Obfsucation.
Following on from an interesting suggestion made by knickerhawk in the comments section of an article a little while back on achieving visual consistency, here’s both a little test/ exercise for you and a little more expounding on the idea of sufficiency. Read on if you think you’re up for the challenge.
Many people claim that they can tell the difference between one camera and another – that ‘Leica look’ for instance, appears to be something fabled and useful for inflating retail prices and seeking personal justification for purchases – but I’m not so sure. If you take away cues such as very shallow or very great depth of field, expose carefully not to clip exposure (or intentionally clip as an artistic choice) – in other words, stay within the capabilities and shooting envelope of the capture device – I really think there is very little to nothing which gives the game away at typical display sizes. Since it’s obviously impractical to do a meaningful comparison of printed media on a website, we’re restricted to web sizes only. Granted, this may change within the next few years as 4k screens become more common and lower initial quality capture devices no longer benefit from downsampling, but for the moment – for all intents and purposes, you’re still looking at a 1000-pixel wide jpeg. At best.
So, without further ado, here’s the test: see if you can identify which camera the following images were shot with. I’ll make it easier and give you multiple choice options below the images; don’t cheat! But once you’re done, click through to the image on flickr to see the exif or scroll down to the very end of the post to see if you were right. Then, if you’re still with me, take the poll at the end. I think the results will be quite interesting. I have some theories, but won’t speculate until the data emerges because it might bias the outcome.
What’s the point of all of this? Since web is the final output for the vast majority of photographers, I think it’s a solid demonstration of both sufficiency and the fact that ultimately, it’s the photographer that makes the difference (or doesn’t make the difference). The camera is merely a tool; use whatever works for you, but don’t get hung up on it.
However: if a significant number of people happened to get all or nearly all of the images right, then perhaps we need to either reconsider the idea of sufficiency, or get you all to take polygraph tests. There will be individuals who will be able to identify with some certainty traits in each image that might give you a suggestion of the output; I’ve deliberately left a clue or two in, but no more than that. Also, just because there’s a camera listed in there you may not think I own – even if I don’t, it doesn’t mean that I haven’t used it or been lent it at some point and captured a worthwhile image or two. 🙂
Enjoy, and please leave your thoughts in the comments – I think this is going to result in an interesting discussion…MT
1. Contax 645/ Acros 100, Pentax 645D, Nikon D800E, iPhone 5s
2. Hasselblad 501CM/ Acros 100, Nikon D800E, Sony RX100, Sigma DP3M
3. Hasselblad 501CM/ CFV-39 digital back, Ricoh GR, Nikon D800E, E-M1, iPhone 5
4. Hasselblad 501CM/ CFV-39 digital back, Nikon D800E, E-M1
5. Leica V-Lux 3, Sony RX10, Nikon D800E/ 85 PCE, E-M1
6. Leica M9-P, Fuji X-E2, iPhone 5, E-M5, Ricoh GR
7. Leica M9-P, Leica M-Monochrom, E-M5, Nikon F2 Titan/ Acros 100
8. Nikon D700, E-M5, Canon IXUS 520HS
9. Canon IXUS 520 HS, Sony RX100, Leica M9-P, iPhone 5
10. Canon IXUS 520HS, Leica V-Lux 3, Nikon D3, Olympus E-410
The answers are at in the very last line of the post.
__________________
H2 2014 workshops now open for booking – Making Outstanding Images San Francisco, Chicago and Venice; Masterclass San Francisco and Venice – click here to book or for more info
____________
Visit the Teaching Store to up your photographic game – including workshop and Photoshop Workflow videos and the customized Email School of Photography; or go mobile with the Photography Compendium for iPad. You can also get your gear from B&H and Amazon. Prices are the same as normal, however a small portion of your purchase value is referred back to me. Thanks!
Don’t forget to like us on Facebook and join the reader Flickr group!
Images and content copyright Ming Thein | mingthein.com 2012 onwards. All rights reserved
Answers: 1. 645D; 2. RX100; 3. GR; 4. 501CM/CFV-39; 5. RX10; 6. iPhone 5; 7. F2 Titan; 8. IXUS 520HS; 9. RX100; 10. V-Lux 3. Surprise surprise: no D800E!
Ming Thein | Photographer 
Interesting – unfortuntaley your Flickr images are small, meaning, i could not zoom, ok perhaps cheating, however this would of helped. I am still of the opinion i can see the 3D pop from leica from the x1 to M10 etc….
http://brianjcosborne.zenfolio.com/blog/2017/3/my-first-leica—the-leica-x1
The Fuji GFX is certainly presenting itself with an amazing “Medium Format” look, the Hasselblad X1D is lacking, a bit flat?
I think the marriage between lens and sensor makes it happen. Anyway, enough engineering talk, i am off to take photos 🙂
Larger images are available if you click on the images through to flickr – but I’ll never post full size because of image theft etc.
GFX vs X1D: could it be a processing thing? I find the Hasselblad images more natural/ linear by default, but that does not mean they can’t be post processed to be punchy; if anything, I’d much rather have the choice of where to put the contrast rather than have it pre-decided for me at the factory 🙂
First of all: Thank you for your amazing site and all the knowledge you share for free!
Love the test and I think the overall conclusion holds: at these sizes and with a photographer who knows what they are doing, you cannot really tell a difference.
I only got two right: #4 (because of the appaling noise) and #7 (because of the very pleasant noise;) ). However, I feel that I was still “better” than 2/10 suggests and I think that might be true for more people here. With #1 I was absolutely certain that it was either the 645D or the D800E because the iPhone wouldn’t focus through the plants and because it looks digital. #2 again, I was certain that it was digital. #3, I admit, really could have been any of the options for me. #4 was definitely given away by the shadow noise. #5 fooled me because I thought the excellent highlight retention meant D800E. #6 was difficult to tell but the DOF did point toward a smaller sensor. #7 I got right because it simply looked analogue. #8 was difficult to tell because the image really doesn’t have any characteristics that would make one of the options impossible. But as someone pointed out, the Ixus was very likely considering the circumstances of the shot. #9 I was sure that it wasn’t the iPhone because of the gradation and detail in the shadows. Also it looks a bit too tele for the native focal length (though it could of course have been cropped). For #10 I was pretty sure that it wasn’t the D3 because… I really don’t know.
What that means is that I think while it is impossible to get them all correct, you can make some observations and exclude one option for most images. I could definitely narrow it down.
Then secondly, there is still the option to tell a camera by the content of an image sometimes. I am thinking of your Cuba photoessay: there it was obvious (or let’s say “highly likely”) that you shot the “forbidden” candids of the military people with the GR. I admit that that is not always possible but sometimes, well, it is.
So here is my conclusion: I think this experiment proves conclusively that under most circumstances, it is near impossible to tell apart the results of very different (and very differently priced) cameras. However, I think that for a lot of the images “out there”, one can at least make an educated guess or narrow it down in the right direction.
And as a post script: There will be people seeing this and trying to convince other that camera XYZ is really not necessary because “as the experiment shows” the smaller/cheaper camera ZYX can do the same thing. Leaving aside the fact that you specifically chose to leave out pictures where one camera would have an obvious advantage and that viewing size can change everything, there is still the matter that a photographer might need the feel of a specific camera to produce a certain image.
Thank you. There is one other critical conclusion: the output of course matters enormously! I think we’d have a much higher average score if judging good prints 🙂
The reason for more-than-expected 0 scores is simple. The instructions were not clear on whether you should skip or guess, when not sure. I decided to skip questions for which I did not have a confident pick. So I only made a wager on two, of which one I got right.
The distribution of the number of correct answers is just barely distinguishable from a binomial p=0.25, n=10. There’s just a bit of overdispersion (a few more people than expected with zero, and a few more people than expected in the 5 to 8 range). But otherwise the distribution is pretty much the same as if people were picking one of the four options at random in each question. This supports two conclusions: people are reporting their results honestly; hardly anyone can tell the difference between the different cameras.
