OLPF filter omission

When Nikon's reps briefed us on the D7100, they were keen to stress that this 'flagship APS-C DSLR' did away with an optical low pass filter (OLPF); a move we've seen Pentax make with the 16MP K-5 IIs, but a first for Nikon. We were of course curious to discover what potential advantages or disadvantages this offered, however, in a 24MP APS-C sensor.

In principle, removing the image-softening OLPF will result in greater resolution, albeit with a potential increase in intensity of moiré patterning. To investigate this we chose to compare the D7100 with Nikon's D5200, which also has a 24MP APS-C sensor but includes an OLPF. Based on our experience with a similar OLPF-effect comparison we conducted in our earlier Nikon D800/800E review, we suspected that to see any differences would require top-quality optics and we would need to pay very careful attention to camera settings. This was borne out in our testing.

Studio scene comparison

In evaluating our familiar studio test scene (shown below) in a very controlled shooting environment using the excellent AF-S Nikkor 50mm f/1.4G lens, we determined that visible differences between the D7100 and the less expensive OLPF-enabled D5200, were only apparent within a narrow range of apertures. The samples below were shot at an aperture of F3.5. In repeated tests with the 50mm F1.4, this gave the greatest amount of center sharpness on the D7100. At apertures wider than F3.5 and narrower than F5, we struggled to see any relevant differences in output.

Nikon D7100: ISO 100, ACR 7.4, Nikkor 50/1.4G @F3.5 Nikon D5200: ISO 100, ACR 7.4, Nikkor 50/1.4G @F3.5
100% crop 100% crop
100% crop 100% crop

The samples above are Raw file conversations made with ACR sharpening turned off and identical amounts of USM applied in Photoshop. Looking carefully, you can see that the D7100 delivers a more crisp file, with some areas of extremely fine detail rendered with greater clarity. Let's be clear though. These distinctions - apparent only at a 100% view - are very, very minor. Its also worth repeating that achieving even this result required shooting at F3.5, which obviously limits the depth of field. And it's hardly an aperture that you'd find yourself using in real world scenarios where edge to edge sharpness was a primary concern.

Real world kit zoom comparison

In real world shooting with the D7100, it's not unreasonable to assume a significant number of owners will be shooting with an affordable zoom like the AF-S DX Nikkor 18-105mm f/3.5-5.6G ED VR kit lens at a more commonly used aperture like F8. So here we've shot the same scene using the D7100 and D5200 with the same 18-105mm kit zoom at an 18mm focal length.

Nikkor 18-105mm kit zoom @18mm, F8, ISO 100
Nikon D7100: ACR 7.4 100% crop Nikon D5200: ACR 7.4 100% crop
100% crop 100% crop

In coomparing the images above, we're hard-pressed to find any differences in sharpness or edge detail among the two cameras. In essence, the optical shortcomings of the kit zoom have an image-softening effect that easily outweighs any advantage to be gained by removing the OLPF in the D7100. You can clearly see, for example that even towards the center of the image at F8, there's no meaningful difference between the two cameras and towards the edges, smeared detail and extreme chromatic aberration is a far greater determinant of image quality.

Real world with the 50mm/1.4 lens

What we've learned so far is that eeking out even minor benefits of the D7100's no-OLPF approach requires great glass shot using careful technique at an aperture range between F3.5 and F5. Here's what this can look like in a real world scene. On this bright cloudless day we mounted the D7100 on a sturdy tripod, and enabled both mirror lockup and a 2 second shutter release. We shot several takes of the scene below with our second copy of the 50/1.4 lens and found that we got our sharpest result at F4.5.

Nikon D7100: Nikkor 50mm/1.4 @F4.5
100% crop 100% crop

Here you're looking at an ACR raw file conversion with no sharpening applied in the converter and a fairly conservative USM amount of 150% with a radius of 0.6. The level of detail and clarity on offer here are quite impressive for any DSLR short of the Nikon D800.

We shot this and several other scenes with the D5200 using the same lens and settings. Our conclusion after viewing dozens of comparisons was that outside the controlled environment of our studio, even extremely minor shifts in focus or (potentially) sensor alignment could trump the absence of an OLPF in accounting for any visible differences between the D7100 and D5200.

In short, even if you were willing to put the best glass available on the D7100 and shoot at a wide aperture, you're not likely, even with a lot of effort, to leverage visible benefits of the OLPF removal. While this may be a bit of a disappointment for some, the very good news is that to date we've seen no practical downside to the filter's removal for still photography. It is essentially neutral with regard to image quality.