« The Worst Photograph Ever Made | Main | Soft-Spoken Deal »

Friday, 19 December 2008


Feed You can follow this conversation by subscribing to the comment feed for this post.

amen to that, Mike, keep going the way you do. nuff talk about gear abound on the web already.

A portrait, or shot like this is not so demanding on a sensor. A 4mp landscape shot may not hold up to quite the same degree of scrutiny at that kind of size where detail is craved for. Sure this has a ton of detail, but on a layered landscape shot, perhaps this may not be the case. Having said that, the 1D is a very sharp sensor, so no doubt someone could prove me wrong on that too.

However, on a lessor camera your theory of size of sensor for size of print may hold more water, certainly for landscapes.

I think there's a relative shortage of intelligent use-testing (which is how I'd describe what you do) of cameras, and a good supply of objective measurement of cameras and results. I think both are pretty important (as in, useful to me).

I also think you down-play the importance of the "conversation" aspect. There are LOTS of people here who actually have something to contribute to such a conversation; and by listening to it I'm learning things I'm unlikely to get from any single author, working alone. This works largely because you have a knowledgeable and articulate group of commenters (one of the most valuable things a blog can have!) and perhaps also because of your moderatorial policies (I don't really know what doesn't show up here).

Not to mention that it's also a lot more FUN to take pictures than to stare at a monitor comparing RAW converters! Don't blame you a bit.

Besides, I've observed something that you've probably experienced: the more detailed, rigorous, and comprehensive any hardware or software tests attempt to be, the more likely they are to be roundly criticized from all directions for overlooking this or that tiny variable. Who needs that? Those critics will never be satisfied by any test that they did not personally oversee anyway....

I just did an High-ISO rant on my own website (www.zone-10.com) where I took images from the horribly noisy Olympus E-1 and challenged the "current think" about how one needs the D700 to take night pictures. Yes, in a way, it was aimed directly at Mike Johnson's D700 reviews and comments.

Just like Stephen Scharf, I had to jump through some serious hoops to accomplish the task. Some of these hoops, just like Stephen's are automated, others are just par for the course in post-processing anything which is of value.

I believe that what Mike is doing in his discussion of the performance of these new, most-excellent, cameras is actually a diservice to the photographic community. This attitude that only the newest, shiny camera can be used to make nighttime pictures or enlargements is pure nonesense. The techniques that Stephen and I use are nothing new and nothing magical. Two years ago, they were the "norm".

I recall the huge hubbub over the introduction of the Canon 20D. This camera was the one that got thousands of people to convert from Nikon and Olympus over to Canon. It was the camera that could do "anything" and do it better than we've ever seen before. It also had incredibly clean High-ISO performance which we all went ape over. It's funny how that camera has been relegated to the dustheap of history.

All the photographic options that were possible then are no longer possible today. We've forgotton how to shoot with what we have and have amnesia as to technique when a new technology comes out.


"But note well, too, that with any camera, from the worst to the best, there will be people who use it well and people who use it poorly. Buyers of a camera aren't buying expertise along with it. That's something they have to earn. As with most things, it's not how big a stick you have, it's how hard you swing it."

That would have made a good preface for your reviews.

" Well, okay then. Stephen gets the prize for eking the most print quality out of 4 MP, that's for sure. My broader point here is that he knows more about printing 1D files than any reviewer was ever going to figure out how to ... "

No, not really - it seems a fairly common & well known workflow to anyone printing their own photos. (Jeff Schewe, Michael Reichmann et al ...)

amen to that, Mike, keep going the way you do. nuff talk about gear abound on the web already.

Ken N,
Anything you could do with an E-1 at 1600, you could do with the D700 at 6400. It's not magic, nor it is it essential. Still, two stops is two stops.

Mike J., doing disservices to photography 24-7

As an off topic aside I'd like to mention that the Inkjet Art Micro Ceramic Lustre paper that Stephen mentions is one of the great cheap inkjet papers out there.

Inexpensive and prints wonderfully.

Hi Mike,
I have to admit that I struggled and thankfully succeeded in suppressing my desire to insist that you use Nikon's Raw converter, because EVERYONE but you, it seems, KNOWS that it is dramatically superior to adobe's but as I said, I held back and said that as long as he uses the same for everything, who cares. I still want to know about the punch, snap and air from each so that I can be sure that you really understand what I am looking for.


Mike, wonderful follow-up to your three way compare. I found your inclusion of Stephen Scharf's "workflow" dead nuts on!