The binomial with p = 0.25 is OK. However, some pictures had three choices, others five, but most four. Qualitiatively the binomial gives a similar result to the Monte Carlo that I reported elsewhere in this thread. Too many folks with zero correct and too many with too many correct.
I got three correct.
No. 2 I had no clue really. This was almost a 100% lucky guess. I just picked a camera that seemed a tad unlikely for the subject (which in my mind begs to be recorded with the most detail possible.)
No. 7 I’m calling this one a “good guess”. It just looked like black and white film to me – and since that was one of the choices I went with that. But I certainly wasn’t sure of it.
No. 8 This guess had nothing to do with image quality and everything to do with context. I considered the subject matter, “telephoto” look, and guessed a likely picture taking situation. I figured if I were at an airport and had a compact “superzoom” that this is the kind of photo I’d take. So getting this right had nothing to do with picture “quality.”
So I got three right, but only one of my correct guesses could reasonably attributed even in part to the technical qualities of the image.
It would be interesting to know which images were guessed right most often. And which the least.
I was better in guessing what camera was definitely not used…
Reminds me of when I used to sell expensive audio equipment. If we got bored we would get the men (they were always men) to listen to 5 amps, or 5 turntables (OK it’s some time ago). We would then send the signal from the cheapest piece of gear in the shop, and tell them we were switching from bauble one to bauble two etc. The results were what you would expect, but still depressing.
I got five right. I put that down to a lot of wine and looking at the images without clicking on them, scrolling on an iPad mini.
Sound facetious but it’s easier to pick signatures at a glance that way I think! I got #3, I admit, because I know you love to wander around photographing trees with that GR! Like others I got #6 via DOF (but also distortion you wouldn’t accept on other lenses with that FOV). #7 I got on rendering. Was that the Noct-Nikkor? #8 and #9 were superb but lacked the enth degree sparkle – I guessed. But I was thrown on the others.
That said, yes, you proved your point, technique, style, consistency above equipment… At least viewed on an iPad.
The learning here are: (and not that how good you are in picking the image taking equipment)
1. If you have a consistent PP style, it is very hard to find out what kind of equipment you used.
2. In case of internet (screen) watching of photo images in low resolution you may not need the most advanced equipment to reach peak quality. (definitely you don’t need actually)
3. Post processing can have more impact than camera.
4. If you want to test yourself get high quality prints of your photographs.
Other associations:
1. There is inspiration coming from the visual topic, there is one coming from your camera, and there must be one coming from the process you choose to develop the final image.
2. The style of the final image comes from this three inspirations. I highly regard consistent style within a set or topic, and less across everything. If you carry on with one style you cannot develop, you stopped. You are static.
Finally, I determined I need to develop further in every ways.
Consistency matters: the feel of the image should not be a property of the camera, but a choice of the photographer.
I got none! I thought the iPhone shot might have been an iPhone shot due to the DOF, but the gorgeous tonality threw me for a loop! Would love to know how you achieved that, I am struggling with getting a decent B&W out of my iPhone (I shot 99% film until last month you see)
Not surprising – I suggest looking at your exposure; white clipping is going to be your main challenge.
Thank you so much, off to practise!
Only got 2. I’m wondering if a similar test could be run based on type of sensor (ccd, cmos), type of filtering and film. In a way, haven’t you proven that all digital capture looks alike?
Even if all digital capture didn’t look alike, we’re still stuck with the same output medium…
Most were toss-ups, but for some reason the RX-100 pics stuck out strongly to me. Can’t quite place why, though.
The difference in all those pictures in my opinion is you. You just know how to get the best out of all these cameras.
Wouldn’t that be the one point of constancy?
Interesting exercise, which I failed miserably, illustrating that a good photo requires talent by the photographer, maybe an element of (possibly manufactured) luck in finding a suitable subject, plus appropriate equipment and the latter, these days, includes a very wide range of possibilities. I conclude that, in many ways, we are in a very fortunate era for photography. On the possible downside, if you feel the need to stand out from the crowd with your photography, you are swimming in a very big pond, since arguably more than half of humankind now have personal access to reasonable quality photographic equipment and merely need to discover their photographic talent. It is the talent that counts now, not so much the equipment. Thank you for challenging me.
First, I tend to agree that it’s very difficult to guess what camera any given shot was made with, absent obvious clues like razor thin DOF and a choice of a full frame vs. an m43 or 1″ camera. Especially on the typical monitor. Especially with my aging eyes. But I do take issue with the idea that web viewing means nothing more than a 1000 pixel wide image. My monitors support 2K resolution (1440×2560), and fairly serious photographers, at least, are likely to have monitors that support much more than 1000 pixels horizontally.
That day may eventually come, but we’re not there yet. Right now, my issue is even with releasing images at the current size, I already have theft and unauthorised use. Imagine how much worse it’d be if I posted even larger images. On top of that, I did run a site experiment at full res 2k – the browsing experience was terrible because of the size of the images; not at all fluid.
Statistically, since the choices per question aren’t the same across all questions, the chances of getting the answer right by shear luck aren’t constant. Poisson is at best an approximation under the situation. Also, there is a difference between knowing that some answers are wrong and knowing the right answer. Even if your readers can’t really tell the difference between cameras, the fact that some of the possible choices can be reasonably eliminated alone would lead to too many people getting too many questions right as you observed. Finally the reason why you get too many people with 0 correct answers is what I call survey-abandonment. If the survey is too difficult, the people taking it would find the quickest way to abandon it. This means picking some numbers more or less by convenience. Zero is at the top, so it’s a natural choice. You might get too many 1s if instead of 0 topping the list you put a 1 instead. Still, I suspect that you’ll get too many responses with zero answer due to guilt-honesty interplay.people who didnlt finish the survey thinking it too difficult would be more inclined to think, “Drats, I don’t have a clue, wouldn’t get any of these right anyway.” Hence, zero correct is always the preferred response. This turns out to be precisely biased. If all you readers without a clue complete the survey randomly, getting all answers wrong is exceedingly difficult.
So what’s your conclusion overall? 🙂
I’d say your readers do have some ability to identify the cameras. Note that this is not the same as saying that they have the ability to tell the camera by the look of the resulting photos alone. The basis of comparison arises from the fact that too many people got too many questions right. The result implies either collectively your survey takers have the ability to judge by look alone or that they merely have the ability to eliminate improbable choices from each photo. Deductive logic, context analysis, and personal knowledge of survey designer’s idiosyncracies can all be used to eliminate improbable choices in addition to “the look”. The way the survey is set up prevents the conclusion from being more specific than that.
The conclusion would be much stronger as a series of yes (it is taken with a Leica) or no (it’s not with a Leica) approach survey. But then it would also be narrower, being Leica’s look versus everything else.
I have a different take on the excess of responses with zero correct. I think that it is a result of Ming making the survey intentionally difficult. That is, he took some pains to “obfuscate” the correct camera by choosing the incorrect alternatives purposefully. I think that my argument is supported by the fact that the alternatives vary from image to image. Also, followers of the site know that Ming often posts images made with the D800, yet Ming himself notes at the bottom that, “surprise”, there were no D800 pictures, even though the D800 frequently one of the possible choices. Of course, making the survey intentionally difficult might very well lead to “survey abandonment”, as suggested by Kirati.
Well, as the author of the survey, I have to say that it wasn’t made intentionally difficult, but it wasn’t made easy, either. Challenging to try and make it ‘fair’ – besides, that wasn’t the point of the exercise at all…
I didn’t count how many I guessed correctly, but it was somewhere close to 3 or 4. That said, I was able to eliminate the large/small sensor cameras for every image, which demonstrates there is a noticeable difference when it comes to sensor size.
