« Open Mike: 'Leadership' | Main | Random Excellence: Benjamin Kanarek »

Tuesday, 17 June 2008


What exactly is meant by the statement that E-3 was designed to be a digital camera from the ground up? The only thing different about this camera is that its lens mount is too big considering its sensor size which means that the lenses do not have to work as hard. Nikon and Canon, on the other hand, have same lens mount for both reduced and full-frame sensor cameras.

A nicely considered, wwell reasoned list. This is why I like TOP so.

Looks like this will be the final push to my purchasing a Zeis Icon.

Why compare the Canon to the Fuji F40fd when the F100fd has Image Stabilization, probably better low light performance and some manual controls (Av, Tv)?....

Thank You Mike...

Looking forwared to your list of "top OLD cameras" -or please make it "top ANALOGUE cameras"

Thanks for your Top 10! Do you really think the image quality of the Pentax K20D is better than the IQ of the Nikon D300? At all ISO? Your remark is a surprise for me - and important, because in a few months I would like to buy into a system. (And the specs of the Pentax are great - especially IS.)

Michael, I think you have included the Canon Ixus but missed one of the best reasons for its inclusion. It can also shoot RAW. There is a whole hacking community online that can do all sorts of wonderful things with this camera range, and all without voiding the warranty. Check http://chdk.wikia.com/wiki/Main_Page
I'd argue that the Sony A700 deserves a closer look. Usually when it is reviewed it is matched with the Sony 17-80 kit lens which was developed a generation earlier with the first Minolta SLrs. The A700 deserves a better lens to strut its stuff. The Zeiss 16-80 was designed for this camera and gives a far fairer indication of the camera's capabilities. I also think the Apical patented DRO system used under licence by Sony (and now Olympus) extend the capabilities of these cameras into another league. The D300 has tried to do the same thing but has their own, and a somewhat limited, Nikon copy of the system Sony and Olympus share. Apical claim that their system allows the iso of every pixel to be individually adjusted in DRO.
Other aspects of the A700 are rarely picked up on. Its frame rate. At first it just looks good, but slightly below the best of other cameras around its level. Then you realize the size of the buffer. With a fast card,the Sony can keep shooting at that rate until the card is full. The hot shoe. Often dismissed in reviews as being Minolta eccentric, reviewers fail to see its wireless flash capabilities.
The in body IS system on the Sony A700 is a real plus, Sony have developed IS considerably over the last two years and the new version on the A700 is very impressive indeed. It allows all those great old Minolta lenses to be used with IS on the alpha mount.

I think the A700 is a direct competitor with the D300 (they share the same Sony sensor, although the Sony has an extra chip on the sensor as part of its DRO system)

"All of the major camera review sites save one (The Luminous Landscape) treat "medium-format" digital backs as if they don't exist. They're never mentioned, much less reviewed. Given the constant disputation (and worse) on many forums about the ultimate in digital image quality, this is strange."

Well, not so strange. Given that they cost quite more than even the top dogs in the "35mm" market, they go well outside the sites' targeted audience. Would a person shooting an entry-level camera think of acquiring a P45+ or H3DII, costing as much as a car? I don't think so. But they can dream about D3 or 1Ds Mark III...

And then, I suppose, the manufacturers are not exactly willing to send their multi-tens-of-thousands-dollars to zillions of sites. I may be wrong and don't quote me on that.

I would love to play a bit with a medium format digital camera, but I'm certainly not a Michael Reichmann and the previous paragraph certainly holds true for me, for all I am a journalist and a photographer. :-)

Thanks a lot for posting, Mike.

I did a longer comment at Pentax forums discussing this list (http://www.pentaxforums.com/forums/pentax-news-rumors/30127-hot-online-photographer-puts-k20d-top-10-better-image-quality-than-d300.html), but let me just copy here my last paragraph:

"Overall, very good news for Pentax and Olympus, excellent news for Nikon (which manages to put in the list three dSLR, D300, D3 and D60 at positions 1, 4 and 6, respectively) and not so good news for Canon, whose representatives in the list are two dSLR (the 450 and 5D at positions 6 and 7, respectively) , and a compact camera, which speaks volumes on how things are going in Canon.... And worst news of all, no dSLR whatsoever from Sony in this list... which finds only this brief mention about their A700 which questions its image quality: "Also well worth considering in this class: the fine, streamlined, ergonomic and easy-to-use Sony A700. Questions about its ultimate image quality, at least with the initial firmware, are what have kept it off our list."

Mike, great article and I can only imagine how much work it was to put together - and how much more work it will entail now that it's been published.

In your recommendation for the Wista 810DX, you note: "a disproportionate number of high-level fine art photographers, such as Steven Shore, Joel Meyerowitz, and Alec Soth, use 8x10 cameras to distinguish their work visually and technically from the great digital flood."

Indeed, much of Steven Shore's work -- such as what went into the book Uncommon Places, mentioned here recently -- was done with view cameras. But I was interested to read a this at the and of an interview posted by Darius Himes: http://dariushimes.com/pages/blog/a-conversation-with-stephen-shore/

DH Which software are you using?

SS iPhoto. The whole Apple publishing is really very diaristic for me. For all that I said about my love of the 8×10, basically what I’m using for the past 3 years or so is a Canon Powershot.

DH Oh really. Have you put the 8×10 away?

SS Yeah. Pretty much.

... so a master of a view camera has changed to a category of cameras best exemplified by another of your recommendations, the Canon SD870IS. These are interesting times.

Ok, this is why I don't get invited to parties anymore. The best non full-frame sensor professionally outfitted DSLR on the market now is not the Nikon D300, it is the Canon 1D Mark III.

By most accounts, its AF problems have been solved, and its resolution, per pixil sharpness, color accuracy, and noise characteristics are superb, and it is a blazingly fast instrument wrapped in a bombproof casing.

A tragic launch has even brought down prices for this black sheep which should be Best in Show.

I'm curious as to the following comment regarding the A700:

"Questions about its ultimate image quality, at least with the initial firmware, are what have kept it off our list."

What's the point of basing the recommendation on the IQ using the initial firmware? You should be able to give an assessment of the IQ with the current firmware, shouldn't you?


Ohhhh. To rank camera brands after lightly criticizing Leica is inviting more rather emotional reactions. I admire people with such thick skin. Anyway, it is an interesting list of choices.
I wish I could have more than just one from the list, and as a long time Olympus film camera fan, I am becoming more and more curious about their present lens quality.

"You should be able to give an assessment of the IQ with the current firmware, shouldn't you?"

...I would if I could get Sony to loan me an A700. It's funny that with any magazine or website I've been associated with, some companies take you seriously and others don't give you the time of day. I would think TOP would be a natural for Sony, since they're advertising all over the web (unlike Pentax, which I think has limited its ads on the web to TWO) and I'm a former Konica-Minolta user who reviewed the 7D in print. But no. Can't raise anybody at Sony. Either that, or they're ignoring my attempts to contact them. And so it goes.

Mike J.

