While image quality and autofocus capabilities largely determine the desirability of a given camera model for bird photography, there are a number of other features that are also desirable, though not all of equal importance.  In this section we will consider many of these other features and explain how they might affect your ability to capture pleasing images of birds.

2.7.1 Drive Speed and Buffer Size

For capturing sharp images of birds in action—i.e., flying, fighting, hunting, mating, running, etc.—a fast drive speed can be very, very useful.  This is the frames per second or “fps” rating that you’ll see in the camera’s specs on the manufacturer’s web site.  When the camera is in continuous drive mode, holding down the shutter-release button causes the camera to take a series of photos in rapid succession.  Precisely how rapid is dictated by the camera’s maximum drive speed, or burst rate.  Most consumer-grade and pro-sumer models these days offer 3 to 5 frames per second (fps), while pro bodies are typically in the 8 to 10 fps range. 
    Shooting action scenes by holding down the button in continuous drive mode is sometimes called spray-and-pray—in analogy to a gunman spraying bullets at the enemy and praying that at least some of the bullets strike their target.  (In fact, it’s often remarked that Canon’s current pro model, the 1D Mark III, sounds like a machine gun when it’s operated at 10 fps in spray-and-pray mode).  By following a bird in flight while continuously shooting (with the autofocus engaged throughout), the hope is that at least one of the resulting images will, by sheer luck, capture the bird in an ideal pose—e.g., with the wings well separated from the body, the bird’s eye visible, the head pointed more-or-less toward the camera, no distracting shadows covering large portions of the bird’s body, etc. 


    For intense action scenes, getting the “ideal” shot by sniping—i.e., taking one shot at a time by deliberately pushing the shutter-release button once for each shot—can require either super-human hand-eye coordination, extraordinary luck, or both.  By employing the spray-and-pray technique, you can substantially increase your luck by giving yourself more chances to get the shot right, and also compensate for a lack of fast-paced timing skills.  Or at least, that’s the theory.
    In practice, much is required of both the camera and its user in order to make spray-and-pray work effectively.  Obviously, the camera’s autofocus system needs to be fast and accurate enough to keep the bird in perfect focus throughout its entire flight.  For birds flying directly toward the camera, this can be especially difficult, since the lens’ focusing element needs to move very quickly to keep the bird in focus.  In spray-and-pray mode, this task is made significantly more difficult for the camera, because during every shot the main mirror is raised during the exposure, thereby blinding the AF system during the shot.  When the mirror is again lowered after the shot, the AF system has to quickly re-aquire the subject and resume tracking in time for the next shot.  When shooting at 10 frames per second, the camera has less than a tenth of a second to re-aquire the subject and establish accurate focus tracking before the next shot.  If it fails to re-aquire the subject, the next shot will obviously be out-of-focus (OOF), and once that happens, it can be extremely difficult for the AF system to get back on track, so most or all of the subsequent shots will probably be out-of-focus as well. 
    In controlled tests by sports photographers using “fast” (f/2.8) lenses and large subjects (humans) running toward them, even modern pro bodies rarely get 100% of the shots in focus at 10 fps.  For bodies with lower frame rates, the camera should have more time to re-establish focus tracking after each shot, so that the keeper rate (percentage of in-focus images within each series) may be higher, though for non-pro bodies the use of a less sophisticated AF system in these bodies may negate any such advantage.  Some pro bodies do allow the fps to be reduced via a custom function in the camera’s menu.
    A fair amount of skill is still required on the part of the user even when using the spray-and-pray technique.  Obviously, you need to keep the bird in frame during the entire flight.  This can be more difficult than many beginners realize (techniques for tracking birds in flight are discussed in Chapter 8).  Just as with the AF system, the raising of the mirror during exposure creates some difficulty by blackening out the viewfinder at regular intervals.  If the AF system does lose focus due to the operation of the mirror (or for any other reason), keeping the bird in frame can become extremely difficult or even impossible, since the bird may become so out-of-focus that it’s no longer visible through the viewfinder (even if the camera is still pointed directly at it). 
    Depending on which AF points you’ve enabled, and on the size of the bird and its distance from the camera, it may not be enough just to keep the bird in-frame.  For single-AF-point focusing, you’ll need to ensure that the selected AF point remains positioned on the bird in the viewfinder at all times; if the AF point drifts off of the bird, depending on the sophistication of the camera’s AF system (and on any special settings you’ve registered in the camera—see section 2.6.3), the camera may quickly re-focus onto some other object in the background, and after that it may be very difficult for you to force the camera to re-aquire focus on the bird.  For this reason, many shooters enable all AF points when tracking birds in flight, so that the AF system can keep tracking the bird even if it drifts from one AF point to another in the viewfinder.
    Having a fast drive speed isn’t very useful if the camera doesn’t also have a large buffer.  Recall that the writing of images to the memory card tends to be slower than the actual image capture process, so that the camera needs to have a special memory buffer to keep images that haven’t yet been written to the card, while newer images are in the process of being captured by the imaging sensor.  For long sequences of spray-and-pray shooting, that buffer will eventually fill up and then shooting will become impossible, forcing you to wait until some of the images in the buffer have been fully transferred to the memory card.  During this time, you may very well be losing opportunities for potentially great shots.  For this reason, it’s good to have either a very fast memory card, or a camera with a very large buffer (or both). 
    Many DLSRs today have buffer capacities of 10 to 20 RAW images; pro bodies usually have larger buffers, with capacities around 30 RAW images.  Although I strongly recommend shooting only in RAW mode, if you shoot JPEGs instead of RAW images you can typically fit many more images in your buffer.  As an example, the Canon 50D, a 15-Megapixel pro-sumer DSLR, can shoot a continuous sequence of up to 16 RAW images or 60 JPEG images with standard memory cards.  In contrast, Canon’s newest consumer-grade DSLR (the T1i) can shoot 9 RAW or 170 JPEGs, while Canon’s current pro body (the 1D Mark III) can shoot 30 RAW or 110 JPEGs under ideal conditions.  Note that certain settings in the camera can cause the buffer size to be effectively reduced.  Common settings that typically reduce buffer sizes are noise reduction processing, the use of high ISO settings, and so-called highlight tone priority, or HTP (a feature in some Canon DSLRs that reduces the incidence of clipped highlights—see section 2.7.10, below, for a discussion of HTP).  When examining the buffer size of a prospective camera body, be sure to find out how severely the buffer is affected by any special menu settings you’re likely to use (especially higher ISO's).