I just tried his "process" using a Canon 50D and the Gimp (no Photoshop here) and the results are stunning. Makes me wonder if I really "need" a 5D MkII. LOL!!!

Thanks for taking equipment off center stage and offering up the fact that some of us are after images. Cameras are just tools, right?

Keep up the good work.

Here is another way of saying the same thing (and something that Thom Hogan has been harping on Nikon about for quite a while now):

Why should it be necessary (or expected) for a user or reviewer of a camera to turn his or her workflow upside down to get acceptable output from such camera? There are lots of raw converters out there, but Lightroom (and I assume Aperture) are just much more efficient to use than most of them. I probably have five or six raw converters and Lightroom is so much faster and easier to use, that it isn't even funny. As a result, I tend to use it for 99% of my work. Occassionally I find it helpful to use one of the other converters, but that is rare. Just as it can take years to master a camera, it can take years to master software. I don't plan on switching software and relearning my workflow every time I buy a new camera, even if I only buy a new camera every 5+ years.

For photographers who have sufficient time and patience to test multiple raw converters, the expertise to recognize the strengths and weaknesses of each and the skill to use each of them well, it may make sense to try multiple converters when trying to extract the last 0.x% of image quality from a raw file. But even if Converter B is theoretically capable of doing a better job converting Camera X's raw files than Converter A, I bet I will still get better results using Converter A (with which I am more familiar) than by using Converter B.

You have to pick your battles when it comes to improving your photography and tackle the biggest issues first. A select few will be so advanced that comparing raw converters is the area that will yield the greatest reward, even if it means using a clunky/awkward/slow converter. For others, and especially for a reviewer with limited time to test a camera, such time is far more productively spent elsewhere.

As long as you are upfront about what converter you use and how you processed your test files, your statements are valid as they stand. In effect, you are implicitly (though obviously not explicitly) saying "Using Converter A in my workflow, the camera that produced the best output was Camera X." Others may achieve different and/or better results using different converters or using your converter differently. So be it. I know that my dSLR would produce better pictures in the hands of someone more talented, no reason why raw converters should be any different. If I thought Mike was using an inadequate program and didn't have any image processing chops, I wouldn't give any weight to his opinion in the first place. Unless we have some measure of faith in Mike's skills and objectivity within the test parameters he describes, why are we reading this blog?

Don't get me wrong. I'm not saying that people shouldn't experiment with other raw converters or that the pursuit of ultimate image quality is a waste of time. I'm just pointing out that everyone's needs and capabilities are different and you don't change your workflow at the drop of a hat.

Of course, there is also the question of whether the developers of raw converters should put more effort into optimizing their programs for specific cameras, or whether camera manufacturers should consider how their cameras' raw files will be interpreted by various converters. Thom Hogan has suggested elsewhere, for example, that it might be helpful if:

-- Camera manufacturers (such as Nikon) were more forthcoming with information such as the exact composition of and transmission data for their sensor filtration, and

-- Raw converter developers used more sample points in extrapolating the effects of various white balance settings, or took ISO into greater account in determining default conversion settings.

Best regards,

I try as best I can to follow the advice David Hurn gives on choosing a subject in 'On being a photographer'. I've been shooting in bars around my region for almost four years. I'm not sure it's worthy of a book but it's worthy of my time.

I've got to take my hat off to you, Mike. I wouldn't know where to begin as I'll demonstrate now.

My new 1Ds mark lll has made it with me to a couple of bars, I've bounced around in the back of a black cab trying to take a portrait after a night on the town with it (three sheets to the wind but I made it). I'll end up riding the bus with it, shooting my fellow passengers and anybody that's with me. I've been told you can use them in a studio and can even use them to shoot celebs (I think Ken Rockwell said that so get a second opinion). I'd say the 1Ds mark lll is best used for taking pictures (of things).

I gotta agree with Stephen here. Having fewer pixels means working very hard to get it right, and 100% right, so you don't waste any. There's no room for sloppiness. I shot a D2H for quite some time...and didn't have a problem getting great 12x18"s out of it. Of course, it's easier with more bokeh-y subjects, studio lighting, etc. What really surprised me though was that the 10MP D200 that replaced it wasn't that much better at those print sizes, given the same workflow and shooting technique - it just meant I could print even bigger. :)

Regarding Mark's comment:
If you saw the actual print that I sent Mike J., you might think differently about having sufficient detail for landscapes. Also, motorsports shots like this are more demanding than you might think; every sticker, graphic, button, switch, etc. are now expected to be rendered in perfect clarity (digital has raised the bar in this respect compared to film). But with respect to landscapes, certainly using a 6 or 8 megapixel camera and this printing workflow, there would a lot of detail.