Having used a number of these cameras (the Pentax 10D, but not the 20D) including the D300, D3, and D60, and one day with the Canon 1DsIII, I've concluded that there's really not a nickel's worth of difference between the comparable levels of cameras; that it all comes down to how each individual fits each camera. I've been Nikon most of my life, but I'm seriously thinking of going to Olympus or Pentax simply for the (small) size of the things. I have a D3 but the thing is the frontal size of, and with a f2.8 zoom, more than the weight of, a 4x5. And as you say, even a D300 is big; we only think it's small in comparison with a D3 or a big Canon. A Pentax Spotmatic was my first serious camera; maybe a Pentax will be my last serious camera. I'm waiting to se if Nikon comes out with a light FF camera, based perhaps on an F6. There are rumors, and the rumors say that we might know by the end of this month.


"Although image quality is not quite as good as the K20D's"

Mike, I'm curious as to your basis for this statement. As one poster noted above, many people will be making purchasing decisions based on your recommendations.

When you refer to image quality, are you referring to resolution, noise characteristics, dynamic range, color fidelity, in raw or jpeg images and if in raw, which converter were you using. My D300 raw image conversions look totally different in Nikon Capture NX vs. ACR and my Pentax K10D conversions look totally different in Capture One 4 vs. ACR.

I own a scary number of the cameras listed. . . so why aren't my photos consistently brilliant? Nahh, don't answer that.

Seriously though, these lists are such an interesting snapshot of where the camera industry is at a given moment. I have old issues of Photo Techniques laying around at home and it is stunning to see the pace of change in this industry over just the last eight years. Even more stunning to think of the R&D required to solve the various technical problems inherent in getting where we are today in terms of image quality, file size and so forth. I guess my question for the group is: when are we "done"? For myself, between the M8 and the D3, I am done (both financially and in terms of image quality, low-light sensitivity, noise levels and so on). In terms of technology, I guess I could see myself replacing the D3 in some high-single-digit number of years if they came out with a D4 that had in-camera image stabilization; but that technology already exists in other platforms -- I chose the D3 because of the high ISO performance and its compatibility with my existing lenses. There is other cool stuff that I could imagine: A tethered VF that integrated seamlessly into my glasses so that I could photograph anything I looked at . . .

But really: any thoughts on when this tech will be mature? Or will the fashion-industry business cycle of tech development continue unabated into the future?

What continues to surprise me is the lack of used digital backs. I've read that one of the things photographers missed when forced to a 35mm-based digital system was the Zeiss/Mamiya/etc. lenses available previously by those who used medium format.

Perhaps it's such a major purchase that users are getting the most out of them, maybe it's that there's so few of them, or maybe the new crop of smaller format digital cameras outperforms older backs but I'd love to try out a digital back, slap it onto a cheap Medium Format body and use the lenses.

K20d image quality better than D300.

Curious about the following. The Pentax DSLR's have been intriging me of late.
Two questions:
Would you happen to know/have an opinion on, the IQ of the K200D vs the K20D? Interested becuse I think the K200D uses AA's.
Second, the new 35mm D macro. Could it function as a high quality "normal" lense? is its IQ high at non-macro distances?



"For myself, between the M8 and the D3, I am done"

Heh! Sure you are.

(I have a number of friends who write me after they buy a new camera to tell me this is it, they're sticking with it this time, getting off the merry-go-round, etc. I've said it myself a time or three.)

The thing about the future is, predictability declines the further out you go. I once attended a lecture by a meteorologist who said, "I can predict with 100% certainty what the weather will be five minutes from now." A study on the Freakonomics blog, by contrast, recently concluded that TV weather predictions are only at all good for 48 hours into the future--the predictions for the next 24 hours are pretty good, and for 24 to 48 hours out they're consistently better than random guessing. But from 48 hours to the end of "five-day outlooks" that TV weathermen routinely present, they'd beat their average, the study showed, if they simply predicted that it would NEVER rain. That is, if they always predicted no rain no matter what, they'd actually be right more often than they are.

I suspect the same might be true for predictions in the camera and photo markets, except perhaps with years as the term rather than days. I can probably present a pretty fair picture of what the camera market will look like in the next year. Between one year and two, I could probably be more right than not (although I'd be wrong a lot, too). For two to five years out, we're all just guessing. Beyond that, and we're into solid unpredictability.

Probably my biggest disappointment with the current state of the camera market is that cameras look exactly like they did twelve years ago. We've gone from 35mm SLRs and 35mm point-and-shoots, though a period of fervid inventiveness, to...digital SLRs that look just like 35mm SLRs and digital point-and-shoots that look just like 35mm point-and-shoots. The lack of creativity and inventiveness in this state of affairs is something I did not predict six and eight years ago, something I couldn't have predicted.

OTOH there are technologies coming that we simply can't imagine yet. Not because people aren't working on them, somewhere, but because we don't know what's going to fly or not. It depends on so many things--timing, backing, public perception and misperception. (I thought that pellicle mirrors would be big by now, for instance. Wrong.) We get glimmers of really fascinating technologies, but there is really no telling which ones will be implementable, popular, profitable to make, useful...and which will prove to be dead ends. Some will be dead ends for a while until new refinements make them feasible once again. Some will take off like wildfire and be ubiquitous in short order. No telling which will be which. And of course there are discoveries that simply haven't been made yet.

So what's going to be the next big thing? We've previewed a few of the possibilities--post-Bayer sensors; focusing in software after the fact; non-glass lenses; still photography becoming just frame captures from video cameras; etc.

I can probably predict one thing, though...at some point, your M8 and D3 will no longer quite do it for you. (Sorry to say.) Just as I can imagine one of our photographic predecessors saying, in about 1850, "That's it. I'm done now. I've finally mastered the Daguerreotype process completely, and I have the exact box camera and Petzval lens I need...."

Mike J.

I own both of the 2nd-place-tie cameras (with an assortment of lenses for both) and have spent a fair amount of time comparing the so-called "image quality" of both.

There have already been a few comments questioning the notion of "image quality" and I'd like to offer my observation on that subject.

In comparing the so-called image quality of the K20D and the E-3, I have found that both cameras are capable of delivering excellent results .... BUT ....

those results are so utterly dependent upon software that I am led to believe that observable camera-native image quality is a thing of the analog past.

I mean, what's the standard by which to judge?

Every RAW converter software that I have used in the aforementioned comparisons - ACR, LR, Aperture, RAW Developer, SilkyPix - produces very different results in image quality from the same file - differences in tone, color, dynamic range, sharpness, noise suppression, etc.

With a lot of screwing around, it is probably possible to obtain somewhat similar results from different converters although it must be stated that not all converters play well with all manufacturer's RAW files - "watercoloring" being the most notable offense.

So, once again, what's the "standard"?

And, just for the hell of it and to really muck up the idea of a standard, let's throw in "operator expertise" - how well any given "operator" uses the converter software, including the so-called "experts" upon whose judgment so many rely for purchase recommendations and equipment advice.

IMO, the elephant in the room that no one talks about, when it comes to image quality, is RAW converter software.

Now there's a comparison I'd like to see.

About the creativity, Mike, you're just right. I'm waiting for a kind of digital 6x6cm with "THE" real live view, not an electronic screen, 10M pixels, (why not a zoom ?; or an interchangeable lens body; a kind of a mix between a Rolleiflex (600X or TLR) and a Hassy...Light enough and durable.. is it impossible ? Where is the fun, nowhere ?