2.7.2 A/D Converter Bit Depth

Since this is a book about digital bird photography, a brief discussion of digital representation of image data will, unfortunately, be necessary.  (We’ll try to keep this as painless as possible.) 
    The smallest piece of digital information is the bit, or binary digit.  A bit can be either 0 or 1.  We can string these values (0’s and 1’s) together into longer binary numbers, like 10111, in which case each successive bit, from right to left, is worth twice as much as the previous bit.  Thus, in 10111 the rightmost bit is worth 1, the bit to its left is worth 2, the next bit is worth 4, the next bit is worth 8, and the far left bit is worth 16.  In this case, the fourth digit (which would have been worth 8 if it had been a 1) is a 0 bit, so we don’t count that.  Thus, 10111 in binary equals 1+2+4+16 = 23 in decimal.  Had the number instead been 11111, the value would have been 1+2+4+8+16 = 31.  That’s the largest number we can represent in five bits.  If we want to represent larger numbers, we’ll have to use more bits. 
    It turns out that the largest number that can be represented in N bits is 2^N-1.  So, counting zero as a valid number, the number of numbers (!) that can be represented by an N-bit number is 2^N.  Since we’re interested in understanding how binary numbers represent colors and images, we’ll belabor the point just a bit with a pertinent analogy.  If we assign a particular hue (color) to each binary number, then with N bits we can represent (or address) at most 2^N different colors.  In other words, if every pixel in an image is represented by an N-bit binary number, then the gamut—or maximum range of colors that can appear in that image—has a “width” of 2^N.  If you remember the earliest personal computers with color monitors, many of those had 8-bit color, so that only 256 (=2⁸) different colors were available for graphics.  In Figure 2.7.2 (below) we show the difference between a 24-bit image and an 8-bit image, just to crudely illustrate the importance of bit depth.