A few examples of what I am talking about:
No. 2: There is not enough fine detail in the image for the photo to have been taken by a medium format Hasselblad or the D800E.
No. 3: This photo actually threw me for a loop. The Ricoh GR does a nice job of clearly reproducing the individual leaves. I knew the image was not taken with the Hasselblad (or at least I believed it wasn’t because of the lack of detail in the tree trunk), but my guess for this photo was the D800E.
No. 4: This one I guessed correctly, but it took a bit of time. This is clearly a natural-light photo, which the sensor captured wonderfully. This allowed me to rule out the EM-1. I almost went with the D800E, but I ended up choosing the CFV-39 because of the shadow noise, which I think would have been better controlled with the D800E.
No. 5: If there’s one thing I knew about this photo, it’s that it wasn’t taken with the D800E and an 85mm PC-E. It simply does not look like a photo that was taken with a tilt-shift lens. Moreover, when viewing the photo at its full size (972 x 1428), you can tell this photo wasn’t with the a M4/3 sensor (the EM-1) as the sensor had some trouble resolving the fine detail in the balcony railings. This is also the only time you mentioned the lens that was paired with the camera—a dead giveaway that the D800E was not the correct choice!
Nos. 6 and 7: Many Leica lenses suffer from field curvature, which, in my opinion, actually creates pleasing photos more often than not. These two images lacked any field curvature, and so I guessed (correctly) that neither were taken by the M9-P or the M Monochrom. Also, the M Monochrom takes photos that are almost painfully sharp, even when viewed on a monitor. Neither of these photos exhibited that quality.
No. 8: I am not familiar with the EM-5 or the images it creates, but I instantly knew that the photo was not taken with a D700.
No. 10: Like you, I’ve taken a lot of shots with the D700, and so I knew that the photo of the flamingo was not taken with the D3. Even though the image is small and the head of the bird is slightly blurry, I knew that the photo was not taken with a full-frame camera. Does it matter that I couldn’t tell whether it was taken with a Four Thirds DSLR or a point-and-shoot? Not really (at least not to me).
#2 – I don’t think this is a valid comment at this size; similarly, just because there’s no detail doesn’t mean I won’t use a larger format camera for the dynamic range/ tonal transitions…
‘Sharpness’/ microcontrast/ acuity at reduced sizes is more a function of the downsizing algorithm used than the camera itself. NONE of these images is soft at 100%.
You could very well be right. That’s just what I was thinking and the MacBook Pro’s retina display may be playing tricks on my eyes.
Link to pic 2 seems to be broken.
Works for me – could be a flickr gremlin?
Strange, got a 404 msg before, now it works – go figure…
A follow-up on my previous analysis of the results with a Poisson distribution. Someone else suggested a Monte Carlo simulation. The results for the MC are basically the same — too many people got none correct, and too many people got too many correct. Here are the results for a MC “experiment” with 1 million “responses”:
0 correct : 5.08%
1 correct: 17.74%
2 correct: 27.67%
3 correct: 25.70%
4 correct: 15.36%
5 correct: 6.29%
6 correct: 1.77%
7 correct: 0.35%
8 correct: 0.04%
9 correct: 0.003%
10 correct: 0%
As with the Poisson analysis, more than twice as many people than expected got zero correct answers in your poll. I have no explanation for that. On the other hand, I am also suspicious about the overrepresentation of individuals who got a large number of correct answers. In my simulation of 1 million random guessers, not one got all 10 correct. And I repeated that with additional simulations. Clearly some people did much better than random expectation. The question is: why, or how?
Experience? Visual clues? Or perhaps the multiple choice-ness skews things?
Great test, thx ! I got 4 right – those marked with “y”. I was hoping for 2-3, since i agree 100% with you – so I’m happy. My experience same with “mediocre” cameras, producing stellar monochrome and colour in sizes up to say 2500p on the long side.
1. iP5s – x 2. RX100 – y 3. Ricoh GR – y 4. E-M1 – x 5. Leica V-Lux 3 – x 6. Fuji X-E2 – x 7. Nikon f2 Titan / Acros 100 – y 8. Canon IIXUS 520 – y 9. Leica M9P – x 10. Olympus E410 – x
For the statistically inclined, I think it is reasonable to model a random outcome as a Poisson distribution. I compared the results (for N = 432) to the Poisson expectation (with mean number of correct choices = 2.8958). The results are highly non-random for the group as a whole. More than twice a many people as expected got zero correct. Also, its very unlikely that 3 / 432 would get all ten correct. There was also a slight excess of folks getting six and seven correct.
So more people got it wrong than random? Hmm, maybe my test isn’t balanced…
Fascinating. Three things:
1. I’m glad you finally did a post like this; should be illuminating for many people.
2. This also helps to demonstrate why so many folks out there now feel as though they have all the camera they need in their smartphone (and, by extension, why camera sales are tanking).
3. I noticed that no Olympus images actually made it into the selection (coincidence, I must assume).
1. No problem.
2. Yes – and the rest of us, myself included, find using one a breath of fresh air. Just expose, focus and shoot. To me that’s more of a revolution in the way we take pictures than anything else in the digital world – just look at the level of adoption to gauge success.
3. Coincidence, as they’re spread through all eras. I was looking for images that were not too distinctive or not distinctive enough…some came to mind immediately; those were the ones I used. Back catalog’s too big to go trawling 🙂
“Back catalog’s too big to go trawling”
No doubt. 🙂
Over on Steve Huff’s website I made a similar — albeit much briefer — argument about vision vs gear in one of the threads … and suggested that I should submit a half dozen images and then let people guess what was shot with what.
Cameras are so good today, that, assuming a certain baseline of “sufficiency”, it’s hard to make a ‘bad choice’, per se.
The trick is to know the limits of your camera’s shooting envelope. Which almost seems funny to say, as all of them have a much greater envelope than we had back in the days of 35mm film.
I remember when I was a kid first starting out, shooting 35mm K64 chromes. OMG, if your exposure was off by more than 1 stop — 2 at most — on either side, you were completely hooped (AND you couldn’t see the image right away, so couldn’t be entirely certain of what you had).
Whereas today’s Olympus E-M1 has 12.7 stops of dynamic range, superior acuity, more accurate color … and today’s lenses are of a [generally] superior optical design (they have to be to keep up with the demands of modern sensors) and can resolve more than lenses from the 1960s-1980s.
We’re so spoiled now.
In fact, I’m very impressed with what the Olympus OM-D can do. Is it for everything? No. I would not use it for…
1. low ambient light shooting beyond ISO 1600, particularly if there isn’t much light/dark contrast
2. astrophotography
3. high speed action requiring tracking (doubly so if the light levels are reduced)
4. if I needed enlargements past 20″x30″
In the case of the first 3 points, I would reach for my D3s. In the case of point number 4, depending on the circumstances (let’s say landscapes) you could actually tilt the OM-D vertically on an L-bracket and shoot some stitched images to up your megapixel count.
Otherwise, the D800 (or an A7r) are probably the best choice … if you wanna go all out and do it right.
People who bought the D800 believing it was a replacement for the D700 quickly realized that they were outmatched, I think. Those who bought to take pictures of their cat … outdoors … at high noon, probably over-bought. You need perfect shot discipline and the best glass available to realize the benefits from that camera, otherwise you run the risk of your photos actually looking worse. I would argue it’s a tool for a very specific set of applications.
On the other hand, for most everything else, and particularly in good light, the Olympus OM-D cameras offer some benefits not present in DSLRs: generally better corner-to-corner sharpness; no back-focusing problems; 81 autofocus points spread nearly across the entire frame; 5-axis image stabilization on all lenses (yes, even at 1/4000 sec it’s still a benefit); a truly outstanding face detect system if you’re shooting portraits, etc., etc.
