photography, so it's what I want that comes first; I'd hope some of my work might be called "good" yet the fact that the equipment on which it's made stays reasonably uptodate technically would only be incidental to the accolade.

Great article. Tech culture (i.e. mainstream culture) has become obsessed with metrics. Enthusiasts have enjoyed the metrics of their respective interests for some time (horsepower, LPPM, megahertz, watts, number of "likes", etc.) but now everyone likes metrics. To paraphrase McLuhan, the media's metrics is the message, and we're another step away from meaning.

DXOmark taps into that perfectly with 4 easy numbers. With all the decisions we're bombarded with, it's so nice to boil things down to a handful of numbers.

But it can't continue like this, the numbers are losing their meaning (the enry level Nikon will have 24 mpix, the mpix race will soon be over and meaningless). We'll have as much interest in bandwidth as we did in the 60s. Metrics will be passé, numbers will become boring (once again).

Blasphemy! But true, IMHO. A good picture taken with a good camera is a good thing. But, a good picture taken with a mediocre camera *always* beats a mediocre picture taken with a good camera. A poor picture taken with a good camera is...junk. If my *vison* is poor it doesn't really matter how good my camera is. So, to get better pictures I need better *vision.* Then--and only then--do I get better value from a better camera. Said a different way, high quality images taken with a top-end camera (i.e., with high DxO scores) do not necessarily make a *good* picture.

Throw out Ansel's 4x5 and 8x10s! Treasure those 35mm frames. Well there were not many of those. A few more 6x6 maybe. Some of this just rings false if you have read the blog long enough to have encountered Mike waxing poetic over a particular Olympus lens a few times. And the comments are thick with irony from people still wanting to prove that film is still better than digital in a column that implies that improvement in digital is unimportant.

My take on this is that the practice of photography is too broad to say what the best tool is for finding the "best" camera . If you're doing high end product or fashion photography where sharpness, detail and dynamic range are demanded, then you need a camera to deliver those qualities. On the other hand its clear that Daido Moriyama would not be well served by the same camera to get his high contrast grainy b/w images. Discretion might be high on his list, a quality not measured by DxO at all. Rinko Kawauchi uses an old medium format Rolleiflex most of the time but her images are generally not sharp nor does she seem to care. There is an unusually enticing pastel like quality to much of her work which many associate with certain Rolleiflex lenses, again, ummeasured by DxO.

Recent work I've seen from Robert Frank looks as though he's using the same b/w film he used in the 1960s and possibly the same camera. The work is grainy and somewhat muddy and still moody and revealing. DxO doesn't do film cameras but even if they did, the logic of what they're trying to accomplish doesn't fit with Frank's style. Egglestone's recent work is certainly not sharp though he uses the best Leica lenses (though DxO generally has not found Leica cameras to be that "good" and Eggleston's prints are rarely rejected at auction houses for their lack of DxO score.

On the other hand there are many fine arts photographers who do need or want tack sharp high dynamic range images ranging from landscape photographers to people like Gregory Crewdson and Alex Soth.

Dxo is a one score glove that doesn't fit all photographers' styles. Unfortunately it's a great tool for forum cowboys who use discussion forums as an extension of the testosterone driven side of their personalities.

After reading all the comments I had an additional thought. DXO scores MAY have mattered once, when digital was struggling to find its feet. But now, the scores are becoming somewhat irrelevant. You now pick cameras like you used to pick your films -- was Tri-X better than Panatomic X? There's no objective answer to that question. Whatever camera you choose (above a certain level) is going to be excellent technically. Whether it's right for you is a question that DXO can't answer. One difference between digital and film, though, is that changing films was fairly trivial in terms of cost; changing cameras is not, and if you buy a D800, you'll never get as usable very-high-ISO photos as you would if you'd chosen a D4, and if you buy a D4, you'll never get the ultimate resolution of a D800. So you look at the Canon that everybody's been whining about, and you could say, 'Okay, here's a camera that doesn't have the resolution of a D800, but it's got a faster frame-rate and a better high ISO. It doesn't have the ISO ability of a D4, but it's got more resolution.' So you could perhaps make the argument that the Canon is a "sweet-spot" camera with the optimum combination of resolution and high-ISO capability. That seems like a legitimate argument to me, but it's not gp0ing to show up in DXO scores.