Kind regards

Good point. Also, what's "image quality"? I've long noticed that with stereo systems, listeners often have specific properties that they respond to, either consciously or subconsciously--qualities that contribute more or less to "realism" for them on an individual basis. I remember auditioning one guy's system. He was completely, and I mean completely, obsessed with bass. He had spent thousands and thousands of dollars on his system and even chosen his house based on his pursuit of the ultimate bass. He had speakers the size of coffins. Well, he described his system-building for me entirely in terms of his search for bass, and when he played it for me, the bass was stunning--huge and powerful, defined, articulated, incredibly deep. And after I had heard a cut or two his very nervous question was, "Well, what did you think of the bass? Is the bass good enough?" The problem, of course, is that the bass was fine, but almost nothing else was.

Sometimes these qualities can depend entirely on the intellectual constructs we wrap around them. Another friend liked what he called "transparency." To me, his system was just very thin-sounding, and dry. To him, that was "transparency," and it's what he was after.

Me, I like "imaging"--i.e., spatial illusions. At least one famous hi-fi writer, Art Dudley, has no interest in imaging and regularly disparages it.

I'm not sure exactly how, or whether, this translates to picture image quality. But I would theorize that something like this goes on, and would have to disprove it before I'd discard the theory.

I'd also take your observations about raw converters one step further, because I don't think it's really relevant what you can get onscreen in your own home on your own monitor. IOW the raw converter software is really just another step in the chain. To me, the print's the end result. So to really get involved in serious appraisals of image quality (assuming you're not just looking for that one property you happen to be moved by, like bass was for my aforementioned friend), you'd have to not only consider software but also all the many variables involved in printmaking--not least of which is operator skill and the operator's artistic intention, too. At which point there are really too many variables to look at with anything like objective scrutiny.

Mike J.

Thanks for your time and efforts on this. A lot of folks are looking for "a list" and it's great to have one so well thought out.


Hi Mike,

Your analogy between audio and photography looks completely spot-on to me, in very different views; and being interested in both fields, I completely agree with you: there is not such a thing as objectivity when considering image quality or audio quality: the personal standpoint is something that affects deeply any consideration.

Same happens with noise, for instance. I am very involved in the world of a certain funky musician, and we have exhausting discussions on audio quality of certain recordings. Noise is an overblown issue in both digital picture and digital audio, and more and more people are coming to a point where noise is not a problem per se: the problem is losing information (be it audio or image information) due to software used to remove noise. As long as the noise is not too much perceptible, and as long as it let you enjoy the song/picture, then why to bother too much? Why to use headphones at very high levels, or why to use big magnifications of your image onscreen, to detect noise?

Pixel peeping is another illness, just as is audio perfectionism: the real thing escapes numbers and specs. ;)

Mike J. wrote:
"At which point there are really too many variables to look at with anything like objective scrutiny."

I agree. Now can we get back to that comment about the relatively lower image quality of the D300 vs. K20D? Several others have asked about it and it really is one of the more intriguing statements in your excellent list to a certain type of user (me included).

Can you at least tell us the variables controlled and the standards applied to reach that conclusion?

Thanks for a great post.


Thank you, Mike. It was well worth the wait for a ranking of current cameras from the unabashedly sane, opinionated and photograph-centric Mike Johnston perspective.

Can't wait for the lens survey. ;-)

You get a lot for your money these days (and one happy result for the cash-challenged like me is good bang for the buck in the used market, too).

I notice that CMOS sensors and IS are prominent on your list.

Jay Moynihan,

I'm sure Mike is reluctant to opine on something he has not examined first hand, but FWIW, I believe the K200D uses the same CCD sensor as the K10D. Carl Weese wrote a lot about that camera on the old T.O.P. site. I recall that in its day the K10D's RAW IQ was regarded highly at most ISO's, while its JPEGs were panned.

Also, FWIW, I note that more DSLRs with celebrated image quality have CMOS rather than CCD sensors (while the E3's is neither). That is not to say that CCD image quality is not excellent, or that it makes any difference for a given output format, but the correlation is there.

A big note of caution for 8x10 shooters hoping to use the Epson V750-M scanner - while described as having an 8x10 area, the film holder area is more like 7 1/2 x 9 1/2 - frustratingly too small for an 8x10 negative.

The K20D is in it's deserved placed in this list - the new CMOS sensor from Samsung is nothing short of mind blowing. Paired with the right lens, like the 77mm f/1.8 limited, it's capable of limitless beauty. Good call, guys, impressive insight.

Curious also about how the 5D compares to the K20D in terms of resolution of detail. I know the K20D has more pixels, but I also know that the 5D has a larger sensor and thus larger pixels.

Has anyone used both of these cameras enough to compare them in terms of IQ, resolution and rendering of fine details?

Ditto to Jeff Glass's question.

I'm not trying to stir things up relating to your statement, but I respect your insights and am very curious to know specifically what aspects of the K20D image quality you like better than the IQ of the D300 (and under what conditions).

As a side note, I would add that since I got my D300, I haven't used my K10D too much because I prefer the quality of the D300 raw files converted with Nikon Capture NX to the IQ of my K10D files converted with Capture One 4.

Jeff K.

Jeffs G. and K.,
I could get around to that, but I know what you're asking for--detailed, experimentally rigorous, point-by-point comparisons with all variables accounted for--and you'll appreciate that I've just put in a whole lot of effort on the last post--which included borrowing cameras, driving around the city to look at cameras, all sorts of reading, discussions with colleagues, weighing of issues and responsibilities, checking of models and specs (and even sales figures), etc., etc.--all manner of shenanigans. Unlike many of my postings, the actual writing of that piece (and the building of the links, and the gathering of the illustrations, and the formatting and proofreading and...) was only the tip of the proverbial iceberg. There are lots of things that I can write off the top of my head. Other things require plain old hard work.

Next in line, I have two paying jobs to attend to, and I already feel behind. So I hope you'll forgive me for not elaborating just yet.

Mike J.

IF you like to shoot in low light without flash, think about this: most wide-normal IS lenses are f/3.2 at best - f/2.8 if you spring for a premium model.

In a dark room with available light you'll be at say 1/10s. The in-lens IS makes that hand-holdable (just) but it won't be sharp unless you're shooting a still life. So you could switch to a 50/1.8 and get 1/30s. Better for catching moving humans.. but then you just lost IS so now you have hand-shake motion blur as well.

With in-body IS (Pentax/Olympus/Sony) that 50/1.8 (if not faster) is stabilised. Big difference.

So the question is which is more useful, more of the time: the slow (f/2.8) zoom, or fast wide prime?

"IMO, the elephant in the room that no one talks about, when it comes to image quality, is RAW converter software. Now there's a comparison I'd like to see."

Here's one. It's from a year ago and only applies to K10d files.


"I note that more DSLRs with celebrated image quality have CMOS rather than CCD sensors (while the E3's is neither). That is not to say that CCD image quality is not excellent,"

Exactly. The only problem, or "problem", with CCD is that they are quite more prone to excessive noise at higher ISO values.

E-3 sensor _is_ a CMOS. It's just a variant that enabled LiveView, called NMOS.

There was an interesting observation over at DPReview Olympus SLR forum - the image from E-1 that had CCD (with excellent image quality at lower ISOs) and the cameras using NMOS is very similar. So it has to be something else, beside the sensor, that creates a particular look. Most probably the image processor. And I think it might be the case with other manufacturers, too.

