Nikon D3x

Noise Reduction (6.51)

The first of our tests in this series looks at the noise levels across the four steps of noise reduction. You can see that software kicks in after ISO 400 for High, and at 800 for the other settings. The Low setting noise reduction has a minimal effect across the board.

We look at image noise results separately for the red, green, blue, yellow and luma (gray) channels. The tight grouping we found here is good, as substantial variation from a tightly grouped pattern would be visually noticeable. More on how we test noise.

In our image noise test, the D700 performed well, but not exceptionally so. It kept noise levels at a reasonable level across all ISOs, especially at ISO 100. For this test, we shoot the X-Rite ColorChecker chart, illuminated to 3000 lux, using the most accurate color mode available (in this case, neutral). The chart is photographed at every ISO setting that Nikon assigns a specific ISO value to, which excludes their Lo and Hi settings, and the images are analyzed using Imatest software to determine the amount of noise in each picture.

ISO Options

When you compare the D3x against the other cameras with noise reduction off, it sits in the middle ranks. It’s a bit lower at ISO 100 and a bit higher at ISO 1600, but overall pretty much right in the middle. The precursor model, the Nikon D3, had an official range of ISO 200 to 6400, with extended ISOs down to 100 and up to 25600.

Focus Performance

As with everything else in this camera, the focusing options offer an in-depth level of control. The D3x has an impressive 51 autofocus points, of which 1, 9, 21 or all 51 can be utilized at a time, for different size focusing areas. When shooting with the full focus-point array you can take advantage of 3D tracking, which is useful for following swiftly moving targets. The autofocus system is rated for a detection range of -1 to 19 EV at ISO 100. You can adjust the brightness of autofocus target illumination, should you need the boost, which is handy under direct bright light, where the standard illumination level might be hard to see.

The autofocus was extremely fast in good illumination, locking on to targets easily. It still focused quickly in our low light tests (20 lux of illumination) but started to really struggle after dark. Street lamps generally provided enough light for it to get a quick fix, but in areas without them, it took significantly longer. Even so, the D3x usually managed to find an appropriate focus eventually, which is good for a camera not designed for low light performance.

The focus modes on the D3x are controlled via a small switch by the lens that toggles between the three settings. On our review unit it felt like there was a non-functioning fourth setting, between continuous and manual modes, which made it difficult to quickly adjust the focus mode. This may have been a problem limited to our particular review unit, though.

The autofocus area can be set to three modes, each with its own options: Single-point, Dynamic-area, and Auto-area.

Long Exposure (11.51)

The long exposure test looks at how the camera performs at shutter speeds ranging from one to 30 seconds, and the D3x handled the challenge admirably, scoring better than any other camera except its full frame sibling, the D700. For this test, we shoot at 20 lux illumination, and test color error and noise at one, five, 10, 15 and 30 seconds, with long exposure noise reduction on and off. More on how we test long exposure.

We looked at the color error across the five shutter speeds we test at. We found that the color error stays pretty even across the test, but jumps at 30 seconds, due to over-exposure with the really long shutter speed.

With these long exposures, the noise levels stay around 0.7%, and fall off with the longer exposures. You’ll notice that long exposure noise reduction doesn’t improve the situation, which is something that we’ve seen across a large number of cameras. Long exposure noise reduction functions by taking a second exposure the same length as the first, but with the shutter closed. The theory is that you can subtract the noise of the latter from the former, but since image noise is inherently random, it isn’t an effective solution, and sometimes actually makes matters worse.

Compared with the other cameras, the D3x does better than any bar the D700. It maintains good color accuracy and low noise across the entire test.