I had a quick idea I wanted to jot down, and I haven’t found anything on it. I feel like someone out there must have thought up something similar already, or it’s already in the works at some black lab of a sensor company or something.

The idea I have is an image sensor that measures light intensity based on time-to-saturate – the time it takes for a particular photowell (representing a pixel) to saturate to its maximum capacity. The concept I’ve come up with has some interesting theoretical advantages in dynamic range over conventional photon-counting designs used today.

Imaging today – photon counting

First, a layman’s overview of how the conventional photon-counting design works in today’s sensors:

The sensor is a light sensitive device, and whenever photons come into contact with it, they are absorbed and a proportional number of photoelectrons are “knocked out” by the photon energy and collected in a photowell. From this photowell, a voltage measurement is taken, and this ultimately translates to a brightness value in the resulting image. In essence: Image brightness ∝ voltage reading ∝ electrons collected ∝ photons collected.

When taking an image, there is a set exposure duration, often referred to as a “shutter speed” in photography terms. This defines the time when a sensor is exposed to and detecting light – the exposure starts, light hits the sensors, exposure stops, and then we count the photons.

A limiting factor in this design is the photowell capacity. The number of electrons that can be stored in a well is finite, and once the photowell capacity is saturated, any additional electrons are not collected and hence the photons they correlate to are not counted. On the flipside, there is also a noise floor, where enough electrons must be gathered to produce a signal that is discernible from the random signal variation due to various forms of dark (thermal), electronics (read), and shot (photon) noise.

These two attributes lead to a problem of dynamic range – in scenes where light intensity differs greatly between the darkest and brightest areas, the sensor is simply unable to measure the full range of brightnesses and must cap measurements above and/or below a certain threshold.  This leads to the “blown highlights” and “crushed shadows” attribute often found in photos of large dynamic range scenes.

Time-to-saturate

The idea behind a time-to-saturate sensor is fairly simple. What we aim to measure in an image is light intensity – the flux of photons per time per area. The area is cancelled out of the equation by the photosite corresponding to a pixel being a certain area, so the measure we are really after is photons per time, for each pixel.

With photon counting, we fix a shutter speed (time duration), and then count the number of photons (via voltage measurement of photoelectrons) captured in that span, and use both to derive the intensity:

Intensity = photons / time = photons recorded / shutter speed

In time-to-saturate, the photon count is fixed at the capacity of the photowell, and the variable we measure is the time it takes for an individual well to saturate fully to the capacity.

Intensity = photons / time = max photon capacity / time-to-reach max-photon-capacity

How would the system work exactly? With a time-to-saturate sensor, we use as long a shutter speed as needed to fully saturate all photowells (in a conventional sensor, this is the minimum shutter speed to generate an all-white (max brightness) image). At the moment a photowell reaches capacity, it records a timestamp which will indicate how long it took to reach capacity. Once the exposure is finished, we are then left with a two-dimensional array of saturation times, rather than photon counts. Rather than recording 100k photons at one photosite, and 50k photons at a neighboring photosite where light was half as intense, the readings we get from this sensor would be along the lines of 1 millisecond time-to-saturate for the first photosite, and 2 millisecond time-to-saturate for the second, half-intensity photosite.

Key Advantages

There are two key advantages in our ability to take light intensity readings, both ultimately advancing dynamic range:

• There is virtually no limitation to the range of highlights we can capture, unlike the limitation imposed by the photowell capacity with photon-counting sensors. In our example, if there was a third photosite which had double the intensity of the first 100k photosite, and was exposed to 200k photons, it would only end up recording 100k photons since this is the capacity of the photosite, and thus both pixels would record the same white (max brightness) value, even though the 200k photosite pixel clearly represents a brighter area in the scene than the 100k photosite. A time-to-saturate measurement, by contrast, would simply produce a shorter time measurement: the 200k photosite saturates in 0.5 milliseconds, which we can compare to the 1 millisecond measurement for the first photosite and clearly conclude that the 200k photosite is twice as bright.
• Noise levels are reduced to the level of a maximally-saturated photowell. In a photon-counting sensor, any photosite that does not record a max white value by definition recorded a fewer number of photons, and thus produces a sub-optimal signal-to-noise ratio (SNR). Photon or “shot” noise has a standard deviation of the square root of the signal – thus for 100k photons we have √(100,000) = 316.2 photons of standard deviation, and a SNR of N/√(N) = √(N) = 316.2. For 50k photons, however, we have an SNR of √(50,000) = 223.6. In contrast, all photosites in a time-to-saturate sensor reach the max well capacity, and will thus all have the max SNR. This ensures that all photosites record values well above the noise floor, and additionally reduces photon noise for all pixels to the level of a maximally saturated photosite (the 100k photon, 316.2 SNR in this example).

In theory, such a sensor would have an infinite dynamic range – the brightest intensities are simply recorded as short time-to-saturate durations, and enough samples are recorded from the darkest areas to place the measurement well above the noise floor.  This would have huge implications for large dynamic range photography and imaging in general, to be able to record the entire dynamic range of a scene in a single exposure, without having to resort to processing tricks like selective shadow/highlight adjustment or high dynamic range (HDR) blending.

Potential Feasibility Issues

I’m not aware of any sources that have thought of this idea before, but if there are then there must be some large feasibility (or perhaps cost) issues that have prevented its development thus far. The few issues that I can imagine, none of which seem like dealbreakers and none of which would place performance any worse than that of photon-counting methods, in theory:

• Timing accuracy/precision of photowell saturation. While photon-counting relies on accurate and precise voltage readings from the photowells, a time-to-saturate sensor would need good accuracy and precision in recording time when a photowell reaches saturation. How precise does the time need to be, to equal the theoretical precision of today’s cameras? Taking the contemporary example of a 100k photon capacity photowell, hooked up to a sensor/imaging pipeline with a 14-bit analog-to-digital converter (found on most high-end cameras today), we would need to quantize measurable photon counts into 2^14 = 16,384 steps. 100,000 / 16,384 = ~6 photons, which is the precision we need to be able to measure time-to-saturation by. Most high-end cameras today operate with a minimum shutter speed of 1/8000 second (125 microseconds) – a 100k photowell that fully saturates in this time (this is the maximum light intensity the photon-counting sensor is able to record, under any settings) is thus 100,000 photons/125 microseconds = 800,000,000 (0.8 billion) photons / second.  Finally, we use this intensity along with our 6 photon steps to arrive at 6 photons / (0.8 billion photons/second) = 7.6 nanoseconds. This is the precision with which a time-to-saturate sensor needs to record time by. Of course, depending on the application the numbers can vary – with fewer bits per pixel, we would need less precision (an 8-bit jpeg in this example would need just ~0.5 microseconds of precision), with lower photowell capacity we would need greater precision, and with a larger minimum exposure time we would need less precision.
• To take advantage of the greater dynamic range capabilities of a time-to-saturate sensor, the exposure duration must be longer than a conventional photon-counting sensor, to capture more light. For static scenes, this is unlikely to be an issue, but for dynamic scenes (e.g. moving subjects), the exposure duration can only be stretched so far before issues such as motion blur or camera shake blur are introduced. At worst, however, the exposure can simply stop after a defined maximum exposure time – at this point any photowells which have not reached capacity simply output a voltage reading like in a conventional sensor – this reading is then used to extrapolate a time-to-saturate which can then be compared with the other photosites. In the worst case, the maximum exposure time is the same as the exposure time in a conventional photon-counting sensor, and would produce an noise level and at least the same dynamic range, if not a greater dynamic range captured in the highlights. For any exposure duration exceeding that of the conventional sensor however, noise levels will be reduced and a greater dynamic range in the shadow regions will be achieved as well.

What do you think?  Any potential pitfalls or feasibility issues I might have missed? I’m especially interested if anyone has come across a source with similar ideas before. Feel free to post links in the comments!

tl;dr: Fujifilm’s EXR sensor is extraordinary, mostly for its dynamic range. If you’re after the best non-DSLR image quality around, your choices start at the Fuji F200EXR, F70EXR, S200EXR, and end there.

Fujifilm has long been a leader in revolutionary sensor technology, particularly at the smaller scale sensor market where the majority of manufacturers have long been content pumping out traditional, vanilla CCD sensors with square grid-based Bayer Filter Arrays.

In September of 2008, announced plans for their latest sensor: the Super CCD EXR, which combines the unique color filter array (CFA) and pixel binning features of various previous sensors into a single “switchable” sensor that can be optimized in one of several areas (which are typically mutually exclusive when designing a sensor): high resolution, high dynamic range, and low noise.

High resolution

High resolution mode is the default mode, which utilizes the full set of photosites on the sensor and produces an image with a corresponding pixel on each photosite – nothing too special here, though Fuji claims the diagonal layout of photosites (as opposed to simple square grid) helps to improve resolution.

