Comparing a Ten-Year-Old Camera to a Modern One
This month marks ten years for the Canon 5D Mark III – a camera once known as the top of the line, now superseded by a couple of different camera generations. With March of 2022 being its ten-year birthday, I thought I’d do a fun little comparison to see how far we’ve come in these ten years, and if having the top-of-the-line camera is as important as many will want you to believe.
What Has Changed In Ten Years?
A lot can happen in the tech industry in ten years, and the photography market is no exception. The biggest change comes in format – DSLR is a dying breed, with mirrorless platforms taking over for all major brands. But format aside, each iteration of new cameras comes with faster shooting speeds, higher resolutions, and chips promising faster and far more accurate focusing. And while these upgrades are universally appreciated, one must ask – Are they really as important as we think they are?
The Competition
To celebrate the ten-year birthday of the Canon 5D Mark III, I decided to put it toe to toe with one of the top of the line cameras today – my beloved Fujifilm GFX 100s. The Fujifilm GFX 100s is a better camera on all accounts – increasing resolution significantly from 22.3mp to 102mp, supplying a much larger sensor, adding built-in image stabilization, and having nearly 7 times as many focus points. And that is just touching on the basics, the Fujifilm has a number of additional features that put it in a class well above what Canon could offer ten years ago – but do those things really add quality to your work or just some conveniences?
To test that, I’ve taken five different photos with the Canon 5d Mark III, and Fujifilm GFX 100s to see if you can pick out which one is which. Sure, there are a million variables added to the equation – from professional lighting to retouching, but the point still stands – Can you really tell the difference between the state of the art camera, to a ten-year-old system?
The Test
Each set of photos was taken on a Canon 5d Mark III with a Canon 100mm f/2.8L Macro IS, as well as a Fujifilm GFX 100s with a Fuji GF 120mm f/4 IS Macro. The images were toned and edited in Capture One Pro and Adobe Photoshop before scaling for the web. I will present all the images below, with the option to click them to display them on a larger scale, followed by a quiz to see if you can guess which one is which.





The Quiz
Differences I Noticed
The biggest difference I noticed from new cameras to old ones came with the contrast in the RAW files. The Canon 5D Mark III comes with considerably more contrast and saturation in the files by default when compared to more modern cameras – particularly in studio lighting. This is kind of expected; if you understand the workflow of RAW video processing, you’ll know that you want to work with a flat image prior to adding color grading either manually or through the use of LUTs. The same applies to photography workflows when shooting with RAWs. Starting with a flat image, you can then go into your RAW image processing software and add back your contrast, saturation, and luminosity – allowing you to get the exact look you’re going for. While this process takes more time, it does generally produce better results, and as sensor technology grew, as did the post-production processing capabilities and workflow.
The second thing I noticed between the two images came with just general sharpness and resolution. This is to be expected, as we’re comparing a 35mm 22-megapixel sensor to a considerably larger sensor offering 102-megapixels. But all in all, I found that the Fuji had better sharpness as a whole. This can be because of a myriad of reasons – from better optics to image stabilization, to a variety of other variables – but it is something I noticed throughout all the images – regardless of whether they were in the studio or on location. And while I absolutely love the 102-megapixel resolution, for the ability to crop in my images and print at a large scale, it does beg the question of how important this resolution is when 99% of photography is displayed on web formats only. Instagram downscales your images to 1080px (long edge), Facebook downscales to 1200px long edge, and most websites will function best if the resolution doesn’t exceed 2500px long edge. All of these are considerably smaller than what both of these sensors produce (Canon 5d Mark III defaults to 5760px long edge at 300dpi, whereas the Fuji GFX 100s defaults to 11648px long edge at 300dpi), so each camera is comfortably over the maximums of most social media platforms.
Finally, the last thing I noticed was that the Fuji GFX 100s was about 1/3rd of a stop underexposed when compared to the Canon 5D Mark III. Now this is likely just some variance in the lenses, but something worth noting nonetheless. I won’t dig too far into it, as it’s complicated, but f-stops by design are theoretical (whereas t-stops are a physically measured transmission). So this differing exposure isn’t a fault on the Fuji GFX 100s or Canon 5d Mark III systems, just a variance that I made note of prior to bumping exposure up a tad in post-production.
So Do You Need the State of the Art Camera?
While counterintuitive to a blog that promotes new gear available for rentals, you probably don’t need the latest and greatest camera to create your art. While a little slower to use, and without all the bells and whistles, I found the Canon 5D Mark III to still be an incredibly capable camera – especially if the work is going to be displayed on web format and not printed. Given that 99% of all photography will probably never hit the printers anyway, it doesn’t seem like it’s entirely necessary to have the highest resolving sensor available.