As you can see, the photo on the right shows distracting bands of homogenous color in the background, whereas the background in the original JPEG image (left) consists of fairly smooth color gradients.  Even in the JPEG image you can see (if you look closely enough) slight posterization (banding) in the background, due to the fact that the JPEG image allocates only 8 bits per channel (i.e., 8 bits for the red channel, 8 bits for the green channel, and 8 bits for the blue channel); the original 36-bit RAW image (12 bits per channel) exhibited even smoother gradations, which were lost upon convertion to JPEG, due to a combination of reduced bit depth and lossy compression (see section 5.4.4).  In this example, the 8-bit image is utilizing an indexed color space, which was optimized by Photoshop so as to retain the maximal amount of image detail; a more uniform 256-hue color space would surely result in a far worse-looking image than that shown above.
    The example above was intended only to crudely illustrate the effect of bit depth on image quality.  While larger Megapixel ratings aren’t always good for image quality (due to the increased potential for noise from smaller or more densely-packed photosites), larger bit depths typically are better.  More bits means more colors—colors that can be used to encode more (sometimes subtle) subject detail, especially in less well-lit areas of the bird.  Of course, more bits per pixel also results in larger file sizes for your images, though as memory cards and hard drives continue to get larger and cheaper, file size is arguably an irrelevant issue.  As some would say, hard drives are cheap, but photographic moments are priceless.
    In terms of choosing a camera, the issue of bit depth comes into play when considering the analog-to-digital (A/D) converter used by a camera’s electronics when encoding image information from the sensor.  Recall that photons striking the silicon matrix of the sensor liberate electrons that then flow across an electric field and into a capacitor for later measuring by the imaging electronics.  The quantity of electrons measured, which should be directly proportional to the number of photons received by the photosite, has to be converted into a digital quantity—a binary number—before a digital image can be formed in the camera from the agglomeration of all the pixel values.  The number of bits making up these individual binary numbers from the photosites is the bit depth of the A/D converter which converts the electron count (an analog signal, as measured by electrical current) into a digital quantity (a binary number). 
    The A/D converters used in today’s DLSRs are a mix of 12-bit and 14-bit converters, with the newer 14-bit units rapidly replacing the older 12-bit ones.  (That’s 12 or 14 bits per channel, mind you).  There are even whispers now of 16-bit A/D converters possibly being introduced in the near future; keep in mind that most of today’s pro Photoshop work is done in 16-bit mode (in software), so that even 14-bit cameras aren’t fully utilizing the color space(s) available in the postprocessing software.  (Unfortunately, this is due primarily to limits on dynamic range imposed by today’s imaging sensor technologies, rather than being simply an arbitrary limitation imposed by the A/D converter—dynamic range is discussed in section 6.9).
    Now, in Figure 2.7.2 we showed that there is a stark difference between 8-bit and 24-bit images.  The important question then is: how much of a difference is there between a 36-bit (12-bits per channel) image and a 42-bit (14-bits per channel) image?  The consensus among armchair imaging “experts” on the internet is that the difference is so negligible as to be virtually imperceptible.  At least one photographer has shown [Johnson, 2008], however, that for underexposed images, the extra 2 bits provided by the 14-bit A/D converter can result in significantly more detail in shadow areas than a comparable 12-bit converter.  More generally, there is (very plausible) speculation that the greater bit depth permits more postprocessing to be performed on images (on average) before posterization/banding becomes noticeable.  Die-hard skeptics insist, however, that 14-bit A/D is no more than a marketing ploy designed to liberate you of your hard-earned cash.  My own experience has been that my 14-bit camera has noticeably better color than my older, 12-bit body, though other differences in the imaging sensors for these two bodies may very well explain much of the difference without invoking the issue of bit depths.
    As a final note about bit depths, we’ll see in section 6.2 that the number of bits used for encoding each pixel (actually, each photosite, but we’ll gloss over that technical detail here) in the RAW file produced by the camera depends on the pixel’s intensity value: brighter pixels are allocated more bits than darker pixels, on average.  This means that lighter areas of an image will tend to show more details than dark, shadowy areas.  For properly exposed images this is a reasonable strategy for maximizing the amount of encoded detail in the areas of the image where it most matters, but for underexposed images, it creates problems when you try to increase the exposure later in postprocessing, because the dark areas will show fewer details and have more noise.  Having an overall greater bit depth (e.g., 14-bit versus 12-bit) means that more bits can be allocated across the entire spectrum, including (especially) the dark end of the spectrum, so that dark areas in underexposed images should retain more detail.  (See the reference at the bottom of this page for some evidence that this is indeed what is seen in practice).