I’ve recently seen two Russian portrait photographers shooting with Olympus m4/3 cameras and getting outstanding results.
Last, but not least, I think 16-24 megapixels is the “goldilocks zone”; more than sufficient for the vast majority of applications.
In conclusion, I’ll only add that, yes, the camera matters, but these days typically only when shooting at the margins, and only to photographers who really understand why, and when, they’re necessary.
We’re definitely spoiled, even at the margins. I know I am the weak link in the imaging chain. I’m not sure those on the original forum you posted on would agree, in my experience they have to defend their purchases…
Agreed. There is one poster in particular over there — a staunch Leica advocate who becomes openly defensive and arrogant if anyone so much as breathes a negative remark about the German brand — who remarked once that while in Africa he was getting better images of the wildlife with his Leica M than the shooters who surrounded him … all of whom were using pro Nikon and Canon kit.
And I entirely believe that’s possible. I also believe that to do so you would have to risk being mauled or eaten.
Unfortunately, none of his superior photos were on hand to support his assertion, however.
As I often say, owning a Stradivarius does not make you a violin player.
Gotta say, the E-M1 with the M.Zuiko 75mm f/1.8 is an amazing bit of kit. I did a series of thematic portraits over the weekend, and the 75mm never came off the camera (poor 12-40mm f/2.8 just sat in my bag). And I don’t think “full frame” would have offered me much benefit here >> http://500px.com/photo/81117395/i've-heard-that-song-before-by-robert-falconer
🙂 We’re all the weak links in the imaging chain at the moment. The point is none of us are going to improve unless we recognise that.
Back in the film days, I had a friend who worked behind the camera at a pro camera shop in L.A. He was a Leicaphile who insisted he could discern the “Leica Look.” To test this assertion, I shot identical images of the same subjects; one set with a Leica M6 and 50mm f/2 Summicron, the other set with a Canon EOS 1n with 50mm f/1.4 EF. I used the same film (Velvia 100) from the same batch, the same apertures and shutter speeds, camera on tripod, same lab, same day processing, etc. Apertures varied from f/2 to three stops down. The result: Even using a 10X loupe, he could not tell the difference between them. He was chagrined to say the least. So the more things change….
I can tell the difference. It’s in my bank account 😛
Fun post. The same could be said for lighting…. can you tell a ProPhoto, from an AB, or an SB900, or a Yongnuo, or Elinchrom? How about a softbox, octobox, or umbrella? Does it make any difference if Tiger Woods uses Nike clubs or Sears &n Roebuck? Probably not… 🙂 That’s why in any disciple, you REALLY REALLY need proshops….
The former – other than power and sync duration cues, no. The latter – light modifiers are very distinctive, and yes, you can tell the difference. If it didn’t, those of us who make a living from it wouldn’t bother buying them – there are commercial decisions involved too. I don’t own any studio lights or monoblocks – I make do with a sack o’speedlights – but I do have an enormous number of light shaping tools, commercial, home-made and somewhere in-between…
I’ve used monos and packs for decades. I have 20 speedlights in a location kit, but with few modifiers. After awhile, you realize you that the shapes you imagined are just limitations imposed by your perception and application of the modifier, but that if you start paying close attention, you can create soft light with a hard source, and hard light with a soft source, and anything in between. Proximity, output, and direction are far more important then the source itself. Workflow and shooting approach are as effective as any physical modifier. Use a compositing technique, I can shoot a car using only soft available light, and unmodified speedlights to sculpt. The modifiers become more relevant if you constrain yourself to single frame capture. But, the time & expense to perfect a single frame capture on a car probably outweigh the time & expense to do the same job via compositing, which need not involve complex masking, but simply & softly painting in the sculpts or reflections.
On the original topic, I would be inclined to say that pixel peeping would tell a far different story then the whole shot at 1000pxl. A D810 with an Otus would likely look different then a 7D with the kit 28-135, and a NEX7 with the 18-105 would look different still. So then, it’s really a question of printing, and at what size the information runs short, given the capture setup… unless you print on coarse canvas, which can level the playing field again. 🙂
I agree on light, for the most part – which is why I don’t own any monoblocks. And same goes for pixel peeping. However, you’re forgetting most people lack the shot discipline to get every last bit out of a 7D anyway, let alone D810/Otus…
Well, I got the iPhone one right at least!
Love the mood in #4. My favourite image of the bunch.
A thought; here you have various subjects and various cameras. What about the same subject with various cameras? (You might need to crop to get roughly the same field of view with each camera.)
Man… None, i thought recognizing the famous “Oly” colors… Guess what?…
Personally i think that postprocessing is much more important than Camera…
Manufacturer color profiles are irrelevant if you use third party software and discard them…
I got three, but only by luck. The IXUS and the iphone I guessed mainly on the basis of those being the cameras you’d be most likely to be carrying in that situation. (Yes, the iphone shot has a very wide depth of field, but I think I could get that with a 1 inch sensor stopped down?)
Anyway, I very much agree with the idea of the post – I browse multiple forums on dpreview looking at the pictures posted, and every single forum features excellent images, regardless of the sensor size or brand. If I pick a photo at random I can almost never guess the origin. (Although, the Nikon 1 forum does feature a disproportionate number of the airshow photos….)
I would venture to say, however, that in some cases your skill as a photographer (including post processing) mitigates some of the weaknesses of these cameras. I can tell the difference in my own photos between those shot with a Nikon 1 and an APS-C Samsung because I’m not able to get as much out of the raw Nikon 1 files in terms of dynamic range or match the white balance perfectly between the two cameras. So I think it’s perfectly reasonable that people could spot the Nikon vs Canon look in OOC JPGs, for example. But this only really applies to judging shots by inexperienced amateurs, which is not where most of the wild claims are made 🙂
Learning the strengths and weakness of one’s equipment can make up for a lot of shortfalls – and save you a lot of money, too…
If i can’t remember what camera i used for my images, how am i going to go with yours..lol….something i have been saying for years, pointed at the modern geek who has no vision but enjoys holding an electronic gizmo in his hands….they’re over at DPR 🙂
well done.
Reading the comments, and thinking about this more, I think multiple choice makes this a little easier and perhaps not completely predictive of whether people can tell a camera’s look apart.
If you had a list of cameras used in one column, and the pictures in another column, and asked people to match them up, I would be surprised if anyone gets higher than random guessing. I might have gotten the GR and the B&W film, but that would be it.
Or perhaps I should just make it easier and say guess the format?
Only 1 correct.
I was amazed at the quality of the B&W images 6 & 7.
Thank you for posting this.
You’re welcome.
I didn’t take this test because I’d fail miserably!
I’m a Leica freak using very old lenses.
Doing everything manually and slowly appeals to me.
For me it’s fun and satisfying.
It’s like preferring a sailboat instead of a motor boat,
a bow-and-arrow instead of a machine gun.
Are my images optically better?
Probably not. Maybe in some rare cases.
Ming Thein’s test was long overdue.
I’m reminded of audio freaks who maintain they can differentiate sonic difference between different speaker wires etc.
Maybe they can.
But let them be tested while wearing a blindfold!
But if you are one of those extraordinarily sensitive individuals who can tell the difference between the different wires, why not? If it makes you happy… 🙂
Ming, thanks for following up on the suggestion. Much appreciated, especially since I know it must have taken considerable time and effort to pull it together. (Minor point: my original comment was to the blog post that used a bunch of spider charts to show different parameters, not the one you linked to.)