MM is right in that this printing workflow is readily available to anyone reading or using techniques developed by Jeff Schewe, Martin Evening, or Greg Gorman (that's who Iearned it from).

For Ken N., regarding the 20D, I couldn't agree more. The 20D was a home run hit for Canon. You should see the B&W architectural photo print I also sent to Mike that was taken with my 20D and printed on Hahnemuhle Fine Art Baryta; I don't Mike would think it was a stretch to label it "Yowza!" (it is the lead-in photo on my website in the Architecture gallery).

Personally, I view the camera's specifications and functionality as "natural talent". How you hone that natural talent is up to the photographer. Micheal Jordan and Mikhail Baryshnikov both had incredible natural talent for their respective disciplines, but both will tell you that regardless of how much natural talent they had, they wouldn't have gotten to where to got to without a lot of hard work and dedication.

As Mike quite appopriately pointed out, it's how hard you swing that stick that's important.

Mike, the way you review is excellent.
In my occupation (Medicine) we make a fair bit of use of the principle of the "limiting factor". In any complex task or situation there are multiple elements which individually and in combination affect your ability to get the desired outcome. Typically, there will be one in particular that sets the limit on what can be acheived, hence the term limiting factor. For example, if a patient has mild arthritis and moderate heart failure we won't be aggressively treating the arthritis, as the arthritis is not what stops them being active, and treating it will aggravate the heart failure (which does limit their activity). For most of us taking photographs to be looked at, the limiting factor(s) will turn out to be features of handling (like the ability to actually see the image clearly in the viewfinder) rather than technical specifications, and that is where your type of review gives insights.
The limiting factor will of course vary between photograhers, due to preferred subject matter, style, hand size, muscle strength, extroversion, etc. so any review that says "This one is the best" is going to be of less real-world value than a review that says "I like this one because ..."
Incidentally, the K100D I bought after you gave it a good wrap is still doing well in the real world.

Eliminating tedious labor and saving precious time are good things, we might agree. If an upgraded tool is developed and released which enables both of those goals, communicating that information is a disservice in what way? Shooting the messenger here?

I suppose yet another disclaimer could be added to every article that says "this is possible with older tools and techniques" or "you don't need X or Y to accomplish Z", but where would one stop with that?

Furthermore, I think Mike is already saying in this article how small an impact brand new shiny equipment can have, and he referenced older equipment.

Where is the disservice here?

I haven't spoken about this at all, but I think that for me the advantages of the smaller (APS-C & 4/3) sensor size outweigh the disadvantages. I think I prefer the smaller sensor, all else being equal.

...Of course, all else is not equal. The smaller sensor cameras still don't have viewfinders as good as the full-frame models, and very few of them have a near equivalent of the 35mm f/2 lenses I was using on the three ff's. That tends to tilt the choice back to ff for me, but for me it's a camera and lens thing, not necessarily a sensor thing.

Mike J.

I have never used an E1, but if depth of field comes into play, it may eliminate most of the 700´s advantage of high iso. I expect that the E1 at 2.8 gives about the same depth as a full frame at 5.6. This is my reasoning to stay with aps-c and save some money, anyway.

"Mike J., doing disservices to photography 24-7"

This, for one, is one of the best reasons to read this blog and this type of review.

Taking one step back from myself being affected by the comparison as user of that or the other brand, it a) is interesting to read about the newest development not from a technical, but a practical side, and b) being constantly reminded that the creation of images is the goal and not the ultimate line-per-mm rendering of any subject on a surface.
It is this combination that makes this blog precious for me, and I do admit that I have a hard time trying to understand the overzealousity with which the personal standpoint is defended in absence of any aggression. Stating 'disservice', hmm, makes me chuckle more than wanting to learn more about the author of this statement.

Although I personally wasnt ecstatic about your canon nikon sony review I found myself satisfied on a different level. I have always prefered Michael Reichmans gestalt type reviews and my new toy the 5D2 has its place in my kit. I found your review well to use your term 'honest'. I now know why MR has 2 sonys now

Well done Mike for putting new camera experiences on a sensible 'how would I actually use it footing' and for (maybe) inadvertently making other techie reviews look like incomprehensible drivel. In younger days as an adventurous outdoors sort there were to me and friends the climbers and the 'equipment men'. We climbers had some stuff we knew worked and could depend upon and had fun. We were told by 'equipment men' how to climb better and quicker and would display their latest crabs (karabiners that is) embarassingly in the bar--their buzzing prattle encroaching on the golden glow of a pint after having a good day on the crags.