However, this really comes down to the law of diminishing return – which states that once an optimal level is reached, efficiency will reduce over time. Or in laymen’s terms, once all of your necessary needs are met from something, those above and beyond moments become harder to reach. A ten-year-old flagship camera still checks all the boxes a professional photographer needs, and while it may not be as flashy and nice as a modern camera, it still does the intended job. With today’s technology, your camera is most likely not the weakest chain in your workflow and isn’t the limiting factor preventing you to get to the next level with your work.

But how did you score on guessing which camera was which? Were there some of the key indicators of the Canon 5D Mark III versus the Fujifilm GFX 100s? Chime in in the comments below.
84 Comments
Andreas Werle ·
Beautiful idea, Zach!
Ok, i was wrong only with the motorcycle. But honestly, i was guessing, that the more intense colours point to the Canon. This was the only idea i had to try to distinguish the photos. As a digital Canon-Shooter since many years, i love the Canon-Colours. 🙂
Greetings from Frankfurt am Main and stay healthy! – Andreas
Zach Sutton Photography ·
Thanks for playing along!
I love Canon’s colors out of the box as well. But there is so much more depth in the Fuji photos once you start really toning them in the post portion. Infact, I had to keep notes with me while I was going through the photos, as some of them were really difficult to distinguish one or the other from. I figured if nothing else, this was a good thought experiment on looking at what your camera is capable of, even if it’s an older system.
macroblue ·
I was wrong on four out of five, lol. I was considering upgrading from my trusty D700 but I guess not! Maybe I should buy a better monitor instead.
Zach Sutton Photography ·
Hahah. If it makes you feel better, as of typing this, 39 people have taken the quiz, and no one has scored 100%. (In fact, the average is nearly 50%).
I always tell people to find the weakest link in your workflow chain and upgrade that first. For some people it’s lenses, some it’s lighting, for others it’s monitors. But it’s almost never the camera – especially if you’re using something from the last few years. I upgraded to the GFX 100s because the resolution made my workflow easier, and I just wanted the dopamine hit of having a new shiny thing….but my upgrade was in no way a requirement. ?
Joshua Richardson ·
I didn’t know you joined the GFX family! The 50s was an early pandemic purchase for me. What a camera.
Zach Sutton Photography ·
Yeah, I was able to get my hands on a GFX100s about a year ago.
gedgetdisqusser ·
Can not take anyone serious who speaks about the “pandemic” anyway…. *rolleyes*
Drew Rick ·
That’s unsurprising when making image choices that don’t push the hardware. The latte is the only image that doesn’t have crushed shadows and every one of these images has blown highlights, as shown by the “image histogram” browser extension. You could probably use a 1 inch camera and reproduce the same thing. Especially in the studio it’s quite easy to reduce DR, even though that wasn’t done here.
I would have preferred to see examples of what has actually changed over the last ten years. Most importantly, it is much easier to obtain accurate AF as well as manual focus results with mirrorless cameras (without dropping back to the slow as molasses contrast AF in the 5D’s live view mode – might as well buy a cheapo mirrorless without a viewfinder if that’s how you use it). Magic lantern is the saving grace of the 5D here, as it adds focus peaking which the 5D is missing.
Ivo Hula ·
Knowing the types of photography you do, Zach, I see why you made the comparison that you did make. If you shot wildlife and sports, I think your “old” camera would be blown away by the new autofocus capabilities of the more recent offerings.
Recently I had to make the hard choice to unload my 1DS MKIII and my 5D MKIV as the keeper rate of acceptable focus with the old vs the new R3 was huge in my hands. I have been shooting Canon since the late 1970’s. Shooting speed skating and pet portraiture as my two main income streams, I saw the writing on the wall. The autofocus on the R3 is phenomenal compared to a 6 year old camera (5d MkIV) and a 14 year old pro body (1DS MkIII).
All of the above is pointless if you miss the focus on the eyes.
Zach Sutton Photography ·
100%.
I wanted to do a test of faster moving objects, but autofocus is such a crapshoot to test and then put to words…so I resorted to taking photos of my motorcycle sitting in my driveway instead. One of the big advances over the last few years has came in the autofocus functionality, and a big reason to upgrade if you are shooting sports, wildlife, or anything else that involves fast moving subjects
Joshua Richardson ·
Got 4/5, but the colors are by far an easier tell with the resolution of online. I’m also lucky to have owned both systems. The colors on the GFX sensors are unequaled (from a lifelong Canon guy).
Zach Sutton Photography ·
I agree. The Canon definitely is richer in tones out of the camera, but you just have so much more control with the Fuji.
earl.dieta ·
got 40% lol and personally for my shooting, newer sensor and AF technology doesnt matter much. Even my EOS R, I’m barely using its full potential and don’t see the need to upgrade, infact I downgraded and got me a 1DS Mark III instead as I’ve come to the conclusion that I enjoy shooting more through an OVF than EVF.