As for the test itself, I relied on my first guess without trying to analyze or rationalize each image closely and ended up getting five right. That’s more than I expected, but I’m confident that if you offered another round my hit rate would be different (and probably lower). To me, the comparisons show that, at these display sizes at least, there really is no general IQ advantage to the larger formats. One is forced to rely on technical “guesstimates” (e.g., DOF and noise patterns) rather than overall effect/impression to gain any statistically meaningful advantage. As far as I’m concerned this blog post beautifully serves as a reminder that the benefits of larger/newer/more expensive formats are almost always to be found at the farthest limits of our shooting and viewing parameters. The next time I’m tempted to “upgrade” my equipment I will remind myself to review this blog post. Thanks again!
I agree completely. I was able to get more right than I expected (7) but that was because I relied on technical items. For instance, the grain in the F2Titan picture or the massive DOF in the iPhone picture. None of the distinguishing qualities remotely affect overall “image quality”, especially when in the hands of a skilled photographer. Great suggestion and great post.
Ah, the shooting envelope? I was trying to use the spider chart to explain what I meant by that term.
I’m pretty sure if I wanted to be unfair I could choose images that were deliberately shot with the characteristics of another format, but that would be disingenuous and rather defeat the point. Even at these sizes, the DOF and noise patterns don’t give away very much, I think.
And yes, that was the point: equipment really doesn’t matter for most uses; the statistics were a red herring. 🙂
Hi Ming, great post and beautiful photographs. I am always amazed at how consistently you capture and showcase your artistic vision, regardless of the image capturing tool.
I was 5 for 10, with correct pics on #1, 3, 6, 7, and 8. I’m not sure what that says about sufficiency, but I do feel like the smaller sensor, more point and shoot images as well as the film images seemed to be more evident. With the P&S/phone images, there just seemed to be less of a 3-dimensional feel and somewhat less richness of contrast. With #6 I also felt like the lens was constraining the true feel of that moment and was showing its limitations near the image corners. With the film, I feel like the images had this really beautiful, smooth grain that really isolated your subjects in focus while still giving character to the out of focus areas. The film shots just have something special about them that just felt…well, “filmic.”
To be honest, I’m actually surprised that #2 was an RX100 and #5 was an RX10. Both of those images are really rich and have a surprising amount of detail in the shadows for 1″ sensors. On the flip side, I surely thought #9 was an even smaller sensor like an iPhone as it seemed more like #6 and #8 to me.
I think for me, maybe what I’ve taken away from this is that film (at least high quality B&W) still has a very important role to play in photography and from the 1″ sensors and up (assuming the lenses aren’t too much of a limitation), the photographer’s art and craft are the major determining factors. For whatever reason, I just can’t seem to find much enjoyment capturing moments with my Samsung Galaxy S3 phone (from the UI to the output). I personally prefer a dedicated camera with full manual controls where I know my only limits are me. I’m loving my Ricoh GR, my Olympus OM-D, E-M5, and my Panasonic GH3 (especially for video and for the great touch AF on the articulating LCD while composing on the EVF…no more fiddling with control pad arrows or dials).
By the way, my favorite was #4. That is classic Ming all the way!
Thanks Hal. Surprising you can see it at these image sizes, because I certainly can’t. And edge limitations aside – #6 specifically – it’s worth remembering that few larger format lenses are perfect; if anything, the corners get worse as lenses get harder to design…
I’m still shooting film to figure out how to get the same tonality out of digital – one day!
Maybe, my selections were lucky guesses or just intuition? I also tried to think a bit about the context of the photos and the situation in which you might be capturing the image. The image of the airplane just seemed like a P&S type of situation and the wedding seemed like one at which you might have been a guest and not on assignment. And there were just so many subtle transitions in the tonality and focus roll off for #7, it just seemed hard to believe it could be anything other than film. So maybe this type of critical (over) thinking of your multiple choices was not staying within the boundaries of your question? And I actually viewed these on my phone while reading in bed…which suggests it was probably heavily luck and these other non-imagery related evaluation concepts.
With regards to your hopes of finding film quality tonality in digital B&W images, please keep trying. If anyone will find a way, it will most assuredly be you my friend! 😉
While I love all of your posts, I am most deeply appreciative of the ones that ask us photographers to question things. You are a fantastic teacher (which is clearly only possible because you are an even better artist and craftsman)! In case these posts are helpful to your other students and readers, I really enjoyed how they complement your technical posts about the tools of the trade. The moral of the story for me…it’s never been a better time to be a photographer and the tools don’t matter much given the gift of magical lighting and creative vision!
The Digital Zone System
http://www.luminous-landscape.com/essays/digital_zone_system.shtml
How Big is a Pixel? (Part 1 of 4)
http://www.pekkapotka.com/journal/2014/6/3/how-big-is-a-pixel
How Sharp is a Camera Pixel? (Part 2 of 4)
http://www.pekkapotka.com/journal/2014/6/4/good-and-bad-pixels
How Deep is a Camera Pixel? (Part 3 of 4)
http://www.pekkapotka.com/journal/2014/6/9/how-deep-is-a-camera-pixel
How Clean is a Camera Pixel? (Part 4 of 4)
oops, forgot the final link.
How Clean is a Camera Pixel? (Part 4 of 4)
http://www.pekkapotka.com/journal/2014/6/12/how-clean-is-a-camera-pixel
Thanks for the explanation and compliments!
I got a ‘0’. I came to a similar ‘conclusion’ earlier this year when I was on Holiday in Hawaii. My wife made me leave everything but a Leica x1 and a Leica c (type112).
Each night i would look at the days images on my ipad and my wife could not tell which camera took the image. (Truth be told sometimes it was hard for me )
Because of the EVF on the C I ended up taking the C and leaving the x1 in the room 90% of the time.
The intended final output matters, as usual. You might not see much of a difference between jpegs on an iPad, but for a large print…
In any case, the X1 is pretty small!
Dear Ming,
As always a very interesting post! I guessed two right; the medium format ones; maybe because I am using my Hasselblad with 50 mB digital back mostly (reason: ergonomics, simplicity, and (I think) subtlety of output on print if the camera is used correctly) even for street photography (going out with camera and tripod on the street is really possible!). On screen output is very difficult to judge; print is the right medium for that. Surprised by the outcome of photograph 6.
Must say that I am now also looking for a second hand D4 (16 mB is enough; this post is showing that also), to get a difference experience (also after your article on the D4!)
Thanks for your blog!
Regards,
Peter
Enjoy. I shot street in Tokyo with the CFV-39 handheld; not easy, but doable. And necessary to have a GR in the pocket for when light gets low…
I didn’t try; I doubt if I could tell, and I’m not concerned that I can’t. Somebody, a few years ago, presented some pics—prints, I think—supposedly taken by a Leica. The fans were ecstatic until they were informed that they were all Canon images. If there is a Leica look, I don’t see it. Some people don’t like the colours from particular makers; again, if there is a difference, it doesn’t worry me.
I had, like another commenter, a Canon 5DII; but it was just too heavy. I sold it, and now have a couple of Leicas (once more). They make me think more about taking a photo; you could argue, despite being essentially very simple cameras, that they are the hardest to use successfully.
The less automation, the better, in my book – less to go wrong, more to make you think. I find my Hasselblads are the most transparent in that regard – you just set focus, aperture, shutter speed. All of it is on the lens barrel – the rest is up to you. At least I don’t have to remember which button does what!
Hello Ming – 5 out of 10. Easy guess with the GR, Blad, RX 100, Ixus. Amazed by the iPhone 5… On my iPhone 5 screen :)…
Sharp and accurate article, as usual.
Cheers
Ming, pure genius, best post ever. When I get some time I want to study this more, learn the nuances, hoping to improve my skills.
Throughout the comments I believe your thoughts on these pictures have been expressed however if you could go through each photo and express why it was chosen, what qualities etc? Please please do this type of post again in the future.
One point on equipment, the quality of the photo is not always a determining factor in choosing a camera, the handling/ergonomics plays an important roll, sometimes the most important. Hasselblad V, Leica M, Iphone to name a few.