2.7.3 AF Microadjust

As we noted in the previous section, today’s autofocus systems are highly complex and require precise calibration during manufacturing in order to work properly.  As many DSLR owners have discovered, this calibration is not always performed properly at the factory before the product is shipped.  Furthermore, lenses and teleconverters also require precise calibration, and can sometimes go out of calibration by being bumped or dropped, or perhaps even through normal use over extended periods of time. 
    In the past, re-calibration of cameras and lenses could only be performed at the factory, so that photographers had to send their equipment in to the manufacturer’s service center and wait a week (or more) for the equipment to be returned.  This service is often not free (except during the warranty period) and is sometimes done incorrectly, so that the equipment, upon being returned to the owner, may actually be worse than it was before being serviced (as has happened to me in the past).
    Fortunately, many of today’s newest cameras provide an AF microadjust feature, which allows the owner to correct minor mis-calibration issues at home.  It is typically buried in the camera’s menu as a so-called custom function.  As an example, on Canon’s 1D Mark III camera, AF microadjust is accessed via custom function III-7 (see the figure below).  When you select this menu option the camera then allows you to dial in a “microadjustment” amount, from -20 to +20. 


The value that you dial in is then used to adjust the camera’s expectations about the separation distance for the twin images on the AF sub-sensors during image registration (see section 2.6.1 for a description of image registration in phase-based autofocus systems).  This results in the camera focusing slightly in front of or behind (depending on whether you dialed in a positive or negative number) where it normally would.  Thus, if the camera or lens has been consistently front-focusing, you can dial in a positive number to force the camera to deliberately focus slightly backward from where it normally would, to compensate for the front-focusing problem.  Similarly, if the camera or lens has been consistently back-focusing, you can dial in a negative number to force the camera to focus slightly forward from where it normally would.
    The trick with AF microadjust is to figure out exactly how much compensation is required; in section 3.11 we discus methods for doing this.  For the purposes of choosing a camera for bird photography, we highly recommend choosing a model that features AF microadjust.  Note however that AF microadjust addresses only one type of autofocusing problem: consistent front/back focusing.  If your camera sometimes front-focuses and sometimes back-focuses (not due to user error, such as incorrectly positioning the active AF point on a tree branch or the background), then AF microadjust won’t help, and you’ll need to send in the camera for repair by the manufacturer.
    Note that on some cameras you can dial in a differet AF microadjust setting for each lens, and even (in some cases) for each lens/teleconverter combination.  Thus, in the figure above, you can see that an adjustment of +5 has been set for my 600mm f/4 lens when it is used with a 2x teleconverter, though when I use just the 600mm lens without any teleconverter the microadjust setting is +0 (not shown in the figure).  These settings are stored in the camera, so you only have to program them once.  On Canon’s current camera models, the camera can’t distinguish between two different lenses of exactly the same type (i.e., both 600mm f/4 Canon lenses) or between two different teleconverters (such as, for example, the defective 1.4x teleconverter I used to have, versus its replacement).  I haven’t yet found this to be an impediment.

2.7.4 Histograms and Highlight Alerts

One of the most important features of a DSLR is its histogram and associated functions such as highlight alerts.  We’ve already seen examples of histograms in section 2.6.1.  The exposure histogram is a graph showing how many pixels of each intensity value are present in the captured image.  Figure 2.7.4 shows some more examples of histograms, taken from the LCD of a DSLR camera.




