Neat list, but What exactly does "Most Commonly Used Camera-Lens Combinations" mean in this context? I find it highly suspect the the most common camera/lens combo is the $3000 US EOS 5D Mk III kit; I'd expect the more affordable D3200/EOS t5 camera range to be more commonly used. Some more information about the methodology used would be helpful.
I had the same thought. Unless I missed it, the site doesn't give the source of the data. I find it very suspect (as in not representative of broader use) that the top two cameras are late model Canon prosumer+ full-frame bodies.
[ADDED: It actually appears that this does align with Flickr stats--which is, of course, not what you'd find on Facebook or other more mass market sites.]
All I could find is "11,495,995 photos from year 2016 which have been posted on an assortment of websites, including Flickr, 500px and Pixabay, with ExifTool technology."
If you look at the source for the images it's public repo's for images (500px.com, flickr...). They usually publish the exif which contains this data.
My guess is that it's based on frequency. What is unknown is whether the sample takes more prolific photographers into account when calculating "common". 5000 people may all own same camera and take single public image as oppose to single photographer who takes 5000 images.
Very cool but would love more details on how they are defining their criteria.
I would assume it's based on the number of uploaded photos taken with those cameras. I'm not sure the Flickr API, for example, even gives you a good way to look at it per user. (You'd basically need to iterate over users and only count a camera model once per user.)
And that they're more inclined to post to sites that tend to be associated with photo enthusiasts.
It does in fact appear that the Canon 5D iii and the 6D are the two most common DSLRs on Flickr based on the number of photos posted. https://www.flickr.com/cameras/canon/
(The leading smartphones are still more popular in terms of total numbers.)
> And that they're more inclined to post to sites that tend to be associated with photo enthusiasts.
This. Of my friends, most people seem to have a bottom- or mid-range APS-size sensor DSLR from Canon or Nikon. But the two of my friends who actually have a 5D iii and a 6D, respectively, are the only ones who a) shoot photos that are interesting to people outside their immediate friends/family and who b) post those photos on flickr.
(While I'm the geek with the superzoom, or the '72 Nikon Nikkormat loaded with Tri-X when I'm feeling analog.)
Probably photos published on "enthusiast" sites where camera/lens info is readily available, as opposed to stock photo agencies or major commercial publishers.
The number of photos taken using phones would no doubt dwarf the data used here, dedicated camera/lens combos are becoming something of a niche market, albeit still a sizeable one.
I use the Olympus EM-1, way down on the list. Nonetheless I found it interesting that the most commonly cited lenses used were the 12-40mm f/2.8, and the 40-150mm f/2.8. These are both excellent quality lenses, but relatively massive and quite expensive, at about $1000 and $1500 respectively.
Saying goes the merit of a photograph depends much more on the person using the camera than the camera being used.
Also note the near total absence of pro gear - no Nikon D3/D4 entries, for instance, still I would expect those cameras _on average_ sees a lot more use than your average D3100, for instance.
(Also, chances are they'd skew the lens numbers significantly - while kit lenses come and go, pro glass tend to live much longer before being upgraded. (The 24-70mm f/2.8 which you often see around single-digit Nikon bodies, for instance, was the standard zoom for 8-9 years before being given an upgrade last year)
Interesting look at some public data but ultimately not all that useful. My dad has done photography for a couple decades and he really swaps lenses for different tasks. The is no single perfect lens.
I think that's what you can deduce about the data as well. I didn't look at Canon figures as I have Nikon myself, but that data clearly showed that entry-level cameras are used with kit lenses, and pro cameras are used with a larger variety of high-quality lenses.
D3100 and D5000 go with the 18-55 kit lens (which is in fact a very good lens considering the price) and have a third of their pictures taken with these. D7000 and D90, being somewhat more advanced hobbyist models, go with their respective kit lens (18-105) with 20-30 % share.
Full-frame bodies (D600, D700, D800 etc) most often have the professional 24-70/2.8 lens, but the share is still smaller, indicating that a larger variety of different lenses is in use, as you would expect with pro/prosumer cameras.
> that data clearly showed that entry-level cameras are used with kit lenses, and pro cameras are used with a larger variety of high-quality lenses.
i.e. perhaps what's more interesting, is the variety of lenses used by a particular photographer. If it's mostly one camera and one lens, we're pretty sure that the photographer has a package, and is just taking simple pictures.
If the photographer has a range of lenses for different shots, we're pretty sure he's choosing the best lenses for the job. Which means concentrating on the distribution per photographer, not per lens.
The page should also show the distribution of combinations among photographers. Most photographers will be at 1-2 lenses. The better photographers will use more lenses.
In all likelihood, the less popular camera + lens combination is the better one.
The blog post about the infographic says that it's a crawl of Flickr, 500px and Pixabay. [1]
If you drill through to the camera/lens combination you actually get to see some of the photos that have been crawled, and then it will take you back to the original source.
