What is a groove?

[Sendy]

Well now we've pretty much covered all angles and sides of what a groove is, could be, will be and is not, we can move onto more pressing questions, such as "What is a juggalo?".

[A.M. Gold]

whatever it is madonna wanted us in it.

And that's when Vurt proved his love to her.

[A.M. Gold]

I knew this question would spark a lot of debate, which is why I didn't just Wiki it and call it a day. I think the wide range of definitions is interesting.

Plus I like how "musicologists" didn't start studying groove until 1990. Guess they weren't groovy before then.

[stringtapper]

Plus I like how "musicologists" didn't start studying groove until 1990. Guess they weren't groovy before then.

And I like how you put the word musicologists in quotes, as if it's not a real thing.

And to put a finer point on that, the fact is musicologists and music theorists didn't start looking seriously at pop music at all until around that time, let alone specific facets of it such as groove.

[A.M. Gold]

I don't know about musicologists but I do know funkologists have been studying groove since way before 1990.

[KoolFartWind]

It's the underlying set of pulses on which events happen the most, each with certain characteristics, such as a probabilistic weight and degree of accent. In short, it's the most basic of rhythmic expectations as based on what the listener has already heard. This is why loops are sometimes called grooves.

Oh, you wanted a simple definition? It's when the stuff happens

What is the difference to meter, then ? I think groove is the temporal template and meter the accentual one.

[foosnark]

What is the difference to meter, then ? I think groove is the temporal template and meter the accentual one.

The way I think of it, meter might tell you how a measure is divided (and where accents tend to be), but groove includes how individual beats are relatively late or early from that division, as well as dynamics.

Groove might even include frequency, to the extent that it's sort of a psychoacoustic accent. Consider these three basslines on 16ths:

C2 C2 C2 C2 C2 C2 C2 C2 C2 C2 C2 C2 C2 C2 G2 Bb2
C2 C3 C2 C3 C2 C3 C2 C3 C2 C3 C2 C3 C2 C3 G2 Bb2
C2 C3 C2 CC C2 C3 C2 CC C2 C3 C2 C2 C2 C3 G2 Bb2

Chances are, a human is going to want to play them all slightly differently in terms of timing and dynamics -- but even if they don't, they'll be perceived as different grooves even if the actual volume of the notes is identical across the three examples.

[KoolFartWind]

What is the difference to meter, then ? I think groove is the temporal template and meter the accentual one.

The way I think of it, meter might tell you how a measure is divided (and where accents tend to be), but groove includes how individual beats are relatively late or early from that division, as well as dynamics.

That is an approach, but when individual beats periodically are relatively late, than this property should reflect in the meter as well. Meter has (theoretically) an infinite amount of levels. For example a 2/4 meter will imply an underlying 6/8 sequence and that will imply other possible divisions (similar to fourier analysis).

Groove might even include frequency, to the extent that it's sort of a psychoacoustic accent. Consider these three basslines on 16ths:

C2 C2 C2 C2 C2 C2 C2 C2 C2 C2 C2 C2 C2 C2 G2 Bb2
C2 C3 C2 C3 C2 C3 C2 C3 C2 C3 C2 C3 C2 C3 G2 Bb2
C2 C3 C2 CC C2 C3 C2 CC C2 C3 C2 C2 C2 C3 G2 Bb2

Chances are, a human is going to want to play them all slightly differently in terms of timing and dynamics -- but even if they don't, they'll be perceived as different grooves even if the actual volume of the notes is identical across the three examples.

Hmm, what you describe, sounds like (sub-)meter to me
First line any meter. Second 2/4. Third 4/4 with a temporarily upbeat accentuation. The user tappermike posted a link in another thread (about meter). In this link the jazz-pianist solely talked about meter. Yet his examples were fundamentally groovy I repost the link when I find it.

Edit :

I think I know what you mean, now. But the concept you are describing seems like a local phenomenon to me, though people seem to have a global "groove" in mind for the song. So groove is, how you syncopate individual notes, in order to make it more interesting. But that can change from measure to measure and should not be a constant pattern, IMHO.

[jancivil]

What is the difference to meter, then ? I think groove is the temporal template and meter the accentual one.

The way I think of it, meter might tell you how a measure is divided (and where accents tend to be), but groove includes how individual beats are relatively late or early from that division, as well as dynamics.

That is an approach, but when individual beats periodically are relatively late, than this property should reflect in the meter as well. Meter has (theoretically) an infinite amount of levels. For example a 2/4 meter will imply an underlying 6/8 sequence and that will imply other possible divisions (similar to fourier analysis).

A bit of lay-back on the 2 and/or 4 of 4/4, or a bit of push @ 2 is not in itself a property to make the meter any different.

One thing I would like to bring in is Dennis Chambers and his personal definition of 'in the pocket', which is about this, particularly a late backbeat. It's not going to be very useful to try and make a predictive algorithm to sort this and enable you to avoid doing the thing. A lot of the things in funk, groove type of rhythm is off the beat in ways that kind of don't make a specific sense in terms of numbers. A lot of it is surprisingly far off the grid where you wouldn't predict it. It is complex; it breathes, it is supposed to move and flow. And respond, as if intimately, to another person's surprises, and references, and sense of humor and sexuality. I'm not making love to no robot y'all.