In the first screenshot (top), you can see the image that has been captured (a Downy Woodpecker, Picoides pubescens) along with several histograms to the right of the woodpecker image.  The top three graphs (red, green, and blue) are the per-channel histograms, while the white graph below these is the overall histogram, which is just the sum of the individual channels (red, green, blue).  For the top woodpecker image, you can see that the overall histogram has most of its mass toward the right end of the x-axis (the horizontal axis).  That means that the image is very bright.  Indeed, looking at the actual woodpecker image, it does appear to be a very bright image—perhaps too bright (though that’s a matter of personal taste in this particular case).  The brightness can, of course, be reduced later in Photoshop, but it’s important to make sure that we’re not shooting so bright that all the details in the brightest regions of the image are lost.  According to the overall histogram, we haven’t blown the highlights, because the intensity curve in the histogram goes down to zero before reaching the right edge of the little histogram window.  The histogram has two strong peaks, and neither of them have been chopped off at either the left or right extreme.  Now look at the individual color channels.  The red and blue channels also look fine, but the green channel looks like it’s slightly clipped: as the green curve comes down from its peak, it suddenly comes to a vertical cliff.  Cliffs in histograms are to be avoided at all cost.  In this case we’ve only clipped one channel (green).
    The next two screenshots (bottom) are for a different image, which is obviously overexposed.  You can see that for this image all of the histograms, including the individual channels, are clipped, because the curves come to a vertical cliff at the right edge of the histogram window.  In the image on the right, you can see that parts of the image are blinking (rendered as solid red in the figure) on the LCD display.  This blinking is the highlight alert, which tells you immediately that you’ve clipped the highlights and need to turn down the exposure before taking any more shots of this scene.  I have my camera set up to show the image at full size (so it takes up the entire LCD screen) with highlight alerts enabled, but with the histograms not visible, because I try to shoot as bright as possible without clipping the highlights.  However, as shown in the first screenshot (top), it is possible for the highlights to be clipped in one channel (green in this case) without setting off the highlight alerts.  Just keep that in mind, if you decide to rely on the highlight alerts alone, without viewing the individual histograms.
    So, what exactly are clipped highlights?  If you think of a white bird, when seen close up the bird may have lots of details visible on it: feather shafts, scutes and scales on the legs or in the sclerotic ring around the eye, the outline of the pupil in the bird’s eye, etc.  But since it’s a white bird, if you overexpose the image, all you’ll see is a big white blob the same shape as the bird, with no detail within that field of white.  The problem is that subtle details in the white regions of the bird often consist, visually, of slightly different shades of white.  When the exposure of the image is increased beyond the clipping point, all of those different shades of white get increased in brightness until they’re all just pure white, with no differences between them.  At that point, all of the subtle shading differences are lost, and can’t be recovered even with the most potent Photoshop magic.  That’s why clipped highlights are to be avoided at all costs—most especially when the clipping occurs in the bird itself (rather than in the background portions of the scene).
    Just about all DSLRs today provide a histogram function, though not all provide a histogram for each of the three color channels, and not all of them will show highlight alerts (i.e., blinking) when the image is shown at full size.  Being able to view highlight alerts on the full-sized image is very useful, because it allows you to instantly check for both exposure problems and focus problems in the shot you just took.  Also note that the sensitivity of the highlight detection function may differ between cameras.  I’ve found that on my Canon 1D Mark III, the highlight alerts are slightly less sensitive than those in Photoshop’s RAW converter, so that I may think (when I’m in the field) that a photo is perfectly exposed, but then find when I get it onto the computer that Photoshop’s highlight alerts are indicating some slight clipping.  However, in these cases I usually find that these “clipped” highlights can be recovered during the RAW conversion process, so in fact my camera’s highlight alerts appear to have just the right sensitivity for helping me avoid irrecoverable clipping.  Other cameras may have higher or lower sensitivity in their highlight alerts, which could make it harder to expose properly without clipping.
    For bird photography, I find the histograms to be relatively unimportant in the field.  The problem is that the histogram shown on the camera’s LCD is for the entire scene, whereas in challenging lighting conditions I’m often forced to overexpose the background in order to properly expose the bird.  Thus, I often only care about the highlight alerts that show up in the bird itself, and largely ignore those in the background of the scene (especially since I typically try to get the background to be as out-of-focus as possible, so as to make the bird stand out more).  Because different birds at different distances will occupy different amounts of space in the image, it’s simply impractical to use the histogram to judge the exposure of the bird, since the histogram may be strongly affected by the background around the bird.  The highlight alerts, on the other hand, give instant notification of any overexposure affecting the bird itself.  The only problem with this is the possibility for clipping in individual channels, which many cameras won’t show in their highlight alerts.  Hopefully, manufacturers will fix this shortcoming in future bodies.

