Any responsible parent should teach their children to count. Most humans will teach their children to count in decimal. I am sure there are a number of computer experts who will teach their children to count in hexadecimal. Old-timers may teach them to count in Octal. I suppose that media engineers should teach their children to count in Drop Frame Timecode. I wonder how that would work?
“OK, Junior, we’re going to count for 11 minutes with high accuracy.”
“OK, dad!” replied Junior enthusiastically.
“0, 0, 1, 1, 2, 2…” says Dad.
“Dad … Miss Maths-Wizard tells us to start at 1 because that’s the smallest number!” said Junior
“Did she?” replied Dad. “Actually the smallest number is zero – you just have to know that. So we have 0, 0, 1, 1, 2, 2 –”
“Daaad, why are you saying each number twice?” asked Junior, confused.
“Because I’m counting in 60p, obviously,” stated Dad. “Each frame is counted twice to make life easy for older equipment. You just have to know that.”
Time passes, “…28, 28, 29, 29, 1:00, 1:00, 1:01, 1:01, 1:02, 1:02 –”
“Dad … Miss Maths-Wizard introduced a new significant digit after she got to 9, not 29!” said Junior.
“Well, Junior. In media, in America, we have about 30 frames per second, so the new digit starts every second – even when there are sixty frames per second. You just have to know that.”
“Okaaay,” said Junior uncertainly as he counted along with Dad for another 58 seconds and 52 frames.
“59:28, 59:28, 59:29, 59:29, 1:00:02, 1:00:02, 1:00:03, 1:00:03 –”
“Umm, Daaaaad!” whined Junior with a look of worry on his brow. “What happened there? Miss Maths-Wizard told us there were 60 seconds in a minute, why did we skip some numbers?”
“Miss Maths-Wizard hasn’t told you about how the world works son. When color was introduced in 1953, the great and the wise chose to change the rate at which frames were displayed and we later changed the way that we count to keep media time and time of day roughly in sync. You just have to know that we skip two counts every minute. Keep up …”
“OoooKaay,” squirmed Junior with caution and worry for another 8 minutes, 59 seconds and 56 frames.
“9:59:29, 9:59:29, 10:00:00, 10:00:00, 10:00:01, 10:00:01, 10:00:02..”, enumerated Dad.
“Waaaaahhhh,” wailed Junior. “You’re supposed to miss two counts.”
“Not on the 10th minute Junior,” said Dad “We count properly on the 10-minute mark. You just have to know that. It’s a bit like leap years – count funny every 4 years but not every 400.”
“Nooo, it’s not! Leap years adjust our time reference due to the quantization error when dividing the earth’s rotation of the sun by the earth’s spin duration,” exclaimed Junior showing remarkable understanding of physics for a 3-year-old. “Whereas this craziness you’re describing adds a whole lot of complexity to a metadata label that is intended to make synchronization simple.”
“I am told by experts in America that this is simple and doesn’t cause problems,” squeaked Dad defensively.
“Simple! SIMPLE!” shouted Junior while changing gear into full blown 3-year-old tantrum mode. “When I grow up, we’re just going to assign a simple number to the indivisible media elements. That number will be allowed to be very big and it will have enough accuracy to be put on a single sample of audio. When I grow up, Timecode will be dead and no-one will be sorry!”
So SMPTE listened to Junior and has started to publish the standards that allow synchronization of media that has arbitrary frame rates and arbitrary sample rates. Carriage and distribution of those signals is being formalized, and young Junior will see that Timecode’s death is imminent.
Or is it? There is huge inertia in the use of Timecode within the broadcast media world. Thousands of people use Timecode every day with great success and it’s hard to see how a grass-roots retooling of the world’s Timecode infrastructure can be paid for. The online media world tends to use media-time (hh:mm:ss:msec) to avoid the problems of Timecode and they have been doing this very effectively. My guess is that if this approach needs more accuracy, then a scheme that uses hh:mm:ss:μsec will pragmatically spring into existence.
Is Timecode dead? No. Will it die? I don’t think so. I think pragmatic software approaches will bypass Timecode in many applications and that pure Timecode or its SMPTE replacement will continue to be used in applications where frame and sample accuracy are vital. The bulk of the world’s media files will quite happily use the concept of media time and very few people will know or care.
In my next post, I’ll be teaching Junior to practice his scales. Enjoy!