Sunday, July 29, 2012

When Your Internal Clock Is Off

Beat time cycles image on Bobby Owsinski's Big Picture production blog
Notice how the time gradually cycles over time
All musicians have a sense of time; some better than others. Why one drummer sounds so in the pocket while another playing exactly the same beat on the same kit yet sounds like moving furniture has been philosophized and analyzed, and I'm not sure than anyone has ever come up with an scientific explanation as to why. The same goes for other musicians who don't have to be as ambidextrous in their execution. Why is a studio musician's time better than other players? Why is machine generated music considered souless?

Maybe we now know, according to a very interesting article in the Harvard Gazette. A study at the Max Planck Institute For Dynamics and Self-Organization in Gottingen, Germany analyzed an expert Ghanaian drummer and discovered some interesting aspects of playing in time.

All humans have imperfect time (we don't we needed a study to tell us that), but it's the way that it's imperfect that makes a difference. When playing to a click, even a great player varies from being ahead or behind by 10 to 20 milliseconds. But what is really interesting is that these variations happen over long periods of time in the song, according to the study. A player may be ahead of the beat for 30 consecutive beats, then gradually fall behind for the next 30 (see the Beat Index on the left).

What's even more interesting is that the whole cycle repeats itself over long periods. In other words, the same cycle of being ahead for a number of beats then behind for a number repeats over and over.

But what's interesting about this is the fact that the beats are consecutive. What sounds bad to us is if we jump back and forth, being ahead for a couple then behind for a couple, which is why the "Humanize" function on a sequencer doesn't sound that human at all. It only randomizes the beats. Humans (at least the ones who are good players) do it in a slow cycle from ahead to behind.

I'm not sure if this study tried more than one drummer or if there were more musicians on different instruments involved, but it certainly all sounds plausible. What I'd love to read is a study on how musicians interact with each other, and what scientifically makes a good "pocket." Any takers?

----------------------------------

You should follow me on Twitter for daily news and updates on production and the music business.

Don't forget to check out my Music 3.0 blog for tips and tricks on navigating social media and the new music business.

4 comments:

Rich Layton said...

Growing up as a musician in Texas, I have had a lifelong fascination with the concept of "the pocket", that mystical elasticity (primarily supplied by the backbeat on snare) that defines the feel of a song. Lots of drummers I've played with have outstanding time, but it's all square and right on top of the beat. When they try to play "in the pocket" the song ends up dragging :) Had an amazing lesson decades ago in the studio when the tracks from a crappy sounding kit were notched to create a trigger signal for a high end drum machine. On a lark, we ran the signal from the snare through a delay, allowing us to move the snare back in time relative to the rest of the kit. Jaw dropping how far it could go before it finally sounded late. Also astounding how much the feel of the song changed as the pocket got "deeper" with each fractional second of delay.

buck baran said...

Look at time in 4/4 in two ways:
#1. As a conductor's standard baton motion with beat 1 at the bottom, 2 at the mid left across to 3 and up to beat 4 dropping back down to beat 1.
#2. Look at it as a clock with 6 o'clock as beat 1, 9 as beat 2, 12 as beat 3, and 3 o'clock as beat 4.
In both cases we slow down as we reach beat 3 then pickup speed into beat 4 and even more speed into the downbeat.
The count fluctuates within itself.
A phenomenon I experience when practicing with a metronome is time appears to slow down. When I stop playing the tempo appears to quicken. Kind of like a "Doppler Effect."

Anonymous said...

So if that is true, the cycling of x number of measures 10-20 msecs behind and then ahead, then it would seem that if you took a perfectly quantized midi drum loop you could manually apply the slowing and speeding in groups of x-measures and produce a more human sounding drum track. Sounds like something worth trying.

Anonymous said...

So if that is true, the cycling of x number of measures 10-20 msecs behind and then ahead, then it would seem that if you took a perfectly quantized midi drum loop you could manually apply the slowing and speeding in groups of x-measures and produce a more human sounding drum track. Sounds like something worth trying.

LinkWithin

Related Posts Plugin for WordPress, Blogger...