JC

Mike-

Amen, Brother! The image quality needs to be sufficient to achieve the intended purpose. 90% of the photographs I love the most would not be considered the highest "image quality" from a contemporary pixel peeping standpoint.

The thing is, this is in no way restricted to the world of photography. You can find the same confusion going on in golf, fly fishing, biking, etc....in any discipline that requires equipment if one wants to participate. There is no way to buy talent or vision. There is no shortcut to the years of practice necessary to the development of good technique. However, if you have the money, you can always buy a piece of equipment that is capable of great performance whether or not you are capable of delivering that performance using that equipment. This leads to teleological thinking. Hence--A. Cartier Bresson used Leicas, and B. I really love Cartier Bresson's photographs, therefore C. If I buy a Leica maybe I can make photographs as great as Cartier Bresson. Feel free to substitute golf club and Tiger Woods, or tennis racket and Rafael Nadal, or flyrod and Steve Rajeff, or bicycle and Lance Armstrong. And the corollary, "the reason my photographs are not as good as Cartier Bresson is because I don't have a Leica". I think that it is a hell a lot easier to argue about chromatic aberration than it is to take a great photograph.

However, it is just human nature to do this. Most of us fall for it in one way or another. It's fun to buy new toys if you can afford them, and, it's great for the economy!

Dear Mike,

I do think that there is essentially the same article that would be titled "Do Film Tests Matter." With essentially the same conclusion: “Usually not… Except when they do.” I tend to be leery when people draw analogies between silver halide and silicon, because they usually mislead in serious ways. But this is one case where I think it's really useful to make the comparison: just think of the DxOMark tests as being the digital equivalent of film tests.

Which leads directly into your main point, which is an unfortunate side effect of computerized photography -- everyone fancies themselves a “film tester” or at least “analyzer” because they now have instrumentality. So they'll pixel-peep and data-digest like mad, failing to understand the all-important distinction between “data” and “valuable information.” (And I mean that both technically and artistically.)

More is not the same as better.


pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
======================================
-- Ctein's Online Gallery http://ctein.com 
-- Digital Restorations http://photo-repair.com 
======================================

'And the comments are thick with irony from people still wanting to prove that film is still better than digital in a column that implies that improvement in digital is unimportant.'

I seem to have missed that. I thoght the comments were saying that, for some photographers, film works better even thogh it provides, in 35mm form at any rate, less 'image quality' than most current digital.

Maybe I'm just missing something.

Mike

Mike, as you know, I've been reading your stuff since your "Camera and Darkroom" days, and I find your perceptions, as in this post, to almost always be right on in all matters photographic. Thanks for your clear insight into a murky subject.

guys, any idea why st. ansel adams didn't use holga?

My ambition is to buy into a system with a camera that hold up for five years! Anyone thinks that is possible?

Hold up to what? Your standards or some arbitrary measurement scheme? If you buy a camera now, and it takes satifying images and performs well enough - what is going to happen in 5 years? Breakage or wear and tear aside, it will be taking the same images.

If this quest for ultimate quality had been so prevelant in the days of film, 35mm would never have survived - we'd all have found it 'essential' to shoot medium format.

I will confess to getting a 5D3 after 'only' 5 years with the original 5D - because I wanted a decent AF system. I have no problem with the old camera's images and will continue to use it when a 2nd body comes in handy.