2.7.5 Weatherproofing

Most nature photography is, of course, done outside.  And outside things have the potential to get wet.  That’s not good for electronic devices, since water conducts electricity and can therefore cause short-circuiting that can fry electronic components.  Pro-grade camera bodies have long been known to better resist the damaging effects of rain and snow than non-pro bodies, and that remains largely true today, despite claims by some manufacturers that they have improved the weatherproofing on their newer pro-sumer bodies. 
    Weather-proofing a camera consists of installing rubber seals and gaskets around all openings and moving parts, such as around the edges of the door to the battery compartment, or around buttons and dials.  These rubber gaskets keep out the stray water droplets that may fall on the camera, but they have little or no capacity (even on the pro bodies) to resist actual water pressure.  That means that they can withstand light rain, but if you drop the unit into a lake or river, the water pressure will almost certainly overcome the resistance of the sealings and cause water to reach the innards of your camera/lens and cause damage.
    Exactly how much damage can result from being submerged in water can depend on the depth of the water, how much of the unit was submerged, whether the unit had any weather sealing at all, and whether the power was turned on at the time.  The following are some anecdotal stories related to the issue of water damage.  I once had a cheap, 300mm zoom lens roll into a pond while I was shooting some ducks with a different lens attached to my camera.  Because no camera was attached to the lens that rolled into the water, there was no electricity, and hence no possibility for frying of electrical circuits.  The lens at first failed to autofocus after I had dried off the outside of the lens and attached it to my camera (a very foolish thing to do, by the way), but after a week or so the insides had apparently dried off as well, and it then continued to work fine, taking sharp images just as before.  It’s possible that over time it may eventually stop working due to rusting of internal components.  I have heard of birders dropping their huge, 500mm or 600mm lenses into a lake and finding that even the manufacturer was unable to repair them—a total loss.  During a recent trip to Florida, as I was photographing herons while lying on on my belly on the beach, I accidentally let my 600mm lens and pro-model camera flop over into about three inches of water.  I quickly retrieved the rig and dried it off, and it has continued to work just fine ever since.
    In terms of precipitation, I’ve used my pro body in light rain with no problems, and I’ve heard of people using them in a torrential downpour with no problems, though I personally always use an umbrella for anything more than a fine mist or drizzle.  Note that some manufacturers, like Canon, only claim their pro bodies to be fully “weatherproof” (not waterproof) when used with a pro lens, since the rubber gasket protecting the lens mount is located on the lens rather than the camera (in the case of Canon's pro lenses).

2.7.6 Viewfinder Features

In addition to the AF points and the framed scene you’re photographing, most DLSRs also display some digital information within the viewfinder, which can be useful during intense shooting, for rapid adjustment of camera settings without having to take your eye away from the viewfinder.  For novices, this isn’t as much of a concern as for more advanced bird photographers, but it is worth noting that some cameras provide more information in the viewfinder than others.  Being able to see the ISO, shutter speed, and aperture all in the viewfinder, all at the same time, is definitely very helpful.  Also useful are the exposure meter and exposure compensation setting, which are illustrated in Figure 2.7.5 and discussed in Chapter 6.  Other information, such as whether you’re shooting in RAW vs. JPEG, or whether the external flash unit is turned on, can very rarely be useful for tipping you off that you’ve somehow messed up some major setting on your camera, but these are in my opinion not essential.



Another potential difference between viewfinders is the fraction of the image frame that is actually visible.  In most pro-sumer bodies, the image that you see through the viewfinder is only a portion—perhaps 80% or so—of the image that will be captured by the camera’s sensor.  This is more of a concern for landscape photographers than bird photographers, since in most situations the bird only takes up a small portion of the image frame.

2.7.7 Memory Cards

Photos obviously need to be written onto some type of removable media in order for you to most conveniently get the photos out of the camera, and to make room for more images after you’ve filled up the current “roll”.  Although a number of different types of memory card are in use today, most DSLRs use either Compact Flash (CF) or Secure Digital (SD).  Some even use both.  In Figure 2.7.6 (below) you can see that the CF cards (left) are much larger than the SD cards (right).  At the far right, you can see the card slots on the Canon 1D Mark III camera, which supports both CF and SD. 