I don't like 'meter, the accentual template' very much. I think I know what you mean but a lot of syncopation in a funk bass part is going to bely it too much. You'll need to be specifying 16/16. OR! Say if you believe this thing is really on the last sixteenth note and quantizing it is going to cut it, that really happening thing might be way off it. And there is no really good way to name it or number that, it is that thing at that time in context with other things, and it also belongs to timbre and attack. But, it doesn't take you away from where 1, 2, 3, 4 are as a ground.

Where I'm at now, a strict grid sounds inept, it sounds off so much of the time for what I'm interested in.

[foosnark]

A couple of neat things:

When the beat goes off: Errors in rhythm follow pattern, physicists find

Science Shows How Drummers' Brains Are Actually Different From Everybody Elses'

[Hink]

Science Shows How Drummers' Brains Are Actually Different From Everybody Elses'

Science Shows How Guitar Players Brains Are Actually Different From Everyone Else

They've been busy

[jancivil]

When the Beat Goes Off

the researchers are listening for his errors. Though the drummer is a professional, like all humans, his rhythm is imperfect. Each time his hand hits the drum, his beat falls ahead or behind the metronome by 10 to 20 milliseconds. On average, he anticipates the beat, and plays ahead of it, 16 milliseconds ahead — less than the time it takes a person to blink, or a dragonfly to flap its wings. What the physicists want to know is: Are these errors random, or correlated in a way that can be expressed by a mathematical law?

So they found a number of things. Chiefly, that there are cycles within cycles and that to find meaning in the variances takes a long sample of what the drummer does. They believe that formally, this is fractal. That structurally the phenomena will be found to be self-similar in layers.
They found an average to indicate pushing the beat, but they found that at one point the drummer is laying back.
errors were correlated across long timescales: tens of seconds to minutes. A given beat depended not just on the timing of the previous beat, but also on beats that occurred minutes before.

“You can have these trends,”

To begin with, they surveyed people and seemed to find that people do not like the randomized humanization results from music software so much. This comes as no surprise. I said above that looking to algorithms to give a feel isn't so useful. So here, it's put to us that maybe this research could be useful. In the abstract, they found particular phenomena and came up with hypotheses as to what the brain does; there is a human component to time. But what do we do as far as making a maths work in order to replace what you do? What is the actual model?

They found things in <this Ghanian drummer's time>. Where is the meaning in the variances? There is musical context and experience and a school of thought, it isn't like there is a simple 'this is what a human does' I don't think.

SO I think this is good stuff to investigate and I'm glad it's done. But I think the patterns of a person are not going to be universal, that the observations are on a sort of macro level and before you think to make models, musical culture and then individual experience factor in. What do you do with this? I've given this some thought. At one point I experimented with making 'groove templates' from my own rhythms. What I found was that a few bars of it reiterated per se was not really worth continuing to print, that very thing they found about needing a longer time to start making sense of the variances, or in order to formulate a desired template. Again, I like the thing to breathe and adapt. Another thing, does this drummer do that same thing when there are other drummers, or other instruments. And another thing, what is the musical content when there is a layback or a push of the time? As given, it's just 'the beat'.

So this is a very subtle thing if you want to get into it.

[trimph1]

My favorite way of describing groove.....

http://www.youtube.com/watch?v=cY7hTT-Zk5M

[KoolFartWind]

When the Beat Goes Off

the researchers are listening for his errors. Though the drummer is a professional, like all humans, his rhythm is imperfect. Each time his hand hits the drum, his beat falls ahead or behind the metronome by 10 to 20 milliseconds. On average, he anticipates the beat, and plays ahead of it, 16 milliseconds ahead — less than the time it takes a person to blink, or a dragonfly to flap its wings. What the physicists want to know is: Are these errors random, or correlated in a way that can be expressed by a mathematical law?

So they found a number of things. Chiefly, that there are cycles within cycles and that to find meaning in the variances takes a long sample of what the drummer does. They believe that formally, this is fractal. That structurally the phenomena will be found to be self-similar in layers.
They found an average to indicate pushing the beat, but they found that at one point the drummer is laying back.
errors were correlated across long timescales: tens of seconds to minutes. A given beat depended not just on the timing of the previous beat, but also on beats that occurred minutes before.

“You can have these trends,”

To begin with, they surveyed people and seemed to find that people do not like the randomized humanization results from music software so much. This comes as no surprise. I said above that looking to algorithms to give a feel isn't so useful. So here, it's put to us that maybe this research could be useful. In the abstract, they found particular phenomena and came up with hypotheses as to what the brain does; there is a human component to time. But what do we do as far as making a maths work in order to replace what you do? What is the actual model?

How I see it, that every listener/sequencer (computed or human) has a periodic wave in mind, that constitutes the meter (or, as you say : meter + groove), the level of accentuation. Humans probably have an addition of sine waves as meter, computers tend to have an addition of triangle waves. As the article stated, the meter has a number of self similar divisions (fractal structure), that is limited by the time resolution of the human hearing mechanisms and mind.

The 16 ms negative delay could be correlated to the distance of the drum player to the audio source.

Now when I put the concept of "Groove" into that framework, a groove would be realized through a temporary phase shift of the meter wave.

[metrognome]

http://www.youtube.com/watch?v=jbN-jO11vKg

Get off the stage and onto the dance floor. That's where 'The Groove' can be felt.

Someone once said, "The drummer moves your feet. The bass moves your hips. That is the groove."

1-bumpity bumpity bumpity bumpity
2-bump bump bump bump
3-BUM-pity BUMP bump

Which one grooves for you?