High sensitivity

A comparison of a typical Bayer CFA (left) and the CFA on Fujifilm's new EXR sensor (right)

The second mode of operation for the EXR sensor is a high-sensitivity mode which Fuji calls “Pixel Fusion Technology”, which is fancy marketspeak for pixel-binning (combining reading from adjacent pixels together to produce a better signal). With the EXR’s pair-based CFA layout, Fujifilm claims that interpolation (and thus color resolution) will be more accurate because the binned pixels are closer together (e.g. the pair blue pixels are pretty much in the same location, while they’re separated by two pixel lengths in a standard square-grid Bayer array. I don’t know that I buy this argument particularly well – it’s true that same-color pixel values will be more accurate since they’re closer, but you can’t get something for nothing: for example, the average distance from red-to-blue is going to be increased, which lowers accuracy for interpolating blue values at red pixels.

Regardless of whether their CFA and photosite layout nets them better interpolation, the key element here is the combination of pixel readings to generate a stronger signal, thus decreasing the proportion of noise. Using microlenses to patch up the fill factor (area of the sensor which is actually responsive to light) and various optimizations to lower read noise will get the high sensitivity mode EXR sensor closer to the noise level of a natively lower resolution sensor.

Wide Dynamic Range

A diagram detailing the two exposures captured by the EXR sensor when operating in large dynamic range mode

The third mode of operation for the EXR sensor uses variable photosite sensitivity to greatly extend dynamic range.  The concept is taken from some of Fuji’s older generation SuperCCD SR sensors – at a given pixel location there are in fact two photosites, one operating at a lower sensitivity and one operating at a higher sensitivity. This essentially produces two images for any particular shot, one at low sensitivity that is underexposed (capturing highlight detail, such as a bright sky), and one at high sensitivity that is overexposed (capturing shadow detail, such as a shaded building face). These images are combined, much like HDR combination is done, to create a single image which captures a much larger dynamic range than a single exposure could.

Edit: dpreview seems to report that the EXR sensor actually achieves this by operating one image at a shorter exposure time (shutter speed) than the other, rather than actually varying the sensitivity. If so this would be even better, as you’d have lower noise due to operating both sets of photosites at the same lower sensitivity.

As with pixel binning for greater sensitivity, the pixel count in the resulting image will have to halve as well.

There are some notable improvements compared to Fuji’s older SR sensors. For starters, the low and high sensitivity photosites are now of equal size, which Fuji claims will allow for a greater dynamic range extension (the SR sensors consisted of mostly “regular” photosites with tiny “low sensitivity” photosites sandwiched in). Furthermore, based on most of the image samples that can be found, the recombination method used for EXR is a bit closer to HDR blending, which doesn’t map values linearly on the same tone curve – this produces a punchier photo with better contrast that still looks natural upon viewing (due to the way human vision judges brightness in relative terms rather than absolute), even if its not quite pixel-accurate. This seems to address one of the complaints about Fuji’s older SR sensors, which provided a large dynamic range but ended up squashing it linearly to the same 12-bit RAWs or 8-bit JPEG images that all other cameras provide – the results were images that did have more highlight detail but looked “flat” and lacked contrast (because that 0.5-1 stop of highlight detail at the top is squashed into a small 250-255 pixel value range).

The EXR sensor has a big advantage over conventional HDR as well (i.e. taking multiple exposures and blending them): it captures an extended range image in a single instance, making it usable for moving subjects (HDR sports photos, yay!).

The Results

The first EXR sensor, the Fujifilm F200EXR, debuted in February 2009, and was followed up not long afterwards by the S200EXR bridge camera and the ultracompact ultrazoom F70EXR, giving us a chance to see some hard results.

Imaging-resource, as always, has perhaps the most comprehensive test bed of images, and samples from their express review of the F200EXR can be found here: http://www.imaging-resource.com/PRODS/F200EXR/F200EXRA7.HTM

Their site isn’t the most comparison-friendly however (though you can give their comparator a shot) so I’ll link to dcresource’s reviews of the F200EXR and F70EXR as well and reference these.

The first thing to note is that Fujifilm hasn’t lost a step in the noise race – in both the standard high resolution (no binning) and high sensitivity (binning, lower resolution) modes, the EXR sensor simply wipes the floor with every camera on the market this side of a full-fledged DSLR.  In the F200EXR review there is a side-by-side comparison (search for the text “Again, things look great through ISO 400″ – it’s right above this) between the 6MP high-sensitivity mode image, and a 12MP high-resolution mode image that is downsized to 6MP – essentially doing the same as pixel binning but off-camera, and digitally, rather than in-camera and analog.  The result is a slightly crisper image but noticeably more noise, though the effect isn’t dramatic.

What’s interesting is that the side-by-side comparison in the F70EXR review shows that the high-resolution mode, downsized to the same resolution as the high-sensitivity mode, actually produces better results – the same amount of noise but much crisper detail. This seems to punch a hole in the effectiveness of the EXR’s in-camera pixel-binning: if the digital data (full of rounding errors, and compressed to 8-bit jpeg) can be averaged and produce more effective results than binning the analog data (the raw readings from the sensor), then we can surmise that having more accurate data on the location of brightness values (i.e. more pixels) helps us produce more accurate images overall than having slightly more accuracy on the actual brightness values.

Further down on the F200EXR review (search “so the two would be the same (6MP) resolution” – right below this), you’ll see a direct comparison using the camera’s wide dynamic range mode. As opposed to the high sensitivity mode, here we can see real, significant benefits – highlight detail that is hopelessly blown out in the left image is very much visible in the wide dynamic range image. For those of you too lazy to navigate the admittedly long and cumbersome dcresource review pages, here’s a marketing image from Fujifilm that gives you the general idea:

Standard dynamic range (left) and wide dynamic range (right) - probably exaggerated a bit

This image gives you a general idea of the difference, though I wouldn’t take it at face value. The image on the right is probably a fair representation of what you’ll get using the wide DR mode (and you can compare this with shaded interior/sunlit exterior/sky photos you’ve probably taken), but the image on the left has way more contrast (and less DR) than any typical camera would, on its default settings at least.

The Triumph of EXR – Dynamic Range

So is Fuji’s EXR sensor a success? It depends on what you’re after. Many diehard Fujifilm Super CCD fans fell in love with the low-resolution F10/11 and F30/31 ultracompacts, both of which came in at just 6MP and absolutely wiped the floor with the competition in terms of noise performance.  And while subsequent SuperCCD iterations have maintained a clear advantage over competitors in this area (and this newest EXR sensor does to it better than its predecessor), the fact is that the high 12MP or so resolutions found on today’s sensors still compromise noise performance, despite any fancy “Pixel Fusion Technology” that Fujifilm tries to market.

The true triumph of the EXR sensor is in its dynamic range capability, and its separate pixel design (it essentially operates two sensors) works not only better than any of its competitors, but far better than even a natively lower resolution sensor.  While a larger photosite does afford more highlight headroom, halving the pixels (doubling the area) affords at most 1 stop. The EXR’s method, which essentially captures two independent exposures, is in theory capable of capturing dynamic range that is infinitely far apart, though for most scenes they’ll likely need to overlap to avoid gaps in coverage, which based on the settings allowed on current cameras is 3 stops.

According to dpreview’s dynamic range test of the F200EXR, the EXR can deliver nearly 11 EV (stops) of dynamic range.  Not only does that far outclass any compact (or even the bulky SLR-like bridges that use the same small sensors) on the market, but exceeds the dynamic range of DSLRs like the Canon 7D, Nikon D300, et. al, which all range around 8 EV for their jpegs. With a bit of tweaking with RAW files in Adobe Camera Raw, the DSLRs just about manage 10EV.

It’s simply remarkable that jpegs from a camera with a pint-sized sensor can beat out RAW images from the highest-end DSLRs, but that’s what innovative technology can do for you over hammering away with sheer physical size and trying small refinements from there (which is how most of the rest of the sensor industry has been operating for years). I can’t begin to fathom how much the Super CCD would change the landscape of photography if Fujifilm ever scaled up the sensor to DSLR size.

So over the past couple of weeks I’ve been working on a little project called Demosaic.  It’s a little online demo that interpolates image data from (simulated) raw sensor output, similar to what almost every digital camera used today has to do.

http://www.thedailynathan.com/demosaic/

The core of the problem is that most digital cameras today make use of a color filter array, or CFA, in order to differentiate intensities based on wavelength, and thereby derive color.  CFAs basically layer different color filters over each photodetector, only allowing that color light through and thereby isolating that intensity.  By far the most common type of CFA used today is the Bayer filter, although many other variants exist.

Diagram of a Bayer pattern sensor

One of the downsides of using a CFA is that each photodetector, which normally corresponds to a pixel location in the final image, only detects light intensity for one wavelength range.  So a blue-filtered detector will record exactly how much blue light was at the location, but has no idea how much red or green light there might be, for instance.

