Zacuto, a maker of filmmaking and video accessories, has once again helped produce and distribute a test of video and film camera capabilities. As with last year's tests, they've included a number of DSLRs, including the Nikon D7000. You can see the 30-minute first episode here.
Even if you're not into video, there are things to be learned from these tests, things that I've been writing about on DSLRs pretty much since the beginning. Let me just mention one of those, which is part of the test in question: the balance of highlight versus shadow detail. (Note: the video in the test was fully "graded," meaning that it has had post processing applied to it to bring color, highlights, shadows, and overall exposure curves into what you'd present in a final film or video. That process is a little different for video than what we do in post processing still images because the final viewable product is displayed a bit differently, for example, prints versus projection.)
One thing that a lot of new DSLR users don't always figure out very quickly is where "middle gray" gets rendered versus the brightest highlight or the deepest shadow detail. It's usually a shock to them when they do find out. You'll see several bar charts of how the various different cameras did in Zacuto's Shootout and it should become more clear. With Nikon DSLRs, for instance, I've been generally saying that they record about three stops above middle gray and five stops below (on a camera where I say there's eight stops of usable dynamic range). You can see that clearly in Zacuto's graph of the D7000 range, which shows a disproportionate capture below the mid-tones compared to above.
Not all cameras are the same. The Fujifilm S5 Pro, for example, was very balanced in the above/below realm, mostly due to its unique dual photosite relationship that recorded more highlight information. (Another important thing to note: the DSLRs in this test run the captured data through their imaging ASIC to produce video, much like producing JPEGs. Some of the video cameras in the test, such as the RED One, can output raw data to be converted later. That difference is important. When you rely upon the camera to allocate where the raw data goes into the pixel data [video output and JPEG] you get the camera maker's interpretation of how that data should be converted into highlight and shadow detail. If you output raw data, you can make a decision on where middle tones are and how the rest of the data is curved around it.)
But again, the important thing to note from this test is the way the D7000 distributes its data in normal recording (video and JPEG): less above mid-tones, more below. In general, all DSLRs have tended to be a bit like this when recording video, and it shows that they're coming from a still world bias (we tend to like brighter mid-tones in stills than their actual relative position in reality, and that probably has something to do with the way we display our images).
The big thing to be gained from looking at tests like Zacuto's is simple, and more than one of the cinematographers alludes to that in their comments on seeing the test footage: know your tool. As was said several times during the test, hand those cameras to a cinematographer making a film and he'll make them all look pretty much the same. That isn't the goal of Zacuto's tests, nor should it be the goal of your testing or your analysis of other people's tests of still camera equipment. What you want to know is how the camera responds to different things. Simply put, a Hollywood cinematographer would light a set differently for a D7000 than he would for the top-of-the-line Alexa. He'd light the range just below the mid-tones brighter and make sure not to push the exposure high enough so that you got channel clipping in the highlights (which produced that yellowness in the AG-AF100 highlight test, for example).
Bottom line: Know Thy Tool. The Zacuto shootouts are helpful for both still and video photographers in that respect (though obviously the video guys will tend to get more out of it).