4K: What's the hurry?
Suppose we're shooting a demonstration where the subject is unpredictable: If there's a shot where we can't know precisely where to aim the camera, then having four times the freedom of framing is useful. We recently recorded a gymnast performing on a horizontal bar to demonstrate a concept for a mechanical engineering course. Out of respect for the student-athlete's time, we wanted to set up the shots wide enough to accommodate any reasonable contingencies and let her get back to her own business as soon as possible. In post, we'd punch in to the framing we would have picked if we'd had multiple takes. A 4K frame would have offered a luxurious comfort zone.
Similarly, while shooting laboratory footage, it wouldn't be extraordinary for something serendipitous to happen that the presenter would explain off the cuff. At 4K, the camera operator need not even attempt to zoom in on what he guesses is relevant. Shooting in 4K, we can "enhance" in post once we've discussed with the expert what to show the students, so we don't have to risk an otherwise solid shot.
Or maybe it's not the subject's position that's unpredictable, but the camera's. We occasionally mount a camera to a drone, where sections of the frame will be sacrificially shot and discarded for image stabilization purposes.
Even for talking-head shots, we can shoot wide at 4K and punch in to the desired framing in post, simulating a multi-camera shoot by cutting between two zoom levels from the same 4K footage to hide edits.
I find these arguments to be compelling, but my team members have yet to make that leap to 4K. Now why is that?
First, the urgency simply isn't there. The difference between HD and 4K isn't as dramatic as it is between SD and HD. We aren't changing the shape of the frame, we're simply recording four dots instead of one. HD footage won't look intolerably anachronistic; even when we switch to 4K, we'll occasionally make a mistake when setting focus and work from slightly blurry footage. My intuition is that, at 4K, we're shooting deeper, not bigger. Most scenarios in educational media production don't require that depth of resolution.
Second, I'm not convinced the 4K shooting technology is mature enough to invest in, and camera manufacturers appear similarly circumspect. Specifically, there are two video codecs efficient enough to store 4K resolution video at bitrates that suit our existing portable media and fileservers: HEVC and AV1. HEVC (or H.265) is the successor to the HD-standard codec H.264, and boasts the ability to halve the file size without compromising picture quality. It is, however, encumbered with a complicated and evolving licensing scheme. AV1 is the successor to the royalty-free VP9 codec (among other projects by Alliance for Open Media participants), and is scheduled to be released by the time this issue is published.
Camera manufacturers are conspicuously not recording video with these video codecs. The only cameras I know of that shoot 4K video using the H.265 codec are sold by Samsung--or at least, they were. CES 2017 came and went without a significant announcement about HEVC adoption in cameras, aside from the notably restrained (albeit impressive) features of the Panasonic GH5's debut firmware.
I'm waiting on the answers the camera manufacturers get to two critical questions: Will AOMedia's AV1 codec be good enough to adopt over HEVC? If not, will it be good enough to push HEVC's licensing costs down enough to cover the difference in quality? My bet is that we can afford to wait another year to find out.
By Liam Moran
Liam Moran (firstname.lastname@example.org) produces curricular digital media and manages the servers to store, process, and deliver it at the Center for Innovation in Teaching and Learning at the University of Illinois-Urbana-Champaign.
Comments? Email us at email@example.com, or check the masthead for other ways to contact us.