That was true for a long time but some sort of medium wide to medium tele zoom has been far more common for ages. I'd be pretty sure that the percentage of people who even own a 50mm prime is pretty low these says.
With the wide assortment of imaging sensor sizes, the "normal" focal length varies accordingly. Probably better to think of it as "angle of view", typically ~45-50 degrees for a normal lens. That's usually the easiest lens to design, so typically the most compact and inexpensive lenses are "normal" primes.
Today the most popular "normals" are moderate, mid-range zooms, equivalent to 28-80mm or so on a 35mm equiv sensor.
For some SLR systems 50 mm (or more often 51.6 mm) isn't actually all that easy (in the 60s), eg. Nikon with a comparatively large flange back distance. We still see some artifacts of this today, e.g. the classic short reproduction macro was always around 55-60 mm, because it's a spot where the lens design is easier.
Macro lenses were more useful with a longer focal length, in 35mm terms, 80-100mm or more longer was desirable to give more working distance to the subject at high ratio, like 1:1. For example, for my MFT system, the macro is 60mm equiv to 120mm for a 35mm frame.
As I understand, lens design in the digital era has changed due to the characteristics of the sensors vs. film. Mirrorless designs have also changed requirements and eliminates some restrictions due to flange distance. I'm not an expert in that field, maybe you know more in that regard.
For most macro uses yes, the longer focal lenses where and are preferred due to working distance.
The shorter macros where developed specifically for reproduction use (photographing records, books, paintings and the like), where the larger reproduction ratios (1:2, 1:1) where rarely if ever needed, so the problem didn't arise.
The shorter flange back distance in mirrorless systems makes some things easier and some things harder (if you want short lenses that is), but overall give more freedom to the lens designer.
Sensor vs film and manual focus / first AF systems vs modern AF systems is a huge difference and for many older lenses esp. in the wide angle spectrum that is the reason for a large performance difference between film and digital.
Even the number "50mm" is massively debated. The 51.6mm number is drawn from the focal length of classic Leica normal lenses, which is 100% arbitrary. There's pretty strong arguments that something more like 43mm is the normal focal length for 35mm.
And designing lenses with adequate coverage for 35mm as well as a sufficient flange distance to clear a SLR mirror was definitely a problem back in the day. Older "normal" lenses for SLRs used to commonly be 55mm or even 58mm. Wide angle-lenses suffered a similar problem, and for a great while rangefinder lenses (or similar lenses that protruded into the mirror cavity of a SLR) were the dominant options for wide angle photography.
It took quite a while to develop good "retrofocal" lenses where the optical center is actually behind the lens (basically the opposite of the telephoto lens where the optical center is in front of the lens). The very first examples of the type are the Angénieux retrofocus and the Voigtländer Prominent's Skoparon, both introduced in 1950. The Skoparon design was essentially a Tessar lens backwards and went on to be one of the earliest wide-angle lenses used in SLRs, notably as the Asahi Pentax 35mm f/3.5. It had an extremely long lifespan, in its medium-format variant (Pentax 75mm f/4.5) it continued to be produced until around 2000. The funny thing is it all happened because Voigtländer patented it for rangefinders but neglected to patent it for SLRs!
The other thing to remember about 50mm-ish macro lenses is that it's still somewhat reasonable to build a helicoid that can get out to 1:1 reproduction without extension tubes. I have a 50mm and a 90mm that will both go to 1:1 reproduction on their own, and I can tell you that while the 50mm may be somewhat ungainly the 90mm is absolutely absurd when you crank it all the way out. The 50mm uses a double helicoid, the 90mm is actually four helicoids running inside each other.
Yes, the old unit focusing macro designs can look rather weird and somewhat unstable when fully "racked out". With internal focusing this has been "fixed", although they still need a lot of displacement of the focusing group (in the Nikon 200/4 AF for example it's about five cm if I remember right, which is a lot for an IF design). Most macros still have a lot of external movement, but internally floating elements reduce the required displacement still. (E.g. Tokina 100/2.8, about 6 cm displacement of the main groups, since helicoil - a cam our two for tiny adjustments of two other groups.
While "angle of view" is the technically correct way to think of this, very few people actually think that way, myself included. It's pretty much limited to people who work heavily with large format and use multiple formats.
"35mm equivalent" focal lengths and "crop factors" are by far the most common ways to mentally visualize this.
It's not entirely a direct conversion since not every format is a 3:2 ratio like 35mm is - for example 6x7 feels tighter because it's much closer to a square format. My 55mm lens is technically closer to a 22-mm equivalent in diagonal FoV but it feels like a 28mm does on 35mm.
I think it's honestly a shame that primes have died out. Dollar for dollar, most primes are massively better performers than zooms in terms of sharpness and aperture.