Zach Sutton Photography ·
As of writing this, 235 people have taken the quiz, and the results are broken down as follows —
100% – 4 People
80% – 35 People
60% – 77 People
40% – 56 People
20% – 52 People
0% – 11 People
Mohamed Boualam ·
Me 100% 🙂
Ian Allen ·
the one I missed was the last one as I couldn’t zoom in and then just went off of colors. TBH, I was looking more at the focal length and DOF than the quality as it was hard to tell the difference at this size. a really good exercise to help not fall into the gear chasing syndrome!
Ian Allen ·
I missed the last one as I couldn’t zoom in and then just went off of colors. TBH, I was looking more at the focal length and DOF than the quality as it was hard to tell the difference at this size. a really good exercise to help not fall into the gear chasing syndrome!
Drew Rick ·
Sadly, that statistic is meaningless due to your image choices.
filmfilmfilm ·
I’ve owned a 5D “classic”, a Mark II, and now an EOS R, and honestly in good light with a good exposure on the same lens there’s no huge difference between the 3, especially at web sizes. Similar to the MK3 in the test the mark 2 is more contrasty than the other 2 and has an issue with oversaturating reds. The biggest things I’ve seen with the EOS R is generally much more accurate autofocus w fast lenses, much less noise at higher ISOs, and noticeably more dynamic range and ability to retain highlights.
Chris Jankowski ·
Looking at the development of high end cameras in the past 10 years a few outcomes seem to be pretty clear:
1.
The progress in sensors characteristics is very slow – sensitivity, noise, etc. This is not all that surprising, as modern sensors already capture more than half of the light falling on the sensor. This puts a hard physical limit on the possible improvement.
2.
On the other hand, there are numerous improvements that are at their core driven by the enormous increase in the processing power built into the new cameras. Several of these improvements changed certain aspects of picture taking not only quantitatively, but qualitatively for certain photographic situations. Several areas stand out:
2.1
PDAF AF on the sensor with large number of points. With lots of processing power and fast actuators in the lenses this brings amazingly fast and precise autofocus for moving subjects. Eye AF is also extremely good and very useful in many applications.
2.2
Sensor based image stabilisation and improved in-lens stabilisation. In bad light conditions this often makes it possible to take acceptable hand-held photos where previously it was just not possible.
2.3
Increased continuous shooting rate with AF maintained between pictures. Invaluable for sports and similar situations.
2.4
Outside pure still photography, the new cameras gained amazing video capabilities.
These improvements are probably qualitatively and quantitatively most significant.
So the outcome is as follows, I think:
1.
If you do landscapes or portraits in a well lit studio and publish on the web you can happily live with you 5D Mk III. A new camera may improve your workflow slightly, but not that much. You can gain more from replacing your old computer.
2.
If you do sports, especially indoor sports you want R3 or A1 now.
And of course there is a whole spectrum of situations in between.
—————
Gazing into the future the trend for increased processing power available to be built into cameras will yield more changes. Other than general quantitative improvements and unavoidable marketing gimmicks, I would expect the following:
1. Global electronic shutters eliminating mechanical shutters entirely.
2. Automatic macro frame stacking for increased depth of field.
3. Automatic frame stacking for astrophotography utilising stabilisation system for less noise and better sharpness.
4. A much better integration with the internet resources e.g. cloud storage. I am still amazed that this has not happened yet. I blame the inherent conservatism of the industry for it.
Achim Schäfer ·
I fully agree with you!
The evolution in sensor IQ is coming to an end now due to physical reasons. I did a lot of concert photos for example with a Sony A65 and A77 II and due to knowledge and post processing (RAW!) they were more than sufficient even in really bad lighting situations. The A7 III and the A1 didn’t bring me here really more – but the improvements in e.g. AF did bring a lot!
But what I still miss is more improvements in UI/UX and the interfaces. For example the WiFi integration is still a mess. Or why did they eliminate GPS? I found it very comfortable to have correct time in the EXIF data without setting it manually. Or 5G (cell phone) integration? Here in Switzerland I can get a SIM card with unlimited data volume for 44 CHF/month (should be around the same in US $) so the cam could send my photos taken on a venue or e.g. demonstration nearly in real time to my office/studio/agency. Cheap smartphones can do this why not professional cameras?
Or usable tethering: should be no problem to incorporate that – why isn’t it there in all models?
I see the same problems of usability in a different but somehow related business: modern TVs – they are crowded with all gimmicks but to use them is just a mess even for an IT professional…
Chik Sum ·
I owned the 5D mk III since it’s launch date and never changed camera ever since, originally having the 16-35 F2.8L II and 70-200 F4L IS before the camera, added the 24-70 F2.8 L II, and a sigma 14mm F1.8 Art and it’s all I would ever need.
Yes reading the release reviews of the subsequent 5D4 and eventually the R5 and better cameras are exciting, but then the diminishing return law kicks hard.