Cheers,

Colin

S. Chris, your comment about "shmoes (me included) who simply don't want to feel foolish for committing a lot of money to something when they could have gotten something else that worked better" hit home ! A lot of people seem to get bothered about those who agonize over how to spend their money. "Your old X is good enough, don't waste your money" or "who cares about the test results, they're all good enough for 90% of photographers" ... you know the arguments. And there's some validity to it all. Mike posted a long time ago about buying digicams ... paraphrased, he said "just pick one; they're all cr*p" (and he's right). But as you indicates, you're still spending your (presumably hard earned) money ! Why buy a Honda Civic instead over a Toyota Corolla or a Ford over a Chevy. They're all "good enough". Why buy a Denon over a Marantz or a KitchenAid over a Kenmore ? Of course it's all "good enough" but when you've got the choice and you've got the money, why NOT look at the tests and pick the one you think gives you the most bang for your buck ? (Of course, dxomark measures aspects of image quality only, and that image quality is close enough on most cameras, that "bang for your buck" should count other factors besides image quality).

The other thing is the obsolescence factor. We all know that the cameras that will be on the market next year or certainly the year after will enable their users to shoot the same things we're shooting at even higher ISOs, or with even more detail. So to stave off the feeling of obsolescence, some will look to be sure that what they're buying today is the best they can get.

I think the dxomark scores are endlessly enjoyable and occasionally helpful to me. On the whole, I don't know that they're terribly important. I think they (along with umpteen other similar sites) probably serve as a red herring to a lot of newbies who don't understand the significance of the measurements and place too much value on the scores, possibly influencing how they spend their money in an unfortunate way. But that's not a new problem.

"the question of whether digital could, or had, matched the quality of film—not a question anybody cares about any more"

OK so we're obviously not reading the same websites :)

Just as an aside to this posting, I've been amazed in the last few days at how many photographers I've known that have mentioned they've ordered the Nikon D800 and are waiting for delivery. Most stated that this camera seems like it's going to be the "magic bullet", and now they can stop thinking about "specs" and the need to replace equipment every few years.

Most of the pros I know wanted to get involved with 120 based digital, not because they didn't know that DSLR's would cover 99.9% of their usages, but because the bigger cameras would cover the last .1% and they wouldn't have to worry about any applications from that point onward. Most of the pros I know couldn't afford to buy them tho, especially in my town, where the professional level payment has plummeted over the last 20 years.

The recent press about the D800 makes it clear that the results from this camera are probably superior to the results from 120 based digital cameras and backs that many photographers I know DID buy, five to seven years ago, and they struggled to pay for and are struggling to eke the last little bit of ROI out of. And all at a sensible price point.

It's good to remember that back in the day, pros bought stuff like Hasselblads and Sinars because once they bought them, they never worried about the technology again, just concentrated on pictures. Every body I ever worked for read all the reviews of view camera lenses they could before they selected a new 360mm or 410mm, they bought from their local dealer with return privledges, tested a few sheets of film, and when they were satisfied, I never heard anything about it again. We just talked about lighting and composition.

5 Years? Well, my D700 is 4 years old. It'll still be just fine at 5, I'm sure. And there's no real upgrade at that level in sight (the D800 is a leap to excellence in a totally different direction: super resolution, but not much improvement in low-light performance and an actual loss in frame rate; the D4 is a clear upgrade, but only at the top price-point, the way the D3 was back when the D700 was new).

There IS something to be said for first-rate equipment. For somebody like me, 40+ years seriously in photography, I'm not really expecting a sudden breakthrough in my own creative vision or shooting skills. I'm sure it's possible, but it's terribly, terribly rare. But if I were using ordinary consumer-class bodies and lenses, I could significantly improve my results with better bodies and lenses (for the kind of shooting I mainly do, faster focus and better low-light capability are key). For me where I am, I'm stuck; kind of plateaued out, where there are no short-term routes to significant improvement, only the long-term route of gradually improving skills. But for a hypothetical me with a much smaller equipment budget, improving the equipment would really give that me a significant boost.

It is remarkable to me how hard this is to learn. I'm still working on it. I acknowledge that most of the photos I admire the most can be beaten technically be my most humble camera, and yet I continue to have a craving for the newest "best".
I guess I can just try to at least separate my ideas about artistic quality and technical quality.

There's a parallel in fine art. Take the ten most successful painters in history. Hardly any of them could draw very well. And of those who could (Picasso for example), few of them used that skill in the art which made them successful.

The comments to this entry are closed.