[ternarizator]

Hi guys !

Though I guess some of you are already aware about it, I have noticed a certain vagueness about the different ways of ternarizing, and of course also about the different ways of playing "between ternary and binary" (very risky term in my opinion).

The best is to take a few examples :

1. The "quaternary passport"

1 e & a 2 e & a
s . . s s . t t

can be ternarized in two ways :

a. "4 → 3" : every beat of 4 pulses (quaternary beat) is turned into a beat of 3 pulses (ternary beat), by removing the second one (every "e" disappears in the counting)

1 & a 2 & a
s . s s t t

By this way, rhythms of Djansa family (16/8) can possibly be turned into a rhythm of Djaa family.
Notice that the number of beats doesn't change.

b. "2 → 3" : every quaternary beat, now regarded as two binary beats, is turned into two beats of 3 pulses (second version below) :

1 - e & - a 2 - e & - a
s . . . . s s . . t . t

1 . . 2 . . 3 . . 4 . .
s . . . . s s . . t . t

You can e.g. find it in some versions of woloso.
Notice that the number of beats is doubled.

2. Moribayassa

On the same model, we can change the djembé pattern of Moribayassa :

1 e & a 2 e & a
b . t t b . s .

By the first way, we get the djembé acc for "Soli des Manians" :

1 & a 2 & a
b t t b s .

The second way would have generated the following :

1 . . 2 . . 3 . . 4 . .
b . . t . t b . . s . .

3. Let's apply this to the third djembé in Makru (special dedication to Jürgen for his terrific work about Yankadi and Makru) :

1 e & a 2 e & a 3 e & a 4 e & a
t . . s . . . t t . s . s . . t

Applying the second way of ternarizing, you obtain :

1 . . 2 . . 3 . . 4 . .|1 . . 2 . . 3 . . 4 . .
t . . . . s . . . . . t t . . s . . s . . . . t

By adding one slap, and regarding it as a dununba/kensereni part, you now have :

1 . . 2 . . 3 . . 4 . .|1 . . 2 . . 3 . . 4 . .
t . . s . s . . . . . t t . . s . . s . . . . t, or

1 . . 2 . . 3 . . 4 . .|1 . . 2 . . 3 . . 4 . .
x .(x)x . x(x). x(x). x x .(x)x .(x)x . x(x). x
D . . K . K . . . . . D D . . K . . K . . . . D

And by adding the right Sangban part, you'll obtain one of the multiple versions of Yankadi :

1 . . 2 . . 3 . . 4 . .|1 . . 2 . . 3 . . 4 . .
x .(x)x . x(x). x(x). x x .(x)x .(x)x . x(x). x
D . . K . K . . . . . D D . . K . . K . . . . D
x .(x)x . x(x). x x .(x)x .(x)x . x(x). x x .(x)
M . . . . . . . S S . .|M . . . . . . . S S . .

(For more clarity, I present below the "quaternary version" of the same. You've just to read it "swing" to see and feel the whole thing :

1 e & a 2 e & a 3 e & a 4 e & a
x . x x . x . x x . x . x x . x
D . K K . . . D D . K . K . . D
x . x x . x x . x . x x . x x .
M . . . . S S . M . . . . S S .)

Now I will show what happens if you play the 3rd acc of Makru applying the other way of ternarizing :

1 e & a 2 e & a 3 e & a 4 e & a
t . . s . . . t t . s . s . . t

becomes :

1 & a 2 & a 3 & a 4 & a
t . s . . t t s . s . t

What happens to other parts of Makru ?

The b . . s . t t . b . s . . t t . can't be ternarized this way, for every first "t" is on a "e" of the counting...

But the b t . t b . s . b . t . b . s . can be changed into :

Ƀ . t b s . b t . b s ., where Ƀ is a b-t flam.

Playing it with a set of dununs (without bell) : D . D D S . D D . D S . can be remarkably groovy (if well played !)

That's all for today...

Vincent.

[korman]

What, then, would be the ways of binarizing?

[djembefeeling]

Just the other way round?

[korman]

Well, I guess, a reductional way ("3 -> 2") would be far less common than additive ("3 -> 4")?

But more important question is - in what context does the ternarization occur? When the speed increases so much that quarternary subdivision becomes too difficult to play?

[ternarizator]

No binarization ! Just quaternarization - or quaternization, the word exists for chemistry e.g.-, for 3 becomes 4, not 2 !!!

If you take for example the variation of Sangban in Djaa :

1 & a 2 & a 3 & a 4 & a|1 & a 2 & a 3 & a 4 & a
x . x . x . x x . x x .|x . x . x . x x . x . x
o . . . o . . . . o o .|. . . . ø . . ø . ø . .

you can "unternarize", or "quaternize" it the following way :

1 e & a 2 e & a 3 e & a 4 e & a|1 e & a 2 e & a 3 e & a 4 e & a
x . x x . x x . x . x . x . x .|x . x x . x x . x . x . x . x .
o . . . . . o . . . . . o . o .|. . . . . . ø . . . ø . ø . . .