Digital cameras must therefore interpolate the missing data based on values recorded at nearby photodetectors, in a process known as “demosaicing”.  Lots of different algorithms exist, with different tradeoffs in speed, quality, and suitability for different subject matter.

I developed the Demosaic demo as a sort of off-shoot of a broader research report I did on color detection for electronic image sensors for my Electrical Engineering 119 (EE119) optical engineering course at UC Berkeley. (That’ll get posted soon as well, probably in a few installments and after I get it reviewed and edited a few times.)  Demosaic takes a test Bayer-pattern image (containing only the red, green, or blue values recorded at each location) and renders it into a final image using a variety of different algorithms.

As an example, here’s an original test image (images magnified 3x for clarity):

And here’s the raw Bayer data for the image. Every pixel is encoded with just a green, red, or blue value denoting the intensity of that color.  To get a final image, we have to guess blue and red values at every green pixel, green and blue values at every red pixel, and so on.

One of the most basic ways to interpolate the information is to simply take an average (arithmetic mean) of the surrounding pixels.  This is known as bilinear interpolatin, and produces a result like this:

Not terrible, especially as bilinear interpolation is just about the fastest algorithm out there (after nearest neighbor interpolation, which simply picks a single nearby pixel).  The demosaicing artifacts are readily noticeable, however – color artifacting at white/black transitions, and “zipper” pattern artifacting along edges.  Needless to say, this really isn’t all that acceptable for serious imaging applications.

Here’s another example with a more advanced algorithm known as Edge-sensing bilinear.  This is an adaptive algorithm, which actually analyzes the image for local spatial features (in this case, harsh intensity transitions, or “edges”) and interpolates the data based on this context.

While not perfect, you can see this algorithm gets rid of many of the artifacts present in simple non-adaptive bilinear interpoltion.  The zipper pattern artifacting has mostly disappeared, and the color artifacts at black/white transitions is also less apparent.  As a tradeoff, however, this edge-sensing algorithm comes at a greater computational cost compared to simple bilinear interpolation.

To learn more about the different algorithms and run side-by-side comparisons, check out the Demosaic site.  Also feel free to upload your own custom images to see how the algorithms perform on different types of image content.

Cover - Daily Cal Elections special issue

As the clock struck midnight and November 5th dawned, I was just headed out of the office after dropping off the rest of my photos from the massive crowd that had gathered down on the streets outside at Bancroft & Telegraph.  The photo editors Anna and Victoria were still in the office (and would be through to the morning) sorting through photos and compiling the photospread (.pdf, 7.2mb, pg 7-8)  that would appear in the next day’s paper, for which the Daily Cal photo staff had already collectively compiled a few hundred photos.

At this point I would say I was pretty content but not particularly happy with the photos I had come back with.  The shots from the viewing party were good – slideshow or photospread worthy – but I wasn’t particularly fond of any of the crowd photos from the celebration on the streets.  Sure they covered the event, but photos like this or this aren’t really going to rock anyone’s socks.

So on my way out I decided to check out the crowd once again.  By now they had vacated the Bancroft & Telegraph intersection (or perhaps moved… a contingent of police officers was now situated at the intersection) and had moved a block down to the Telegraph & Durant intersection.  The crowd wasn’t the raging horde of celebration it was, but there were still plenty of people, shouting or running about, climbing up streetlights and buildings, setting off fireworks, and the general stuff celebratory crowds do.

It happened to be that the very first subject I happened upon was this girl carrying an American flag, riding on someone’s shoulders above the crowd.  I snapped a quick shot, and here’s the straight-out-of-box result:

The original of the eventual front cover image

It proved to be one of the nicer images from the scene – at this point there wasn’t too much going on besides general celebration, and while those subjects can provide usable images:

A man sets off fireworks in the middle of the crowd celebrating President-elect Barack Obama

Members of the crowd climbed atop the Noah

Both images told the story of the scene perhaps as well as the first image, but the issue with these photos (as with many photos photojournalists take, in general) is that they are completely generic.  The photo of the man with fireworks could be from the Fourth of July.  Are the men above Noah’s Bagels some eco-protestors demanding the company only use free-range bagel trees?  Or what’s to even denote photos like this, from the earlier crowd, as celebratory, rather than the start of some massive and destructive race riot?

The girl with the flag was far from the ideal subject, but it was better than nothing.  At least we can tell she is celebrating, and the presence of the flag clues us in that the nature of the celebration is perhaps something related to nationalism or politics.  Given an appropriate publication time context (day after the U.S. election), a typical viewer can extrapolate the information given in the photo into the full story, which isn’t so clear-cut with either of the other photos.

So eventually we decided to run with that first photo I took, which was really a simple snapshot (literally the first shot I had from when I arrived on the scene).  I was long gone from the office by then, so I can’t elaborate on the exact decision process and rationale, but the thing with these special issue, featurey-type cover images is that you want them to be somewhat monolithic.  Their point is not to cover or tell the story, but to simply give the viewer some visual element that, in a very abstract sense, summarizes the story. Now, whether you want to interpret that cover image as some wishy-washy Lady Liberty looking forward to new hope/taking a sigh of relief, I’ll leave up to the individual viewer; it’s a fuzzy thing and as a photographer who doesn’t deal with the realm of art I will say for certain I had no thoughts of conveying any particular emotion or symbolism with the image.

A brief aside to the technical aspect – the lighting situation in this particular scene was garish.  As you can no doubt tell from the original photo, the white balance is off by a lot, with the image taking on an extremely strong orange cast.  Normally, you’d want to make sure to correct your white balance before shooting, by taking out some grey card or just applying a custom white balance based off the street lamp itself.  The alternative, of course, is simply taking your photos in some RAW format, which allows you to apply the exact same white balance correction in post-processing.  In this photo, the white stripes on the U.S. flag provided a perfect neutral reference, so I used the dropper/sampler tool on that area and voila, perfect white balance!

White-balanced

This looks a bit dull and lacking in contrast (note especially the lack of any true black), so I restored the black point and made a few other minor curve adjustments, and here’s the final image:

Final Processed Image

Interestingly enough, the above shot wasn’t the one I had tabbed to be “the image” for inclusion in whatever slideshow or photospread we would be doing (I had no idea we would be doing a full-page cover; I probably would have spent some more time on this shot and others if I did).  For the last series of shots I took before I left, I had a slightly more abstract vision in mind, taking the same girl and her flag (which was by far the most prominent symbol of “American politics” on display by this time of the celebration) but focusing much more on the flag in the midst of the celebration in general.  I thought these photos might work out as a better general symbol, where as the above photo hones in a lot more on this particular woman, who aside from being one of the participants in the celebration had nothing to do with the event or the story.  Of course, if you as the viewer choose to interpret this graceful-looking caucasian woman as perhaps some allusion to Lady Liberty or something similar, I suppose the symbolism still holds and the photo is abstract enough. But in my view this is somewhat of a risky proposition, as a very plausible scenario could involve some uninformed reader stumbling upon the paper, seeing a huge full-page spread of some woman in the front, and wondering what story there is concerning this particular person.  Of course, this could very well be the unwarranted paranoia of the fuzzy coming from this engineer’s worldview of photographic meaning.

For the Daily Cal’s election night coverage, I was initially assigned to cover the results watching party held at the Institute of Government Studies (IGS) Library at Moses Hall.  After I had gotten back to the office and was in the midst of sorting through and cutting down my images, we started hearing a crowd gathering outside.  So most of the people in the office went over to the Bancroft-facing windows and found this:

Part II: Berkeley Street Celebration

Election night crowd on Bancroft

Like any good photojournalist, the next thing I did was to hop in the elevator and leap down into the fray.

Working the Ultrawide

When dealing with huge crowds and cramped spaces, your biggest friend will be an ultrawide lens.  Usually you’ll want to show the vastness of the surrounding crowd, and most of the time that’s not quite possible with even a moderate wide-angle.  Compare this shot, for example, which is taken at a 12mm equivalent:

Election night crowd at 12mm

To this crop of the same photo, which represents a 24mm equivalent (which is still a fair bit wider than the standard lenses most photogs use).  In this second shot you get the sense that it’s very cramped and crowded, but that’s it.  You don’t get any idea about the vastness of the crowd, and since there’s so many people around, they take up virtually the entire frame, leaving no room to see what exactly the context or the surroundings are.

Election night crowd at 24mm

The ultrawides also offer a few more advantages.  Oftentimes in these situations you’ll be forced to simply hold the camera as high above your head as you can, and unless you’re equipped with live view and a rotatable LCD screen, there’s simply no way to see what exactly you’re getting.  The technique here is to simply spray ‘n pray – set the camera on continuous high and just keep clicking away, with the hopes that at least one of the shots will be the actual framing you want.  The ultrawide helps here by capturing more of the scene, thereby giving greater latitude in cropping down to the framing you actually want.  Ultrawides are also a lot less susceptible to shake, which is useful for these low-light situations (although this advantage disappears if you start cropping in).

