For quite some time, I have been wanting to gather the EXIF data from all of my published images, and produce some statistics on various criteria such as exposure settings and equipment used.
During a recent period of renewed creative energy in the form of computer programming rather than photography, over the past month I finally developed automated code-based functionality in my site to extract, analyse and present what I consider to be some interesting statistical data about my collection of published images.
By its very nature, some of this data can be quite dry and geeky; but the results do show some interesting facts in terms of exposure and equipment preferences, and in some cases the results were surprising to me.
Note: If you don’t care what equipment I use, what exposure or stylistic photographic preferences I have, or you don’t care for numeric data generally, it is advisable to skp this long article; but if that kind of geeky thing does appeal to you, keep reading.
Presently, my site gathers data and presents tabular data, which lists from most frequent to least frequent, the numbers of images captured:
- with a particular camera;
- with a particular lens;
- with a particular telephoto extender;
- within a particular focal range;
- at a particular focal length;
- at a particular shutter speed;
- at a particular aperture;
- at a particular ISO sensitivity;
- during a particular year;
- during a particular month; and
- on a particular day.
I have also produced what I call a ‘time trend’, in the form of a table showing the number of images shot by month and year.
In this article I am going to break it down and explain the trends I see.
What this exercise has shown me is that the raw numeric data alone does not tell the whole story; there are some ‘human’ elements which can explain why a certain exposure setting or piece of camera equipment rates higher or lower in popularity than would otherwise be expected based on my knowledge of my preferences.
So, let’s look at the trends.
I have owned three digital SLR cameras in that time. Here is the breakdown by camera model:
|1||Canon EOS 5D||893|
|2||Canon EOS 5D Mark II||519|
|3||Canon EOS 20D||188|
The Canon EOS 20D was my first DSLR, and I owned it for over three years. I shot many images with it, but over the years I have deleted some of the images I published, as in those early years, I was an ‘anything and everything’ kind of photographer without much discernment.
My Canon EOS 5D lost its life to a rogue wave during a seascape shoot in 2010, and on the same day I replaced it with the Canon EOS 5D Mark II. I owned it longer than my Canon EOS 20D, and owned it for a similar amount of time to my current camera (the antiquated but still as-useful-as-it-ever-was Canon EOS 5D Mark II), but much of what I shot with it was during a period of frequent activity and rapid development. In other words, I shot a lot, and got better at it.
Most of my best work has been shot with my Canon EOS 5D Mark II, but over the past few years, as further analysis will show, my output in terms of quantity has significantly decreased, but my output in terms of quality has significantly increased.
This explains why my current camera is the least used.
I have owned many Canon EF lenses over the years, some of which I have sold. I have also ocasionally used a few lenses I have never owned.
To me, the lens usage statistics are more interesting, and there are some surprising results, which can again be explained by ‘human’ factors or circumstance.
Here is the data:
|1||Canon EF 16-35mm f/2.8L II USM||562|
|2||Canon EF 24-105mm f/4L IS USM||181|
|3||Canon EF 300mm f/2.8L IS USM||173|
|4||Canon EF 70-200mm f/2.8L IS USM||161|
|5||Canon EF 85mm f/1.2L II USM||100|
|6||Canon EF 35mm f/1.4L USM||98|
|7||Canon EF 135mm f/2L USM||93|
|8||Canon EF 400mm f/2.8L IS USM||78|
|9||Canon EF 300mm f/4L IS USM||36|
|10||Canon EF 180mm f/3.5L Macro USM||34|
|11||Canon EF 50mm f/1.8 II||31|
|12||Canon EF 85mm f/1.8 USM||24|
|13||Canon EF 70-200mm f/4L USM||17|
|14||Canon EF 17-40mm f/4L USM||3|
|15||Canon EF 24-70mm f/2.8L USM||3|
|16||Canon EF 400mm f/5.6L USM||2|
|17||Canon EF-S 17-85mm f/4-5.6 IS USM||1|
|18||Canon TS-E 17mm f/4L||1|
|19||Canon EF 200mm f/2L IS USM||1|
|20||Canon EF 100mm f/2.8 Macro USM||1|
It’s no surprise to me that my Canon EF 16-35mm f/2.8L II USM is the most frequently used lens of the 20 featured, as my main forms of photography over the years have been those which lend themselves well to ultra-wide focal lengths.
