Perfect Rhythm

[Hornbein]

Everyone knows about perfect pitch, what about perfect rhythm? I became aware of this when overlaying the live version of a song over the studio recording. As far as tempo goes they were the same. That is, after four minutes of music they two versions were as close to the same tempo as I could measure, which is within about 30 milliseconds. Maybe not perfect but close enough. To get that close of a match the tempos have to be within 0.001% of one another.

Wait, you may say, couldn't they have been using a metronome/clicktrack on the live version? They might except for one thing. There was a section where everyone else dropped out and the singer went completely solo. She took 60 milliseconds more on the live version. I had to cut that out to get them to match. No clicktrack.

In another event, these two completely separate versions by different artists fit together with no editing whatsoever.

 

[.Scott]

I would say "wait - are you sure that's an untouched 'live version'?.
You may not have done editing, but the studio almost certainly did.
See here for details.

Also, the way that a metronome is used my an instrumentalist is discussed here. Musicians seem to do differentiate between locking onto the phase vs. locking onto the frequency. So it is certainly possible to use the same metronome signal before and after the vocalist - just choose a different phase.

On a related note (no pun intended), a person can train themselves to develop some useful level of echolocation. This ability is described in a National Academy of Sciences article. The Google AI describes this a being potentially good to about 6msec (about a 1 meter round trip). That's probably about the best I could do in estimating distances to walls of buildings on a quiet, snow-covered, night-time street.