This often works for rhythms of djaa-family, especially for kensereni and sangban, but is not always possible. The following dunumba pattern doesn't work for unternarization, due to the "impossibility" of the bell :

1 & a 2 & a 3 & a 4 & a
o . o o . o . o . o . o

1 e & a 2 e & a 3 e & a 4 e & a
x ? ? x x . x x . x x . x x . x
o . . o o . . o . . o . o . . o

If you consider a rhythm such as wolosodon or yankadi, where all strokes are on the beat or just before the beat, you only need to "unswing" the pattern. In this case, two measures of 12/8 turn into one measure of 16/8 (here you could talk about binarization, cause 3 turns into 2, but the result is still quaternary) :

1 . . 2 . . 3 . . 4 . .|1 . . 2 . . 3 . . 4 . .
o . o . . ø . . ø . . o o . . o . . ø . . ø . .

1 e & a 2 e & a 3 e & a 4 e & a
o o . ø . ø . o o . o . ø . ø .

Regarding rhythms of the other ternary feeling, you need to shift them by 4 pulses before unternarizing, in order to take into account the beat-shift theory. For example mendiani :

1 . . 2 . . 3 . . 4 . .
o . x x . o o . x x . o (kk)
x . x x . x x . x . x .
o . o . . ø . . o . o . (sgb)

turns first into

1 . . 2 . . 3 . . 4 . .
x x . o o . x x . o o . (kk)
x . x . x . x x . x x .
o . o . o . o . . ø . . (sgb)

Then you get the following :

1 . . . 2 . . . 3 . . . 4 . . .
x . x . o . o . x . x . o . o . (kk)
x . x x . x x . x . x . x . x .
o . . o . . o . o . . . ø . . . (sgb)

Does it answer your question ?

Vincent

[ternarizator]

Our posts have been sent at same time, so I try to answer :

Of course ternarization (4-->3) can be used when you need to play faster and faster. For example, at a time, we (my group) were trying to play Sunu for our dancers, but it was never fast enough. Fortunately, there exists a ternary version of Sunu, so it solved the problem. Nowadays, we begin with the slow version (quaternary), with breaks, djembé solos, song by the dancer, another break, and then accelerate whith a quaternary chauffe, followed by a specific call of our own, and then the dance begins on the ternary version.

We do the same with a ternarized version of Djolé, and we did it with Tiriba, beginning with an unternarized version, followed by the ternary one (the "normal" Tiriba) when the chauffe occurs.

On the contrary, you can slow down a (quaternary) rhythm by applying the 2 --> 3 ternarization, like for woloso and yankadi.

[korman]

Thanks, that makes sense!

For example, at a time, we (my group) were trying to play Sunu for our dancers, but it was never fast enough. Fortunately, there exists a ternary version of Sunu, so it solved the problem.

Just curious, how fast (in terms of bpm) did they want it?

On the contrary, you can slow down a (quaternary) rhythm by applying the 2 --> 3 ternarization, like for woloso and yankadi.

So in terms of speed progression it could work like this?
Slow - Yankadi 12/8
Faster - Yankadi 4/4 (w/o swing)
Even faster - Makru (4/4 double tempo)

[djembefeeling]

So in terms of speed progression it could work like this?Slow - Yankadi 12/8Faster - Yankadi 4/4 (w/o swing)Even faster - Makru (4/4 double tempo)

Never heard that. And I think Makuru is Yankadi without swing... sorry for interrupting...

[ternarizator]

Just curious, how fast (in terms of bpm) did they want it?

At least 160 bpm, up to 180...

So in terms of speed progression it could work like this?
Slow - Yankadi 12/8
Faster - Yankadi 4/4 (w/o swing)
Even faster - Makru (4/4 double tempo)

Yes, it could. And even faster, a possibly ternarized Makru, as I mentioned above (not traditionnal)...

Vincent

[ternarizator]

Another example : koreduga

We use to play it "normally", with song, break, and dance. Then, while dancers are leaving the stage, we play an unternarized version of the rhythm :

1 e & a 2 e & a 3 e & a|1 e & a 2 e & a 3 e & a
b s . s b . t t b . s .|b s . s b . t t b . s . (dj#1)
t . s . t . s . t . s .|t . s . t . s . t . s . (dj#2)
o . o . ø . . . o . o .|ø . . . o . o . ø . . . (kk)
x . x x . x x . x . x .|x . x x . x x . x . x . (sgb bell)
o . . o . . ø . ø . . .|o . . o . . ø . ø . . . (sgb)
. . . . . . o . o . o .|. . . . . . . . . . . . (dnb)

Concerning the djembé 1, one can also consider that the original (ternary) consists of a superposition of the beat (basses) and a truncated "standard pattern" (slaps and tones) :

1 & a 2 & a 3 & a
s . s . t t . s . (= truncated s . s . t t . s . s . ? )
b . . b . . b . .

In this case, the quaternary version should be :

1 e & a 2 e & a 3 e & a
฿ . . s b . t t b . s ., where ฿ is a bass-slap flam

Vincent