What could be considered surprising, but in hindsight not so surprising, is the presence of the Canon EF 24-105mm f/4L IS USM lens in second place.
I sold that lens nearly seven years ago (and ceased using it up to a year before I sold it), but I bought it in 2005 on the day of its release, and, as explained earlier, I shot anything and everything in those days. It was my ‘walk-around’ lens, and hence it got a lot of use. The last image I published was one I shot in November of 2007, which was a long time ago now; yet it still features prominently due to my experience and the development stage at the time.
I will not go into detail about all 20 of these lenses, particularly those with which I shot less than a handful of images, as those lenses were borrowed or tried; but I did want to draw attention to my two super-telephoto lenses, as by nature of my overall subject matter, they’re not commonly used, but some particular circumstances have made them surprisingly prominent.
The first is my Canon EF 300mm f/2.8L IS USM. I have owned it since 2007, but it gained its most significant usage half a decade later in 2012 when we visited Africa for the first time and fell in love with wildlife and the photography thereof.
In the early days it was used sparingly and for nothing particularly serious. A few years later I went through a phase of aviation photography, during which it was used more; but our South Africa trip in 2012 was where that lens came to shine.
That leads me to my Canon EF 400mm f/2.8L IS USM. It presently sits in 8th place, yet it is my most recently-purchased lens — I’ve only owned it for over a year. What drove it so high up, relatively, was our trip to Kenya, where I used it, by my rough estimate, 98% of the time. I took two other lenses (16-35/2.8L II and 70-200/2.8L IS) and also used a 300/2.8L IS for a few shots, but the 400/2.8L IS pretty much lived on the camera.
There isn’t much to say about the prevalence of my Canon telephoto extenders, as there are only two models, they work only with particular Canon EF L-series lenses, and in my case, I tend to only use them with my two longest lenses.
All that can be said is that of the two telephoto extenders, the Canon Extender EF 2x II features more prominently, which can be explained by the two Africa trips we have done. On the first trip to South Africa, I used it with my 300/2.8L IS, giving me 600mm of focal length (ideal for wildlife photography); and on the second trip to Kenya, I used it with my 400/2.8L IS, giving me an even better 800mm of focal length — absolutely wonderful for frame-filling images of big cats and birds.
|2||Canon Extender EF 2x II||98|
|3||Canon Extender EF 1.4x II||54|
This particular category is amongst the most interesting to me.
In developing my statistics gathering and reporting system, I initially only looked at actual focal lengths; but further into the project (actually, after ‘completion’), it dawned on me that breaking down focal lengths into distinct focal ranges would be more useful and revealing than reporting on single focal lengths alone, especially as the presence of zooms can mean lots of isolated cases of esoteric ‘no man’s land’ focal lengths appearing (such as 33mm or 98mm).
I used Canon’s definitions of focal ranges as the baseline for my own categorisation of focal ranges, and crunched the data to produce the following table:
|2||Short Telephoto (70-135mm)||350|
|4||Medium Telephoto (136-399mm)||227|
|5||Long Telephoto (400-800mm)||196|
I have known for years where my preferences lie in terms of focal lengths.
I like ultra-wide and ultra-long focal lengths, and not much in between.
The second spot is occupied by short telephoto ranges, which I specifically use (particularly the 85mm and 135mm focal lengths) for portraiture and still-life. I have many portrait images in my gallery, which allows that focal range its prominence.
The standard focal range is my least used focal range, which is no surprise, as for years I have regarded the 50mm focal length with complete disdain, and have also boycotted ‘standard’ focal lengths, as, quite frankly, they bore me. Given the standard focal length of 50mm more or less replicates what the human eye can natively see, it has no interest, as it’s neither wide and expansive, or long and detail-revealingly tight.
While I have owned both a standard zoom and a 50mm lens in the past, for most of the time I have had no lenses with focal lengths between 35mm and 70mm. I just have no use for them, and my subject interests do not require or lend themselves to such focal lengths.
As outlined in the previous section, statistics about focal length ranges are more useful and revealing than statistics about individual focal lengths, particularly as there are over 700 available amongst the lenses I own or have owned; but for the sake of completeness, here are the statistics:
The main points of interest in this table are the double-digit and triple-digit occurrances towards the top of the table.
