D750: Some Details

The usual misinformation (and real information) is starting to appear about Nikon’s latest DSLR. I don’t have a camera to play with, only images shot by others who’ve had access to the camera, so certainly can’t answer all the detailed questions some have at the moment. That said, things are becoming clearer.

In terms of buffer, the D750 looks like a winner. 87 shots JPEG Fine Large and 15 shots NEF 14-bit Lossless Compressed. At 6.5 fps, that’s 10+ seconds of JPEG minimum and almost 3 of raw minimum, both decent. If you’re worried about raw buffer, you can always choose 12-bit Lossless Compressed at 25 buffered shots, or 12-bit Compressed at 33. Shoot in DX mode on the D750 (~9mp) and basically max out the buffer at 100 images for every setting except 14-bit Lossless Compressed (48). Those numbers are all with 16GB SanDisk Extreme Pro cards, by the way. If your card isn’t UHS-I at 95MB/s, you might see lower buffer levels. File sizes range from 19.2MB to 26.9MB for FX raw files, by the way. 

D750 24 120 top 1.high.jpg

The D750, despite the number, has the consumer build body. Which means that it takes the MC-DC2 remote cable, has Scene exposure modes and two user User Settings positions on the Mode dial, plus uses overloaded buttons for things like WB and ISO instead of dedicated ones. You also get the consumer rectangular eyepiece instead of the pro circular one with shutter, though what’s behind the eyepiece is mostly the same as D810 in terms of view (magnification is the same, eyepoint has increased to 21mm). 

There’s a new MB-D16 grip (US$370), which has a full set of controls (dials, direction pad, AE/AF lock button) and allows use of an additional EN-EL15 battery or six AA batteries. 

The built-in WiFi is essentially the same as the optional WU-1a/b units we’ve seen in the past, and requires Nikon’s WMU (Wireless Mobile Utility) on an iOS or Android device to operate. Nothing really new there: you have some basic control over the camera (shoot, live view, self timer, etc.), while downloads to the device are JPEG only and subject to some “gotchas” (interrupted connections mean the image being downloaded at that time will never get downloaded automatically, despite resuming the connection. Nikon claims a transmission distance of 98 feet (30m), but I’ve not been able to consistently get a locked connection at anything approaching that with my iPhone on the WU units.

In short, the built-in WiFi is a basic option, and not the one that pros are looking for. For that, you need the expensive and cumbersome UT-1 and WT-5 combo, at which point you can do real tethering and direct-to-computer work. 

The tilting LCD doesn’t tilt quite as far as you might think: 90° up or 75° down only. The good news is that the tilting shouldn’t impact tripods or camera plates, as it happens away from the camera body itself. The LCD itself can be “calibrated,” basically allowing you to adjust the AB/GM color axis, much like the WB settings allow.

Nikon has once again changed the way the Information display on the back LCD works and the way the “i” button lets you set things fast. Instead of options along the bottom that highlight, we now have them along the right side in a scrolling list. Nikon can’t seem to make two cameras the same in this respect. Apparently there are extra engineers in Tokyo spending all their time figuring out how to modify something that already works just fine. 

I mentioned the SB-500 in a separate article, but this is a nice option, I think, despite the fact that built-in flash has the usual two-group commander mode the internal flash that we’ve seen on the pro models before. The SB-500 more than doubles the output of the internal flash, gives you tilt/swivel, and doubles the number of groups you can control from the camera. It seems like a pretty reasonable option for quick and dirty wireless, since it can also act as a remote for the camera’s internal flash, as well. At its reasonable price, I see it as a no-brainer purchase for a D750 user. 

While I won’t go into the video aspects here, one thing that didn’t get mentioned much in Nikon’s marketing is that we now have Zebra capability while shooting video. Timelapse also gets an exposure smoothing function.

Another mostly unmentioned aspect of the D750 is that it has 2-5 frame bracketing at 2 and 3EV (9 frame bracketing up to 1EV). HDR shooters will be happy.

Finally, I get to the two contentious details about the D750: noise and focus. 

Already we’re hearing anecdotal reports about "improved high ISO capabilities." These are due to the changes of processing in the EXPEED4 chip, not the sensor. In other words, Nikon has further tweaked the noise reduction applied to JPEG images from previous iterations. It remains to be seen what that really means. Noise reduction is always a trade off with edge acuity. I don’t know anyone that’s managed to find a way around that. So more noise reduction generally is going to mean edges that are tougher to sharpen. Raw shooters shouldn’t expect any real changes from the D610 (though I suspect that Capture NX-D might also be processing D750 images differently). 

Then there’s the -3EV number associated with low light autofocus on the D750. This is lower than the -2EV number published for the D810 and -1EV for the D4s. Thus, a lot of folk are already trying to gush over “better autofocus.” I’m far less sure of that and need to actually put a D750 through its paces to figure out what that number really means. I’ve seen Nikon DSLRs with lower EV numbers for AF perform worse than ones with higher numbers before. That number seems a bit meaningless to me, though it is an attempt to say at what light level the focus sensors can still operate. But it’s probably a “managed” number. In other words, you could use the same part and just put an arbitrary cutoff on one camera using it versus another. Worse still, if you’re using fast lenses (e.g. f/1.4) in low light on your Nikon DSLR, it’s more likely that the lens is the gating element in focus performance. That’s because the contrast/flare of the fast lens is making the data seen by the phase detect sensors less reliable.  


© Thom Hogan 2014 / All Rights Reserved bythom.com  @bythom #bythom