Attaining Focus

Before I get to my next point, here is a smattering of other crowd shots I took earlier on when I first got down to the street and started making my way through the crowd.  Needless to say, these are the ones I didn’t even bother turning in to the editor.

The biggest issue with all of these images is that none of them have any particular subject that captures the viewer’s attention.  All the viewer sees in these images is a huge collection of people.  Beyond the existence of this very large mob, there is really very little information that these images reveal – why are all these people gathered here?  what is the sentiment of the crowd – are they celebratory or riled in anger?  Most of these images feature a fairly bland and random crowd (no real consensus of ethnicity or age or gender or dress style) and are taken far away enough that the facial expressions of the individuals aren’t easily noticeable, thus providing us no context whatsoever.  If you weren’t viewing them here in the context of a photo blog about election night coverage, you wouldn’t be able to tell this crowd from a riotous mob taking to the streets, an organized political protest on the march, or perhaps even the sold-out crowd at some outdoor street concert?

From a journalistic standpoint, the last image here is the closest to what we want, even though aesthetically it may be the worst of the bunch.  Nevertheless, this photo prominently features (somewhat) the faces of some members of the crowd up front; we can clearly see that they’re cheering and having a good time, which tells us this is a celebration of some sort.  Furthermore the flag being held up in the background tells us that there is some element of politics or perhaps patriotism involved in this gathering.  If the average viewer can put two and two together (perhaps in conjunction with the fact that this is running on the day after the election, near some headline related to the election), it becomes obvious that this is a crowd that is celebrating the election results, which isn’t something that’s very obvious in any of the other images.

Of course, this image isn’t perfect either.  It’s not very interesting aesthetically; there’s not quite enough focus on the individuals up front (though this is a delicate balance between also showing enough context (i.e. the massive crowd behind them)); and since the background is basically pitch black, we don’t have a clue where the heck we are.

So as I was taking and reviewing my shots in the field, surrounded by this crowd, I began to realize how fruitless my current approach of spray ‘n pray to get a good variety of crowd shots was – no matter where you’re pointing the camera, the endless sea of nondescript people will look the same everywhere, which doesn’t provide much variety at all.  My focus turned to picking out elements that would stick out from the crowd and provide a good focus for the viewer.

I found my subject in the form of a parked police car at the Bancroft and Telegraph intersection.  The crowd had essentially encircled it, but there really wasn’t much going on – the officer was mostly just standing there by his car and occasionally stopping the overzealous individual from hopping on it.  Other than that there was really very little interaction – the police car simply served as an interesting visual element to break up the monotonous sea of people.  Here are the two images I ended up bringing back (take a guess at which one ended up printed on the front page?)

Election night crowd - girl and flag

Election night crowd - police car

The first shot here has basically all the elements we’re looking for.  The aesthetic is there – we have some clear subjects in the cheering girl, the flag, and the police car to draw the viewer’s attention and tell us about the nature of the people here and what they’re doing.  We also still see plenty of the vast crowd, and the ultrawide angle gives us a view down Bancroft and Telegraph, which provides the scene-setting context.  Everything about the event (who was there, what they were doing, where it happened) can pretty much be told by this photo.

The second photo actually isn’t all that much different from the first few photos.  Sure there’s a waving flag (although it’s not really as prominent here as in previous ones), but otherwise a fairly nondescript crowd for which we have no idea what they’re doing here.  The image does have an aesthetic draw, which is the police car, and the officer talking with some members of the crowd, but in the grand scheme of this celebration, this basically amounts to a photographic non sequitur.  There wasn’t anything special about the car or the officer – they didn’t participate or interact with the crowd; they were just there, and showing the viewer that a police officer was at the scene doesn’t tell them anything at all about the event (it may even be misleading, depending on how deeply a viewer reads into it).

It turns out the second image was chosen anyway, which I sort of chalk up to a case of aesthetics winning out over storytelling.  Sometimes that happens, but in any case I think this is a pretty good example demonstrating the differences between covering the event as a regular photog (looking for the best-looking image) and covering the event as a photojournalist (looking for the most comprehensively descriptive image, usually).  Sometimes you’ll have to sacrifice aesthetics in order to tell a more accurate or more complete story.  And sometimes the image with better aesthetics will run anyway, if it tells the story “good enough” (which I don’t necessarily disagree with – readers will get bored if you simply run the safe, descriptive image every time).

Members of the public watch a broadcast of Senator Barack Obama's victory speech at the IGS Library in Moses Hall

Part 1: Results Viewing at Moses Hall

I had classes throughout the day on Tuesday, so the first event I caught was the viewing party for members of the public (mostly Cal Dems) at the Institute of Government Studies (IGS) at Moses Hall.  There was a large screen set up inside the IGS library, where a crowd of mostly tepid adults were watching.  I stayed here for a few minutes, but it quickly became apparent that this was probably the least energetic group of election results viewers in the entire city – the entire crowd sat throughout the broadcast in absolutely silence, without a single cheer or handclap as each of the states in the presidential race were called.  Meanwhile, the group of students watching under an outdoor tent set up in the courtyard between Moses and Stephens Hall were letting out whoops and hollers that could be heard through the windows, so I decided to head out there in hopes of catching a little more enthusiasm.

Usually, outdoor streetlights are bad enough, since they’re usually old and emit a heavily off-white yellow color, and aren’t bright enough to light anything adequately.  Making matters worse, the organizers had pitched up a tent which blocked off what little light the streetlamps gave, leaving the faint projector light reflected off a screen as the predominant source of illumination in the room.

Cal Dems viewing tent outside Moses Hall

Even throwing aside all the issues with blurry long exposures and noisy high-ISOs, your typical “natural light” photo here would look atrocious due to the quality of light given.  Issue 1 is the terrible contrast between the fairly neutral white light emitted by the projector, and the distinctly yellow glow of the rest of the scene that was illuminated by the surrounding street lamps.  With a single light source type emitting a certain color of light, color-balancing a RAW file will do the trick, but this is impossible with multiple light sources.  In this case, we’re either stuck with a photo like this, or a photo balanced for the surrounding yellow areas, which would leave the projector screen and the subjects next to it unnaturally blue.

Issue 2 is the large dynamic range in this image, which goes from a detail-less bright white screen to detail-less blacks in the shadow areas.

Given these issues, I chose a spot on the left side of the projector screen, focusing mostly on the few people sitting in the front row.

Front row, from standing chest level height

Taking the projector screen out of the photo, the worst of the dynamic range issues disappear.  We still have problems with crushed shadows, but no highlights are “blown out”, and the most important elements – people’s faces, or perhaps the flags – are correctly exposed and perfectly visible.

This photo was taken at about chest level, which is still high up enough to see most of the background.  While background elements are usually nice for context, in this case it’s somewhat hurting the photo because the back is a lot darker and less detailed (on account of our primary light source, the projector screen, illuminating the front row only). The background is also mostly illuminated by light from the surrounding yellow lamps, strongly clashing with the neutral white light from the projector screen.

To solve this, I ended up taking most photos while sitting/kneeling to achieve a lower vantage point.  This gives a slight perspective difference and makes your subjects appear bigger, but more importantly removes the dim 3rd row+ crowd and most of the ugly yellow background and replaces it with the nice, fairly neutral white tent ceiling.  The effect is a cleaner photo with no ugly distractions, allowing viewers to focus attention on the front row people and their reactions/expressions.

Front row, waist-level height

While taking these shots I also noticed the light emanating from the laptop of the fellow on the left, which provided another interesting light source to take advantage of:

Laptop light source

Laptop light source

Laptop light source

After spending some time taking these natural light photos, I decided to try busting out the flash.  Typically flash doesn’t work so well in outdoor nighttime environments, since there’s no practical surface to bounce off of, and direct flash (even the sto-fen “diffused” kind) will black out any background beyond a couple of meters (think your typical flash party pictures), providing no context for whatever primary subjects are in the front.

Since we were under a tent (see first image) with a white canopy, there *was* an actual ceiling to bounce off of.  While this worked to illuminate the room, it did so too well – the whole scene looked like it was taken indoors or in daylight, with a bright-white ceiling. I’ve actually deleted all of the ceiling bounce photos, but they looked something like this, which is to say not like a dark outdoor tent whatsoever.

This flash was actually bounced behind me at a slightly upward angle. A direct ceiling bounce looked far worse (somewhat like this, but with a lot less fill)

Obviously this wasn’t going to work, and I thought I’d be stuck with ambient lighting for the night.  Fortunately though, the tent wasn’t exactly an open-air tent – it was encased with a transparent plastic curtain.

The mostly transparent but somewhat reflective plastic curtain

Obviously a sheet of see-through plastic is not the most ideal surface for bouncing, but it was just enough to do the trick here, at about 1/4 flash power on my Sigma EF-500 DG Super (GN 165ft) with ISO1600 and f/2.8.  I angled the flash to fire backwards, reflecting off the back curtain and illuminating the people in the front.