Moderate telephoto and long telephoto focal lengths also tend to dominate, which is driven by my subject choice, and also the fact that most of my lenses are primes. The use of telephoto extenders on the big lenses, and the use of focal lengths at the extreme ends of a few zooms, also contrbiute to the particular focal lengths and frequencies documented in the top 20 rows of the table.
My shutter speed usage is reported in a rather long table, too.
I haven’t drawn any analytical conclusions from this particular data, but apparently I favour 1/200th of a second.
Of interest towards the bottom of the table is the occasional very long shutter speeds I have used.
My longest exposure was just over an hour, and there have been a few over the years than ran for single-digit minutes, and one which ran for ten minutes.
The 30-second shutter speed has achieved sufficient prominence, and otherwise most shutter speeds are in the middle-of-the-road territory of up to 1/500th of a second.
|1||1/200th of a second||114|
|2||1/125th of a second||77|
|3||1/160th of a second||77|
|5||1/500th of a second||55|
|6||1/250th of a second||52|
|7||1/100th of a second||52|
|8||1/320th of a second||50|
|9||1/640th of a second||44|
|10||1/400th of a second||43|
|11||1/50th of a second||41|
|12||1/60th of a second||37|
|13||1/1,600th of a second||36|
|14||1/80th of a second||36|
|15||1/1,250th of a second||34|
|16||1/2,000th of a second||29|
|17||1/40th of a second||29|
|19||1/800th of a second||28|
|20||1/1,000th of a second||28|
|21||1/2,500th of a second||27|
|26||1/25th of a second||23|
|33||1/8th of a second||18|
|34||1/30th of a second||18|
|36||1/15th of a second||17|
|37||1/20th of a second||17|
|38||1/4th of a second||17|
|42||1/6th of a second||15|
|47||1/13th of a second||13|
|49||1/3,200th of a second||13|
|50||1/10th of a second||11|
|53||1/8,000th of a second||8|
|54||1/4,000th of a second||8|
|55||1/5th of a second||7|
|56||1/5,000th of a second||6|
|61||1/6,400th of a second||3|
|63||1 minute and 20 seconds||2|
|66||1 minute and 30 seconds||2|
|67||1 hour and 7 seconds||1|
|68||3 minutes and 32 seconds||1|
|69||2 minutes and 28 seconds||1|
|71||6 minutes and 15 seconds||1|
|72||15 minutes and 24 seconds||1|
|73||15 minutes and 1 second||1|
|74||2 minutes and 43 seconds||1|
|76||10 minutes and 1 second||1|
|79||3 minutes and 7 seconds||1|
The statstics about my aperture choices are quite interesting — to me, at least — as they show that despite my preference for having the widest aperture in a given focal length, I don’t always use those wide apertures.
Here is the breakdown:
The fastest lens I have is f/1.2, and the two longest focal lengths I have don’t get any faster than f/2.8; yet f/8 seems to be my most frequent aperture.
Again, this is easily explained by my proliferation of wide-angle scenic images. Even though I shoot these with my 16-35mm f/2.8 lens, I rarely ever use that lens’s maximum aperture; for in shooting architecture, cityscape, landscape and seascape images, I want a wide view with a deep depth of field and rich, sharp details from foreground to background, which calls for narrower apertures.
The difference between f/8 and f/11 is not great, but unsurprisingly f/11 is in the top five.
Another common aperture I use is f/2.8. I have four f/2.8 lenses, two of which are super-teles, which I almost always shoot wide-open. I didn’t buy fast super-tele lenses to shoot at f/8! My Canon EF 70-200mm f/2.8L IS USM also ostensibly gets used wide-open more often than not.
The moderate-to-slow aperture of f/5.6 also appears rather frequently, which I figure is likely to be the result of using my Canon Extender EF 2x II on my 300/2.8 and 400/2.8, giving me 600mm at f/5.6 and 800mm at f/5.6 respectively.
I also use f/5.6 frequently for portraiture. Even though my Canon EF 85mm f/1.2L II USM lens is my go-to lens for portraits, I don’t frequently shoot it wide-open. Just because a lens can be shot at extremely wide apertures does not mean it always should be shot at such apertures.