I was interested in your comment on Sony 700 image quality as I own a K-M 7D, which I like very much, especially now that I bought the vertical grip. But what really came to mind was a thought from the film world. When I replaced my Minolta 9000 (film camera) with a Minolta 7 (also film camera) I gave no thought to the idea that the image quality may not be as good. After all I would be using the same lenses, film (Ilford), developer and enlarger, lens and paper. What I was gaining was a better designed body (and vertical grip). But the digital world has thrown into the mix a new set of variables: sensor, camera firmware, raw convertor software, etc.

Very much so. Complicates camera reviewing greatly, too.

Mike J.

I like the list! Personally would have put the E-420 and 25mm f/2.8 pancake on there since it is so much smaller than any competing DSLR combination. Similar size and weight to the Zeiss Ikon and 50 Planar. I posted some size comparisons here -> http://www.seriouscompacts.com/2008/06/olympus-e-420-with-zd-25mm-f28-size.html

Regarding the 5D, I'm one of those who doesn't need really anything more. I've had the camera for three years now, and never have I wished for more AF points, more frames per second, etc. Your article "Me and My D3" nicely summed how I feel about the 5D.


Dear Mike,

I can see the point that some of the readers are making. I agree with you that image quality is a seriously subjective and arbitrary thing, but we all have factors we weight more or less heavily. If you gave people an idea of what you consider more/less important in the mix, they'd have a better sense where you're coming from.

For instance, at today's level of quality in the better cameras, I consider exposure range to be a lot more important than noise. If camera A has better range but worse noise than camera B, I'm likely to say camera A has better image quality. I give some weight to sharpness and detail, but (this may surprise some people) I give none to delta-E (color fidelity). Why? Because (a) among the better cameras, the color fidelity is so good overall (far better than film ever was)I think it's quibbling, (b) it's highly manipulable and correctable in the computer, and (c) technically accurate color is not the be-all and end-all of color photography. (Anyone remember the "gamma one" notion in color photography of about 20 years ago? What a technically correct and artistically misguided idea that was.) And pixel count counts for zip.

My point is not that my particular weights and biases are objectively correct, but if I lay them out even minimally it gives people a much better idea of what my rankings mean.

In the same vein, I think you should be listing the software you use, not because I think it makes any big difference as some other people do, but because every so often there will be a glitch (like some raw converter that plays especially badly with one particular camera) and if people know what you're doing your work with, they can alert you to that if it affects your results.

~ pax \ Ctein
[ please excuse any word salad. MacSpeech in training! ]
-- Ctein's online Gallery http://ctein.com
-- Digital restorations http://photo-repair.com

Dear Mark,

A definitive RAW converter comparison?

Don't hold your breath.

It would be an incredibly large amount of work and I don't know anyone who would pay for it. It's not that it's technically difficult, but you're talking about putting in effort equivalent to a dozen full-length normal review articles. Very rarely will you find anyone willing to come up with that kind of money, and there are even fewer people who will do it for love.

Especially considering that the constant flood of new cameras will date the article, and any time a new version of a converter comes out it will obsolete a big chunk.

(Now watch; next month someone will publish exactly such an article, proving how cloudy my crystal ball is.)

By comparison, testing an assortment of films in various developers is a cakewalk.

Give me $10,000 and I'll take on the task. Not for a cent less. Any takers out there?

(That's not an especially outrageous sum: I got paid about two thirds that much in constant dollars for the comprehensive enlarger lens test I did for DARKROOM magazine back in the early 80s.)

I am not yet convinced that RAW converters make a big difference in relative rankings of cameras. Yes, we all know of occasional specific cases where the converter and the camera don't play well together, but those seem to be so much the exception than the rule that I'd need to be shown it mattered among the cameras Mike included. Which is a good reason for him to list what software he used. But if it's merely that converter A is consistently better than converter B, that doesn't affect his list.

What is the "standard?" Whatever gets decided as the standard by consensus or fiat. For example, in black and white film processing the established industry-standard is a formula similar to D-76. Is D-76 the best black-and-white developer out there? Not by any objective or subjective standard; there are some that consistently perform better. It's merely a good representative developer. But by virtue of agreeing on a standard, it lets all the manufacturers do useful tests and comparisons.

That doesn't prevent a film manufacturer from coming out with a particular film and developer combination that they think works much better. Those are rare, but not unheard of.

Personally, for any camera tests I do I use the latest version of Adobe Camera RAW. Why? Because it's the 500 pound gorilla on the block. If a particular camera doesn't play well with the current version of ACR, it's up to the camera maker and Adobe, both of them, to fix that. Because the reality of the world is that the majority of serious players are going to be using ACR.

I'm still willing to do the definitive RAW comparison if someone wants to hire me for it, hint hint [ greedy grin ].

~ pax \ Ctein
[ please excuse any word salad. MacSpeech in training! ]
-- Ctein's online Gallery http://ctein.com
-- Digital restorations http://photo-repair.com

"What continues to surprise me is the lack of used digital backs.[...]"

The few that are out there come up occasionally on eBay and elsewhere, but by the time they do, they usually don't have drivers for the current popular operating system, so they often entail taking someone's three-generations-old computer as well, and the file size is likely to be comparable to an entry-level DSLR. If you're interested in trying a high-end digital back, and if you live in or near a major city, you can rent a system for a weekend, usually for the single-day rate Friday to Monday.

Ctein, I find your statement interesting:

"Personally, for any camera tests I do I use the latest version of Adobe Camera RAW. Why? Because it's the 500 pound gorilla on the block."

Your experience dwarfs my own. But I have spent a lot of time and energy over the last five years learning to use three different raw converters (ACR, Nikon Capture and Capture NX, and Capture One 4) with two different camera systems (Nikon and Pentax).

Whenever I get a new camera, I try processing the images in ACR. It has a great and intuitive interface, is fast and does a great job with batch processing. However, I find the colors from the ACR raw conversions to be inferior to Nikon Capture and Capture One and always end up going back to those converters for important images. Now with a lot of work I can get my ACR conversions to look pretty good, but never as good as with the other two converters. I just want to share my opinion, but from surfing a number of Nikon and Pentax forums, I think there are many others who agree with my findings.

Jeff K.

Jeff K. : "However, I find the colors from the ACR raw conversions to be inferior to Nikon Capture and Capture One and always end up going back to those converters for important images."

A serious, well-intentioned question: what does this mean? What are inferior colors? Inaccurate? Less/more saturated? Less amenable to tweaking? Is it a matter of taste or a measurable "wrongness?" I went from an earlier version of Nikon Capture to an earlier version of ACR because Capture seemed limited and kludgy, and while the colors between the two were different, it didn't seem to me that either was "inferior," just different, and the differences could be processed away...


Dear Jeff,

What do you mean by 'inferior' color? Can you be more specific? I'm curious!

(BTW, I presume you're running the latest version of ACR.)

Until Xtol came along I used D-76 for all my B&W film. It was good enough.

Give me reason to think that that ACR isn't 'good enough,' not just that something else might be better.

Anyway, my important point was that 'better' and 'standard' are not the same things. I consider Microsoft Word to be one of the least competently designed word processors I've used. But woe unto anyone who needs to deal with the outside world and can't manage to work with Word docs or produce files that are Word compatible.

pax / Ctein

About the olympus: the difference in sensor-size between 4/3rd and APS-C/DX is not so apparent in image-quality after the event, as in ability to choose a narrow DoF at time of shooting.

