Watch any film, and a whole stack of equipment made by competing companies has quietly cooperated to put it in front of you. The camera, the microphones, the editing software, and the speakers in the theater all came from different manufacturers who have no particular reason to like each other. Something had to make them all work together.
A lot of that something is an organization with a name nobody can pronounce at first: SMPTE, said “simp-tee.”
Who (and what) is SMPTE?
It stands for the Society of Motion Picture and Television Engineers, founded in 1916. It’s a standards body, which means it doesn’t make products. It writes down agreements so that everyone else’s products fit together. Think of it as the people who decided what size the plug is so that every lamp works in every wall.
SMPTE has written hundreds of standards. But when musicians, editors, and sound engineers say “SMPTE,” they almost always mean one specific thing: SMPTE timecode.
The problem: two machines, one performance
Video is not a smooth stream. It’s a rapid slideshow of still photographs called frames, typically 24, 25, or 30 of them every second, fast enough that your brain sees motion.
On a film set, the camera records the pictures, and a separate recorder captures the sound. Two machines, each with its own internal clock. Here’s the catch: no two clocks run at exactly the same speed. The difference is microscopic, but let both run for an hour and they’ve wandered apart. The actor’s lips stop matching their voice.
You could say “the scene starts at about a minute and twelve seconds in.” That’s fine for friends planning a song. A machine needs better than “about.”
The fix: give every frame an address
SMPTE timecode stamps every single frame with an exact address, written like a stopwatch: hours, minutes, seconds, frames.
So 01:14:32:12 means 1 hour, 14 minutes, 32 seconds, and the 12th frame. No ambiguity. An editor can say “cut at 01:14:32:12” and every machine in the building knows the precise frame.
With a shared address system, machines no longer have to guess where they are. One device generates the timecode, the others read it and follow along, and everyone stays on the same page. A film composer uses it the same way: to land a cymbal crash on the exact frame an explosion hits the screen.
How the address travels
The original method is wonderfully low-tech. LTC, or Linear Timecode, records the timecode as an audio signal on a spare track, a buzzing, screechy sound that’s unpleasant to listen to but easy for machines to read. It’s still used all over the industry today.
Digital setups can also carry the same information as data over a network cable, or through MIDI as MTC, which has its own article in this series.
Different route, same address.
The thing everyone gets wrong: “drop-frame”
Spend any time around video people and you’ll hear the term drop-frame timecode, which sounds alarming. Are frames being thrown away?
No. Nothing is deleted from the picture.
Here’s the quirk. Older American color television ran at a hair under 30 frames per second, about 29.97. Count those frames as if they were exactly 30 and your timecode clock slowly falls behind the real wall clock, until a one-hour show shows up a few seconds too long on paper. For broadcasters selling airtime by the second, that’s a real problem.
The fix is to skip some numbers in the labeling, not any actual frames. At the start of most minutes, two frame labels are quietly skipped, like a calendar that skips a few dates so it matches the seasons. The video is untouched. Only the numbering changes, so the timecode agrees with the clock on the wall.
So, to sum up
SMPTE is a standards organization, and SMPTE timecode is its best-known gift to the creative world: an exact address for every frame of picture and moment of sound, so machines that don’t share a clock can still agree on where they are.
You’ll never see it on screen. But every time dialogue lines up with lips, or a score hits an explosion on the exact frame, a very old, very boring, very reliable clock is doing its job.