Diffused direct flash

Diffused direct flash

So the moral of the story here is, explore your scene and get creative with your light sources.  In low-light conditions, chances are that the lighting given is going to be crappy – there is very little you’re going to be able to do to capture the entire scene (in this case, the entire room of viewers) with sufficient quality.

Looking for isolated subjects that are particularly illuminated by a light source is a good start – in this case it was taking only the front row of students who were illuminated by the projector screen, and later it was the handful of people huddled around a laptop.  Sure, focusing on such a select group doesn’t adequately “tell the whole story”, but then again the single photo that usually runs in a paper never does.

Bringing a flash with you opens up far more options.  Pop-up or embedded flashes are useful mostly as a last-resort tool, when there’s no possible ISO/shutter/aperture combination that will give you a sufficient exposure and acceptable blur.  With an external flash that’s capable of swiveling, however, nearly anything in your immediate proximity can be turned into a light source.  It’s still a good idea to have a relatively dialed-down flash output, and a longer shutter to include more ambient light – this gives you most of the positive effects flash can offer – sufficient illumination, frozen motion – while still preserving the overall mood of the ambient light (both of the final images still look like they were taken in a dark tent, rather than a bright indoor room, like the ceiling bounce photo).

The two things you want to keep in mind when choosing a surface to be your new light “source” via bounce are direction and surface area.  Surface area is of course going to determine the level of diffusion – the larger the surface area, the softer and more diffuse light will be (generally a good thing).  This is perhaps the primary factor in producing a pleasing flash effect rather than an atrocious one.  Despite having flash come from a directly forward direction like a typical direct flash, the two photos above still appear pleasing, primarily because that light was reflected over a much larger area on the plastic curtain first.

Direction has much more to do with the “feel” of the lighting, and typically if you want to preserve a similar feel to the ambient lighting, you want the direction of the light source to match that of ambient lighting.  This is why most indoor events will see light bounced off of the ceiling, emulating ceiling lights.  And light bounced off the floor looks like campfire lighting.  And sidewall-bounced light never looks realistic.  In my case, the light was bounced from behind me, which was where light from the outside streetlamps was entering anyway.

That about wraps up the photos from the results viewing party.  If you were wondering, the last photo was the one chosen for publication, otensibly because of the expression/reactions that I had no part of, but I like to think because of good technical lighting quality.  Next post I’ll talk about the ensuing craziness in the Berkeley streets and the photos for that, including the eventual Daily Cal front cover.  If you’ve got any other questions about the shots above, or the thought process through the assignment, feel free to post it in the comments.

Clear skies in Palouse, WA at Washington State

Since the wide and mid-range setup worked just fine, I decided to stick with that, but on the long end I brought a Nikon D200 and Nikon 400mm f/2.8 non-VR lens (Nikon, yucky!).

Long range: Nikon D200 with 400 f/2.8 (35mm equivalent: 600mm f/4.2)
Midrange:  Canon 1D Mark II with 70-200 f/2.8 IS (35mm equivalent: 91-260mm f/3.6)
Wide: Canon 5D with 24-70 f/2.8 (35mm equivalent: 24-70mm f/2.8)
Ultrawide: Canon 5D with 12-24 f/4.5-5.6 (35mm equivalent: 12-24 f/4.5-5.6

Handling the 600mm Beast

Compared to the Canon 40D with 1.4x teleconverter and 70-200 f/2.8 IS I had last time, the D200 equipped with a big prime like the 400 f/2.8 was a very different kind of beast.  To start off, the setup was far more clunky – while a 70-200 and 40D can easily sling over your shoulder or around your neck, and can be handheld without a problem, the 400mm is heavy and on top of that really needs to be used with a monopod.  This is problematic in a few ways:

• Monopods need to be set up.  Sure, you can set it to a certain height, but this also somewhat restricts your vantage point.  You might want to take some shots close to the ground to give the “imposing giant” perspective when looking upward at players.  Or you may want to go up higher so that you can see “over” the offensive and defensive lines and actually get a shot of the quarterback.  Using a monopod (same issue with a tripod) requires one to extend or contract the height in order to get the vantage point you want.
• It’s hard to run with a 10 pound lens.  Football games require a lot of running around, moreso than most other sports.  While baseball or association football (soccer) might sit all the photographers into a designated press box, football allows for some modest free reign back and forth along the sidelines and behind the end zones.  This means you’ll do a fair bit of running back and forth, especially after any big plays that move the ball deep downfield, or any change of possession.  With my more moderate setup the previous week, I could just barely make my way across the field after a change of possession, and it was impossible to switchover in time after a big offensive play.  With a huge 400mm weighing me down, it was pretty much guaranteed that I couldn’t make my way across the field without missing at least half of the next play, and that was with a little extra hustle and sprinting than I’d normally put in.
• The 400mm can’t be dropped.  One of the facts of sports photography is that no one lens will really cover it all (well, if you’re using quality primes/short zooms rather than cheap ultrazooms).  Inevitably, you’ll need to switch from your long lens to your short, especially on a big play that breaks from say a far-away quarterback to a wide receiver who’s racing down your end of the sideline.  Because my long-range was a prime (unlike the 70-200 zoom I had last time), the need to switch lenses on the fly was even more critical.  Unfortunately, the weight of the 400 and its attached monopod means I couldn’t just toss it on the ground or flip it over my shoulder like I could with the 70-200 setup.  I eventually learned to just sit the 400 gently on my lap when I needed to switch, but this solution resulted in a few missed frames and series, so it wasn’t ideal.

Beyond physical lens-handling issue, actually shooting with an effective 600mm angle of view wasn’t a walk in the park either.  Obviously, the angle of view you’ll get is extremely narrow – just 4.1 angular degrees across the diagonal.  Shooting from the sidelines, you’re typically relegated to focusing on just a single player, which can produce some cool shots but most of the time leaves you without any context whatsoever:

Cal quarterback Kevin Riley... uh... holds the ball? Not really sure what's going on in this photo.

Another issue with a 600mm equivalent is having far too narrow an angle of view – there simply isn’t much margin for error at all with regards to framing, which means a lot of cropped-off appendages and helmets.

Washington State's Chantz Staden can't quite get his feet and legs all in the frame.

In addition, the “viewfiner vision” with such a narrow angle of view means you’re primarily locked onto a single player and have no reliable way of anticipating where other players are or where they might interact, and the result is many frames of funky composition focused on a single player.

Christopher Ivory faces unknown out-of-frame obstacles

Cal quarterback Kevin Riley hands the ball off to someone awkwardly out of frame

To be fair, this was my first real experience using the Nikon 400, and I got much better as the game progressed, so this is certainly a learnable skill.  One of the workarounds, for instance, is to shoot from the endzone rather than the sideline.  This applies even when the line of scrimmage is way off on the other side of the field – the 600mm equivalent will reach far enough, but the cross-field distance now gives you more flexible working room.  You don’t get the same close-up effect, but the results are still very good (and you’ll get a lot more keepers):

Cal cornerback Syd'Quan Thompson zeroes in on Wazzu receiver Brandon Gibson

All in all, the D200 and 400mm setup was a pain to use – its weight leaves you more worn out in the latter stages of the game, takes time to set up and takes time more to switch over to another zoom range, all of which result in missed shots and a more rushed pace while shooting.

The extremely narrow angle of view, while delivering fantastic close-ups, also lowers the keeper rate significantly.  There were dozens of shots like those above which would have looked fantastic, save for the small framing errors that render them effectively useless.

Despite all this, there’s no denying that when it gets it right, shots from the D200+400 setup blew away anything I got with the 1dm2+70-200 midrange kit:

1D Mark II, 70-200: Cal defenders team up to bring down Washington State RB Christopher Ivory

Nikon D200, 400: Cal defenders team up to bring down a Washington State player

1D Mark II, 70-200: Cal quarterback Nate Longshore gets ready to hike the ball.

Nikon D200, 400: Washington State quarterback Gary Rogers talks to the sideline.

By the numbers

So numbers-wise, how did the game pan out by equipment?

324 total shots
Nikon D200 with 400 f/2.8: 187 shots (57.7%)
Canon 1D Mark II with 70-200 f/2.8 IS: 96 shots (29.6%)
Canon 5D with 24-70 f/2.8: 26 shots (8.0%)
Canon 5D with 12-24 f/4.5-5.6: 15 shots (4.6%) 

Wowsers – despite all of the troubles I had with the 400 and all the shots I threw away, a majority of the “good” shots came from the long-range setup.  This is in huge contrast to the 40% long-range proportion I got in my first game using the 40D and 1.4x+70-200 f/2.8 as my long-range lens.

The story isn’t dramatically different when taking at the selection of “best” shots that I included into my gallery for the game.

