(commentary)
A lot of folk are asking me about the Nikon EU product manager's comments about Nikon "will be looking into [4K video] for the future." It seems that a number of folk think that this will get Nikon "back into" the video game.
Technology pursued for technology's sake sometimes eventually yields useful results, but not always. Let me state things a bit differently than you've probably heard: if Nikon really wants to make a splash in video, then doing better 1080P would be far more useful than giving us 4K.
Indeed, Nikon did a bit of that with the D4 and D800, giving us the ability to send uncompressed video out the HDMI port and capture it with something externally. I'd call that a baby step.
Here's the thing, broadcasting still is standardized on HD video (all the variations up to 1080P). If you want really great broadcast quality material, you want at least 50Mbps bandwidth, and a minimum of 4:2:2 color (we really want 4:4:4), preferably in 10 bits so you can avoid posterization when grading the video in post production. What's coming out of still cameras that do video is typically 20-28Mbps, 4:2:0 color, and 8 bits.
Having an uncompressed video stream out of the camera is nice, but I still need to record it with something external, which just adds cables and complications. What we really want is something like Apple ProRes compression instead of AVCHD, and at one of the higher bandwidths with as many bits as possible. Of course this puts immense bandwidth burdens internally on the camera to achieve, but so does 4K video ;~).
So which would we rather have? 1080P (1920x1080) done right, or 4K (3840 x 2160) done wrong? The answer is simple to anyone doing video with DSLRs: 1080P done right. Again, credit to Nikon for giving us that uncompressed HDMI output, which allows us to do things mostly right externally, but it would be so much better to have this ability in camera.
Okay, then what's with the push to 4K? Is Netflix going to convert to all 4K downloads next week or something?
Blame the TV makers. They had a nice run up from CRTs to flat panel HD screens. Then there was a race to the bottom that drove prices down to insane and unprofitable levels and made the primary brands look for anything that might drive new higher-end volume. Remember 3D? That technology was going to be the savior for TV makers and get margins back. Didn't happen. Guess what this year's emphasis is? You guessed it: 4K TVs. And again, it won't happen, and for much the same reason: the content isn't there yet.
But wouldn't putting 4K into cameras enable the content? Well, that's what Sony is hoping (remember, they make TVs, image sensors, and cameras, so where do you think some of the 4K push is coming from? ;~). But 4K currently isn't standardized as a broadcast medium, and all that extra data would tax the broadband capabilities in a lot of countries if iTunes, Netflix, YouTube and others all converted to it (Netflix and YouTube are said to make up half the peak Internet traffic in the US, for example; and Netflix generally downloads at 720P with heavy compression, not even 1080P). Yes, Netflix is experimenting with a few 4K downloads, but they're super compressed down to 3Mbps, which takes away a lot of the advantage of having those extra pixels.
So my message to Nikon is this: sure, look at 4K all you want. But in the meantime, improve the 720P and 1080P capabilities so that we have better original data to work with.