When I had an e-500, I used to switch to a 45-150 (think 90-300mm equiv) in order to get differential focussing. That was coming from a D70 so I know I was experiencing the difference in sensor size.

It was quite funny: I once had a landscape scene with heather in the foreground (say 6' away) and trees fading into the mist about 40' away. I could get everything focussed on the e-500 (4/3rds sensor, DoF, natch) or the Linhof 5x4" LF camera (tilt movements) but not on the MF (hasselblad - insufficient DoF even down at f/22).

C'mon Chaps, pull yourselves together !

(Those of you, I mean, who are asking for deep insights and chapter and verse on IQ and IQ testing procedures for sundry different cameras.)

1) Think about yourselves. Even if you're hovering nervously on the edge of buying something, the differences between cameras abilities and specifications are surely tiny compared with the importance of matching your personal requirements. As long as they're on or nearly on Mike's list, you aren't going to get a bad one. As we recently established - if you're (primarily) a war photog (does that happen ?!) don't get an M8. If you're primarily a sports shooter, get something with fps. Et cetera ad nauseam. If you actually need a single camera that's really good at everything, well, I extend my sympathy.

2) Do some thinking FOR yourselves. I don't mean that to be unnecessarily provocative; if you've been a long-time reader of Mike's, you know more-or-less what he's into and what he rates to be important. It's a tough one, because he's pretty catholic about giving generosity to every light-tight box that's competent for some genre of image-formation. If you haven't been a long-term reader, you shouldn't be taking his word as gospel, surely ?

3) Actually, if you are taking his word as gospel, here's how to do it. None of the holy books I'm aware of allow you to go back to the writer and ask "Oi, what EXACTLY did you mean by that ?" For that you have a community of scholars to work over the material. You can be sure that this'll happen on t'internet, and it's never pretty. "Galbraith says.... but he's a...."; "Reichmann says.... but he's a well-known first-grade..." etc ad nauseam. I don't think we should encourage it.

But we can do better. For we can go to lots of rather high-quality source material and puzzle out for ourselves WHY Ye Master (Mike, we DO highly value your writings observations and are affectionate about your style and personality - and the rest of your team too) might have reached the conclusion he did. We'll learn something that way.

4) All (or most) present acknowledge that there are still some fundamentals that make big old real life differences. As exasperated engineers sometimes say - "constants aren't, and variables don't." Ctein, I never thought I'd say you are significantly undervaluing yourself, but ONLY 10K to untangle the Raw converters question ? I guess I'm letting my aerospace OEM background show through...

5) The number of people writing well about artistic and historical aspects of photography is way fewer than those maundering about hardware. Don't distract Mike too much from MY subject of parochial interest !

Right, that's better. British Tongue-in-Cheek humour alert can be switched off now.


Jose wrote: "Noise is an overblown issue (cut): the problem is losing information (be it audio or image information) due to software used to remove noise".

I agree wholeheartedly. I have a copy of Noise Ninja Pro which is widely regarded as one of the top 2 or 3 noise reduction programs. But I most commonly use it on ISO 400+ files from my digicam. Even at ISO 1600 on my old 7D, unless the picture was underexposed to begin with, I prefer the noise that's there in an out-of-camera jpeg or an ACR-converted file with low NR to the loss of fine, textural detail I get with any settings that cause a noticable decrease in noise. With my A700, viewing a pictures at 25% pretty well fills my 17" monitor and represents a good-sized print, so I don't even consider noise, which is absent at 25% at ISO 1600 (I haven't shot it at 3200 or 6400 yet).

I just about agree with your TOP 10. I think the D300 deserves the no.1 spot, but I think it may well be soon replaced by the D700


An upscaled D300 with the D3 sensor would be a worthy successor, well that’s if it’s for real & I think it might be.

I’ve had a 10x8 Wista for a couple of decades, so I’m pleased to see you have included it in your TOP ten. It’s an ideal camera to get you started in 10x8 and can be bought new or s/h off eBay. Indeed there are lots of good LF deals to be made on eBay, recently I bought a collection of LF gear comprising of 5x4 & 2 10x8 Horseman monorails, spare bellows, rails & a 5x4 reducing back, 2 Rodenstock lenses, a heavy duty Manfrotto tripod, 5x4 & 10x8 holders, 5 10 x8 deep tanks & holders, plus flight cases. New this would have cost thousands of pounds, but I only paid just over £400 for the whole lot, that’s less than the price of an entry level DSLR. Has the world has gone mad, no just digital, which means dinosaurs like me can get some great gear for ridiculously low prices.

It was nice to see you mention Steve Szabo. His book “The Eastern Shore” is one of my favourites. Perhaps you should write a series of postings on photographers who are mainly unknown to the masses, but in your opinion they should gain wider recognition.

I've owned the mid-grade DSLRs from Nikon, Canon, Olympus and Pentax (including the D300 and K20D), and eventually sold them all as I came to my senses.

When you have a $1.00 instant Polaroid rendering color, tone and highlights more beautifully than a D3, it starts to dawn on you that pouring your finances into digital is a waste of money until the manufacturers stop obsessing over noise and start delivering images that are actually pleasing to look at.

I'd spare my wallet and go with your basic Rebel kit, in silver so no one looks twice at you. It'll do just about everything most photographers in the world need, with about the best image quality available in consumer digital unless you're splitting hairs (and sadly, many of us are.)

In reply to several questions about Pentax K10/K20 "image quality"--it's "subtly improved in a number of ways."

I can't think of any way in which the new sensor is worse, but it's not radically, shockingly, different either. There's more resolution, a little more ability to retain highlights and an unmistakable increase in tonal "smoothness" throughout the tone scale.

This is working with RAW files processed through ACR. On the notion of how important software has become, I'll add this. If a camera system I were considering turned out to play really badly with ACR, that would be cause for me to reject the system, just as much as I might reject it for not having the exact lens types I want, or because it just felt bad in my hands. Software, including personal workflow and preferred RAW processor, is every bit as important as hardware in digital photography.

About the 35mm 2.8 macro Mike and I will report on in detail at some point: it makes a superb general purpose "normal" lens (except it's a little longer than I'd wish for in a normal) that happens also to focus in to 1:1. It's easily the best more-or-less-normal focal length lens I've used so far for digital capture.

Thanks for mentioning Steve. He was one of my teachers and a good friend. I dedicated my book to him. He died of MS.

His work is included in a current show at the Kathleen Ewing Gallery in Washington, D.C. It's her 2008 Summer Exhibition, called "At the Water's Edge," and will feature 9 photographers and selected 19th century vintage prints. The show will run from now until July 31st.

Mike J.

Re the raw converter comparison, the most sensible advice I've seen is just to try the demos and pick one - things like colour are a subjective thing after all.

When I did that for D300 raws I found ACR introduced horrible artifacts on high iso images (someone else noticed too:


ACR has had a few updates since which makes me think that this just isn't a "bug" which'll be corrected. This is a big shame because Lightroom does have a lovely interface. Obviously for low-ISO shooters it wouldn't be a problem.

