Experiment: graph of lens characteristics on log scale

[gabriel_l1]

<p>Hello all,</p>

<p>I was thinking that using the linearly measured focal length to classify lenses is not completely intuitive when the resultant field of view changes logarithmically (i.e., adding 10mm to a 10mm lens results in a far narrower 20mm, but adding 10mm to a 500mm lens is negligible). On a whim I decided to graph an assortment of EOS-mount lenses by focal length on a log base 1.4 scale (actually log base square root of 1.4, artificially starting at 10mm and rounding to integers to give the "nicest" tick marks). This gives some convenient lens milestones (e.g. 105mm, ~200, ~400) but even more conveniently makes it trivially easy to see what happens when the lens is mounted on a 1.4X extender - assuming compatibility of course. Simply shift a given lens's point (or range if a zoom) two guidelines to the right, which has the effect of multiplying the range by 1.4.</p>

<p>Naturally, experienced photographers know quite well what is meant by "17-55" or "100-400" so this isn't exactly news. I also don't expect this to be particularly helpful to anyone who knows Canon's lineup inside and out, but I found the results interesting nonetheless. Please note that I did this the worst way possible - manually - so I do NOT guarantee perfect accuracy. Obviously the Canon, Voigtländer, and Sigma brand names are the properties of their respective companies and this graphic is purely for personal use.</p>

<p>Questions/comments welcome.<br /> Cheers,<br /> -GLL</p>

<p>PS - forgot to mention, the numbers at the start and stop of zoom ranges and at the points of prime lenses are max/min apertures, in case that wasn't obvious.<br />PPS - oh! Also, red points / ranges are L lenses. Yeah, I know.<br>

PPPS - the forum shrinks the image, making it less readable. If you right-click and select "view image" (firefox), or "copy image location" and go to that link, you can read a slightly larger version.<br>

<img src="http://www.georgetownphantoms.com/images/lenses.gif" alt="" /></p>

[alan_bryant1]

<p>Cool. I've made similar plots by focal length -vs- f/stop, but never did one vs cost, interesting.</p>

[gabriel_l1]

<p>Thanks Alan. I had briefly considered making aperture a "3D" value, but lacked the time/effort/knowledge to render that accurately or via software. I'd be interested in seeing your plots if you have them and don't mind sharing. I chose cost as the y-axis because I initially hoped it could help people (read: me) make more logical decisions about what to buy based purely on coverage, especially taking a 1.4X extender into account. Of course, this graph is best augmented by a working knowledge of these lenses' reputations since optical quality / reviews are completely lacking on this graph - unless you take cost or that red "L" color as an indicator. ;-)</p>

[gabriel_l1]

<p>Hah, already found a mistake, I labeled the EF 24-105 f/4L IS USM as an EF-S lens. D'oh.</p>

[gabriel_l1]

<p>Fixed.</p>

[bjscharp]

Another mistake: all fifties except the 2.5 compact macro are listed as `sixties', and the 60mm EF-S macro is at 64...

 

Nice work though!

[gabriel_l1]

<p>Thanks B.J. Scharp. Fixed that as well as the proper aperture for the Voigtländer.</p>

[allardk]

<p>Nice bit of work. What do you mean by <em>I did this the worst way possible - manually -</em></p>

[alan_bryant1]

<p>Here's a little grid of available primes. This was mostly done from memory, so there probably are omissions. I adopted your standard of putting the L's in red. Some special lenses, like fisheyes and some macros, were left out intentionally.</p>

<p><img src="http://img.photobucket.com/albums/v490/alanpix/CanonPrimes.gif" alt="" /></p>

[sarah_fox]

<p>I think this should be a log-log plot, as price also seems to vary exponentially with fl! ;-)</p>

[gabriel_l1]

<p>Allard, I say that manually was not a great way to do it as it involves too much work, too much possibility of error, and too little flexibility to change. Ideally this would all be done automatically by some software solution (e.g. rendering it in Processing or Mathematica) but as I said I don't have the requisite knowledge.</p>

<p>Alan, that's a useful chart, thanks for sharing.</p>

<p>Sarah: haha, I know what you mean. A 500mm is over twice the max value I used for price, and an 800mm over four times. I decided to keep the stratospheric lenses out of it though. It seems to me much more honest to keep price linear - makes you realize just how very expensive/inexpensive some of these lenses truly are. I know I have a very bad habit of "thinking logarithmically" when it comes to price (i.e., how much do you care regarding $1300 vs $1100? $150 vs $80?) so looking at it linearly helps keep my buying habits more rational.</p>

[allardk]

<blockquote>

<p> price also seems to vary exponentially with fl! ;-)</p>

</blockquote>

<p>If only that were true! I'd go out and buy that 14 2.8 and 17 T/S right now.. :-)<br>

(nerdy comments about alternative formulas to calculate price omitted by my own internal moderator)</p>

[KenPapai]

<p>Cool graph Gabriel.</p>

<p>Another good one could be plotting 3 or 4 separate lines:</p>

<p>XXD series from the lowly D30 & D60, thru the usable 10D up to the 50D -- cost vs. MPs<br>

XXXD series similarly<br>

1D series along with the 5D series.</p>

[gabriel_l1]

<p>Hi all,</p>

<p>Added a few more lenses. If I am not mistaken, the graph now includes ALL sub-$2300 Canon primes. Still a bunch of medium zooms not included, and of course anything over $2300 I simply didn't leave room for. Maybe later.</p>

<p>-GL</p>

[gabriel_l1]

<p>Image is no longer hosted, so I'm uploading it to this post.</p>