The top lens for the top camera (Canon 5D Mark III with Canon EF 24-105mm f/4L IS) is sold in a packaged kit with the body.
This lens is...not very good.
It's only popular because they're bundled.
I think this site is useful, however. It gives you a very quick way to navigate to a whole collection of photos of any body/lens combination. If you're in the market for a body/lens, this seems like a good source to go to. No side-by-side comparisons like an edited review, but at least an overview.
Do professionals actually post photos to the sites this scrapes from, or is it mostly amateurs using expensive gear?
What? the 24-105 is an L lens and a solid walk around choice. I bought the 24-105 well before upgrading my body. Are you thinking of the 24-105mm f/3.5-5.6?
They are not wrong in that it is bundled with the MkIII. But I feel like the demographic that purchases the MkIII aren't getting the 24-105 with it 'just cause'
It was a common kit lens for the 5D Mk III, and AFAIK in the pairing the lens limits the camera (edit: from a pure optical quality perspective). It's not a bad lens at all, but clearly an allrounder with compromises and not a specialist. (And thus a good kit lens and it's not a surprise many people use it, since getting a zoo of better lenses is really expensive and a lot to carry around)
IMO it's just the opposite. It's L-quality glass, and the IS makes it a decent performer for low-light and video as well. With the exception of pixel peeping, I'd feel far more limited with a single fast prime in most situations.
I meant "weaker" purely from an optical perspective: the camera has a higher resolution than the lens can deliver, at which point there's the question if the money isn't better spent elsewhere, assuming you don't have other reasons to go for the expensive high-end body. Generally a characteristic of bundled lenses: they are flexible and (with the better cameras) high in quality, but generally they don't reach the body in quality. (Although looking at tests again, it seems better than I remember the one I borrowed a while back)
And e.g. if i were to switch to full-frame, I don't need the extra length of the 24-105 and would go for a sharper off-brand 24-70 2.8 (although there is a new revision of the 24-105 around the corner), since I'd want to cover longer ranges with additional lenses anyways. But that option wasn't available when the Mk3 came out, and I totally get why people choose the 24-105.
> This is a $1,000 lens--definitely not a kit lens.
It is also definitely not an f/2.8 lens like all the other top lenses associated with this camera. I too have found it puzzling just like the parent did.
Lens speed is not the end-all be-all of photography. The 24-105L is significantly longer and has image stabilization. Sometimes the ability to handhold with a shutter speed 2-3 stops slower trumps the ability to increase the shutter speed 1 stop.
IS lenses are generally considered better for all-around photography, particularly for subjects which do not move like landscapes. Faster lenses give you more ability to stop the action of a moving subject and a shallower depth of field (given equivalent focal length).
The 24-105L is probably the most solid general-purpose lens for a full-frame in Canon's lineup, it's silly to disregard it just because they offer a bundle with full-frame bodies.
All you say is correct and I agree that lens speed is not the only element to consider. As you noted, it is a trade-off between different factors.
My main gripe with the 24-105L vs. the 24-70L (i.e. the lens that came second in that chart) is not just speed in itself, but the fact that the former has a zoom extension of more than 4X (my previous comment should have been more exhaustive).
As an old-style guy, I usually am uneasy with zoom ratios over 3X because I believe they are detrimental to quality. So when I have to trade between zoom extension vs. something else, I often go for the latter. The MTF charts for the 24-105 f/4 L [1] and the 24-70 f/2.8 L II [2] seem to confirm my gut feeling...
If you can only afford one lens, it's a fine all-around piece. It just isn't as good as some other choices. Many have complained that it is too soft. I was explaining why this combo is at the top of the list and commenting that it's popular, but not "the best" combo.
I suspect the frequency at which zoom lenses appear at the top of the data is in part related to their inclusion as 'kit' lenses. I suspect that it is also in part a natural consequence of their versatility.
By which I mean that because a 24-105mm zoom can be set to a suitable focal length for everything from landscape to portraiture to a bit of sport, in aggregate it will be used for more pictures than any of the prime lenses one would use in each of those situations.
To put it another way zooms that cover wide angle to telephoto get used because they're really useful. Particularly when taking pictures under uncertain conditions. Any lens suitable for one of those conditions will stay in the bag in other contexts.
Yup, you don't even see the "holy trinity"(35L/85L/135L) anywhere in the top 10. Neat data aggregation but there's a lot more that goes on in lens selection.
If I know I'm going to need multiple focal lengths I'll usually use two bodies but I find the constraints of a prime lens to force me to think about composition and give better results overall.
> If I know I'm going to need multiple focal lengths I'll usually use two bodies
Sure but the amount of people using and carrying two bodies is very limited compared to the ones who have only one or only want to travel light.
>I find the constraints of a prime lens to force me to think about composition and give better results overall.