Well done Mike! - the usual torrent of interesting and varied comments. We are learning a lot about different folks priorities when they choose a camera, though whether this could help the manufacturers is anybody's guess.
Presently I only use film SLRs but have owned Fuji S1 & S2, Pentax *istD and Olympus E300 in the past. My personal gripes have mostly been about colour and detail resolving with DSLRs - in each case I found the best results were always obtained by using the camera manufacturer's raw conversion but the original Pentax software was the only one I found "unclunky" enough to stick with.
So ACR it has to be, that way you do at least get some continuity when changing camera or doing comparisons.
Maybe ACR is set up for best colour with each camera used in Californian light - I don't know but it never seems to give me the "right" colour with the shots I take in Western European light.
When considering a new camera it is always worth searching for a downloadable raw file to do your own experiments with - far more instructive than the usual jpg samples found on the web.
Back to the audio parallels - maybe my susceptibility to colour balance and detail resolution is similar to my listening for image placement and clean bass power with speakers, it's a personal thing and I shouldn't expect others to see or hear in the same way that I do.

Cheers, Robin

I have listened since a few posts to all this voices discussing image quality, raw, jpeg and so on.

Someone mentioned this elephant, the software. And I'd go further and mention the even bigger elephant, the camera calibration. Does anyone use this in ACR?

Though color accuracy may not be an artistic necessity, I find that the raw conversions simply look better with calibration, giving a result that can't be obtained - at least not easily - without calibration. (btw, I am talking about the simple method with the color checker described on luminous landscape).

That calibration in ACR is good has proven true for me for any camera I have used so far.

Another thing are maker's raw converters. I find that the Pentax software does a better job color-wise on K10d and K20d raw files than does ACR even with calibration. The Photolab doesn't need nor support calibration.

Then there are jpegs. K20d jpegs look good to me, similar to Photolab conversions, very differnt from ACR conversions. Olympus E-1 jpegs are superb and exactly similar to what ACR delivers on E-1 raw files with calibration. Nikon jpegs are superb. Don't know about Capture NX. ACR gives a color mess with Nikon raw files. Olympus C5050Z jpegs are bad IMO, raw files from ACR great.

And another difference is made by ISO settings. I only ovbserved this with K20D raw files. ACR does mess with noise from ISO 400 on. Photolab gives a much better result, but especially the dng-conversion from photobrowser do very well with these files. The differences to ACR are comparable to the Nikon samples posted by Charith: http://lagemaat.blogspot.com/2008/04/high-iso-noise-in-lightroom-and-capture.html

Benjamin Kanarek uses ACR happily, but he shoots at base ISO which might explain his preference.

So there are many many variables and it is not easy to draw reliable conclusions.

BUT all this distracts from the real thing, i.e. what does the image say. Too bad I am such a geek like many others here ;-)

I can only congratulate triplight for mentioning the pleasing experience in viewing polaroids. No problem with color or noise, they just look good, and I frequently enjoy to look at mine...

best always

Incidentally, the software I use with the K20D is ACR, but, as Carl taught me, it's very important to shoot in PEF and then convert to DNG in Adobe DNG Converter rather than shoot in native DNG and open directly in ACR.

Mike J.

I am by no means expert, but I am very surprised by Mike's comment on the importance of shooting in PEF. I thought both formats should give the same information as RAW, hence would like to get some further reasons to understand this issue: Mike, Carl, could you please enlighten?


Thanks for that good information and your opinions. You and triplight and others provoke an intriguing thought: That we geeks enjoy Polaroids for their qualities, but maybe also for the fact that we can't do anything about them. Since there is no way to "tweak" a Polaroid, there is no point in agonizing and analyzing; we are forced to set aside the creator's angst and open up to the appreciator's delight. For a process geek who also loves just looking at photographs (they reason many of us became process geeks in the first place), a Polaroid is a guilty pleasure.

Having shot film for years, I just bought a second-hand DSLR, determined to shoot RAW, master the nuances of RAW conversion, and realize the maximum quality from every image. But these comments have altered my agenda. Of course I will learn RAW, but now, whenever I have the luxury, I will shoot a la Polaroid--jpegs only, to be accepted or rejected as they are, only pre-capture tweaking allowed.

Thanks, guys. Thanks, Mike. I believe I'm happier for reading this blog today.


robert e

"About the olympus: the difference in sensor-size between 4/3rd and APS-C/DX is not so apparent in image-quality after the event, as in ability to choose a narrow DoF at time of shooting."

Tim, this is true. However, the very compact E-420 & 25/2.8 combo I tote can achieve just as shallow a DOF as an APS-C camera with a typical f/3.5 or f/4 zoom attached, the latter being significantly bulkier. Thus, if 50mm is a focal length you can live with, you're not giving up any DOF control to the kit zooms most people use. The fastest 4/3 zooms (f/2) actually give a slightly shallower DOF possibility than the fastest APS-C/DX zooms (f/2.8), though the former are huge and expensive.

Granted, there is no 4/3 DOF match for the fastest primes used on APS-C systems. Four Thirds is not a format for those who want super shallow DOF, but it does allow for more DOF control than some folks realize.


All in good time.

Mike J.

Polariods, and the tweaking thereof.

I used to do manipulation of the long gone SX-70 polariod film, and miss it. Actually, I have 4 packs in a freezer waiting for the muse to bite.
Sometimes I would start with a B&W print, get medieval on it with oil paints, pastel chalks, razor blades (whatever else the music on the stereo made me think of). then do a SX-70 of it, manipulate that. Then Lastly, macro the SX-70 print using an now gone Agfa film designed for advertising shots of plastic and metal products, and then get big enlargements made, of that. Fun. sigh.

Gotta love Google, look what I found :)


Dear folks,

Jao's examples are very interesting, because they do illuminate the big differences in noise reduction between software.

It also brings home to me how much it is a matter of taste. The Lightroom conversion that he considers properly optimized looks like utter crap to me. The grain no longer looks like fine sand, it looks like coarse granite. Technically the noise may measure as lower, but visually it's just horrible. Artifacts overwhelm any sense of visual quality.

For me, it's a tossup between the Capture conversion and the MacOS X conversion.

For what it's worth, I've noticed that the software provided with my Fuji camera, FinePix Viewer, also does a better job of handling high ISO noise than ACR... but the software is so appallingly slow even on a very fast computer that it's unusable (I am not the only reviewer to observe this).

Personally, I keep my noise reduction during conversion to a minimum and prefer to deal with it with heavy-duty tools down the line in Photoshop.

Still doesn't answer the more important question for making a "camera list" which is does the choice of converter change the ranking? If I rank cameras as 1,2,3, and 4 using ACR, will I *typically* (not just occasionally) get a different ordering if I use another program or will the relative positions usually stay the same? If the latter is true, then picking one conversion program as a standard ( doesn't matter which one) serves the purpose instead of haggling over which one is the best.

~ pax \ Ctein
[ please excuse any word salad. MacSpeech in training! ]
-- Ctein's online Gallery http://ctein.com
-- Digital restorations http://photo-repair.com


You keep bringing up the most appealing camera I am currently aware of. The Zeiss Ikon. It makes no sense as I have 35mm digital, an AE1 and a Kiev 4k, but it just appeals.

Please tell me what it's really like.