In this particular camera model you can use either memory card or both simultaneously.  The camera can be configured to first write all images to the one card and then to switch to the other card when the first card is full (“switchover” mode).  Or it can be configured to write all images to both cards simultaneously, so that one card acts as a backup for the other.  Or you can even set it up to write RAW images to one card and JPEG’s to the other.  Non-pro bodies typically don’t offer so many options, and in fact most don’t have dual card slots like the pro bodies.  Dual slots definitely offer greater flexibility, as well as increased memory capacity (when used in the “switchover” mode described above).  There is little agreement among users, however, on whether it’s best to have a dual configuration with two different types of cards (i.e., CF and SD) or to have a dual configuration featuring two of the same type of card. 
    In terms of CF versus SD, there is little agreement among users as to which is fundamentally better.  I like CF because it seems more durable to me than the “flimsy” SD cards, though I like the mechanical write-protect switch that most SD cards have.  While some CF cards are faster than some SD cards (in terms of how fast the camera can write images onto the card, thereby clearing the buffer to accept more shots), there are also some SD cards that are faster than some CF cards.  More expensive cards of either type are typically (though not always) faster than cheaper cards of the same type, though some cards rely on special data transfer protocols, such as UDMA, for their speed, and not all cameras can take advantage of that speed, either because the camera doesn’t support the required protocol, or because the camera’s processor is slower than the card’s write speed.  Due to the variety of different CF and SD cards available (i.e., different capacities and different high-speed protocols), it’s best to find out for certain which capacities and transfer protocols are supported by your camera before buying an expensive card, and likewise it’s worth looking into these issues when choosing a camera model.
   

2.7.8 Built-in Flash

Most consumer and pro-sumer camera models have a built-in flash, while also accepting an optional external flash unit.  For bird photography I’ve virtually never used the built-in flash on my Canon 30D’s and 40D’s, due to their lack of power, and indeed, Canon’s pro bodies don’t even have a built-in flash.  In Figure 2.7.7 you can see the difference between the built-in flash unit on the 30D and the external Speedlight 580EX II flash unit attached to the same camera.






The external flash is much larger and bulkier, but it provides much more power, and I consider it essential for most forms of bird photography (except for birds further away than about 200 feet).  As we’ll discuss in section 4.3.2, even these external flash units don’t provide enough flash power in many situations, necessitating the use of a fresnel flash extender to magnify the light from the flash.

2.7.9 Flash Sync Speed

Although we won’t be discussing the details of how flash works until Chapter 7, it may be important when shopping for a birding camera to take into consideration a prospective camera’s flash sync speed, or maximum sync speed.  This is the fastest shutter speed that the camera can use when the flash unit is enabled (unless high-speed sync is being used—we’ll discuss high-speed sync in section 7.3).  Most cameras have a maximum sync speed of 1/200 sec to 1/300 sec.  Obviously, since birds tend to move fast, the faster shutter speeds are generally better for freezing motion.  Unfortunately, even 1/300 sec is typically too slow for shooting birds in flight, and for this reason I either use high-speed sync, which gets around the maximum sync speed limitation of the camera but unfortunately reduces flash power, or I turn off the flash altogether when shooting very fast motion.  (One exception to the latter rule is when shooting hummingbirds, in which case a slow shutter speed with a fast flash and a dark background can tyipcally freeze the bird better than even shutter speeds of 1/1000 sec or 1/4000 sec or more—hummingbird photography is discussed in section 7.6).  Thus, while a max sync speed of 1/300 sec is preferable to 1/200 sec, in my opinion this isn’t the most important aspect to look at when choosing a camera.