And so to this wasp's nest of reviewing equipment, I don't get the oneupmanship, the smugness at pointing out you've not done everything possible by the equipment men (interestingly not so much the ladies). The point is you have a camera, you use it--hopefully have a blast--you tell people what you felt about it not what the numbers add up to, that's just drivel.
Oh, by the way, being predominantly a user of film and finally after 3 years having achieved a slightly skittish workflow of getting great results scanning black and white (older emulsions scan best for me) and printing on the B9180--choice 50% down to you--does the Sony produce a more rounded, airy, 'film like' image? Its the 'wire sharp', occasionally flattish, stepped highlight look of DSLR images that has discouraged me from replacing my Nikon D80, which I don't use so often now. Sorry to ramble on.

"You have to pick your battles when it comes to improving your photography and tackle the biggest issues first. A select few will be so advanced that comparing raw converters is the area that will yield the greatest reward, even if it means using a clunky/awkward/slow converter. For others..."

I like the bit about picking your battles. The shots you don't take are probably better than the ones you spend hours on "in post", which is a good argument for taking MJ's handling reviews more seriously than the scientific image quality ones. The problem, of course, is that arguing on the internet about the shots you missed is significantly less entertaining than arguing about pixels noise.

But then Thom veers off into equating virtue with technical mastery. Maybe you discover that the pictures you want to take do need every ounce of detail, in which case best you learn about that. But for many of us, the pictures we most want to make don't rest so heavily on technical quality. There's nothing lazy or "non-elite" about choosing software which helps you with choosing, and visualising series & layout, even if it hurts technical quality. But, again, the number of people you can argue with on the internet about which 10 out of 100 pictures you should show, and in which order, is rather limited.

Adam, back up there a ways --

It would certainly be a good thing if the camera manufacturers were more forthcoming about exactly how their raw images work, but that is unlikely to happen because it would (or so they believe) reveal a great deal about their internal image processing pipelines, and I don't think there's any way they could get good publicity out of it. (Pentax has been getting some long-term stick on various fora over the K20D not allowing in-camera noise reduction to be set to "actually none" when shooting raw, for example.)

From my own idiosyncratic point of view, Adobe doesn't make any of their image management products for Linux, so I don't have the opportunity to use them. Since this is true for pretty much every raw conversion program out there, most discussions of raw conversion and work flow have a mildly surreal tinge from my point of view; they come from an obviously passionate and concerned but decidedly alien reality.

I use ufraw, which is a GUI front end for Dave Coffin's excellent dcraw. I'd suggest it to Mike as his second test converter (should he decide to adopt one), since it is both free and purely reverse engineered and thus very consistent in its biases, but it has the modest disadvantage of being relatively unknown.

But, anyway -- I think part of the point you're making, or, at least, the point I'd like to make riffing off the one you're making, is that how the camera interacts with the raw converter and other software you're already used to matters, and it is, up to a point, an issue with the camera design.

A camera with utterly splendid image characteristics so long as it is only used with the proprietary converter shipped with it is a different camera from one with equivalently splendid image characteristics when used with any old image converter. (Presumably the makers of the second camera are much more open with their raw image specifications.)

So I think Mike is right to test cameras with whatever raw converter he is in the existing habit of using; where the camera is on the spectrum between "works only with the proprietary software" and "works with anything" is an important thing to know. Since the converter he's using is the one with the lion's share of the market, especially so.

(I don't qualify as "experienced", but I think I get quite a lot out of Mike's camera reviews; a big chunk of it for me is some notion of what one can reasonably expect to get out of the camera, since not being experienced I don't have much basis of my own to conclude anything about that.)

Mike, oh you really can be a fabulous old curmudgeon sometimes! (In Oz we would say 'ratbag').

I was interested in the Scharfe sharpening protocol mentioned in the previous post. For instance:

1.What perceived/actual advantages does Photokit have over the standard sharpening tools available in PS. (In my case CS3).
2.What does 'unpressed' in Photoshop mean exactly?
3.Why 360dpi instead of the generally accepted standard of 300dpi? Can it really be discernible I wonder?
Thanks. Dennis F.