Towards the bottom of the table are extremely narrow apertures, which I rarely ever use. One must be careful to avoid diffraction caused by very narrow apertures, which leads to softening of fine details. If a deep depth of field is required, there is little benefit in stopping down beyond f/11 or f/16 at wide focal lengths, as apertures of f/8 and f/11 tend to achieve a very acceptable compromise of sharpness and depth of field.
The statistical breakdown of my usage of different ISO sensitivity ratings was no surprise at all, but it remained interesting to see.
The top five spots are occupied, in order, a stop apart, starting from the lowest native ISO my cameras have offered.
I have known for years that I favour as low an ISO as possible, and for most of my photography, I can get away with this; but for wildlife and some other low-light photography (such as band/musician performance photography), I have simply needed to push the ISO to achieve acceptable shutter speeds to freeze movement in low light.
According to the data, the most extreme ISO sensitivity rating I have used is 6,400.
Year of Capture
The next set of criteria I examined in my image statistics was time; specifically, which years, months and days of the week were more popular.
Again there are some unsurprising results, but a few others stood out as interesting.
|#||Year of Capture||Images|
Over a ten-year period, my most frequent activity happened in 2006, 2007 and 2008. It was during this time that I rapidly developed my interest in photography, and spent many weekends shooting, as well as travelling domestically on photography-centric trips.
I was also a more generalist photographer, shooting all manner of random subjects. My style and subject matter preferences were only in their infancy of development. All of this explains my proliferation at the time.
The year 2005 features as the least frequent year. It was the year in which I first bought a DSLR. I had been shooting digitally since 2002, but had not really established much of a presence online, and I didn’t consider photography a serious or even semi-serious interest.
From 2011 to 2015, my output took a serious decline in frequency. I became less interested, de-motivated, occupied by other interests and priorities, and I had become extremely fussy about what I would shoot and when.
With the exception of 2005, for which very few images are still published online, my record low year is 2014. I really shot very little, and my peaks of interest were few and far between, with only a few images here and there, most of which were captured during domestic trips we took, and during two photographic workshops I attended during the year.
I had stopped shooting seascapes — in fact, my last seascape image was shot in 2013. Even now, I remain disinterested in shooting seascapes, and it feels like a photographically significant chapter of life which has more or less closed. I know I can always return, but for now the desire remains lacking.
The year 2012 is the exception to this long-running ‘lull’ period, in that a massive trip to South Africa was the main contributor towards my output that year. I also had some frequent and productive seascape shoots that year as a result of a photography group with which I was quite actively involved at the time.
The present year, 2015, remains a low-output year, except for one event: our Kenya trip. I shot thousands of images in Kenya, and have published 95 images at the time of writing. Otherwise, I have shot only six images, and it was late in April before I shot the first!
Month of Capture
The breakdown of months is interesting from an academic viewpoint.
Here are the details:
|#||Month of Capture||Images|
I cannot offer any explanation as to why June is in the top spot; even though most of 2015‘s images were shot in June, it would still be a popular month. In my part of the world, it’s winter, and pre-dawn seascape shoots during winter mean being cold and wet. Unappealing!
January’s number-two place makes sense, as I tend to take time off work in December and January, and apparently I’ve been out and about over the years.
November is the least popular month, which is interesting, as 2015 has been a low-output year in terms of quantity, and November is only a month away!
Day of Capture
There isn’t much of interest or surprise in the statistical breakdown of days.
When I was a frequent (weekly, even!) seascape shooter, Sunday was my day. Saturday follows very closely behind due to its status as a weekend day which is generally busier than Sunday.
|#||Day of Capture||Images|
Lastly, I have produced a table showing the number of images shot by month and year.
I haven’t looked much into what this indicates, as the year and month breakdowns already discussed paint a better picture; but it remains never the less an interesting exercise to see how my actvity during the same month over an eleven-year period can experience peaks and troughs.
In conclusion, it has been an interesting exercise to gather and analyse my image data to see which cameras, lenses, exposure settings and points in time were most and least significant; and if you’ve read this far, congratulations, as this indeed is a long, dry article which analyses the sorts of information about which most people would not care.
For me, the data I now gather and report can be useful in recognising trends in my photography over the years from an EXIF data viewpoint, but it essentally remains little more than an academic exercise which was rewarding from a creative and geeky viewpoint.