Yes and no. There are pictures you just cannot take without a zoom - you can only zoom with your feet if the subject is on a path where you can actually walk.
> Yes and no. There are pictures you just cannot take without a zoom - you can only zoom with your feet if the subject is on a path where you can actually walk.
Well, if you want to get that pedantic the longest focal length lenses Canon makes aren't zooms, they're primes ;).
If we're talking professional photographers all of them are going to be using two bodies(aside from the craigslist specials) simply because having your one body die on-shoot is going to make for a very unhappy clients. Like a parent comment mentioned earlier, this dataset looks shifted towards the non-professional dataset which is important to consider when looking at the data.
Its a great lens. It would be my pick for the best Canon lens as an all-arounder (either that or the 24-70L, but iirc that one is much heavier). The sensors are so good nowadays that you can shoot well in low light even with the f/4 aperture.
I have that lens and its one of the best Canon lenses. Not much distorsion, not bad vignetting, very good resolution and with an included IS for video shooting as well. Its not without reason that most serious Canon pro photographers recommend it.
Some strange data appears on this site, the second most popular for the full-frame D600 is a 2004 DX (notably bad) lens. No mentally sane person would ever commit such as sacrilege.
Didn't go further to investigate but this puts a serious doubt on the representativity of the data.
I've seen plenty of EOS 7Ds and 5D Mk II/IIIs with Canon kit lenses on them (of the ~$200 3.5-6.3 variety). I've personally had to tell people who came to me for purchase advise to buy either the 'enthusiast' or 'pro' tier lenses and completely ignore the bottom end.
Too many people "blow their wad" on the body and then find themselves only able to afford a couple of hundred more for a lens, rather than investing in the glass and upgrading the body (which will get upgrades every few years).
I'm surprised the 40mm f2.8 isn't more popular, it takes great pictures, it's cheap, and physically using it is great (it really transforms how the camera feels because it's so small). Strongly prefer it to the 50mm f1.8, unless it's super dark.
If you read pretty much any guide online about 'what lens should i get for my dslr' you'll pretty much always see whatever the 50mm f1.8 for your camera model at the top of the list. So im guessing lots of people just buy them due to that
A 50mm f1.8 or f1.4 was pretty much the historic "normal" lens to get. The 40mm pancake parent mentioned is a nice lens. I'm not sure lens speed matter as much as it once did though. And I'm also not convinced that a pancake lens makes all that much a difference when the body is already pretty big and heavy.
That said. It's a decent lens. If I could only find what gear pocket mine got shoved into :-)
It's my opinion that the 40mm f/2.8 from Canon is a much better lens than either the 50mm f/1.8 II or the 50mm f/1.4 (haven't tried the STM version) in terms of color, contrast, distortion, and as the GGP said, size. Although it's not a very fair comparison, because these 50mm lenses are both much older. This is one of my favorite travel lenses, and I do agree that it changes the camera handling (for the better). I love how it goes with a 5D.
However, I think one of the biggest reasons the 50mm gets recommended so much is because it is such a huge difference for someone to go from a f/3.5-5.6 kit lens to an f/1.4 or an f/1.8 in terms of DoF, and for beginners trying to move up this can be a huge eye opener.
Personally I think the 40mm is much easier to use, and it's certainly much sharper wide open.
The reason the 50/1.8 gets recommended is cost, full stop. It's the only prime in Canon's lineup that you can get for $100, and as a prime it is naturally better than all but the most expensive zooms. 80mm equivalent is a good length for portraiture but it's not particularly good as an all-around focal length.
Nikon has an APS-C format normal lens and that is the standard recommendation for that system. Pentax has a 35/2.4 lens and that is the standard recommendation for that system. Canon's lack of an affordable normal prime is a glaring weakness in the lineup and one of the reasons I eventually moved on to a NEX - you could get a 35mm and a 19mm prime for less than I could get a single 40/2.8 in the Canon system.
The 40/2.8 is the closest thing to a normal lens in Canon's lineup. But 40mm is still significantly longer than 35mm, and since Canon uses smaller 1.6x crop sensors instead of the industry-standard 1.5x crop sensors we are talking about going from a 50mm equivalent on other systems to a 64mm equivalent on Canon. That's still too long for a normal lens, it's not particularly fast at f/2.8, and it's more expensive than a Nikon 35/1.8 DX or a Pentax DA 35/2.4, but it's the least-bad option in the Canon stable.
My personal opinion is that Canon's lineup is heavily geared towards professional users, particularly full-frame users. They have nice L-series glass and their primes are well-placed for full frame users, but they just don't have many low-cost options for entry level users. There's no EF-S 12mm, 14mm 17mm, 22mm, or 32mm, all crucial focal lengths. Nor are there any cheap APS-C long primes. And their electronic aperture impairs the ability to use manual lenses like the excellent Samyang series, which could otherwise fill in some of these gaps.