47 total shots
Nikon D200 with 400 f/2.8: 23 shots (48.9%)
Canon 1D Mark II with 70-200 f/2.8 IS: 16 shots (34.0%)
Canon 5D with 24-70 f/2.8: 6 shots (12.8%)
Canon 5D with 12-24 f/4.5-5.6: 2 shots (4.3%)

The long-range setup is still favored by a wide margin, especially compared to the long-range results from the week 1 game (just 22.2% of the gallery shots).

So after two games, my general conclusion is that the focal range/working distance tradeoff doesn’t matter too much – with a midrange 70-200ish zoom, you’ll be up-close at the sideline.  With a huge 600mm equivalent, you’ll simply sit further back and shoot from the corners or back of the endzone, rather than the sideline.

The key here then is simply the depth of field and quality of the out-of-focus regions.  In the first game, the midrange 91-260mm f/3.6 setup simply gave a shallower depth of field and better isolative effect than the long-range 157-448mm f/6.4. (If you want to boil it down to equivalencies, they’d be the same at 260mm f/3.6 vs. 448mm f/6.4, but at any focal length below the maximum 448mm, the midrange would have the advantage.  For example, at 250mm you’d still be getting f/3.6 with the midrange, while you’d be stuck with f/6.4 on the long-range).

In the second game, of course, the 600mm f/4.2 long-range delivered far higher bokeh quality in most long-range situations, since most of the midrange lens’ shots would have to be cropped in further, quickly multiplying the f/3.6 aperture (if you wanted to crop down to a 600mm equivalent, the equivalent aperture would be f/8.3).

Zach Follett returns a fumble for a touchdown. This image was initially taken at 260mm f/3.6 equivalent, but after cropping becomes a 407mm, f/5.6 equivalent

Of course in close-range situations where the 90-260mm could provide a close-up without any further cropping, the f/3.6 aperture would actually edge the f/4.2 slightly, which is why the midrange is still capable of delivering plenty of shots that look equally as good from a technical standpoint:

Washington State halfback Christopher Ivory looks for a hole

California quarterback Kevin Riley jogs back to the sideline

The Midrange vs. Longrange Tradeoff

The problem with using the midrange lens and achieving high-quality isolating shots, of course, is getting close enough to the subject in the first place.  With the huge expanse of a football field, chances are that the major action or the player you’re trying to focus on won’t be close enough for a 260mm equivalent lens, so you’ll have to crop down.

Of course, there will be situations where the subject is too close for a long-range 600mm equivalent, which is no good either.  However, you can always back up or move further downfield to give you more working room.  With a midrange lens that doesn’t reach far enough, there’s no way to get any closer to the subject than your spot on the sideline. (This is the exact opposite problem faced by most photojournalists, landscape photogs, or travelers – oftentimes in these cases you’ll be able to get as close as you want, but you’ll often find situations with very cramped working room, so a wide-angle there is actually the most versatile.)

Of course, the reduced quality from wider shots can also be a blessing in disguise.  While close-ups are the only way to maximize the shallow depth of field isolative quality, it also severely limits your margin for framing error, as well as awareness of outside-of-frame events, as seen in the 400 shots above.  Having a wider lens gives you room to crop the perfect framing afterwards, and also catch action that suddenly moves up close to you.

So the tradeoff again boils down to that age-old equipment debate: quality or versatility?  In most cases a large aperture long-range lens like the 400mm f/2.8 will deliver higher quality, except for rare cases where the subjects come close enough that a similarly large aperture 70-200 f/2.8 can match it.  On the other hand, the more versatile midrange lens will always afford you greater versatility and margin for error (with cropping, its framing possibilities are a superset of a longer-range lens).

As a photojournalist, and entering my first game and a half with that mindset, my initial preference would be for the midrange lens.  True, quality won’t be as consistently good, but the journalist’s mindset is to simply cover every facet of the game, which means getting a usable shot from every play.  For the photojournalist, it’s more important to get the shot of the crucial game-changing interception, than it is to come back with five generic portrait-like images that could appear on a Sports Illustrated cover.  With this kind of approach, you’ll come back with a huge volume of photos, and chances are if the sports writer wants an accompany photo of specific play or player, you’ve got it.  Job accomplished.

Towards the end of this second game, however, and partially with the assurance that I had already bagged plenty of usable images during the first half, I began to experiment much more with the long-range and getting better, more selective shots.  Yes, it meant missing probably dozens of shots due to bad framing, or players racing too close to my end of the field, but the few shots you do end up with using the long-range lens are most likely the ones that will end up on a magazine cover or in your portfolio.

While I’m still fairly a novice at this, I think the second approach is ultimately the solution to becoming a good sports photographer.  Most of the struggles, especially with regard to framing errors, are just that – errors that can be improved upon with training and experience.  As you get better, the keeper rate goes up, eventually (hopefully) to a point where you’ll be able to sufficiently cover all the major events or action despite using a more restrictive long telephoto.

The other equipment-related solution is a massively large aperture zoom lens.  Hopefully one of these days I’ll get a chance to play with a Sigma 120-300 f/2.8 or Sigma 200-500 f/2.8 for sports photography – either of these would be an absolute dream lens for sports photography.

This is part 4 of 4 in Football Photography X’s and O’s, a 4-part series of insights on shooting football.

Part 1: Equipment Analysis 1 (Michigan State game)
Part 2: Aperture-priority Exposure Technique
Part 3: Lighting Situations
Part 4: Equipment Analysis 2 (Washington State game)

Daylight

Anyhow, full daylight creates problems with extremely harsh shadows, particularly on player’s faces under helmets and such.  The problem is exacerbated if you’re shooting at an angle where the player is backlit or even severely sidelit.  For example, this might be a perfectly usable photo, but it doesn’t quite have the instant eye-catchiness of better sports photos – the entire image is really busy because the brightest-lit areas are the least detailed (field, crowd in background), and the important areas (player’s faces, bodies) are masked in shadow.

Washington State's Christopher Ivory about to collide with California linebacker Anthony Felder

Contrast this to this fully-lit photo (actually sidelighting from the right, but the player is facing that direction anyway).  Now the brightest (and instantly eye-catching) area of the image is the player’s body and face, which also happens to be the most detailed area and the focus of the image:

California running back Jahvid Best evades Michigan State tacklers

The key for daylight then (without clouds) is to get into a position where the sun is coming from behind or to the side of you (but still relatively behind, if possible).  Since most games are noon or later, the sun will tend to be on the west side where it sets, so the preference would be to shoot from the west end of the stadium (of course, you may not have this luxury, as sometimes they restrict you to the visiting team’s sideline).  Another strategy is to simply shoot from the endzone - most all football stadiums are oriented facing north-south to avoid playing directly into the sun during sunlight games, so at the very worst you’ll have a side-lit image, which often isn’t bad at all, as the above photo shows.

A closer look at more difficult shaded and nighttime lighting conditions after the jump.

Shade and Cloud Cover

As the sun begins to set, or if clouds start rolling in, the dynamic of the light changes completely.  Rather than having harsh shadows from a single point light source (the sun), under shadow conditions the entire sky becomes your (much less intense) light source, with the atmosphere reflecting light and illuminating the field.  Or with cloud cover, the entire cloud (significantly bigger than the sun in angular area terms) becomes the light source.  In either case, the lighting becomes much more diffuse, which essentially eliminates shadow problems and having to take lighting direction into consideration.

Michigan State running back Javon Ringer looks for a hole to make the quick break

I like this photo a lot, possibly moreso from a non-sports standpoint – because of the diffuse lighting you can see all the detail in the image extremely well – there’s barely anything blown out at all.  It’s exactly the attributes you’d want from a sports photo if you were a landscape or interior building photographer.

One thing you’ll notice, though, is that despite the details and how easy the image is on the eyes, it definitely lacks some of the hard contrast or “bite” that one finds in bright daylight photos, like the Jahvid Best photo above.  While it was definitely easy to process, once the entire stadium is shrouded in shade like it is here, it’s a bit harder to create the contrast that separates the main subject from say the background, since everything in the stadium is evenly lit.

Dealing with semi-lit half-shade

Another issue that crops up with shade as the sun sets is the in-between period when the sun has only partially set, thus leaving part of the stadium shaded and the other side still exposed.  One can imagine this wreaks havoc with exposure, and lighting in general… well, it does:

Bears defensive back Michael Mohamed is backlit by the shade

Autoexposure here obviously locked onto the background and tried to bias it in favor of that.  This situation is pretty bad, because the background is going to be far brighter no matter what you do, which is going to distract attention away from your primary subjects.  Of course, depending on where you’re shooting this can also work in your favor, as a shaded seating area can provide a dark backdrop that focuses attention on the players:

Defensive lineman Tyson Alualu emerges from the shadows

This isn’t really a fantastic photo, but it looks a heck of a lot better than if the background was completely exposed and distracting.  Other issues including subjects being half-lit:

Zach Follet attempts to tackle Michigan St. quarterback Brian Hoyer through the shade

Some might say this photo looks cool or interesting, and the half-lighting effect is pretty… different.  In general I’m not a huge fan because it makes the photos far less useful – you can forget about using any of these photos as a cutout because of the huge exposure difference, and it’s impossible to get the exposure just right so that you can see all of the detail – here the image is exposed for the upper bodies, and as a result you don’t see much detail or definition in their legs.  Of course exposing for their lower bodies would completely blow out the top half.  It’s a photo you might try publishing once purely because of the interesting lighting effect, but otherwise as a sports photo it can’t really be used for much.