Thanks. Of course the ergonomics are important: you’re not going to carry/use something that you don’t like, and therefore no photographs at all will be made…
I got 2. But I have a question? If you post processing with Adobe Photoshop or Lightroom all this pictures, is impossible to guess wich camera took it.
I think that would depend very much on your PS skills 🙂
Hallo Ming,
with joy and without cheating; I got five.
(Being a Ming-Thein-Fan helps a little.)
For me the easyed was No. 7; the F2. Takes me 1 sec.
Owning only Digital-Cameras and lightyears behind your mastership… my cat knows…
this shows another sideway-discussion, that is still not closed for me: The-Analog-vs-Digital-Tale.
Analog taken by a master rocks.
Thanks for another enjoyable post.
Well done!
Seven correct, though my feeling is that has more to do with me remembering places you have visited. Somewhere in my mind that information is stored, and my picks felt intuitive. The iPhone and Canon IXUS were surprises.
Glad to see I can still surprise the pros occasionally 🙂
Nope: I also got 2. I do, though, strongly admire your post-processing abilities, as well as your distinctive visual style.
Thank you.
Well, a big zero – tried only five (4 through 8). But I don’t know if that means anything cause all look so Ming Thein’s… Now days I believe equipment are not so relevant from a IQ point of view- but skills, a particular way of view, and post processing abilities are much more prominent. I sold my 5d-II not because I was unhappy wth pics, but it had became to heavy and cumbersome for me.
Rationally one may choose equipment for ergonomics and easy of use – but few of us – no professional people – are rational. There is a kind of “brand supporting” (oh my english – soccer supporters) that obscures any rational decision.
I sold my 5d-II not because I was unhappy with pics, but it had became to heavy and cumbersome for me.
Sorry, Reposted accidentally – my internet is rolling me this days..
If the cameras are within their shooting envelope the pictures are ok for the web – but it is amazing how often this is not the case. Today it was dynamic range that was a challenge.
Less so these days, actually. Even the iPhone’s envelope is significantly wider than you’d expect because of the multi-shot HDR function…
I got 2 of 10, purely by guessing. But the interesting thing is that I was going to pick the 800E for #7, because of its particular form of sharpness, but we weren’t offered that choice; then I looked at the flickr and saw that an 800E did indeed take the picture! (For reasons you explained in answer to an earlier comment.)
As noted by many earlier commenters, it’s not the cameras . . . lit’s you. I find #4, the tea ceremony, particularly wonderful.
Thank you.
Sufficiency is key, and you rightly stated that up front. Rarely do contemporary photographers output to print. A print, sized appropriately, is the only way an artist can control the viewing experience and his end product. And knowing the output source helps us choose the appropriate imaging tool.
Precisely – hence my recent obsession with printed output. I want to make sure that as an idea, the audience sees the same thing…so at least any losses in translation can be solidly attributed to something compositional…
The image that caught my eye was No. 6. Then I saw that it was taken with the iPhone. I was very surprised. But I have long thought that much of the sound and fury expressed on the gear forums regarding comparative differences in formats or camera really is irrelevant to how most of us (including many of the most vocal forum protagonists) view our images. How many of us actually print at all, let alone at sizes whereby we can see differences in sensor size or brand of lens? Thanks for the comparisons, they are enlightening.
A pleasure.
Can you guess if this soup is cooked on a gas stove or an induction stove?
I was once part of a wine tasting quiz, where teams of 3 were given 3 wines to taste (times 3), and 5 choices for each set where they should place the wines. A team with 2 wine buffs got such a low score that it was almost improbable, below what could be expected based on pure random guessing. Does this make them less knowledgeable about wine?
There are of course certain properties that can be analyzed objectively (DOF, resolution etc.), but ultimately, I think it is more a question of what kind of tool that works for you, what kind of tool that gets the pictures you want to create. Regarding the “look” (Leica, or other) I think the only way to test this is to compare a series of images of the same subject/light/motif/style, taken with with different tools/cameras/lenses. I have heard Leica aficionados say that their Leica images are rated higher than those taken with other equipment, but again, if this is the tool they prefer, maybe they make better pictures with it?
Fun, but hard to draw conclusions.
Bingo. The whole point is you are not supposed to be able to draw conclusions for all of the reasons you state (and more). 🙂
I make scrambled eggs on different stoves, and yes there is a difference in taste from gas/electric and induction. But mostly difference is noticeable.
Good photos rescaled from 1:1 to web size, so difference starts to be unnoticeable as long as specific parametrs start to be visible (like bokeh: onion ring bokeh, FOV, DOF).
Your photos Ming, shows that you are controlling your gear and you always got great effect.
p.s. Thanks for you email and clues. I have to look deeper and read about Malaysia but I have hard time in job.
An impossible task Ming! I got 2 correct and that was from looking at content rather than quality. I wonder if at that size there would be any difference in No.6 if it had been taken with a Leica rather than an iPhone? What about if it was a 15 inch print? I would hope so or maybe it doesn’t matter. I think now is the time for using the tool your happiest using as long as it’s sufficient for the job. A Nikon D3200 with the right lenses could have done most of the above photos and they’re as cheap as chips and look like toys but if it gives you that creative urge…
No reason why content should be limited by equipment – long tele shots of wildlife and sport may be an obvious exception, of course.
At these sizes? No difference to a Leica. A normal 12×18″? Probably not much difference, either. Only if you go much larger or Ultrparint.
Output intention matters. I agree the D3200 could have done the job – the iPhone clearly did – but given my output is for high end clients and high resolution prints, there’s quite a difference between that and medium format…that said, if your output is web only, then even a D3200 is overkill.
Well, a big zero – tried only five (4 through 8). But I don’t know if that means anything cause all look so Ming Thein’s… Now days I believe equipment are not so relevant from a IQ point of view- but skills, a particular way of view, and post abilities are much more prominent. I sold my 5d-II not because I was unhappy with pics, but it had became to heavy and cumbersome for me.
Rationally one may choose equipment for ergonomics and easy of use – but few of us – no professional people – are rational. There is a kind of “brand supporting” (oh my english – soccer supporters) that obscures any rational decision.
By the way you can get really good pics from any equipment – that’s an exception?
I’d be more worried if they didn’t look like mine!
Even irrationally we choose equipment based on ergonomics and ease of use: you’re far more likely to buy something that you like to use rather than what is technically ‘best’…
Composition and light are all equipment-independent. Every angle of view behaves the same regardless of camera or format. So theoretically, yes – you can make good images from any equipment…
I got 7 right. It’s surprise for me. It might be due to my Nec monitor.
Not necessarily since many are monochrome – might just be your eyes…
Well sad to say I got 0 right. However, the takeaway, which follows what I have told people I have mentored: do not get hung up on your equipment. If it can take a picture, focus on the photograph you want to capture and use the device to capture it. You can become an awesome photographer using whatever tool you are able to afford. You do not need to wait until you can spend thousands of dollars on equipment before you can call yourself a photographer. Just get out and take photos and learn and grow along the way. Thank you very much for this eye opening exercise. I am going to share it with my group.
very interesting! the Link to Flickr for Image #2 is not working. I would have gotten it wrong if I could have seen it anyway. I seem to be able to pick out the Sony RX cameras mostly.
Works for me, but it might be a Flickr thing. I’ll check.
I didn’t take the test, simply enjoyed – and in parts re-enjoyed – the pictures. The thing with this test on your site is, that for me there is too much context to base a guess simply on image quality or camera/sensor/lens characteristic. Where would you shoot, when under what circumstance, etc… all this gives clues. Sure, the iPhone is always there, but not all these shots are ‘iPhoney’.