Thanks for all the hard work Mike (and others)

Can I assume you (like many) that you give the nod to D300 because it is a more well rounded product? You never mention image quality in regards to D300 but did with the #2 cams. I have considered an upgrade to E3 or changing all together to D300 mostly for the what seems to be much better flash system as well as low light/higher ISO over the Oly. The other thing that struck me is that you have always championed in camera IS and D300 lacks that but both #2 cams utilize it.

Most all these cams will give superlative results under the best of conditions. It's the sketchy stuff when doing event work with my E300 and slowish and unpredictable flash that I think Nikon.

It's not often I get this work and I make do. I just can't help thinking the D300 (and flash system) a much better choice than E3 for such work.

Thanks again

>> Still doesn't answer the more important question for making a "camera list" which is does the choice of converter change the ranking? <<

Not quite the same thing, but I can think of a couple of special cases where the user's tolerance (or lack of same) for the manufacturer-provided raw converter can make or break a camera. The first is the Fuji S-series DSLRs, and the second is the Sigma SD-series DSLRs and also the DP-1. In both cases the sensor is distinctive, and it's arguable that the manufacturer's converter extracts a little extra something compared to the aftermarket converters. I've not used a Foveon-sensor camera, but I certainly felt that way about the Fuji S5 and Hyper-Utility.


"If I rank cameras as 1,2,3, and 4 using ACR, will I *typically* (not just occasionally) get a different ordering if I use another program or will the relative positions usually stay the same? If the latter is true, then picking one conversion program as a standard ( doesn't matter which one) serves the purpose instead of haggling over which one is the best."

Based on my limited experience with the D200, D300 and K10D using ACR, Nikon Capture and Capture One 4, the order would change.

As far as your question to me above about what exactly I find objectionable about ACR color, it's difficult for me to describe but easy to see. I think if anyone takes 25 raw images with a Nikon camera in different conditions and converts them in ACR and Capture NX and compares the results, the difference in colors will be stark and the choice as to which app produces a better looking file would be clear.

The above is my subjective, unquantifiable opinion.

Best regards,

Jeff K.

I agree with Jeff Kott when it comes to ACR and how it renders Nikon (and other files). The Nikon software is clearly superior to ACR..spot on.

I haven't seen many indications on the internet that say any different. From a general color and IQ standpoint, many, many folks prefer the latest Nikon Capture to ACR by a longshot.

I would use the Nikon software exclusively on my tankin G5 dual 2.0 with a lot of RAM but it's just so honkin' putt-putty...why cant they make this thing run smooth and fast like ACR?

For images that I really care about (high class uppity fiiine art stuff) I do use the Nikon Capture to get an initial TIFF.

ACR's rendering bugs the carp out of me. The carp is big and finny and it's a painful departure. I do like it's batch processing though.

"E-3 sensor _is_ a CMOS. It's just a variant that enabled LiveView, called NMOS."

For the sake of accuracy, we should say that NMOS "is" CMOS in the sense of a common manufacturing process and digital nature, but "is not" CMOS in terms of logical and electrical strategy and operation. I don't know which of those aspects has more impact on ultimate image quality, but I concede that "Live MOS", while it employs some CCD-like tactics, is still an apple (a digital MOSFET) and not an orange (an analog CCD) (AFAIK, please correct me if I'm wrong).

Yesterday I wrote "When considering a new camera it is always worth searching for a downloadable raw file to do your own experiments with" - and forgot to include this useful link to a collection of raw files. Give it a try and if you have a camera that's not on the list (I had four) then send it to Jakob, he will be grateful and you'll be helping the rest of us.

Cheers, Robin

You hit the nail on the head regarding the Nikon/Canon lens lineups. If Nikon announces another 18-xx consumer zoom within the next year I might lose my mind. The number of lenses people want Nikon to update for the digital age is staggering. And then they drop head-scratchers like the revised 60mm macro -- the new one is superb, but I fail to see how that was an urgently needed update compared to a new 50, new wide primes, etc. Right now we Nikon shooters are caught between amazing, exotic glass normal folks can't afford, DX consumer zooms, and old primes that are cheap (especially used) and pretty good, but which we know could be much better.

Dear Folks,

This RAW business presents an interesting problem for reviewers and editors. I've noticed a consistent (though not universal) pattern, which is that the camera makers' converters generally have a superior rendering engine, an inferior feature set and unacceptable performance. Much (though not all) of the time, using their RAW converter will produce the best looking photo... but if the software is so lousy that serious workers are likely to find it unacceptable, is it proper to use it for testing and comparison?

The Fuji FinePix Studio software's a good example. It's 10-25 times slower than ACR in every step and operation, even screen preview rendering. On my MacBook Pro, we're looking at 10-20 seconds to update the preview every time I make an adjustment to the settings, vs less than 1 second in ACR. Once I've decided to commit, it's 2 minutes to create the TIFF, vs 10 seconds in ACR.

That's horrible enough that I'd never be able to use Finepix Studio routinely; I just don't have enough hours in the day. Plus, its lack of the advanced controls in ACR means that it does a substantially worse job on a minority of subjects. For example, converting high luminance range scenes: it clips highlights and shadows that ACR has no trouble retaining.

BUT... it's fine for dealing with normal luminance ranges. And for low-light, high ISO images, especially ones photographed under indoor lighting, it's amazing. I'd even say indispensable. The extremely high noise level in those files, especially in the blue, flummoxes ACR. Color rendition and color noise get wonky; there's a lot of yellow-blue color crossover that's very difficult or to correct. The FS results are astonishingly better. It's not remotely a subtle difference.

So, what's the right answer? Test with a converter that produces superior results, but is so clunky that serious workers can't use it?

I'll throw out another question to you all. Is there a third party (universal) RAW converter that folks feel consistently does better than the others *AND* performs well?

pax / Ctein

"I'll throw out another question to you all. Is there a third party (universal) RAW converter that folks feel consistently does better than the others *AND* performs well?"

The closest I've found (not that I've looked very hard...) is LightZone. Fabio insists it is *not* a raw converter, that it is an editor intended to open and edit all sorts of files with equal ease and convenience, including raw, and I have no quarrel with that. But it also does a good job as a raw converter, and I admit I use it that way fairly often.

Mike J.

>> I'll throw out another question to you all. Is there a third party (universal) RAW converter that folks feel consistently does better than the others *AND* performs well? <<

If you want it to cover odd-sensor cameras like the Fujis as well as the more mainstream models from Canon, Nikon et al, the answer is no.

"I'll throw out another question to you all. Is there a third party (universal) RAW converter that folks feel consistently does better than the others *AND* performs well?"

Ctein, I've had good luck with Caputure One 4:


When I was unhappy with how ACR was treating my Pentax PEF raw files, I tried a number of converters and thought this had the best combination of image quality and workflow usability. I do not have a Nikon D3, but have noted that on the Nikon forums several people have said the image quality from Capture One 4 conversions of their D3 NEF files is even better than from Nikon Capture NX.

As a side note, I believe Sean Reid of Reid reviews (who is also a real working photographer) uses Capture One as his converter for all of the cameras he tests.

Jeff K.

I'll second the recommendation for Capture One v4. I shoot Pentax, and when I want to do a basic, as the jpg would have been conversion for quick sharing with others, I use the supplied Pentax/Silkypix software. When I really want to work a photo, I use C1 v4.
(Some have complained about the interface of it, which I find odd - I like it. Sometimes people are funny- complaining about something being too complex and then praising another for giving you a control for every single conceivable adjustment... )

anyway, I like the exposure controls, noise reduction, and the colors it gives are consistently good, if cooler than the Pentax software. Plus, it's a bargain.