For these half-shade situations, the key here, in initial or exposure or even post-processing, is to screw the background.  Your focus here is on the players, so you need them to be properly exposed, clipping or blown highlights in the background be damned.  Just compare these images:

    

Jahvid Best runs to the endzone in the shade

Yes, the background is completely blown out, and now it’s really bright and attracting even more attention before.  This is true, but there are several reasons why the second image is better.  Firstly, the subject now has actual, visible detail in the image, instead of being some shadowy figure that partially blended in with the ground.  Secondly, those of you who know what it actually means to clip will realize that, since the second photo really has begun to clip to white, proportionally the background is now less brighter than the subject.  And lastly, despite making the background brighter, blowing the highlights here serves to *decrease* detail in the background, rendering it less appealing than the well-exposed and very detailed image in the first image.

Now that we’ve ascertained that exposing for the player is of the utmost importance, how do we go about doing that on the field?  The answer here is partial or spot metering.  Unlike evaluative/matrix metering, partial or spot looks exclusively at the center of the image (or where ever your AF/metering point is set to) – thus it’s not affected at all by super-dark or super-bright backgrounds, and if you’re tracking the player correctly, the player will always be correctly exposed, which returns the same correct result in fully-lit situations, and the correct result you wouldn’t get otherwise in situations with a much darker or brighter background.  The potential pitfall here is the possibility of the spot meter being thrown off completely by jersey brightness, since this is where you’ll most likely point at, and pretty much all the meter will see using spot.  I would use something slightly larger than a spot meter – maybe the partial meter or at worst, the center-weighted average metering, which tends to be more consistent and balanced out if the players are wearing heavily contrasting jerseys (dark blue vs. white).

Nighttime and Stadium Lights

Once the sunlight levels really start to fall, stadium lights become your primary source of lighting.  Depending on the stadium, this might force you to start taking drastic measures to maintain decent motion-freezing exposures – start ramping up the ISO or even shooting in RAW and underexposing if need be.  Fortunately the lights at Memorial Stadium (and probably most stadiums at the college and professional level) are plenty sufficient – on a Canon 1D Mark II with 70-200 f/2.8 IS lens I was getting perfect exposures at f/2.8, 1/500s, and another Canon 40D with 70-200 f/2.8 IS + 1.4x extender setup, I simply boosted the settings to f/4.0, ISO3200 to achieve the same 1/500s shutter speeds.  With less powerful lights at high school stadiums or community fields, you’ll likely need to boost up to a higher ISO (if you don’t have a higher ISO option, you can use RAW, underexpose by one stop, and boost the exposure in post-processing afterwards, which gives you nearly the same effect), or live with a little motion blur at 1/320s or 1/250s (anything below that and most of your shots will likely turn up unusable).

With most modern DSLRs, even maxed-out noise at 3200 really isn’t bad at all, especially considering that the noise will barely show up on most of the mediums for your images (web-resolution, small prints, or newspaper-quality paper).  For example, here’s an ISO3200 image taken on a Canon 40D:

f/4.0, 1/500s, ISO3200

There’s barely any noise here – certainly nothing that makes the image unusable, and not really even enough that a non-photographer would notice.  Even if you’re really a stickler for noise, you can always run your images through a noise reduction program like Neat Image to filter out what little noise there is:

f/4.0, 1/500s, ISO3200 filtered through Neat Image

Voila, an image that has about as much noise as you’d get shooting ISO400 in daylight, which is to say no noticeable noise at all.  Though we’ve solved the noise problem most often associated with low-light photography, you’ll notice that the photo above still doesn’t have the same aesthetic appeal as any of the photos above taken in daylight or shaded lighting.  I only have a handful of photos from a quarter of a football game to base my observations on, but the lighting quality on all of them look fairly similar to this example, which is to say the lighting is uniformly bad no matter what you do.

The first problem with stadium lighting (as with any artificial lighting) is that the color temperature of the light isn’t usually even across the visual spectrum, like daylight is (daylight will emit relatively equal amounts of red thru blue light, or an overall neutral white).  Classically, tungsten lights have tended to produce more light in the red spectrum and less in the blue, while new fluorescent lights may have even weirder green-magenta shifts. Especially with some fluorescent lights that don’t even emit a continuous spectrum but simply certain specific wavelengths of light, this creates a problem where certain colored subjects aren’t as detailed or richly saturated, simply because the light source isn’t emitting the kind of wavelengths of light that they reflect.

For example, in the night photos above, notice how we have a rich and very detailed yellow, while the blue parts of the jerseys are much darker/muddier, and lack any strong blue saturation – they pretty much just look black, with a hint of blue.  This is what we might expect from a light source that reflects relatively less blue wavelength light.  Contrast that to the much richer blue tones we were getting in the daylight photos, where our light source had an even continuous spectrum of wavelengths:

Kevin Riley looks rich and saturated while immersed in lighting that evenly spans the visible spectrum

Unfortunately, there’s no real way to get around this problem – it’s just the nature of the lighting you have.

Another potential issue that crops up due to the uneven light spectrum problem is white balance – since the light emitted isn’t actually white, and if taken for face value (1 red photon + 1 blue photon + 1 green photon = white) would actually produce a very off-color image (usually orangish or yellowish with artificial lighting), the camera has to compensate and restore a correct “white balance” point.  In my case, the opposing team wore white jerseys that provided a perfect white balance reference, so most of the images turned out fine.  Even so, I’d get an off-color image every once in a while, and the problem would probably have been exacerbated if the other team was wearing a colored jersey as well:

Jahvid Best runs dodges Michigan St. linebackers but can't escape the incorrect white balance.

In most cases, some color balancing in post-processing should be able to easily fix any color cast issues (the example above would take only a very small adjustment), but to fix things in camera, simply define a custom white balance in the camera.  If you happen to have a grey card with you, put it under the stadium lights and you’re gold – otherwise a white jersey or pretty much anything neutral should work fine as well.

In conclusion…

Each of the three main lighting situations (four if you count the hybrid directly lit/shade situation) has its own characteristics, and accordingly advantages and disadvantages.

In terms of quality images, full daylight is probably the best to shoot in – if you can get the subject to be directly lit, you have a very well-lit images with strong, contrasty shadows on areas you don’t care about.  The difficulty with daylight is that you have to keep the direction of the sun in mind – if you wind up in a situation shooting the subject while its’s sidelit or backlit, you end up with harsh shadows obscuring the detailed/important elements of the image, which can all but ruin a great picture.

On the other side, a shaded lighting situation (whether it be from a setting sun or cloud cover) constitutes probably the easiest lighting situation – with diffuse light coming in from virtually every direction, shadows are all but eliminated, allowing you to shoot from anwyhere and not worry about backlighting.  While you’ll certainly get more keepers, you may never get a single shot as good as the best direct daylight shot, because the lack of shadows strips the image of the natural contrast you get from harsh shadows.

A half-lit, half-shadow situation produces interesting results – it can help the photo if you’ve got a shadow in the background, but at the same time it can also hurt the photo if you’ve got a lit area in the background.  The key here is to use a center-weighted metering mode, such as center-weighted average, partial, or spot, to ensure that the main subject is always properly exposed – having the background too dark or too bright is really a secondary concern in this case.

Toughest of all is nighttime situations, simply because the quality of light is so crummy.  Omni-directional lighting means filled-in shadows and less shadow contrast, while non-neutral light sources mean certain colors will simply not render as richly as they’re supposed to.  The one good thing about stadium lighting, however, is that you can now switch to manual exposure, since the output of the stadium lights isn’t changing, and in a decent stadium, the entire playing area should be evenly lit (possibly not the case with lower-budget fields used for high school football or other events). Using manual exposure helps by removing autoexposure issues from the equation completely – the camera should now nail the exposure every time, without regard to jersey brightness or the illumination of the crowd in the background.

This is part 3 of 4 in Football Photography X’s and O’s, a 4-part series of insights on shooting football.

Part 1: Equipment Analysis 1 (Michigan State game)
Part 2: Aperture-priority Exposure Technique
Part 3: Lighting Situations
Part 4: Equipment Analysis 2 (Washington State game)

Use the largest aperture. This goes almost without saying – you want to isolate your subjects in sports photography, and the best way to do that (given a certain camera/lens setup) is to use a wide-open aperture. This also has the advantage of letting in as much light as possible. The margin of error for focus *will* be thinner, but this really shouldn’t be an excuse or barrier to return inferior shots taken at smaller apertures just because it’s easier. Take the out-of-focus shots as they come – every ruined shot should just be an incentive to learn how to effectively track subjects and utilize your camera’s AI Servo/continuous focusing abilities.