Ironically, Canon was the first company to really release an affordable consumer DSLR. I suspect their lens lineup choices are probably biased toward pro/prosumer full frame and mass market APS-C. Relatively inexpensive zooms are going to be a lot more popular on Rebels than primes are.
My post was intended to support the original post in the thread and to gently disagree with the post I was responding to.
While you bring up very good points I'm personally inclined to disagree on the "cost" argument when comparing 50mm f/1.8 and 40mm f/2.8, simply because by the time the 40mm pancake was released the 50mm f/1.8 went up in price from $80 to $125. Very shortly after it was released the price for 40mm went from $199 down to $149 for a while and at that price the difference between $125 and $149 is negligible. At this price point the 40mm is arguably of a higher "value" than the 50mm unless you really need f/1.8.
In the first sentence of your post you used present tense and I'd like to point out that the 50mm f/1.8 has been discontinued, so anyone continuing to recommend this lens is giving dated advice (not saying you are).
Currently the 3 cheapest Canon primes are the 50mm f/1.8 STM at $125, the EF-S 24mm f/2.8 STM at $149, and the 40mm f/2.8 at $199.
I currently recommend people the 40mm if asked about what to get for a full frame or the 24mm f/2.8 STM for APS-C. The 24mm on APS-C gives you a nice 38.4mm equivalent and I don't know why you've dismissed this fine lens (I hope that's no too harsh).
Interestingly Yongnuo released the YN 50mm f/1.8 for $50 that looks almost identical to the Canon 50mm f/1.8 II that's been discontinued.
The biggest issue I see with this data is that it gives me a snapshot of which camera systems have been most popular with photographers in the past, which doesn't necessarily reflect what photographers want to buy today.
For example, many mirrorless camera systems are gaining market share but are badly under-represented on this website.[1]
It would be more useful to see how the figures have changed over time.
[1] For example, right now, five out of the top 10 best-selling cameras at B&H Photo & Video are mirrorless.
Not sure. I have both but I know people who have sold their DSLR kit and now use mirrorless exclusively. Especially if you don't do a lot of sports or other action photography using long lenses and benefitting from the best autofocus.
For much of the photography I do, I find the Fujifilm cameras/lenses a better match. If I were starting out I'm not sure I'd make the investment in full-frame Canon gear.
Hmmm. Most full frame cameras are not medium format[1] ( ~ 24x36 to 100x125)
And most medium format cameras aren't large format [2] ( > 100x125)
Full frame has its fans, and there's certainly some photography styles that benefit from it, but all these formats are fairly arbitrary - especially full frame.
Highly recommend Full Frame Wars [3] that includes a history (and some myth-busting) of the format - all the way back to a 1891, when an interest in having 16 frames per foot of rolling film was the basis of this particular (and now 125 year old) standard.
Another possible interpretation of the data (as presented) is that it reflects the interest / ability of those particular camera owners in stripping EXIF data. Or, from the other side, the interest / ability of owners of low- or non-listed cameras in doing same.
Consider DPReview's reader's awards for 2015[1] - Sony A7R II, Olympus E-M5 Mk II, Nikon D7200. Not suggesting this is a qualitatively better measure, given it's on a photography-centric web site, but it's significantly contradictory to the results in this article.
Anyway, it'd be nice to see what subset of these 11,000,000 images contained no EXIF data at all (the page doesn't include the string 'exif', and there doesn't appear to be a 'no information found' category).
Facebook reports ~300,000,000 new photos uploaded daily - obviously 99% of those are likely camera phones, but it does put this 11 million number in some perspective.
Pages like this make me really convinced I need to print up frame cards for the various lenses I use. I use a lot of vintage glass when I'm shooting that obviously won't have Exif data. It would be useful for later browsing to know at a glance whether I was shooting with a Jupiter 3 or a Pentax Super Takumar.
This measurement is going to be biased somewhat towards zoom lenses because they cover multiple focal lengths in a single lens. In other words you might have a 24mm, a 50mm, and a 70mm prime lens that each get 1/3 of your time but if you were using a 24-70mm lens instead then all of those would count as one lens.
It's by no means entirely inaccurate, zoom lenses are very popular now, and there's still some merit in seeing which lenses are actually on cameras most of the time. Just bear in mind that slicing them in this way is inherently a little biased to certain lenses rather than slicing them based on some other dimension like focal length of the shot.
70 comments
[ 3.3 ms ] story [ 139 ms ] thread[ADDED: It actually appears that this does align with Flickr stats--which is, of course, not what you'd find on Facebook or other more mass market sites.]
https://explorecams.com/stats/global
My guess is that it's based on frequency. What is unknown is whether the sample takes more prolific photographers into account when calculating "common". 5000 people may all own same camera and take single public image as oppose to single photographer who takes 5000 images.