While I don’t even need to think twice when upgrading the 40D to 5D3 which cost like 60% of my monthly income back then where both the speed, OVF quality and most importantly, hand held night time snap shot capable/usable ISO to 6400 or even 12800 with heavy NR and resize, the newer cameras though having ever better DR and ISO, isn’t that night and day anymore, no more the zing “man I have to get these!!!”
YS ·
I owned the 5D mk III since it's launch date and never changed camera ever since, originally having the 16-35 F2.8L II and 70-200 F4L IS before the camera, added the 24-70 F2.8 L II, and a sigma 14mm F1.8 Art and it's all I would ever need.
Yes reading the release reviews of the subsequent 5D4 and eventually the R5 and better cameras are exciting, but then the diminishing return law kicks hard.
While I don't even need to think twice when upgrading the 40D to 5D3 which cost like 60% of my monthly income back then where both the speed, OVF quality and most importantly, hand held night time snap shot capable/usable ISO to 6400 or even 12800 with heavy NR and resize, the newer cameras though having ever better DR and ISO, isn't that night and day anymore, no more the zing "man I have to get these!!!"
Supreme Dalek ·
Of COURSE it’s hard to tell the difference when you’re comparing images on a web page. You could have done the same demonstration with a 6-megapixel camera from 20 years ago and with similar poll results. Where the difference comes in is when your client says, “Hey, we love that feature image you did for our website so much that we want to use it as at 96×120 inches on our trade show booth. Can you send me a file that will work for that?” It doesn’t happen often, but you kick yourself when it does.
Zach Sutton Photography ·
With over 1000 people taking the quiz, the average score is 44.7%. Pretty wild.
https://uploads.disquscdn.com/images/798dcec21b562cb18560ceaffbd6792b7d34b4f137821f60250ee5551dbe47e7.png
Kers ·
I did not realize you culd see the photos a little bigger ( except the last one?)
As soon as you do that it is much more easy to see and then still 1200PX
Class A ·
These numbers are completely meaningless as the quiz is useless. I’ve posted my detailed comments in another comment.
Drew Rick ·
Lol, downvoted by the author. I have rarely seen anything more pathetic, and I usually try to avoid hash words.
For the record, cI absolutely agree with Class A.
Jacknife Skiggs ·
Are you claiming the people who participated closed their eyes as they made their guesses? What could possibly disqualify a BLIND TEST?
This whole argument that the GFX was intentionally handicapped to allow the Canon more chance to compare competitively is the only thing I’ve seen here that is pathetic.
Drew Rick ·
No, people had open eyes and couldn’t see the difference because the photographer was incapable of producing one by a series of technical blunders that both Class A and I laid out in comments below.
You didn’t seem to have the patience to scroll down to actually read “detailed comments in another comment” before commenting yourself and making accusations on an 8 month old thread, why is that?
Henry Winokur ·
I took the quiz and failed–what I’d describe–as miserably, LOL. OTOH, I didn’t upgrade my 5d4 to an R5 for the sensor size. I needed better auto-focus and lighter weight. I felt that my 5d4, which I *loved* had gotten too heavy and I could no longer hold it steady at speeds I sometimes shoot at. Also, the autofocus on the 5d4 simply can’t compete with mirrorless. With the IBIS in the R5, I am again able to hand-hold shots at speeds I haven’t been able to get to in the last 10-15 years! For me, it was worth the upgrade.
It’s an interesting article, Zack. But when I read the headline, and then the first paragraph, I thought you were going to compare the 5d3 to a similar camera, perhaps the R5. Was it “fair” to compare a 35mm format to the Fuji, considering the differentiation in formats and therefore file sizes? I, for one, would expect that a camera with a larger and newer sensor would do a better job.
Fritz Rutz ·
One reason for the “sharper” perception of the Fuji might be the missing AA filter, especially compared to the 5D MK3 which seems to have a quite strong AA filter to avoid aliasing in videos. There are some pages which describes that the sharpness noticeable improves when the AA filter is removed.
Jalan ·
Thanks Zach – very interesting comparison. I have a 5D III and IV and the III is still a great camera for what I do (weddings, portraits and some video). The most important practical difference is the better high ISO performance of the IV. Weddings are often very low light so being able to use 12,800 ISO versus 3200 is a big boost. R5 upgrade in the future but don’t see parting with the 5D III as a backup or “B” camera!
rhoel6 ·
I got 4/5 right, and I based it mainly from the larger sensor having the shallower DOF. Not sure if that’s right.
thebitterfig ·
Depends on the lenses. DOF isn’t exactly from sensor size: it’s from focal length, aperture, and distance to the subject. Larger sensors typically make it easier to use longer focal lengths and shorter working distances. There are a lot of variables on maximum aperture, but at least with GFX, most lenses have narrower apertures than Canon’s EF lenses.
The Fuji lens has slightly longer focal length, which helps, and actually has a wider field of view so you’d need to be closer, but it’s a full stop darker, and that claws back the gains. Ought to be about equal. But that’s just the theory.
