Review: Nikon D800 DSLR, Part 2: Usability and Image Quality

Adjusting to DSLR Focus Issues

No matter how you handle the camera, what follow focus system you plan on getting, or even what lens, if you’re moving to a DSLR from a camcorder, adjusting to limitations in focusing will be a big part of the adapting how you operate the camera. It is not simply a matter of “We'll just replace our camcorder with a DSLR and everything will suddenly look great.”

If you're coming from on-shoulder ENG cameras with manual focus, you also need to understand something about depth of field. The much smaller sensor in the ENG camera means you naturally have a depth of field measured in feet, or tens of feet. With a full-frame camera like the D800, your ability to capture a depth of field at the lower apertures that gives you the shallow DoF that you’re after means your DoF is measured in inches.

It can be mastered. Shooters do it every day. They shoot live action sports, like fishing on a boat, where space to stand is at a premium, and the fish isn’t going to wait. They nail the focus because, with enough practice, they “get it.” Just be aware that getting it takes time and practice, and that a paying gig is not the time to practice.

D800 Audio

The D800 is one of the first DSLRs to offer both a 1/8" stereo microphone input and a 1/8" stereo headphone jack. This 1-2 punch has the potential to save users from spending time or money on a separate audio recorder, and from extra steps syncing up separate audio in post. In testing, however, I found the features to be not as useful as they could be.

First of all, you can't vary the recording level or headphone audio while recording. This simple oversight makes using the D800 for live events, run-and-gun, and documentary work a lot harder than it should be. Moreover, the quality of the in-camera audio leaves much to be desired.

I recorded some audio in the camera, first with an external audio amplifier, and then without. I raised the in-camera audio levels by 5 steps at a time. Lastly, I connect that same microphone to an external audio recorder for comparison. Check out my results in the video below.

As you can hear in the audio test I performed, the noise floor in the D800 quickly becomes easily audible. It rises noticeably as you increase the audio level in the camera. Recording with a level higher than 10 yields audio that I would want to run through noise reduction software. This audio you hear first in the video above was captured with an external microphone. 

Using the internal microphone is even worse. Levels lower than 10 are usable. The hiss is there, but it’s not very obtrusive. This may be passable if you have microphones where you can crank up their output level, and the audio you record will have a music bed under it, masking the hiss. However, it’s nowhere near as clean as a dedicated external audio recorder--the last audio sample in the test.

If you’re going to use an external mixer to raise the level of audio going in to the D800, you may as well just use an external audio recorder and get the best possible audio with no hiss. If you’re pressed to use the audio capabilities in the camera (because it's so easy to plug a mic in, and to monitor the audio) the D800 does provide accurate audio monitoring via headphones and the on-screen meters so you can easily see the levels, and hear any in-camera audio issues like hiss, clothing noise, or clipping.

Video Quality Comparisons: Nikon D800 vs. Panasonic GH2

I put the Nikon D800 up against the Panasonic GH2, a little powerhouse that delivers video far above its class (Figure 2, below). The GH2 has been compared favorably to several much more expensive large sensor motion picture cameras in several camera shootouts. While the GH2 can be hacked to utilize data rates far exceeding 100 Mbitsps, the D800 has not (yet). So, to compare them on a more level playing field, I shot the same scene with both cameras, in 1080p30, shutter of 1/60, using the stock 24 Mbitps each camera offers. I used an ISO of 1250 on both cameras. The GH2 was set to Nostalgic (-1, -2, 0, -2). The Nikon was using Standard with no adjustment to sub-settings. The kit zoom lenses on both were set to around f4. The Nikon records a generic H.264 clip as a .mov file. The GH2 records an AVCHD clip as an .mts file. In the end, it's just a question of how effectively does each camera use those bits, and how much quality can be squeezed into such a paltry data rate.

Figure 2. Clash of the Titans: The Panasonic GH2 vs. Nikon D800.

This test was to shoot a fairly high-contrast shot of a person wearing a dark shirt working on a computer. The subject is also sitting in a black office chair. I exposed for the highlight of the computer screen--a typical “protect the highlights” method of shooting. I endeavored to then compare how well the footage from each camera could be graded to pull a usable image out of the blacks, and how the compression on each camera would affect the image. You can ignore the white balance because, despite being set to tungsten, the D800 still ended up with an orange cast. I later found that the fluorescent setting produced a better white balance for this shot, but the white balance does not affect the results of the test.

You can see in the ungraded footage in the video below that the chair and shirt really are nearly lost in the blacks. You might consider this hopeless with a prosumer camcorder, but today's large-sensor cameras can really reach down there and provide some usable video. Lifting the mids and blacks reveals a startling amount of detail in the D800 footage. The GH2’s high-resolution image is marred by the blocky compression artifacts. The D800’s compression looks more like soft film noise and it is quite usable deep into the blacks.

When I tried to bring down the mids to recover highlight detail in the computer screen (which was deliberately not clipped) I found they both handled it fairly similarly. Still, though, the GH2’s blockier compression could be seen in banding on the computer screen. The D800 also had some banding, but it was less visible.

The GH2 can be hacked to use a much higher data rate. This would dramatically reduce the artifacts created by the low data rate. But the point here is that the D800 produces impressive footage right out of the camera. I later realized that the Natural setting, as opposed to Standard, may have provided even more latitude in the D800 footage. But this is not a test of every setting, and sub-setting adjustment that could possibly be used. The D800 captures deep shadow detail you wouldn't think possible. It does it with a low data rate. And the D800's data rate does not visibly hamper the quality of the footage with blocky artifacts. Protect for the highlights and you can push and pull this footage to produce some amazing results.

Even when I selected the ‘Normal’ video quality--a lower data rate than "High Quality"--I did not see heavy compression artifacts in typical “corporate-type” footage. I suppose if you were to run a stress test, you could break it down, but the D800’s video quality is beautiful. I might only wish that it could offer higher-bitrate capture options like Canon models do, but what Nikon has made possible with the limited bits the D800 uses, especially in terms of keeping the blacks without compression artifacts, is amazing.