I have often wondered what it must be like to drive a Ferrari, or a Lamborghini, or even just a Porche.  I've yet to find out.  I have, however, had the opportunity to use (and now own) the camera-equivalent of a turbo-charged, top-of-the-line, luxury sports car (and an expensive one at that!), in the form of the flagship Canon "pro" body, the infamous EOS 1D Mark III.

Photo of 1D Mark III, taken with a Canon 30D.

I honestly never thought I would ever own such an expensive and sophisticated piece of equipment.  I've taken pride over the years in my (modest) ability to push cheaper equipment to the extremes, to extract the highest quality performance from less-than-perfect tools.  I did this for many years with cheap electric guitars, cheap binoculars, cheap computers, cheap software.  And in the arena of digital photography I think I've extracted some number of fairly high-quality photos from non-pro bodies, especially from the lowly Canon EOS 30D, which despite its age and sub-$1000 retail price still produces amazing images from time to time.  Indeed, I had grown so fond of my well-worn pair of 30D's and their remarkable image quality (when properly operated), that I had begun to doubt whether higher-quality images were even obtainable with any earthly gear of any brand or price. 

I was wrong.

The image quality from my new EOS 1D Mark III is truly a level above that of any camera I've used before.  The noise levels in image backgrounds are so low at normal ISO speeds that I wonder whether I'm looking at a computer image at all, or whether somebody has scotch-taped a premium-quality print from a film camera onto the front of my computer screen.


Because the 1D series all employ a 1.3x crop factor, the depth-of-field is reduced somewhat relative to that of a 1.6x crop-factor camera such as the 30D or the newer 40D.  What this means for bird photography (my personal passion) is that at any given aperture the subject tends to be more isolated from the background, resulting in more striking images.


It should be noted that I've also noticed a very significant difference between the image quality of out-of-camera JPEGs versus RAW images as displayed by Canon's "DPP" software.  The RAWs have far less noise than the JPEGs, even before any noise-reduction software is applied, which is something that I was not previously aware of.  I knew that JPEG encoding involved some degree of image compression, which could conceivably reduce image quality in subtle ways, but I was not prepared for the magnitude of the difference between the RAW images and the out-of-camera JPEGs.

Unfortunately, RAW images from a 10 megapixel camera run about 12 Mb in size, which places rather stricter requirements on one's in-field memory resources than when using JPEGs.  Previously I carried a total of 12 Mb of Compact Flash (CF) cards with me at all times, and had never as of yet run out of memory in the field with my current setup (which included two 8.2 megapixel cameras, both Canon EOS 30D's).  For the 1D Mark III in RAW mode I am figuring I will need at most 48 Gb (!) worth of memory cards in the field; this is based on my maximum memory-card usage during intense photography trips to such places as San Diego or the Atlantic flyway during fall migration.  With high-capacity 8 Mb cards this is entirely feasible, though with a significant outlay of cash ($500 or so, based on current CF card pricing).

Carolina Chickadee (Parus carolinus).

Canon 1D Mark III + Sigma 800mm f/5.6 lens.
800mm, f/8, ISO 400, 1/400 sec, JPEG-L

Fortunately, the 1D Mark III accepts Secure Digital (SD) cards in addition to Compact Flash (CF); SD cards tend to be 25% - 50% cheaper than CF cards of the same capacity.  Which of CF or SD is faster is not an easily resolved question, however, as each card has a different maximum bit transfer rate, as does each camera model.  In the case of the Canon EOS 1D Mark III, the maximum buffer capacity of 66 RAW images (or 30 in a single "burst" when the shutter button is held down continuously) means that in the field I am ulikely to ever be limited by the speed of my memory card, even if I use the slowest card available on the market today. 

That's a good thing, since it means I can buy the cheapest memory cards available.  Note that in the past I've always bought the cheapest cards I could find, except for one instance in which I "splurged" and bought a super-expensive "premium" card (a Lexar 133x CF card).  That's the only card that ever failed in the field.  So, the moral of the story is: with the Mark III (assuming you shoot like I do --- one carefully calculated frame at a time) you can use the cheapest memory cards available (whether SD or CF).  The most important thing is to have enough memory (even if cheap memory) to shoot everything in RAW.

House Finch (Carpodacus mexicanus).
Canon 1D Mark III + Sigma 800mm f/5.6 lens.
800mm, f/8, ISO 400, 1/640 sec, JPEG-L

An obvious question for any prospective owner of a 1D camera is whether the pixel count is high enough for his/her needs---i.e., is the higher megapixel count of the (much more expensive) 1Ds model worth the roughly 2x price as compared to the corresponding 1D model?  I could not say for sure, without using a 1Ds first-hand, whether the higher megapixel count of that model would translate into better bird photos. What I can say for certain is that the 10 megapixels of the 1D Mark III have been enough to capture the subjects I've worked with so far.  Although my 30D has a significantly higher pixel density (in terms of pixels per unit area on the image sensor), I'm far happier with the output of my Mark III than my 30D (though I'm keeping both my 30D's as backups). 

For some pretty damn convincing evidence for the utility of the Mark III for serious bird photography, click HERE .