Great list, as usual. It's nice to know there's options to my box camera and Petzval lens. ;~)

A decent enough list but quite a bit I can't agree with.

Olympus 4/3rds as a format is inferior to APS-C. Full stop and at little more than half the size, only to be expected. Born out of marketing and Olympus' desire to get help from other makers to compete with Canikon. They've put size above outright IQ and even failed at that (witness the size of the 300/2.8 compared to Canons for example). Its a shame, but they're easy to dismiss. Sonys A700 sits above it. The Pentax is very decent but its IQ isn't better than the D300 and certainly not the Canon 40D. Nicely featured and built, but operationally, still a generation behind.
Outright the D300 deserves its accolades, but at the expense of expense. A Canon 40D is barely 60% of the cost and is 95% the camera. Nicer low-ISO IQ, better handling of high-ISO noise, AF thats better in as many situations as vice-versa (massive myth that 51 point AF is better than 9. Its not a numbers game).

I completely don't understand the comment about lens lines. Admittedly, Nikons is a mess of suffixes, but Canons is clear, wide-ranging and high quality. A few EF-S lenses isn't cause for confusion. By contrast, Pentaxs' range is small, dated and unambitious. Olympus in contrast, are expensive and limited. Granted you can mount OM lenses with an adaptor, but you can mount virtually anything to an EOS in exactly the same way. The policy of expensive and esoteric lenses combined with cheap, small sensored bodies absolutely defies common sense. Is any E510 user seriously going to buy a £4k 300/2.8? Olympus, Pentax and Sony need something to push them past Canon and Nikon in the features or quality arenas. Thats not happening so they're failing. Digital SLRs are one of the few consumer industries in which the most common products are also the best. Its good to champion the underdog, but not just for the sake of it. These companies need to give us a reason. The A700, K20D and E-3 are well short of being reason enough.

You also missed the Alpha 350 in your budget cameras. Better than the E520 and Nikons crippled D60, I'd say a clear joint 2nd with the K200D behind the EOS 450D.

Not meaning to be argumentative and I'm genuinely unbiased, but these recommendations will possibly be acted on by many people. At this moment in time, is it fair to offer a firm nod to an Olympus range blighted by an expensive and limited lens and accessory range and undistinguished bodies?

I hesitate to stick my oar in here, because I doubt I qualify (especially in this company!) even as far as "dilettante" as a photographer, but, well, here goes. (Since I am at least somewhere past that with computer stuff. :)

I have a K20D, the which I like very much, and I shoot pretty much exclusively DNG raw files. It's the only DSLR I've ever used.

I'm also one of those pesky Linux users, and I have to admit that I have absolutely no idea how any of the converter programs y'all are talking about work; I've never used any of them, and I'm not all that likely to do so.

What I use is Dave Coffin's dcraw -- http://www.cybercom.net/~dcoffin/dcraw/ -- typically via a plugin for the GIMP (and Cine Paint) called ufraw. Dcraw itself is a command line program, but it might be a good candidate for Ctein's 'pick one!' standard converter all the same; it has broad RAW format coverage, is available cross-platform, and it's simple to process all the test files in a consistent way by using the same command line.

Ufraw effectively presents manipulable values and manipulable curves next to a preview image. It has autofit functions, but I wind up moving exposure values and/or sticking a point in the brightness or saturation curves and nudging until I think the image looks right, most of the time. (The exceptions are those rare and shining moments when anything I did to what's right there would make it worse.)

I really like this. It's wildly subjective, to the point where I notice that having recently cleaned my glasses affects my opinion of whether or not something looks right (and I could probably benefit from a better monitor), but, well. "It looks right" is what I'm after; getting 'this is what was actually there' is not of practical interest.

So I very much appreciate a work flow that enables a desire to make the image look right.

I am pretty sure that quantifying "looks right" is effectively impossible to do in a useful way; it's either an individual idiosyncrasy or a broad statistical measure in a population, it's not a characteristic of the photograph as such.

"Olympus 4/3rds as a format is inferior to APS-C. Full stop and at little more than half the size, only to be expected."

I realize that the above is the "received wisdom" on various Canon and Nikon forums, but it's just false. The 40D sensor is 14.8x22.2mm and the E-3 sensor is 13.5x18mm. You can do the arithmetic yourself, but linearly the Canon sensor is ~1.23X the Olympus's, and area is 1.35X.

Any way you cut it, that's not "a little more than half the size."

The "JND" (just noticeable difference) for all sorts of format-related parameters is about 1.5X. That is, the ability to detect which is the larger and which is the smaller sensor WHEN LOOKING AT PICTURES is *at least* 1.5X. (2X is more reliable.) Unless you use other clues, you will not be able to tell the difference between pictures taken with an APS-C-sensor camera and pictures taken with a 4/3rds-sensor camera. And nobody else will either. There's simply not enough difference in size.

There are plenty of legitimate reasons to choose something other than Olympus. Sensor size isn't one of them.

Please see my longer article in the March 2008 issue of "Black & White Photography" (issue #83) for a more detailed explanation.

Mike J.

"(witness the size of the 300/2.8 compared to Canons for example)"

This seems to be deliberately obtuse, since the 300/2.8 on the olympus has more or less the same, field of view as the 400/2.8 from Canon on the 40D. They both give the field of view of a 600/2.8 on a 35mm camera.

In other words, when comparing it to a Canon lens for the 40D you should compare it with the HUGE 400/2.8

The Olympus is much smaller and much lighter: 7.3 pounds as opposed to almost 12 pounds. it's also cheaper.

I'm sorry, but the comparison remains valid. The fact remains that both lenses have identical focal lengths and maximum apertures. Sure you get a little extra reach on the Olympus but that doesn't alter the fact the lenses themselves are identical. Four-thirds was put forward as a system capable of miniaturisation. It isn't and citing the sensor multiplier is disingenuous. Theres also the question of light transmission if we're going to be bringing the sensors into the equation and maybe asking why Nikon and Canons 300/2.8's can light up a FF sensor and yet be physically smaller than the Zuiko that lights up one-third the area..

MikeJ, you're right. I mistakenly took the 1D3 sensor as comparison. Don't totally agree on your other point though. In fine conditions certainly, but the more difficult the conditions, the greater the difference (dim lighting etc) surely?

TBH, the whole point of 4/3rds has completely alluded me. Everyone else has settled on an APS-C size format and there is now a gentle trend towards the larger. A comparison of the sensors in the E3 and the 5D is frankly ridiculous. Olympus tried to different for the sake of difference, continuing a line of blatantly unsuccessful dead-ends since the advent of auto-focus 25 years ago. Would it have hurt them to innovate in the same room as everyone else instead of tinkering outside?
Nobody wants this once great manufacturer to fail but does anyone truly see them competing with Pentax and Sony if they carry on along this path?

I understand that many people seem to be happy with the Nikon D300, but that being so, why are all of my friends buying the Canon 40D?

I think that I know 6 people with Canon 40D (including myself now) and Zero people with D300. Maybe the Nikon users that I know cannot afford or justify the "jump"?

The comments to this entry are closed.