In perception, the sharpness of in-focus areas will increase the appearance of shallow depth of field. That was one reason old Medium Format film images had the feeling of shallow depth of field. With the same print size, the much higher resolution 120 film over 135 film meant that there was a lot more detail in the in-focus areas. Even though these are all downscaled images, the higher detail of the GFX100s might increase the perception of shallow depth of field.
Fredrick H. ·
Zach, I saw your comparison. It looks like Fujifilm GFX 100s photography is really good. And, I agree with you, “Fujifilm GFX 100s is a better camera on all accounts “.
Richard Walker ·
I opted to just pick the picture I liked better for each pair, rather than attempt to pick which camera took which picture. I opted to say that Canon took the better picture, just to track. I got every one wrong, which tells me that I prefer Fuji’s results across the board. I definitely see an improvement over time, but, like picking out a TV, if I didn’t see the images next to each other, I probably would be totally fine with either.
miao ·
4/5, yay
spider-mario ·
To 1080 in width, but in fact it can be up to 1080×1350.
Nope, 2048.
Zach Sutton Photography ·
Facebook says 1200 on their official website.
https://www.facebook.com/business/help/469767027114079?id=271710926837064
spider-mario ·
You can also easily test it by uploading a 2048×2048 image to Facebook and seeing that it stays that way.
Drew Rick ·
Omg the cluelessness intensifies. You should rename your account to “Alan Smithee”, this can’t be good for your business.
pop ·
I missed the 2nd one.
Rich Seiling ·
If your end goal is making fine art prints, particularly large ones, there is no comparison, the Fuji will win hands down. I have worked many 5D Mk III files for professionals and printed to large sizes, and even at 100% on screen you can see the issues with that sensor. I would not want to be using a ten year old camera from any manufacturer. Major improvements in sensor quality were made around 2014-2018 with the big three companies. Cameras before that really show their age. You only get one chance at most photos, so there is a value in capturing it the best you can. In 2022, that means not using a ten year old Canon body if at all possible.
Zé De Boni ·
The cameras chosen don’t fit well for such comparison. If you put side by side the 5D and the R5, then the gap will be much bigger for 2 top products from the same manufacturer. But as we are talking about generic technology evolution, then we should choose the newest state of the art, the Sony A1.
Your lovely Fuji GFX 100 should have as reference the Hasselblad H4D-60 or the H4d-200MS, or maybe the Phase One IQ 180. Price inclusive.
Zé De Boni ·
Yes, web publishing downgrades the advantages of more capable equipment, but…
But image capture is a fundamental property of those tools, whether you compare just sensors or fully equipped cameras. The choice of hundreds of active focus points, subject tracking, eye-focus, life-view, 30fps, silent shutter, makes a huge difference in the resulting image. Resolution-wise, the posted image may look similar, but the content (what really matters), and the photographer’s success are clearly boosted by all the technological improvements.
The title of the original article is: “Comparing a Ten-Year-Old Camera to a Modern One”
The foreword proposes a comparison between cameras, not just sensors from 10 years range. Some of the MF from 2012 still used CCD’s and cost $40K, so that is what should be compared with the GFX. On the FF side, BSI stacked sensor with silent shutter, 8K capable with IBIS, higher resolution finders, faster processing chips, rule today, 2022. To leave the advantages of such features out of context of any comparison is a biased short sighted failed attempt.
Ash ·
Adding a Fuji APS-C to the comparison will be also interesting as I guess both Fujis would have similar color processing, esp skin tones ?
Class A ·
The quiz makes very little sense. The framing and subject difference is often different. The subject sometimes is moved, causing different lighting. You really could have taken a picture of an apple with one camera and an orange with the other and then asked us to have a guess. You don’t tell us the f-ratios used. Did you use equivalent settings? Your focal lengths are not equivalent to start with. You tell us in the text that the Canon images are more saturated, yet in the (different!) processing of the image pairs, the more saturated shot is almost always from the Fuji.
As long as no conclusions are drawn from this, it’s all just a waste of time, but I’m afraid someone will conclude from this that viewers cannot tell the difference.
Finally, given that the Fuji only has a “baby-MF” (crop!) sensor, it is not “much larger”. There is just a factor of 1.3 between FF and the baby-MF format, that’s less than the difference between APS-C and FF.
Zach Sutton Photography ·
Hey Class A,
There few errors on your end, along with some pretty brash misconceptions. First, yes, all of the photos were taken with the same settings, and I figured that was assumed based on the (nearly) equal exposure throughout the text images. As for the focal lengths, they are incredibly similar, as the 120mm attached to the GFX system would be about 96mm on a 35mm equivalent. To put it simply, I don’t believe there is any lenses that meet a 1:1 ration on focal lengths, and this comparison has (using 35mm sensors as the basis) a 4mm variance between the two focal lengths. As for saturation, I specifically say that the saturation was way more noticeable…in the studio lighting shots, which this test one has one of.