2.7.10 In-camera Processing

In their never-ending quest to increase sales and beat the competition, camera manufacturers have begun adding various image-processing capabilities to their cameras.  In my opinion, these are largely useless for serious bird photography, and I never use any of them.  However, if you don’t have time to postprocess your images on the computer, you might possibly benefit from some in-camera processing.  Just don’t expect the automatic in-camera processing to come anywhere near to what can be done manually in Photoshop. 
    The most tempting of the in-camera processing options I’ve seen, for bird photography, are noise reduction, highlight tone priority, and red-eye removal.  I’ve never tried red-eye removal in-camera, though I doubt it works well, with any consistency, for birds.  The noise reduction option is potentially useful if you’re shooting at high ISO settings, though as discussed in section 2.5, aggressive noise reduction is likely to remove some detail when it reduces the noise.  Some of the more recent models allow you to specify different levels of “aggressiveness” for the noise reduction function.  Finally, Canon’s “highlight tone priority” (other manufacturers may offer a similar function under a different name) is a “trick” whereby the camera shoots at a lower ISO setting than the one you’ve requested, and then increases the brightness post-capture to simulate the brightness that you would have got with your requested ISO setting, but with an added filter that detects blown highlights and tries to recover them during the artificial brightening process.  Another popular type of in-camera processing is white balance adjustment, which is targeted largely at those who shoot indoors, in artificial light, though this could conceivably be useful during early morning or late afternoon shooting outdoors to reduce the yellow cast produced by low-angle sunlight.
    Just to reiterate, my recommendation is to turn off all such in-camera image-processing functions, since the same processing can be done later in Photoshop, where you’ll have far more control over the final result.  Also, some of these in-camera processing functions reduce the camera’s frame-capture rate or effective buffer size, so that for continuous shooting (“spray-and-pray”) of intense action scenes, enabling the in-camera processing can result in missed shots as you wait for the buffer to clear.

2.7.11 Integrated Sensor Cleaning

One type of “noise” that we didn’t talk about in earlier sections is the type that results from particles of dirt that have become stuck to your imaging sensor.  These particles show up as large black spots in your image.  In the next section we’ll discuss methods for removing dirt from the sesor using air or chemicals.  Many of today’s DSLRs now incorporate a mechanism which tries to remove the dirt automatically, so that you don’t have to use the chemicals as often.  These “integrated sensor cleaning” functions generally work by vibrating the imaging sensor while electrically charging a dust collector inside the mirror box.  The vibrations in the sensor are intended to jar the particles free, while the charged collector is intended to attract those particles away from the sensor. 





Different models appear to differ in their ability to remove sensor dust in this way, and manual cleaning (i.e., using blown air or wiping with chemicals) seems to still be necessary from time to time for the current generation of DSLR’s.  As manufacturers continue to innovate, these automatic dust-removal systems may continue to improve.  The newest designs employ special materials in the manufacture of the sensor’s protective screen (the anti-aliasing filter—see section 11.3) that reportedly attract less dust due to their reduced potential for holding minute eletrical charges.  Only time will tell how well these various attempts fare.

2.7.12 Plastic Versus Metal

Though it’s not one of the most important issues when choosing a camera, the ruggedness of the external body can be an issue if you shoot in sandy conditions, such as when shooting shorebirds at eye-level—i.e., lying on your belly in the sand.  In these situations, sand abrasion can cause significant damage to your camera.  Whereas consumer and pro-sumer bodies are typically made of plastic, the pro models are often made of metal, which resists abrasion much better.  I’ve also heard of plastic camera bodies getting melted when sunlight inadvertently passes through a fresnel flash extender and is focused to a point on the camera body.  I’ve actually seen this happen (sunlight getting focused to a point on my camera by the flash extender) though since the camera I was using had a metal exterior, no damage was done.  However, my flash unit, which is made of plastic, has sustained damage from focused sunlight in this way.

2.7.13 Expanded ISO

Just like highlight tone priority, expanded ISO is another “software trick” used by the camera to simulate ISO settings not natively supported by the camera’s imaging circuitry.  These expanded ISO settings are usually below 100 or above 1600.  In these cases, the camera uses a native ISO setting close to the one you’ve requested, and then artificially increases or decreases the brightness of the image, via in-camera software, to simulate the expected exposure from the requested ISO.  Recall, however, that even “native” ISO amplification is itself somewhat of a “trick” (as we described in section 2.5), in that it involves the boosting, via hardware circuitry, of the signal and any attendant noise, after the image has been captured by the actual photosites on the sensor (i.e., it doesn’t actually change the sensitivity of silicon atoms to light, despite popular belief).  Thus, the artificial nature of “expanded ISO” is less of a concern than the much bigger problem of high-ISO noise that results from using extreme ISO settings like 6400 or higher.  I’ve yet to see a pleasing bird image taken at such extreme ISO settings.

References for this page