Very cool but would love more details on how they are defining their criteria.
It does in fact appear that the Canon 5D iii and the 6D are the two most common DSLRs on Flickr based on the number of photos posted. https://www.flickr.com/cameras/canon/
(The leading smartphones are still more popular in terms of total numbers.)
This. Of my friends, most people seem to have a bottom- or mid-range APS-size sensor DSLR from Canon or Nikon. But the two of my friends who actually have a 5D iii and a 6D, respectively, are the only ones who a) shoot photos that are interesting to people outside their immediate friends/family and who b) post those photos on flickr.
(While I'm the geek with the superzoom, or the '72 Nikon Nikkormat loaded with Tri-X when I'm feeling analog.)
The number of photos taken using phones would no doubt dwarf the data used here, dedicated camera/lens combos are becoming something of a niche market, albeit still a sizeable one.
I use the Olympus EM-1, way down on the list. Nonetheless I found it interesting that the most commonly cited lenses used were the 12-40mm f/2.8, and the 40-150mm f/2.8. These are both excellent quality lenses, but relatively massive and quite expensive, at about $1000 and $1500 respectively.
Saying goes the merit of a photograph depends much more on the person using the camera than the camera being used.
(Also, chances are they'd skew the lens numbers significantly - while kit lenses come and go, pro glass tend to live much longer before being upgraded. (The 24-70mm f/2.8 which you often see around single-digit Nikon bodies, for instance, was the standard zoom for 8-9 years before being given an upgrade last year)
D3100 and D5000 go with the 18-55 kit lens (which is in fact a very good lens considering the price) and have a third of their pictures taken with these. D7000 and D90, being somewhat more advanced hobbyist models, go with their respective kit lens (18-105) with 20-30 % share.
Full-frame bodies (D600, D700, D800 etc) most often have the professional 24-70/2.8 lens, but the share is still smaller, indicating that a larger variety of different lenses is in use, as you would expect with pro/prosumer cameras.
i.e. perhaps what's more interesting, is the variety of lenses used by a particular photographer. If it's mostly one camera and one lens, we're pretty sure that the photographer has a package, and is just taking simple pictures.
If the photographer has a range of lenses for different shots, we're pretty sure he's choosing the best lenses for the job. Which means concentrating on the distribution per photographer, not per lens.
The page should also show the distribution of combinations among photographers. Most photographers will be at 1-2 lenses. The better photographers will use more lenses.
In all likelihood, the less popular camera + lens combination is the better one.
If you drill through to the camera/lens combination you actually get to see some of the photos that have been crawled, and then it will take you back to the original source.
[1] https://explorecams.com/blog/
Today the most popular "normals" are moderate, mid-range zooms, equivalent to 28-80mm or so on a 35mm equiv sensor.
As I understand, lens design in the digital era has changed due to the characteristics of the sensors vs. film. Mirrorless designs have also changed requirements and eliminates some restrictions due to flange distance. I'm not an expert in that field, maybe you know more in that regard.
The shorter macros where developed specifically for reproduction use (photographing records, books, paintings and the like), where the larger reproduction ratios (1:2, 1:1) where rarely if ever needed, so the problem didn't arise.
The shorter flange back distance in mirrorless systems makes some things easier and some things harder (if you want short lenses that is), but overall give more freedom to the lens designer.
Sensor vs film and manual focus / first AF systems vs modern AF systems is a huge difference and for many older lenses esp. in the wide angle spectrum that is the reason for a large performance difference between film and digital.
And designing lenses with adequate coverage for 35mm as well as a sufficient flange distance to clear a SLR mirror was definitely a problem back in the day. Older "normal" lenses for SLRs used to commonly be 55mm or even 58mm. Wide angle-lenses suffered a similar problem, and for a great while rangefinder lenses (or similar lenses that protruded into the mirror cavity of a SLR) were the dominant options for wide angle photography.
It took quite a while to develop good "retrofocal" lenses where the optical center is actually behind the lens (basically the opposite of the telephoto lens where the optical center is in front of the lens). The very first examples of the type are the Angénieux retrofocus and the Voigtländer Prominent's Skoparon, both introduced in 1950. The Skoparon design was essentially a Tessar lens backwards and went on to be one of the earliest wide-angle lenses used in SLRs, notably as the Asahi Pentax 35mm f/3.5. It had an extremely long lifespan, in its medium-format variant (Pentax 75mm f/4.5) it continued to be produced until around 2000. The funny thing is it all happened because Voigtländer patented it for rangefinders but neglected to patent it for SLRs!