What about the Sigma 20/1.8? Or 24/1.8 if you prefer. That matches up (on APS-C) with your 35/2 on FF pretty darn well IMO.

I'm actually looking at getting it for my FF Sony Alpha 900, but that's because I am a wide addict and don't feel I need anything for available light work between the 50/1.4 and the 20/1.8.

But if I want something "in between" there's always the Sigma 20/1.8 or Minolta 28/2. . .

Certainly Mr. Scharf can make a 4 MP image work, but I am curious as to why? If he is Track Photog for Laguna Seca, why the labor with a setup that if things aren't perfect, no shot? Why not get a 1Dmkll, $1200,
have double the resolution and a little leeway in the workflow, maybe more keepers?
Just wondering....

On this raw converter subject - how about looking at it this way?
When you buy a camera today the manufacturer's raw converter is part of the overall package, it's like buying a film camera that will work with any film but will only give of its best when using the film recommended by the maker.
Yes, many of the raw converters that come "with the camera" are slow & clunky but personally I get around this by just quickly adjusting exposure, DR and colour balance - anything else is done later in my normal image editing programs.
We would only get a level playing field on raw comparisons if all cameras output dng and only dng - no wonder most reviews confine themselves to jpg results. As soon as you start discussing raw there are too many variables.

Cheers, Robin

"I believe that what Mike is doing in his discussion of the performance of these new, most-excellent, cameras is actually a disservice to the photographic community. This attitude that only the newest, shiny camera can be used to make nighttime pictures or enlargements is pure nonsense."

That's not his attitude, as I read him, or the attitude of most people here, many of whom still celebrate the use of relatively slow, grainy film, even in nighttime shots. You've set up and are knocking down a straw man.

But guess what -- the newest, shiniest equipment makes it easier to take these shots, and in some circumstances, makes the technical quality of the shots better. There's nothing wrong with printing 13x19, if you like small prints -- and there's nothing wrong printing 36x60. But if you printed the motorcycle shot at 36x60, it'd look a lot better with a 1DsIII, because with the 4mp camera, most of the information would be invented, and it'd look like it.

Or, try this with a low ISO camera -- go out and shoot the night sky, the Milky Way, and get good high-res "pin-point" stars, rather than star-tracks. Can't do it, unless you're using a tracking computer...not enough light gets to the sensor fast enough.

The argument isn't that the new equipment is perfect, or that you can't take good photographs with lesser equipment, it's just that the new equipment is *better.*

Stephen takes great shots with his 4mp camera; but so much depends on use. Because 4mp is fine for *one* person, doesn't mean that it's fine for everybody, and everything. Landscape artists are the most demanding: start uprezzing a tree line, with billions of individual leaves out there, and you'll get something that looks like it was painted by Picasso during his cubist period. And I would be interested to see what the uprezzing technique would do if a rider were wearing a *plaid" uniform.

When Mike expresses amazement at what the D700 can do, it's because (the way I read him) he comes from a film background, as I did. The early digital cameras didn't impress me that much, because they were sorta like film. But the D3 is a different world for people who grew up with a choice between ISO 100 and 400, and you only used the 400 in a pinch...I worked the Republican National Convention with cops all over the place and crazy people running through the streets, at night, where you *really* didn't want to use a flash, and you could actually take usable shots by available streetlight, and do it routinely...with a friggin' zoom lens.

"Having fewer pixels means working very hard to get it right, and 100% right, so you don't waste any. "...it just meant I could print even bigger. :)"

Guess what? Being able to print bigger is (and was) pretty important to a lot of people, including some who are generally accepted as the greatest photographers of the Twentieth Century. That's the basis of Ansel Adams' comment about "taking the biggest camera I could carry." The "having fewer pixels means working very hard to get it right" argument reminds me of the old BS discipline of not cropping the photo -- it puts discipline, or even the accident of position, before image quality (image quality in the widest sense, including composition.) Don't want to do that...


And she said, "Is that a stick in your pocket or are you glad to see me?"

I'm a new reader of your site, thanks for the great content!

I really appreciate your review style and the way you've described it reminds me of my favorite gaming site Kotaku. (http://kotaku.com/5012473/about-kotaku-reviews)

Merry Christmas!

PS: My budding photographer sister just got a Xsi for her birthday and I've been showing her as much as I can. I'm basically teaching her what I've learned as an SLR benchwarmer since I can't afford one for myself yet.

The comments to this entry are closed.



Blog powered by Typepad
Member since 06/2007