There was an article on PetaPixel recently, where a photog used a D800 and a D3100 with lenses of corresponding quality to shoot similar scenes. It’s amazing how close in visual appearance on the screen some of those images are:
http://petapixel.com/2014/08/23/cheap-shot-challenge-photos-taken-expensive-gear-recreated-cheap/
Anyway, always enjoying your posts, Ming!
I think part of the reason the two Nikons looked the same is because most manufacturers try very hard to give their cameras a consistent look – in jpeg, at least. In raw with processing, all bets are off.
As for the camera phone – we tend to pick those images out (typically) because they are shot generally with a kind of looseness and lack of attention to technique and composition that is a consequence of not being seen as a ‘serious’ tool, even if they are technically capable of good results. The other thing is that nobody would think of shooting seriously with an iPhone if something more capable was to hand…
I got 4, but have to admit it’s not because of the tonal range dof midtones etc, but through your photographic style + your photographic era (your style evolves) + some blurry memory of the location of photos during which camera review.
All wonderful photos and a wonderful article to demonstrate your point. Well done!
Thank you Ming for this one. For what it counts, I only got 2 and both because a bit of my brain recognized the images and connected to a camera (#9 and #10 to be precise, the heron was clearly memorable being, what, a couple of years at least?). The rest, even with some help from perceived perspective and DOF, was a pathetic wildass list of guesses. Perhaps the fact I was using an iPad2 in broad dailight didn’t help…
But at the end you are spot on, for most consumption purposes and for the time being 4k screens not being the norm, there is a good likelihood that a camera/lens/sensor/or/film/scanner combination will be trumped by quality of processing (your unique signature style counts more than your equipment, for one) and quality of viewing medium. Which in turn means that (1) one should choose a camera on usability factors more than IQ and (2) this will change, to our shocked surprise, when better screens will spotlight the differences between a poorly executed shot and a good one, and good camera/lenses will have their revenge day…
Will redo te exercise on a better screen indoors, but I doubt I’ll fare much better, the only diference will be how even more shocked I’ll be at the iPhone picture’s quality…
The iPad is actually pretty good color-wise. All that will change is physical size, because here the limitation isn’t the output medium but the original size of the uploaded image…
To be honest, I won’t even try. Because I totally agree with your statement that “there is very little to nothing which gives the game away at typical display sizes” (those four words being about the most important of this article). However, what I dò see in all of those pictures, unmistakably, is Ming Thein! 🙂
So maybe we need to do another test: placing a number of well chosen pictures, giving us some five different possible photographers and then let’s see if we can identify the right photographer…
Perhaps that would make more sense.
Thanks Dirk. The challenge with the test you propose is how would we decide what was ‘representative’ of that photographer? Often the work we remember or is famous is either too easily recognizable already, or there’s no access to the rest of their catalogs…
Absolutely! While writing down my earlier reply, I thought it wouldn’t be easy at all for exactly those reasons. But, maybe a small variation could be done. 10 pictures out of which 5 are Ming Thein’s. Indicate which ones.
Could still be very difficult, when you’d select “strange” pictures that come very close to what you do… But, you know, it would be a lot of fun as kind of a game. And… it would become very interesting if the contestants would have to write down under each of the chosen pictures why they think it is yours! I think I’d participate in this game (knowing that maybe I still would make a lot of mistakes).
I’m not even going to try, I’m just enjoying the pictures!
I love this article/exercise Ming and I’m going to take the test right after my Starbucks kicks in. For me, the operative phrase in your article is “at typical display sizes”. I’ve just gone through days of testing my a7r against my a7s to find that the 12MP gem crushes the 36MP colossus on my 2560 x 1600 hi-res display. a7r files look flat compared to the a7s on the hi-res screen however on my laptop I can’t tell the difference. Since I spend 80% of my picture viewing time on the hi-res display, the a7r is going to start collecting dust.
Output size matters!
Yes it certainly does and wrestling with a 36 megapixel file unnecessarily is a waste of my time. There is some unexplained tech going on with the 12MP a7s sensor -it is a wolf in sheep’s clothing. As a side benefit, its curing me of another waste of my time; pixel peeping allowing me to spend more time on a much more useful endeavor; shot discipline 🙂
I got 6, purely through assuming that a small sensor was used when there was a large dof. I thought the first was Arcos, but that’s how I got the F2 Titan, nothing to do with the camera. If you had not given multiple choices, then I probably would have 0.
Good test.
Michael
We’re so used to assuming large sensor/format = shallow DOF these days; unfortunately it’s a consequence of the bokeh trend…
Or the other way around: Small sensor = large depth of field.
Then again, it’s the only thing to work with (most noticeable difference anyway) when telling one camera from the next.
Michael
Not if you rarely use shallow DOF anyway 🙂
I didn’t take the test as I know nothing of these things, for me a picture is a picture and either it’s a good picture that I like or varying downward degrees of not so good. While it’s true people have differences in their senses and some people are adept at distinguishing visual clues,Tastes and sounds etc I think these people are few and far between. As I normally shoot with bottom to mid range consumer equipment I can sometimes see the difference with higher end equipment but then post processing also comes into play; I’m not a big fan of this but I use Darktable on linux sometimes and noticed that the base tone curve can be adjusted to something that approximates other makes and models of camera, so I can make a photo from my Canon Eos1100d have a tone curve that is Leica like. All that said it’s an interesting experiment and I’d be interested to hear/see the results.
I got 5, the leica was the easiest to pick, the iPhone was a surprise. Was not easy on a ipad 2
Not any easier on a computer, nor is it meant to be 🙂
As was foreseeable, I was only able to approximate sometimes (i.e. I was far from a final conclusion – so it was really guesswork), but I got 4 right (one more if I had acted straight away on first observation – I really thought No. 7 had that film look …). Very interesting, the whole thing, but for me, also revealing – because even if I got it wrong, I was pretty much able to tell if it was a more sophisticated or less sophisticated device (I’m talking image quality, not features), except for two images (one I should have known, at that) – but then, I knew many of them, and sometimes the story and thinking behind them, so it may or may not be a high “success” rate.
In my eyes, it’s not so much the overall impression, but the interplay of certain factors (gradation, detail, perspective, sometimes format) that points in a certain direction. But, Ming: You’re that good at post processing that it’s no wonder you actually get the very highest level of consistency – I’m still amazed at No. 6. For me as a clumsy amateur in that respect (when it comes to image processing, at the very least), I think the quality of what the camera delivers is much more important than for you – in other words, I’m more likely to actually get and then stick to the “look”, and I’d probably subconsciously enhance it rather than normalise it in post. It’s a piece of very important information that if you know what you’re doing, you don’t have to. Most of us could replace lots of spending by learning and practising.
It’s not just postprocessing, it’s familiarity with equipment – no matter how good your PS skills, there are some things the hardware physically cannot do; you eventually learn not to even attempt these situations, so there’s nothing to have to recover afterwards. Working within the shooting envelope and all that…
Makes a lot of sense, thanks. I’m just back from a photo walk, and as much as I still like shooting with my PM1, its DR is limited, and it blows highlights too easily, too, so one has to watch out for that or choose a different camera – I should have reckoned with that more consciously.
But by the same token, it’s good for very low key moody work precisely because it already crushes the shadows for you if you spot meter for the highlights…know your equipment and all that.
2 – but only through guesswork on my part!
The difference is that you take the photographs and the camera doesn’t take them for you.
Your photographs have and esthetic consistence from your post-processing and your style of composition.
From the photographs I chose correctly three of them and I recognized three of them and, perhaps it’s useful the data to you:
-The picture taken with the Sony RX10, I remember that from your review so it was not my merit.
-The one with the Nikon F2 but that was a cheap guessing so it was not my merit.
-The one with the Canon Ixus. I think I can see the softness of the lens that I used to see when I had an Olympus superzoom, I mean, it seems the rendering given by a lens for a consumer camera, but it could be atmospheric so it’s more a bias than a certainty.