The other thing to remember about 50mm-ish macro lenses is that it's still somewhat reasonable to build a helicoid that can get out to 1:1 reproduction without extension tubes. I have a 50mm and a 90mm that will both go to 1:1 reproduction on their own, and I can tell you that while the 50mm may be somewhat ungainly the 90mm is absolutely absurd when you crank it all the way out. The 50mm uses a double helicoid, the 90mm is actually four helicoids running inside each other.
"35mm equivalent" focal lengths and "crop factors" are by far the most common ways to mentally visualize this.
It's not entirely a direct conversion since not every format is a 3:2 ratio like 35mm is - for example 6x7 feels tighter because it's much closer to a square format. My 55mm lens is technically closer to a 22-mm equivalent in diagonal FoV but it feels like a 28mm does on 35mm.
I think it's honestly a shame that primes have died out. Dollar for dollar, most primes are massively better performers than zooms in terms of sharpness and aperture.
This lens is...not very good.
It's only popular because they're bundled.
I think this site is useful, however. It gives you a very quick way to navigate to a whole collection of photos of any body/lens combination. If you're in the market for a body/lens, this seems like a good source to go to. No side-by-side comparisons like an edited review, but at least an overview.
Do professionals actually post photos to the sites this scrapes from, or is it mostly amateurs using expensive gear?
[EDIT: Apparently it was available as a bundle at one point.]
IMO it's just the opposite. It's L-quality glass, and the IS makes it a decent performer for low-light and video as well. With the exception of pixel peeping, I'd feel far more limited with a single fast prime in most situations.
And e.g. if i were to switch to full-frame, I don't need the extra length of the 24-105 and would go for a sharper off-brand 24-70 2.8 (although there is a new revision of the 24-105 around the corner), since I'd want to cover longer ranges with additional lenses anyways. But that option wasn't available when the Mk3 came out, and I totally get why people choose the 24-105.
It is also definitely not an f/2.8 lens like all the other top lenses associated with this camera. I too have found it puzzling just like the parent did.
IS lenses are generally considered better for all-around photography, particularly for subjects which do not move like landscapes. Faster lenses give you more ability to stop the action of a moving subject and a shallower depth of field (given equivalent focal length).
The 24-105L is probably the most solid general-purpose lens for a full-frame in Canon's lineup, it's silly to disregard it just because they offer a bundle with full-frame bodies.
My main gripe with the 24-105L vs. the 24-70L (i.e. the lens that came second in that chart) is not just speed in itself, but the fact that the former has a zoom extension of more than 4X (my previous comment should have been more exhaustive).
As an old-style guy, I usually am uneasy with zoom ratios over 3X because I believe they are detrimental to quality. So when I have to trade between zoom extension vs. something else, I often go for the latter. The MTF charts for the 24-105 f/4 L [1] and the 24-70 f/2.8 L II [2] seem to confirm my gut feeling...
[1] https://www.usa.canon.com/internet/portal/us/home/products/d... (click on "Read more..." to see the MTF graph)
[2] https://www.usa.canon.com/internet/portal/us/home/products/d... (click on "Read more..." to see the MTF graph)
http://www.kenrockwell.com/canon/lenses/24-105mm.htm
http://www.the-digital-picture.com/Reviews/Canon-EF-24-105mm...
By which I mean that because a 24-105mm zoom can be set to a suitable focal length for everything from landscape to portraiture to a bit of sport, in aggregate it will be used for more pictures than any of the prime lenses one would use in each of those situations.
To put it another way zooms that cover wide angle to telephoto get used because they're really useful. Particularly when taking pictures under uncertain conditions. Any lens suitable for one of those conditions will stay in the bag in other contexts.
If I know I'm going to need multiple focal lengths I'll usually use two bodies but I find the constraints of a prime lens to force me to think about composition and give better results overall.
Sure but the amount of people using and carrying two bodies is very limited compared to the ones who have only one or only want to travel light.
>I find the constraints of a prime lens to force me to think about composition and give better results overall.
Yes and no. There are pictures you just cannot take without a zoom - you can only zoom with your feet if the subject is on a path where you can actually walk.
That's what crop is for ;)
Well, if you want to get that pedantic the longest focal length lenses Canon makes aren't zooms, they're primes ;).
If we're talking professional photographers all of them are going to be using two bodies(aside from the craigslist specials) simply because having your one body die on-shoot is going to make for a very unhappy clients. Like a parent comment mentioned earlier, this dataset looks shifted towards the non-professional dataset which is important to consider when looking at the data.
Didn't go further to investigate but this puts a serious doubt on the representativity of the data.
Too many people "blow their wad" on the body and then find themselves only able to afford a couple of hundred more for a lens, rather than investing in the glass and upgrading the body (which will get upgrades every few years).
So unfortunately, I can absolutely believe that.
