The last part of this post is a look at some analysis results from the Z9 using FoCal. As mentioned at the start of the post, the primary motivation was the testing of the support for the Z9 within FoCal rather than a detailed analysis of the camera itself, but at some point in future we’ll specifically dig deeper.
[Note: Since this article was originally written, FoCal IQ has been updated with a camera database which includes stabilisation system performance as a metric along with many other measurements. Take a look at the Camera Database here, and there’s lots more information in our recent posts here and here.]
The latest release of FoCal has had some significant improvements made to the Stabilisation Test, making it easier to analyse the performance of the stabilisation system on your camera. As we had both the Z9 and the Z7 in the lab, it seemed logical to compare the two.
The following are some results from the testing of 3 lenses: the native Z-mount 24-70mm f/4S, and 2 adapted lenses: the AF-S Nikkor 50mm f/1.4G and the AF Nikkor 35mm f/2D.
For each set of data, the Z9 results appear on the left, and the Z7 results on the right.
Nikkor 24-70mm f/4 S
The thing to compare with the Quality Value charts is curve shape rather than absolute values – a slight variation in y-axis values between cameras is not important as we’re comparing the relative quality with stabilisation on and off.
The red line shows the quality with VR disabled, and the green with VR enabled. You can clearly see in both cases that VR is having a significant effect from around 1/80s on the Z9 and 1/50s on the Z7, as the green (stabilised) line stays high while the red quality drops.
Take a look at the Z7 results for shutter speeds faster than 1/50s (the grey area on the right-hand chart above) – notice how the red line is actually higher quality than the green. This indicates that the VR system of the Z7 is actually degrading the image quality at faster shutter speeds – something that doesn’t happen in the Z9.
The charts below show the number of exposure stops between points of the same quality. For example, take 1/2s on the Z9 chart below – this shows an improvement of just about 4 stops, which means with the Z9 you’re getting the same sort of quality with a 1/2s exposure using VR as you were with a 1/30s exposure with VR disabled.
Just to show exactly what this looks like, here’s crops from the test. On the left, you have a 1/30s exposure with VR disabled, and on the right a 1/2s exposure (4 stops slower) with VR enabled – they’re not identical, but they are pretty similar quality:
And here’s the 1/2s exposure with VR disabled – the stabilisation really is quite effective!
AF-S Nikkor 50mm f/1.4G
Moving to an F-mount lens on the FTZ adapter – the 50mm f/1.4G. This lens dates from 2008 so have a relatively modern optical design and includes an autofocus motor, meaning AF is available using the FTZ adapter.
Again, on the left below we have the Z9 and on the right the Z7. The first pair of charts again show the slight degradation on the Z7 with VR enabled as you have shutter speeds faster than 1/100s, and no change in quality at these faster speeds on the Z9.
Looking at the number of stops improvement, it looks like the VR system of the Z9 doesn’t perform quite as well as the Z7 across the active range. For the Z9, there’s a gradually improving performance from about 2 stops at 1/13s through to almost 4 stops at 2 seconds. But the Z7 seems to settle in to a 3 stop improvement by 1/25s and hold between 3 and 4 stops for the remainder of the longer shutter speeds.
To be clear, this is a small difference, and realistically you’d probably not notice any worse performance with the Z9 VR system than the Z7 in real use.
AF Nikkor 35mm f/2D
The last tested lens is another adapter one, this time from 1989. The 35mm f/2D lens has no focus motor, so the test was run in manual focus mode where you pre-focus the lens before running the test – this makes no difference to the test results, just adds a small extra step before you start the test.
The now familiar pattern is shown again – slight degradation on the Z7 with faster shutter speeds and VR enabled, where the Z9 keeps the quality very similar between VR on and VR off.
And again, the performance of the Z9 is flat at an average of a little under 2 stops for this focal length, whereas the Z7 rises to an almost 4 stop improvement around 1/4s, averaging a little under 3 stops.
VR Testing Conclusions
With limited time using the Z9, this wasn’t a carefully planned and rigid set of tests, but the results above do clearly show two things:
- if you’re using the Z7, it’s probably best to switch VR off at higher shutter speeds, whereas with the Z9 there’s no need.
- the VR performance of the Z7 is slightly better than that of the Z9.
When we have access to the Z9 again, we’ll put the various VR modes (Off, Normal and Sport) through their paces in a more controlled way, comparing with various other cameras and give a more detailed view of how the VR performs.
[Since we wrote this post, we’ve released the FoCal IQ Camera Database which has this sort of information for a wide range of cameras. You can, for example, do a direct comparison of the Nikon Z7 and Nikon Z9]