Maintain a fast shutter speed in the shaded region. The goal here is to have a fast enough shutter speed to avoid blur, and with your aperture stuck at wide-open, you’ll do that by manipulating your ISO sensitivity. Since you’re shooting in a situation with dynamic lighting, you want to choose the ISO that will give you the necessary speed in the darkest area (i.e. in the shaded portion of the field, or when a cloud rolls by and blocks the sun). If you’re maintaining a decent enough speed in the dark areas (say 1/500s), then you’re guaranteed to get a decent speed in any other area, since it’ll be brighter (if you’re getting 1/500s in the shade, you might get 1/2000s in the sun). Does this mean you’ll be using a higher-than-necessary ISO when you’re in the brigher areas? Yes (if you’re getting 1/2000s in the sun, that means you could drop the ISO 2 stops and still get your 1/500s minimum). However, the noise is going to be a minor problem at the lower ISOs where you might deal with this half-lit, half-shaded situation (the difference between ISO100 and ISO400 is virtually indistinguishable), and in any case you’re only over-using high-ISO in the brighter area, where your noise is going to be less (due to greater amount of light) than whatever you’ve already accepted putting up with in the darker area.

The real important point here is that at all costs, you want to avoid slow shutter speeds, since blurred out pictures are completely unusable and unsalvageable, while most agencies (and any skilled photoprocessor) can put up with a relatively huge amount of noise. So take the noise hit in the brighter situations (which is not going to be that much) if it will help you get rid of blur in the darker regions (which is going to be a huge problem)

Keep track of your shutter speeds as lighting dims. Over the course of a late-afternoon to evening game, the sun is going to set and you’ll gradually see light levels drop, and concurrently, the need for longer shutter speeds and higher-ISOs to compensate. If you’ve got some sort of auto-ISO feature on your camera that helps to maintain a specified shutter speed by adjusting the ISO, use it. Otherwise you’ll have to monitor your shutter speeds as the game goes along and bump up your ISO periodically as you see the shutter speeds dip below the motion blur threshold you want.

Consider center-weighted metering with dynamic secondary elements. In most situations, the default evaluative/matrix/segmeneted metering mode on the camera will do a fantastic job of determing correct exposure.  Where these metering modes often get confused is with highly dynamic secondary elements in the image – very dark shadowed stands in the background, or very bright field in the foreground of a shadowed area.  This throws off the metering and makes the camera think the scene is darker or brighter than it really is in situations like rolling clouds or sunset, where the field (or parts of it) may rapidly become lit or unlit.  The solution for this is to use a center-weighted metering mode that will bias the exposure towards your selected subject.  This way things like a dark background or very bright foreground won’t have any effect on exposure – the camera only looks at your primary subject and determines the correct exposure for that, which is all we care about.  Of course, the potential danger in this is that an athlete’s dark or white jersey will similarly throw the camera’s metering off – the best compromise is probably to use a broader center-weighted focusing mode, such as partial metering or center-weighted average, to include more of the surrounding area and balance out extreme variations.

This is part 2 of 4 in Football Photography X’s and O’s, a 4-part series of insights on shooting football.

Part 1: Equipment Analysis 1 (Michigan State game)
Part 2: Aperture-priority Exposure Technique
Part 3: Lighting Situations
Part 4: Equipment Analysis 2 (Washington State game)

I recently had the opportunity to shoot a couple of football games for the paper I work at, The Daily Californian.  It was my first time shooting football game, and as someone who’s generally not been very good at sports photography, I was definitely a bit nervous.

Both games I shot were in pretty good light – the UC Berkeley/California vs. Michigan State game started at 5pm, so it played from daylight through to just about dusk in the 4th quarter.  The Washington State game began at 3pm, so it was pretty much daylight except for a bit of (rather nice) sunset light at the end.

Sunset at halftime at Memorial Stadium, UC Berkeley

Equipment analysis – Week 1 vs. Michigan State

The biggest difficulty with football with regards to equipment is covering action that happens over a vast expanse (over 5000 m² of field area) that can be traversed by speedy athletes in a matter of seconds.  So while you might be sitting nice and cozy with a 300mm lens that perfectly covers the action mid-way across the field, all of a sudden the quarterback can fire off a deep pass or the running back finds a hole and flies off, and you’re stuck without the ability to get the shot.

I went into my very first football (2008 Aug 30 against the Michigan State Spartans at home) with the mentality of a photojournalist – cover the event in its entirety, capturing every player, every moment, and not missing anything critical.  To that end, I decided on the following setup the first time out:

Long range: Canon 40D with 1.4x and 70-200 f/2.8 IS (35mm equivalent: 157-448mm f/6.4)
Midrange:  Canon 1D Mark II with 70-200 f/2.8 IS (35mm equivalent: 91-260mm f/3.6)
Wide: Canon 5D with 24-70 f/2.8 (35mm equivalent: 24-70mm f/2.8) 
Ultrawide: Canon 5D with 12-24 f/4.5-5.6 (35mm equivalent: 12-24mm f/4.5-5.6)  

Between the equipment I had on hand and pool equipment from The Daily Cal, this was probably the best setup I could scrounge up.  During the game I found the setup to be fairly versatile – I could cover the action on the other end of the field fairly adequately (although at an f/6.4 equivalent, the quality wasn’t fantastic), and by the end of the game I was getting a handle on the technique of dropping the long-range lens onto my shoulder and quickly slinging the midrange setup out whenever the big plays broke.  Here’s some stats on the shots I turned out from the game (after cutting down to the ones I’d actually upload to the Daily Cal):

344 total shots
Canon 1D Mark II with 70-200 f/2.8 IS: 139 shots (40.4%)
Canon 40D with 70-200 f/2.8 IS and 1.4x: 148 shots (43.0%)
Canon 5D with 24-70 f/2.8: 54 shots (15.7%)
Canon 5D with 12-24 f/4.5-5.6: 3 shots (0.9%) 

The numbers change somewhat dramatically, however, when you look at the subset of best shots I actually pulled for inclusion into my gallery for the game:

36 total shots
Canon 1D Mark II with 70-200 f/2.8 IS: 19 shots (52.8%)
Canon 40D with 70-200 f/2.8 IS and 1.4x: 8 shots (22.2%)
Canon 5D with 24-70 f/2.8: 6 shots (16.7%)
Canon 5D with 12-24 f/4.5-5.6: 1 shot (2.8%)

Taking a look at the best shots, the proportion shifts dramatically in favor of the midrange setup.  Reviewing the photos, I think the primary reason for this is simply the far shallower depth of field at an f/3.6 vs f/6.4 equivalent.  While a ton of the long-range photos like the one on the left were usable, the midrange setup with larger aperture produced a much cleaner background and isolated the subject better, which ends up being the key to differentiating usable/publishable photos from the best photos you’ll actually end up printing out or putting in your portfolio.

Kevin Riley prepares for the snap - effective 448mm, f/6.4 (left) vs. 230mm, f/3.6 (right)

Despite more of the “best” shots coming from the midrange setup, the extra reach on the long-range lens turned out to be just as versatile, considering it was used on 43% of the shots, even if they weren’t the absolute best ones.  So how to make the long range setup more effective?

The smallish aperture was one of the culprits.  There were numerous shots like these, that were *almost there* but just needed a touch shallower depth of field to seal the deal and make it a great shot:

Michigan St. QB Brian Hoyer hands off to halfback Javon Ringer

Looking over the shots afterwards, I also realized that even though the effective 448mm was long, it still wasn’t quite long enough in a lot of situations, which resulted in a bunch of mediocre-but-could’ve-been-good shots like these:

Defensive lineman Tyson Alualu goes after Michigan St. QB Brian Hoyer

Jahvid Best turns up the afterburners as he eludes a pack of Michigan St. defenders

After this first game, I was fairly happy with my midrange setup, which returned a lot of good quality shots.  The 5D and 24-70 setup wasn’t really useful for sports, but as always it worked great for covering the various other more phtojournalistic aspects of the game: 

The Golden Bears make their entrance into Memorial Stadium for the season opener against Michigan State

Zach Follett plays a joke during Darian Hagan's post-game interview

The only setup I knew I really wanted to improve was the long-range.  It needed to be longer, and depth of field needed to be way shallower.  So for the Washington State away game the next week, I would finally bring the big Nikon 400mm f/2.8 out on an assignment.

This is part 1 of 4 in Football Photography X’s and O’s, a 4-part series of insights on shooting football.

Part 1: Equipment Analysis 1 (Michigan State game)
Part 2: Aperture-priority Exposure Technique
Part 3: Lighting Situations
Part 4: Equipment Analysis 2 (Washington State game)