As for the Fuji sensor compared to the 35mm sensor on the Canon, it’s actually 1.7 times larger than the Canon. It seems you’ve done the math wrong. The Fuji has a sensor measured at 43.8 x 32.9mm, whereas the Canon has a sensor at 36×24mm in size.
Finally, no one is suggesting that this is a test encouraging people to conclude that a 10 year old sensor is as good as a modern one, and I make mention of that several times throughout the piece. It seems that you walked into this article with a frustrated frame of mind, when it was intended on being lighthearted and a fun exercise.
Class A ·
Hi Zach,
I did not walk into this article with “a frustrated frame of mind”. I became frustrated when I saw what could have been an interesting exercise in determining how relevant certain IQ differences are for a certain output size and whether there are any discernable rendering differences, turned out to be entirely useless for that purpose.
I did not do “the math wrong” either. I wrote about a “factor” and meant the “crop factor” that is commonly used to compare sensor sizes. The crop factor (~1.3) is the square root of your 1.7 ratio of the sensor sizes. Just like the crop factor between (Canon) APS-C and FF is not 2.56 (sensor size ratio) but 1.6.
The focal lengths are similar but they are not exactly equivalent. Doesn’t matter anyhow because you partially used different subject distances which renders small differences between focal lengths moot due to the significant impact on the amount of background blur.
Using the same settings on two different formats is entirely inadequate, as you not only need to convert focal lengths between formats but f-stops as well. Please see the DPReview articles on equivalence or use some other source. The “same exposure” could be achieved by all sorts of settings (-> reciprocity) so a same or similar exposure provides no clues whatsoever about the all important f-stops used.
As for what is a “fun exercise” we have different ideas. Using a reductio ad absurdum technique, I could say an article that asks “Which takes the better portraits, a phone or a GFX 100?” and then use studio lighting for the phone while the GFX 100 gets only moon light through thick curtains to work with, is “fun”. If the conditions (in your exercise, the shooting parameters) are not equivalent, all bets are off. Where is the “fun” in doing something that is pointless? I’m not saying you cannot write a fun article about the old Canon vs the new Fuji, but an apples to oranges quiz is not what I call “fun”.
Finally, you must know that some, if not most, people will not view your numbers just as “light-hearted fun” but will claim that some kind of conclusions can be drawn from them. No conclusions can be drawn, so what is the point of showing them?
Zé De Boni ·
@Class A – BRAVO!
Jacknife Skiggs ·
But how does any of this specific technical talk disqualify the fact that comparing the images on the average computer monitor even a trained eye cannot discern the difference accurately? I think that clearly is “the point of showing” here.
Also this point you made “studio lighting for the phone, none for the GFX – so its unfair” has no grounds whatsoever, the first female model shot clearly has very comparable lighting in both shots, as do the rest of them honestly. This initial shot is a studio setting as well with professional controlled lighting – I would think if the difference is not noticeable here that is a very interesting result.
Bottom line, on 99% of consumer systems the difference (in these instances) is not discernable even by professionals, I don’t see how that can’t be taken at face value when a blind test literally gets at the very bottom line (that’s the point of a blind test). Perhaps if the article was called “You can’t tell the difference between a 10 year old canon and 7,000$ Fuji on your regular computer” you wouldn’t be objecting as much?
Drew Rick ·
Yet you keep triumphantly posting stats from your survey (that you’ve designed to yield random results) that you were in fact successful in making the survey yield random results.
grubernd ·
Thanks for the challenge.
Of the three I got wrong (both portraits and the drink) I was going by the colors and the reasoning that the GFX should have the better ones. But it shows your edits of the Canon files shine from your experience and the ones from the GFX are bland with hot hightlights and aweful gradients.
Thanks for proofing once again that the editor is more important than the camera or the software.
Matthew Willis ·
I got 4/5 but the one I got wrong was the only one I felt confident about – lol.
Mike Earussi ·
I got three out of five but I was looking primarily at contrast and color, not resolution as none of them seemed very sharp.
AlexisZ ·
I scored 100% (in shock about it), but I think it’s because I read your article pretty closely!
JB ·
I scored 3/5. The 3 I got correct I totally guessed on. The 2 I got wrong I was totally sure I had right. My a7R IV will be in my trash can if anyone wants it. I’ll be out with my Nikon D100.
ProfHankD ·
I am shocked to see that I scored 80%. At first glance, it seemed like I would just be guessing, and I honestly expected I’d score lower given these are low-resolution and significantly postprocessed. However, there were somewhat consistent color and tone differences that made things pretty clear except for the last shot, even though I have never directly used either of these cameras and obviously didn’t personally see the original scenes. At least on my (not well calibrated) display, both images in last case look a little off, so that was a guess I happened to get wrong. My take-away on this is actually that color science advancements have been more significant than I thought… although I’m still pretty happy with the technical quality of images I shot with a Sony A100.
