Nikon D7100 test report: Imatest measurements

Measurements for our Nikon D7100 review have been carried out with the aid of Imatest. For the test method and explanation of terms, see FAQ. Conclusion and sample images are shown in our Nikon D7100 review.

Resolution

The sensor of the Nikon D7100 does not have an optical low-pass or moiré filter (OLPF). In theory this could lead to a higher resolution for the Nikon D7100 with respect to the Nikon D5200 or the D3200 Nikon, because these two cameras do have an OLPF.
To achieve the highest possible resolution from an APS-C / DX sensor you need a great lens: the pixels on a 24 megapixel DX sensor are as close together as the pixels on a 56 megapixel FX sensor camera.

In our measurements we find, both for in Lightroom developed RAW files and for in-camera jpg files, no significant differences between the Nikon D7100 (RAW: 2900 LW / PH) and the Nikon D5200 (RAW: 2800 LW / PH). The Nikon D3200 scored, with 2600 LW / PH, slightly lower, but the difference is still so small that you will not see it with the naked eye. All test images were shot at f/5.6, with a very good lens. Perhaps that the differences will be larger if we repeat the measurements with the same lens at f/4 or f/2.8, because in theory for a perfect lens the MTF50 will be higher at larger apertures. Anyway, the Nikon D7100 belongs, like the Nikon D5200, in terms of resolution to the absolute top of what is for sale at the moment.

A disadvantage of the omission of a low-pass filter could be, that you will see some moire, an interference phenomenon that is shown to the right for illustration purposes. Both in our studio images and in our images shot outdoors we encountered no clearly visible moiré. Read the article:  Do Sensors “Outresolve” Lenses? by Ruben Osuna and Efraín García on Luminous Landscape, if you want to dive deeper into this subject.

Now we come back to the 100% crop of a jpg image of a dandelion. Our measurements show that the jpg files are sharpened too much with the settings that we have chosen. Too much sharpening is identified by a “hump” at a sharp transition from black to white in the Imatest results. In the bottom left image you see the properly sharpened RAW file. The bottom right image shows a too much sharpened jpg file.
Maybe it’s a good idea to publish an article sharpening images, in a few weeks. This issue seems to become even more important in order to get the best out of your high resolution modern cameras and lenses.

Dynamic range

In theory, it can be expected that the omission of a low-pass filter results in a lower noise, and therefore a higher dynamic range. Fujifilm mentions this as an explanation for the high performance of the Fujifilm X-E1 sensor. But we did not measure a higher dynamic range for the Nikon D7100 (without filter) compared to the Nikon D5200 (with OLPF). The measured total dynamic range is 12 stops for an ISO 100 RAW file and 11 stops for a jpg file. That’s very high, but before the RAW file slightly lower than the dynamic range that we have measured. For the Nikon D5200 Also DxO-mark reports a slightly higher dynamic range before the Nikon D5200 with respect to the Nikon D7100. Beware of comparison: DxO measures in a manner other than CameraStuffReview, you can also absolute numbers of DXO and CameraStuffReview not compare directly. In practice, you will not notice the difference in dynamic range of the Nikon D7100 and Nikon D5200. Both cameras are simply very good in terms of dynamic range. The score for dynamic range of both cameras in our camera test summary is therefore equal.
The dynamic range of the Nikon D7100 is so high, that even if you shoot straight into the sun, where the automatic exposure chooses a fast shutter speed in order to limit the amount of blown out “highlights”, there is still no clipping in the shadows. Lightroom marks the overexposed areas in red (the sun) and underexposed parts blue (a branch right). The signal/noise ratio is very good at low ISO, so you can easily use the shadows/highlight function in Photoshop to lighten the image without seeing noise or banding.

Noise

Our Imatest measurements for the signal / noise ratio (in db; the higher the better) at ISO 100 and ISO 6400, are shown here. The first thing you notice is how small the differences are at 100 ISO. Two cameras with an FX sensor, the Nikon D600 and the Nikon D700, score better than the other Nikon cameras at 100 ISO. But the differences are small. At higher ISO values, larger pixels clearly result in a better signal to noise ratio, making the Nikon D600 and Nikon D800E score better than the other cameras.

Nikon sensors also are getting better over the years. The older Nikon D700 with a FX sensor is at ISO 6400 almost equaled by more recent Nikon D7100, D5200 and Nikon D3200 cameras with a DX sensor. And the Nikon D7000, which in terms of signal / noise ratio, clearly lags behind its successor, the Nikon D7100. 

Color accuracy @ daylight

The Nikon D7100 delivers RAW files and jpg files with good color rendering. The illustration shows the color errors of a 100 ISO jpg image shot in daylight. The further the ideal color (square) is removed from the color of the camera (round), the greater the color difference. Here is nothing to be desired. For the test method and explanation of terms, see FAQ

Color accuracy @ tungsten light

In artificial light you can, both for jpg and RAW files, not blindly rely on the auto white balance, as is clearly seen in the results shown for a 200 ISO jpg file. This is nothing special, this color shift is just as large at the other SLRs we’ve tested to date, such as the Nikon D3200 and Nikon D5200. When you shoot in RAW, you can easily adjust the white balance almost without loss of quality.

Conclusion and sample images are shown in our Nikon D7100 review.