All of the examples given in this article are from real technical specification sheets. Add your own in the comments!
A long time ago, your narrator caused confusion by walking into a production office and asking at what frame rate a feature film was to be shot. This seemed like a reasonable question, speaking as the person who would have to put that setting in the camera. Knowing exactly how much movement we wanted in our movie seemed pretty fundamental. No matter the answer, it is not a complicated question.
The problem is, having a script before starting the shoot is also pretty fundamental, and some movies notoriously don’t manage that, so it should be no surprise that getting an answer took two days. The difference between 23.98 and 24 is unlikely to invalidate anyone’s multi-million investment in film production, but this sort of slack-jawed production office confusion is hardly unfamiliar.
For another example, not long ago, yours truly also consulted on a development project involving a requirement for a video input according to RS-170M. This sounds good – a decision has been made and a controlling standard clearly identified. Some of our more, eheh, experienced readers might wonder if that sounds like a specification for 30-ish-frame, standard definition analogue video.
Prehistoric standards
The problem is, it isn’t. Thumbing through the literature, we find lots of standards have titles a bit like “RS-170M”, but it’s quite hard to find anything actually and specifically called RS-170M. The RS sounds like an EIA standard, but the EIA has not existed in pushing fifteen years and its standards were mostly superseded in the 1970s. EIA RS-170 describes the first monochrome 30-frame television system. RS-170M also sounds a bit like SMPTE ST 170M, which specifies 29.97-frame colour TV. EIA RS-170 and SMPTE ST 170M therefore specify two different things with important incompatibilities.
Anyone with experience in defence will be completely unsurprised that this situation arose from a military contract. Technological archaeology like this is much more common under circumstances where the original requirement was punched out on a mid-century typewriter by someone terrified that the red menace might already be lurking in a closet with a Nagra SN. At best, updates to that kind of thing are limited to scanning the original, at a slight angle, into a marginally-readable, low-resolution PDF.
It would be nice to be able to say that this sort of dedication to Luddism is reserved for government work. It turns out, though, that clinging to something that worked last time is a fairly universal human instinct, one with a strong foothold in media and entertainment. The word “worked” can be a dangerous catchall. Shooting a whole movie in 8 bit works. Many things that worked demanded late nights, gnawed fingernails and the sort of technological workarounds that we agree never to discuss again.
That makes people impatient with specifications calling for, say, log raw, drop-frame 24p, or 4:2:2 colour space, especially when those things worked last time. Last time, we deduct, someone didn’t follow the rules, because the rules were ridiculous. Much as it is everyone’s job to be helpful, papering over the cracks lays a minefield for future people.
So in reality, a past plan might have worked as intended only sometimes. Or once. Or, in fact, never, because every person ever asked to shoot a movie at 50Mbps has suppressed the tears of rage and disappointment long enough that the people in charge never knew they were being ignored.
Mikey the digital loader
At least 50Mbps was once a realistic broadcast specification, even if a slightly misguided one (it was the highest data rate supported by early solid state onboard recorders). We must take some responsibility for encouraging this sort of maximalism among admin people, having spent decades inculcating the idea that more is more and information is easier to discard than to recover.
Even so, whenever someone decides that today’s cellphone-bound corporate video requires 6K uncompressed raw, the most immediate consequences fall on the shoulders of Mikey the digital loader and his sweating Macbook. Worse, though, is another cellphone-relevant idea: that 60 frames per second is more than 24, and therefore better. Much as it provides a lot of useful employment for optical flow software, that is inevitably a destructive mistake.
Yes, seriously, these are all real-world examples, but okay: office people are not camera specialists. They might insist on an approach which will make lots of technical people grind their collective teeth, but we should not be unfair to them. Possibly, instead, we should make them write out, one hundred times, the words “I must not ask the camera department to bake protanopia.cube into every frame we shoot.”
Given the impossibility of proving a negative proposition, if anyone can actually prove that a standard called RS-170M does exist, we would be pleased to hear about it in the comments. In the meantime, let’s be magnanimous next time production demands we shoot in “4K UHD” and argues passionately in defence of the idea, even though it should not take much domain-specific knowledge to understand that 4096 is more than 3840.
The answer, incidentally, was 23.976fps. Nobody seemed to know why, and a year later, the blu-ray was in the bargain bins, but it’s not clear whether there is any connection between those two things.