Class A ·
According to the author, he used the same settings on both cameras. So if the subject distances, lighting, etc. had been controlled, it would have been easy to pick out the larger format camera every time. The fact that not nearly everyone got a 100% right, given the same settings were used on both cameras, shows how disruptive the variabilities allowed were for any sort of significant result.
ProfHankD ·
In 2012, I created a computational repair for the rather blatant Fuji X10 "white orbs" image defect. The slides for my Electronic Imaging presentation on it are at http://aggregate.org/DIT/SP... . The interesting thing was that I had people using the online tool to repair their images rank the repair quality. While 22% gave 100% quality for the repair, 40% of scores were below 50%! Why? It turns out that the majority of low scores came from people who submitted images that did NOT actually contain any white orbs, and my DeOrbIt tool correctly did nothing to those images. So much for humans as image quality judges.... ;-)
Fred Mueller ·
100%. Not hard at all if you just look for the better gradation in the roll off to highlight clip. This has nothing to do with resolution or lens attributes. It’s just the better tonal response of the Fuji sensor. I did the test on my iPhone 11. If you were doing high end portraiture or fashion the Canon would be a mediocre choice. The final frontier in digital imaging is dynamic response in both capture and display response. There is yet a long way to go.
Bobby Joe ·
I shot with 5dmk3 for a long time. Taking shots at even f2 with 85L was pain. With GFX50s/110mm f2 combo I could nail the same shots at 1/125 while with 5dmk3 for whatever reason I needed 1/200 or faster. Now with GFX100s there is no comparison. If you show small size prints like in the examples posted, even old 5dc will look the same, why even use 5dmk3.
OLDFASHIONED ·
the DOF is VERY different. I know, I own a GFX100 (not s) and can’t use it for products because of that. and with practically any lens. but the resolution difference is out of the question.
Frank H. ·
Got four out of 5 right. Key things I looked for were how the “shoulder” to highlight compared, and what the transition from one plane to the next one immediately behind looked like.
The GFX photos seem to be smoother in both cases, but I would never say it was enough to force the Canon out of use.
Frank ·
Got four out of 5 right. Key things I looked for were how the "shoulder" to highlight compared, and what the transition from one plane to the next one immediately behind looked like.
The GFX photos seem to be smoother in both cases, but I would never say it was enough to force the Canon out of use.
Chris ·
I scored 50%… Well actually the result says 40% but the last picture was a pure guess since the zoom function did not work and therefore I didn’t bother to check for details.
However, there was not much difference in detail – I mainly went for ths Bokeh as I guessed the larger sensor will have more pronounced unsharpness despite the f/4 lens.
Regarding the visible detail and main look and feel of the final pics, I could have flipped a coin, too… wouldn’t have made much of a difference… thank you for that comparison.
On my Mk4 I mainly prefer the – visible – difference in High ISO, the GPS function and the touch screen (you definitly get used to it!!) over my 5D3 – that’s why I switched. No need for the 30MP
Jonathan Laufersweiler ·
4/5 right here. For me, the tell was the colors and tone, probably corresponding to the exposure difference he mentioned. Apropos, I usually set my Canon to -1/3 as well.
Spaturnio ·
I was right on the first 4 shots and only missed the last one.
The field of view gives away the lens (and subsequently the camera) with which the shots were taken.
All the pictures from the Fuji show a slightly less distorted face aspect, but I wouldn’t be able to tell if not side by side and not knowing I only had two option to chose from.
nadir carollo ·
Still using the mark 3, I have done a wedding Saturday, still hold perfectly for my use
Drew Rick ·
Completely static subjects – I’d be curious about the procedure. Was focus accuracy improved on the Canon by using live view? DR challenges were avoided with compositions that have no detail in the shadows and/or allowing highlights to blow out (specular in studio, shirt, flower).
Drew Rick ·
Author doesn’t know Fuji uses a different ISO scale? Oh boy. Just use the histogram and ETTR.
Florent ·
I fully agree, who really needs such high resolutions? Very few photographers, really.
Also, I’d argue the same about sensor size: unless one shoots at very high ISO or needs super extra thin DOF, APS-C or even MFT sensors are more than good enough for most tasks.
BTW, I just scored 80% (2nd one guessed wrong) but I think it was probably just more luck than anything else 😉
Daniel South ·
I owned (and loved) the 5D Mark III and its predecessor, the Mark II. I traveled with them extensively, and to this day, the images that I captured with those cameras are among my favorites. I had two minor complaints, both related to dynamic range. First, it’s easy to blow out the red channel as when photographing a red flower. Second, shadow noise was a common problem. If I photographed someone in bright sunlight, for example, there was a lot of noise in the shadows under their arms, for instance. Otherwise, both bodies were excellent.
