One of the main reasons why I bought a D600 when it came out a couple of years ago, was its amazing 24 MP sensor. I really liked the sensor, because it gave a great balance of pixels vs noise when compared to 16 MP and 36 MP sensors. My initial conclusion about the D750 was that its image quality was similar to that of the D600 / D610 cameras. However, after spending two days testing the cameras, I realized that Adobe’s beta DNG Converter and Lightroom’s DNG rendering did quite a poor job at ISOs above 3200.
Some of the color and shadow areas appeared washed out in comparison to the D600 / D610. I then re-ran the analysis using Nikon’s Capture NX-D 1.0.3 software, which showed a totally different picture – the D750 appeared to have visibly better performance starting from ISO 800, where it showed less noise in comparison. At high ISOs above 3200, the D750 showed between 1/2 to a full stop of difference in performance (depending on where you look), which was rather surprising. You can see this on the Camera Comparisons page of this review.
There is an interesting observation though – the difference in performance does not appear from newer sensor technology, but rather from a more complex and smarter way to reduce noise as part of the image processing pipeline. I initially noticed similar behavior when testing the Nikon D810 versus the D800E – Nikon seems to be applying a different tone curve to RAW images, essentially darkening the shadow areas, then applying a more effective noise suppression algorithm that reduces chroma noise. This explains why images from the latest Nikon cameras appear a bit “darker” in some areas and the colors do not quite match…
In terms of big improvements, I think we have more or less reached the maximum potential of the current CMOS sensor technology because the noise performance has not drastically improved in the last 2-3 years. Manufacturers are simply tweaking the output of sensors and applying different noise reduction technologies.
The dynamic range on the D750 is superb, just like it is on other Nikon DSLRs, including the D610. While I have not done any scientific measurements to evaluate the dynamic range of the D750 yet, I have tried recovering both highlights and shadows from RAW images and I have been impressed by the results. Take a look at the below underexposed image that was recovered later in Lightroom (+2.1 Exposure, +90 Shadows, +30 Whites, -10 Blacks, +10 Saturation):
When it comes to dynamic range, I found DxOMark to be quite accurate in their assessments when compared to my lab tests. So if you want to see the exact dynamic range in EV numbers and compare the D750 to other cameras, I would wait until DxOMark evaluates the D750 (which will probably happen when full RAW support for the D750 is provided).
Quality Assurance and AF Point Accuracy
As I mentioned at the beginning of this review, I have been using three different samples of the Nikon D750. All three are solid in terms of overall build/construction and I do not see any differences between them after a month of rather heavy use. No rubber parts are peeling off and all buttons are functioning normally. Autofocus performance and accuracy are excellent (as reported on the next page) and I did not see any deviation or misalignment of focus points as we had seen on the original D800 / D800E cameras.
Sensor Dust Report
The Nikon D600 was plagued with dust issues due to its poor shutter construction, so many Nikon users grew weary of this problem and might be interested to find out if the D750 has a similar issue or not. I am happy to say that all three units that I have been using have a solid shutter mechanism that is devoid of this problem. Seems like Nikon has addressed this problem once and for all, which is great.
Table of Contents