That said. It's a decent lens. If I could only find what gear pocket mine got shoved into :-)
However, I think one of the biggest reasons the 50mm gets recommended so much is because it is such a huge difference for someone to go from a f/3.5-5.6 kit lens to an f/1.4 or an f/1.8 in terms of DoF, and for beginners trying to move up this can be a huge eye opener.
Personally I think the 40mm is much easier to use, and it's certainly much sharper wide open.
Nikon has an APS-C format normal lens and that is the standard recommendation for that system. Pentax has a 35/2.4 lens and that is the standard recommendation for that system. Canon's lack of an affordable normal prime is a glaring weakness in the lineup and one of the reasons I eventually moved on to a NEX - you could get a 35mm and a 19mm prime for less than I could get a single 40/2.8 in the Canon system.
The 40/2.8 is the closest thing to a normal lens in Canon's lineup. But 40mm is still significantly longer than 35mm, and since Canon uses smaller 1.6x crop sensors instead of the industry-standard 1.5x crop sensors we are talking about going from a 50mm equivalent on other systems to a 64mm equivalent on Canon. That's still too long for a normal lens, it's not particularly fast at f/2.8, and it's more expensive than a Nikon 35/1.8 DX or a Pentax DA 35/2.4, but it's the least-bad option in the Canon stable.
My personal opinion is that Canon's lineup is heavily geared towards professional users, particularly full-frame users. They have nice L-series glass and their primes are well-placed for full frame users, but they just don't have many low-cost options for entry level users. There's no EF-S 12mm, 14mm 17mm, 22mm, or 32mm, all crucial focal lengths. Nor are there any cheap APS-C long primes. And their electronic aperture impairs the ability to use manual lenses like the excellent Samyang series, which could otherwise fill in some of these gaps.
While you bring up very good points I'm personally inclined to disagree on the "cost" argument when comparing 50mm f/1.8 and 40mm f/2.8, simply because by the time the 40mm pancake was released the 50mm f/1.8 went up in price from $80 to $125. Very shortly after it was released the price for 40mm went from $199 down to $149 for a while and at that price the difference between $125 and $149 is negligible. At this price point the 40mm is arguably of a higher "value" than the 50mm unless you really need f/1.8.
In the first sentence of your post you used present tense and I'd like to point out that the 50mm f/1.8 has been discontinued, so anyone continuing to recommend this lens is giving dated advice (not saying you are).
Currently the 3 cheapest Canon primes are the 50mm f/1.8 STM at $125, the EF-S 24mm f/2.8 STM at $149, and the 40mm f/2.8 at $199.
I currently recommend people the 40mm if asked about what to get for a full frame or the 24mm f/2.8 STM for APS-C. The 24mm on APS-C gives you a nice 38.4mm equivalent and I don't know why you've dismissed this fine lens (I hope that's no too harsh).
Interestingly Yongnuo released the YN 50mm f/1.8 for $50 that looks almost identical to the Canon 50mm f/1.8 II that's been discontinued.
For example, many mirrorless camera systems are gaining market share but are badly under-represented on this website.[1]
It would be more useful to see how the figures have changed over time.
[1] For example, right now, five out of the top 10 best-selling cameras at B&H Photo & Video are mirrorless.
For much of the photography I do, I find the Fujifilm cameras/lenses a better match. If I were starting out I'm not sure I'd make the investment in full-frame Canon gear.
And most medium format cameras aren't large format [2] ( > 100x125)
Full frame has its fans, and there's certainly some photography styles that benefit from it, but all these formats are fairly arbitrary - especially full frame.
Highly recommend Full Frame Wars [3] that includes a history (and some myth-busting) of the format - all the way back to a 1891, when an interest in having 16 frames per foot of rolling film was the basis of this particular (and now 125 year old) standard.
[1] https://en.wikipedia.org/wiki/Medium_format_(film) [2] https://en.wikipedia.org/wiki/Large_format [3] http://www.digitalsecrets.net/secrets/FullFrameWars.html
It really depends on what kinds of photography you do and how much you have to carry your gear around.
Consider DPReview's reader's awards for 2015[1] - Sony A7R II, Olympus E-M5 Mk II, Nikon D7200. Not suggesting this is a qualitatively better measure, given it's on a photography-centric web site, but it's significantly contradictory to the results in this article.
Anyway, it'd be nice to see what subset of these 11,000,000 images contained no EXIF data at all (the page doesn't include the string 'exif', and there doesn't appear to be a 'no information found' category).
Facebook reports ~300,000,000 new photos uploaded daily - obviously 99% of those are likely camera phones, but it does put this 11 million number in some perspective.
[1] https://www.dpreview.com/articles/2761741499/readers-choice-...
It's by no means entirely inaccurate, zoom lenses are very popular now, and there's still some merit in seeing which lenses are actually on cameras most of the time. Just bear in mind that slicing them in this way is inherently a little biased to certain lenses rather than slicing them based on some other dimension like focal length of the shot.