(click to zoom)

Carolina Chickadee (Parus Carolinus).
Canon 1D Mark III + Sigma 800mm f/5.6 lens.
800mm, f/8, ISO 400, 1/800 sec, JPEG-L

At this point, a note or two about megapixel ratings is probably warranted.  Although megapixels might seem (to the uninitiated) to serve as a perfectly objective measure of a camera's capacity to produce high-quality images, the relation between total megapixels, pixel density, and image quality is somewhat confounded by other factors not reflected (directly) in the megapixel count.  As I explained in my Megapixels Demystified article, the pixel density---i.e., the number of pixels per unit area on the image sensor---will (in theory) serve as a better indication of the detail-capturing potential of a particular camera model, since the density of pixels limits the theoretical maximum of the number of distinct image details which can be represented.  In practice, however, the empirical resolution of an image sensor emerges as the observed tradeoff between pixel density and pixel fidelity---that is, the perceived resolution of the resulting image will vary in proportion to the pixel density and in inverse proportion to the average noise level of the individual pixels.  Thus, the optimal megapixel count in terms of final image resolution will depend on both the pixel density (i.e., the megapixel rating in combination with the crop factor) and on the per-pixel noise level.  Unfortunately, there appears to be a tradeoff between pixel densities and noise level, so that for larger densities (such as on the newly announced Canon EOS XSi model) there will tend to be greater noise, simply due to the physics of light-collection at small photosites on a CMOS sensor.

Yellow-rumped Warbler (Dendroica coronata).
Canon 1D Mark III + Sigma 800mm f/5.6 lens.
800mm, f/8, ISO 400, 1/1000 sec, RAW

In the case of the EOS 1D Mark III, the 10 megapixels delivered by the camera's imaging sensor have quite excellent noise characteristics, due to the use of larger micro-lenses at each photosite.  Although the actual area of light-gathering sensitivity for each pixel occupies only a small area of the photosite allocated to each pixel, photons which would otherwise elude the sensitive area of the photosite are in many cases redirected toward the sensitive area by the tiny micro-lens which is positioned over each photosite for the purpose of focusing light rays within the photosite onto the much smaller pixel sensor in the middle of the photosite.  The result is that individual pixels are able to gather more photons per unit time, so that when averaged over the entire image we see less "noise" due to sampling error of small pixel sensors active over finite sampling periods.  In terms of image quality, the backgrounds tend to be smoother, with less speckling than in images produced by inferior sensors.

House Finch (Carpodacus mexicanus).
Canon 1D Mark III + Sigma 800mm f/5.6 lens.
800mm, f/8, ISO 400, 1/800 sec, RAW

Some other nice features of the Mark III are the 14-bit color, the self-cleaning sensor, the autofocus "microadjust" capability, and the 300,000-cycle shutter rating. The 14-bit color allows for smoother color gradations --- I remarked on this feature in my review of the Canon EOS 40D, since that model also features the new 14-bit A/D conversion hardware.  Another feature shared by these two models is the self-cleaning image sensor, which utilizes a vibrating filter over the sensor to reduce the incidence of dust specks in images.

The autofocus "microadjust" feature was one of the primary reasons for my purchasing the Mark III. Over the past year or so I have been noticing subtle differences in focusing accuracy of different lenses when attached to different camera bodies. The problem is that cameras and lenses can become uncalibrated (and unfortunately, many of them arrive direct from the factory in a poorly calibrated state, due to Canon's terrible quality-control at their factory).

More precisely, a camera and lens need to be properly calibrated to each other in order for the autofocus function to produce perfect in-focus photos. Otherwise, the camera can end up focusing slightly in front of or behind the subject. These front-focus and back-focus problems, when severe enough, can ruin an otherwise perfect photo. Sometimes a front-focusing lens in combination with a back-focusing camera can result in good images, since the front-focusing and back-focusing cancel each other out. The idea behind AF microadjust is to allow you to fine-tune your camera to correct for lenses which have a slight front- or back-focusing problem. Each of your lenses can be separately adjusted; the Mark III recognizes which lens is attached and automatically applies the preset microadjust during shooting. I haven't had time to do the AF microadjust on my camera yet, but I've read accounts of others (including pro bird photographer, Art Morris) who have noticed a significant improvement in autofocus accuracy after using this feature.

Yellow-rumped Warbler (Dendroica coronata).
Canon 1D Mark III + Sigma 800mm f/5.6 lens.
800mm, f/8, ISO 400, 1/1000 sec, RAW

(click to zoom)

Eastern Bluebird (Sialia sialis).
Canon 1D Mark III + Sigma 800mm f/5.6 lens.
800mm, f/8, ISO 800, 1/400 sec, JPEG-L

(click to zoom)

Blue Jay (Cyanocitta Cristata).
Canon 1D Mark III + Sigma 800mm f/5.6 lens.
800mm, f/8, ISO 800, 1/800 sec, JPEG-L

(click to zoom)

  American Robin (Turdus migratorius).
Canon 1D Mark III + Sigma 800mm f/5.6 lens.
800mm, f/8, ISO 800, 1/800 sec, JPEG-L


American Robin (Turdus migratorius).
Canon 1D Mark III + Sigma 800mm f/5.6 lens.
800mm, f/8, ISO 800, 1/1000 sec, JPEG-L

So, what don't I like about the camera?  So far, nothing.  I did have a brief issue with the flash exposure, but after researching this on the internet I found that the "high-speed synch" mode which I was using on my 580EX II flash unit was the culprit.  Apparently, the high-speed synch mode requires rather more user intervention in setting the appropriate power level.  I don't recall this being the case with my 30D, but for now I've turned off the high-speed synch on the Mark III and haven't had any problems whatsoever.  The Mark III has a maximum synch speed of 1/300 sec, which has been more than adequate for my birding needs.  Indeed, on a number of occasions I've had to stop down or reduce the ISO speed to avoid over-exposure, and this has only improved the image quality of my photos.

Be sure to check out the additional sample images linked below (all taken within my first 48 hours with the camera).  Better yet, check out my newer photo album of 125 bird photos taken during my first month with the camera.

---

Some More Sample Images