Bingo – thank you.
Alright, now we need a monte carlo simulation to see what the random selections would look like and assess whether these observations indicate with statistical significance that individuals can determine what camera was used to generate each photo. I think the lack of shallow depth of field made this much harder. If indeed people can differentiate between cameras, you can shrink the images down and determine the minimum resolution for camera identification. However, with no prior knowledge of cameras used for your various projects, I think most people wouldn’t know which camera was used to generate an image. Great point!
You’re welcome to try 🙂
Four (clicked three on accident) but not by any clues in the images. 1) I guessed because of the unusual camera, 3) because I have the Tokyo video and saw you shoot it. 4) I actually did choose based on photo 5) because I remember the RX10 set and 9) because you were in San Francisco and you posted RX100 photo from that trip (although it was not a primary camera you used).
I should stop recycling images…:)
I got 3 correct (# 1,4 & 9) ..probably lucky guesses since i’m not a pro..:-)
Amazed with the shot using the iphone5!
Thanks.
This was an interesting exercise.
There were a few that were much easier, 4. was too easy courtesy or the chroma noise in the shadows, and 7. quite obviously has film grain even at these display sizes. 1. required slightly more thought, but the DOF and distance to the reflection was clearly more than a iPhone would provide, and it just seemed to be medium format for some reason. The absence of grain (a dangerous idea with Arcos and small display sizes) felt like the medium digital was the way to go.
Now, I didn’t award myself 6. because of my own laziness – I was tied between the iPhone and the GR, and I never really decided before checking. The instant I did however I realized how clear the couple sitting only centimeters away were, clearly? iPhone.
8. I feel I would have had correct had I been familiar with the reach of the IXUS paired with its convenient size.
2. Was more than likely to be digital, as you make clear the limitations of films non-structured grain with presenting a sharp architectural image.
After these, I wasn’t doing much other than narrowing down what systems it wasn’t – and even then one or two I was surprised by.
4: Not necessarily, it could have been (and was) extremely dark; you could go to 3200 or 6400 on the smaller sensors and easily see that much noise or more – especially if the white balance is off.
2: Yes and no – depends on the size of the film 🙂
I got 6 out of 10 !
To be honest, I didn’t guess on any of the images. To me, it makes no difference what the camera was. The only question to me is “Did I like the image and enjoy looking at it”. The camera is a tool, the ending result is the image. I know that is somewhat simplified, but I either like the image or I don’t no matter what camera was used.
Cheers…
Bingo!
Great article Ming. I think the thing a lot of people forget when picking a camera is where the images will end up. Why do a lot of people think they need a D800E? Because they see that megapixel count or sensor size. They forget that the images will most likely be scaled down either by image size, resolution, or, more than likely, both. Whether you put the images on Facebook, Flickr or your own personal blog, they will all end up the same (most of the time).
There’s a guy on Flickr ( https://www.flickr.com/photos/izumiflowers/ ) who pastes this huge essay that he updates regularly, on ALL of his images complaining about each of the cameras he buys (and eventually returns) not being up to his standards. And yet when you look at all of his photos, they’re nothing but snapshots that YOU (Ming) could have done better with an iPhone.
Thanks Vernon. Maybe he derives enjoyment out of it?
Well, I got 7. But I’m not conceited enough to claim it was entirely due to visual perception. At least 2, or3, I remember from previous posts. Others were more interpreting the style/perspective/”unusual camera inserted here” (Ixus 520 😉 ) than purely “seeing” the difference. So it’s an honest result, but gained by methods that don’t fit the criteria of the question. Sorry!
Must agree with the previous poster, #6 got me completely – iPhone. Yes, wow indeed. And I couldn’t spot the F2 either…although maybe if I look closer…nope, wouldn’t have got it except by guessing.
Nice challenge. 🙂
I was counting on people not having that long memories. I suppose it’s a good thing because it means my work is at least somewhat memorable!
I guessed 4 correctly (1,3,6,7). Key word here being “guessed”. In fact, the only reason I got number 6 correctly is because you had posted some photos from that set earlier and I could not believe that those pictures had been taken with an iPhone.
I should have used 100% fresh images 🙂
the girls-in-white picture stood out to me and screamed “tiny sensor” (so i figured “smartphone”) because the dof really goes all the way through the frame, including the super-close foreground person.
everything else would have been random guesses for me.
I admit, I was trying to use the tonal map to throw viewers off…extended DOF could have been achieved by say a 1″ sensor though.
I picked two, but it seemed more like random luck than skill. Now, if we were looking at large prints in person, I do not know if I would be able to identify the cameras (and film emulsion where applicable), but I suspect that my preferences would be a bit more narrowed from among the list of equipment in the images above (and no not necessarily towards any particular brand or format). But, you are correct, the skill and post processing techniques of the photographer have a much greater impact on the final product (print or screen) than the camera/film selected.
Up until two years ago, I had been shooting with Nikon digital bodies exclusively for about six years, mostly because of my lenses rather than brand loyalty. I remember seeing an exhibit one day, shot with some digital Canon body, and the consistent style and “look” to all of the images really caught my eye. I was never able to get hold of the photographer to find out more about how they achieved what they did, but it still haunts me, not unlike your images often due (regardless of equipment). I tend to process each image as I see fit, and I am sure my tastes are somewhat similar from image to image, but I have not really focused my output to have a definitive signature. I applaud those who can do it well, because when it is done poorly, it is not unlike adding the same hot sauce or spice to every dish in exactly the same manner.
Thanks for the challenge,
–Ken
You’d have to actually go fairly large (assuming we’re not Ultraprinting) to be able to tell.
Consistency: practice, and reworking the images at the same time or immediately concurrent, I suspect…
I’m confused. Image 7 says D800 on Flickr
If you read earlier posts – I think thats how he scans his negatives [note the 60 macro and flash fired]
Film scanned with a D800E. The rest of the exposure information is impossible for the image.
Got it, thanks. Didn’t look at EXIF! Woops
5 for me (correct: 1, 3, 4, 5, 7) but to be honest, I mostly used perspective to guess at the camera. 3 was obvious because I’m pretty used to the GR’s look and perspective. The 645D picture I guessed because I remembered you were doing pictures of that style with it. The Japan picture I guess Hasselblad because you brought that with you. The RX10 I guessed because it looked like a compressed perspective so I went with the one with the longest lens. The film image might have been the only one where I used the intirinsic look of the tones to get it, but it wasn’t super confident. I looked at them on an iPad 3.
But really, at these sizes there is very little to differentiate them.
Actually, that’s pretty well done!
I got 6, 7, and 8. Definitely thought there were at least a couple D800Es in there! Interesting and fun quiz – looking forward to seeing what other people think!
Got five right. I thought #4 was pretty obvious judging by the low light performance. Got #6 right by DOF, #10 by instinct, #1 and #3 by luck. Nice post Ming, I completely agree that nowadays the difference in image quality is so minuscule that sometimes it does;t really matter what camera you use. However I do think that the variety in the handling of film cameras still sets cameras apart. (Mainly the differences between a SRL, a rangefinders, a TLR etc.)
Well done!
EXIF data won’t necessarily work for the scanned film images among them.
Actually, that’d be a dead giveaway because they’ll all be 1/250s f8 60mm ISO 1000 D800E 🙂
Why ISO 1000 for your scanning rig?
100, definitely not 1000.
Interesting idea Ming…number 6… I phone…wow
No kidding!
Indeed – that one fooled me the most, even though there’s a certain flatness to the borders (especially right, left’s better, but still …).
Thanks. Watch the exposure and midtones 😉
It is easier to distinguish good photographer vs bad photographer from the web sized pictures than differentiating equipment.
Loved the challenge. Didn’t take it since it would have been mostly wild guessing.