That said, the experience seems so long ago. I’ve used mirrorless cameras for the past five years. I remember how challenging it was to focus in darkness through an optical viewfinder, using mirror lockup, using depth of field preview to position ND grad filters, etc. Mirrorless photography is a different experience.
I have been using the GFX 100S since it was first released a year ago. Working with is is quite different from the 5D series not just because it has all of the innovations of mirrorless cameras, but because medium format is a different experience altogether. I wouldn’t use the GFX for sports or events; it’s more of a niche camera, whereas the 5D series was a jack of all trades. That said, the first thing that I noticed when I took the GFX out of the box was how much the body reminded me of a Canon DSLR. It has a very similar size and feel, right down to the buttons and the texture.
The image quality of the GFX is outstanding, leaps and bounds over even 50 and 60 MP cameras, but as your quiz demonstrates, it’s really not a factor when sharing images on social media. But open a raw file from a GFX against any ten-year-old camera – even a Phase One – and the difference is huge. The GFX is inspiring me to revisit the places where I traveled with my beloved Canons. It’s time to fall in love all over again. 🙂
Mark Rustad ·
4 of 5 correct. I struggled with the last shot Zach as was temped by the eyes on the right for Fuji but when I looked at the fabric crenellations/details plus eyes on the left I thought it had to be the larger sensor res…wrong. Nice work. Great juxtaposition. We can still create great images with older, in spec cameras with great glass–no need to go broke chasing the latestest and purportedly greatest gear!
Michael Henke ·
Didn’t view them at full size, and got all five wrong!
Theodor of Defeat ·
I still own a 5D3 and it is still all the camera I need and then some.
sperho ·
I only got 40% correct. I’m pretty sure online photography doesn’t benefit significantly from a 100 Mp camera. 🙂 I shoot a lot for a large marching band in the Southeast. Not a paid gig, but a few things I’ve learned after posting thousands of pictures online for friends and family consumption: 1) down to 4 Mp, camera resolution doesn’t matter. 2) Sensor noise is almost completely irrelevant, 3) light and composition rule all. Bonus learning: 90% of the general population looking at pictures online do so with their phones (I learned this through a poll a few years ago. I bet that number has only gone up a bit by now). Probably why points 1 and 2 are the top two learnings.
Shai Yammanee ·
I was shooting professionally with the 5Dmk2 for over a decade. I only had to upgrade when the weather sealing died on location during an unexpected tropical storm.
The one thing I found is that the image quality improvements have only been incremental over the years. But, the tools to make my shooting faster, more accurate, and more ergonomic, and lighter, have increased dramatically.
Still….. I wouldn’t say no to any of the MF cameras on the market.
Even the older models ?
Nemo Niemann ·
Zach, you really must be kidding. This is not even a close nor fair and accurate comparison. This is like comparing your results with Fuji Provia 100 shot with a Canon SLR and the same film with a Mamiya 645, or really an RZ67. Essentially, Apples and oranges. For a true comparison of the state of technology, compare the 5D3 to Canon EOS R3 or perhaps a Sony a7 IV. THEN you would truly be comparing the advances in technology. You can’t just compare a small sensor to a large sensor and say this is 1:1, then-and-now. I say this as someone who has shot professionally, commercially since 1976, with almost every Canon film camera from 1980 on, almost every Canon pro digital camera since 2001, as well as owned Hasselblad 6×6 film cameras, Mamiya 645 and 4×5 view cameras. BTW, got 40% viewing on my phone. I probably would have done better on my computer.
Bear Pat ·
Got 3/5 right for the test, just trying to guess by the colour science. Honestly in such small thumbnails and composition, the sensor size and resolution doesn’t matter, 100MP is pure overkill.
Usagi ·
Got 3/5 right for the test, just trying to guess by the colour science. Honestly in such small thumbnails and composition, the sensor size and resolution doesn't matter, 100MP is pure overkill.
Comparing a ten-year old camera may be much more meaningful if you compare it under the same sensor size and Company like 5dIII vs R5 or 7dii vs. R7, against the RAW file's latitude, High ISO noise performance, the camera bodies' AF speed and accuracy under continuous burst, etc.
Muresan Mihai Daniel ·
80%, i got motorcycle wrong
paulraphael ·
Dude. You keep writing things like “Starting with a flat image….” There is no such thing. Every raw interpretation is an interpretation. What you’re calling “flat” is just a bunch of arbitrary settings that someone chose as a default based on some preference. The way to go “flat” or neutral is to create a custom profile for each camera under each lighting situation. If you’d done that, these pictures would not be distinguishable in this format.
Likewise your observation of the Fuji underexposing by 1/3 stop. No, it isn’t. The default raw settings you’re using are simply giving an interpretation that’s a 1/3 stop darker than … what you happen to expect.
These differences are meaningless. The real differences are in resolution, noise, actual color bit depth, and actual raw dynamic range. None of which will be visible in a picture shot under reasonable lighting and reduced to tiny dimensions for a blog post. This comparison is meaningless.