In Search of the Perfect Slap...

When you listen to a really good West African djembe fola play for the first time it is jaw dropping. My first experience of this was in The Gambia in 2002, watching Maitre Samsou perform with his group Jolibah. Yes, he was an amazing showman and did all the actions with his hands whilst playing but what fascinated me the most was the sound...I'd never heard anyone play with good technique before and I couldn't believe how he made it sound like two drums at the same time! I remember the time when he played at a near-by compound and his ex-student who had stolen from him turned up to play. After half an hour or so of swapping solos Waido said he was tired and wanted to play dunduns, so Samsou took the solo and continued to perform for the next hour and a half! Samsou played like a demon that day (he clearly had a point to make), and as I watched the blur of his hands my jaw literally dropped and I watched with my mouth open. It's a bit like your first love...in your life you'll see many djembe folas playing, but nothing will quite come close to the first time!

Over the years I've studied with several djembe folas and my technique has improved (sometimes painfully slowly) to the point where I can, at least in practice, definitely make the right sounds. I know this both through listening to recordings of my playing and through analyzing the wave-spectrum of my slaps and tones (more of this to follow) Iya Sako has helped me a lot in this respect, though my problem now is one of consistency (more about that later too)

As a student of the djembe it is both magical and frustrating in equal measures. When I play under the right circumstances, at the right speed, and well within my ability range, then I can sometimes sound like a djembe fola. And when this happens, often when I'm practicing down the studio, it is the greatest pleasure I have ever experienced; suddenly I am playing effortlessly and deep rich tones are blended with melodic slaps to make a melody that can sound like two djembes. It feels a bit like one of those flying dreams I used to have as a kid. Just magical.

But under pressure that magic can be shattered (the 'look mum no hands' syndrome...) and my good technique can disappear. It doesn't feel nice and it doesn't give me pleasure to play when this happens. Very frustrating...

Now then, while as a djembe player I practiced and practiced, the scientist in me (I have a Ph.D in Cognitive Psychology) wanted to know how on earth djmebe folas made the drum produce two such distinct and melodic sounds, and why they were so different to my sounds (and those of most non-africans I have heard play)

When you learn you are taught to close your fingers for a tone, and to let them open and bounce to get a slap, but clearly it's not that simple, or everyone would sound like a djembe fola! Furthermore, every djembe fola you learn from has a slightly different technique, and worse, most of them don't even close their fingers for a tone, making it very difficult to work out what on earth is going on!

Then one day I joined the website djembefola.com and read some discussions on technique. I chatted to a guy called Michi Henning and he put me on to an interesting paper written by a guy called 'Albert Prak' in about 1998; it was called 'The Physics of the Djembe' and in it Mr Prak uses a number of techniques to investigate how well the djembe conforms to the theoretical numbers and how djembe folas produce the sounds that they do.

OK this bit is going to be a bit techy but I'll make it as simple as possible. Deep breath...

When you hit a djembe properly towards the middle of the drum you get the bass note. This is mostly produced by the air inside the drum getting squashed, then vibrating up and down (it's called 'Helmholtz resonance'). Much harder to achieve are the two sounds you can make on the edge of the drum, the slap and the tone. These sounds are unconnected to the bass (if you tighten the djembe the bass will stay at roughly the same pitch) but rather are produced by the skin (mostly). Luckily physicists, engineers and mathematicians have studied this; it is called the physics of circular membranes if you want to look it up...

Just as the way in which a string vibrates is governed by the harmonic series, the way a circular membrane vibrates is very predictable: it vibrates with particular modes each of which has a frequency or pitch. So if you hit the edge of the djembe willy-nilly then you'll get a combination of different notes. Clearly djembe folas manage to precisely control which notes are produced and which are suppressed...

Albert Prak found that the frequencies produced by the djembe closely match the theoretical values, and that djembe folas control the sound in the following way (deep breath again...):

When they play a tone djembe folas manage to suppress all of the higher notes (called partials) and emphasize the lowest (called the fundamental). Prak called the fundamental M1, and then the higher frequencies M2, M3. M4 and so on. So a tone is effectively all M1, with hardly any M2, M3 or M4. This is very interesting in itself since the most well-studied drum that is a circular membrane is the timpani; according to the wonderful article entitled, 'The Well Tempered Timpani: In Search of the Missing Fundamental', the note that rings out on a timpani is M2! M1 is considered too fleeting and short-lived to contribute to the sound of a timpani yet somehow djembe folas manage to get M1 to ring so loud it can cut through an entire drum ensemble! Indeed Prak notes that the M1 partial has a very fast decay and kind of collapses in on itself like a pudding (I love that choice of words...)

Next Prak found that the slap is mostly M3 and M4 with a bit of M2 and hardly any M1. It sounds complicated but there is an easy way to show it visually. I recorded myself playing, then dumped the file into 'Audacity' and plotted something called a 'spectrum-analysis'. This is basically a graph, showing peaks and troughs that relate to the different notes you are producing in your slaps and tones. Here's a graph of one of my typical slaps:

The first rounded bump on the left is an echo of the bass. The next hump is the tone frequency (about 400Hz). The next 2 or 3 sharper peaks are M2, M3 and M4, then the peaks tail off towards the higher frequencies.

As a comparison here's one of Mamady Keita's slaps. The fact that it's more spiky is probably due to differences in recording and EQ but the general shape is telling:

Again, from left to right we have a bit of bass, then a double spike around the M1 area. The interesting thing here is that the next peak, the M2, is quite small (as low as the bass actually), leaving M3, M4 and M5 poking the top of the graph.

So this is the difference between my slap and Mamady's slap: While my slap is made up of mostly M2, M3 and M4 Mamady's slap is made up out of M3, M4 and M5! Moreover I have analysed the slaps of many different djembe folas and all of them show the same pattern. Eureka! This is the secret to the slap!

What really surprised me though was the variation between different slaps by the same djembe fola. Even the great Mamady Keita doesn't always produce slaps of this caliber. Here's another spectrum of one of his slaps (taken from the same album):

Here all the spikes are equal except for one really high one. This one has a slightly muffled sound.

Most djembe folas have at least two different slaps, so they can control the balance of these harmonics to a certain extent. For example what Mamady calls the third slap (or tonpalo) is achieved by getting almost entirely M2, whilst suppressing all the other frequencies. Here's my effort at the third slap:

Almost entirely M2! Not bad eh? The problem is I can't get it with anything like the consistency I need to use it effectively.

Here's a graph of my tone:

And one of Mamady's:

Quite similar I think you'll agree. He does a slightly better job of suppressing a couple of the higher frequencies, but my tone is pretty close...

 This is all well and good, I hear you say, but how do you actually do it? Well, Mr Prak provided some valuable insights here too: He found that a slap is produced by a small surface area (the finger tips) contacting the skin for a very short period of time (just a few milliseconds). A tone, on the other hand, is produced by a larger surface area (the whole of the underside of the fingers) contacting for about 1.5 times as long as for a slap. Prak suggests that the trick is to 'press' the higher frequencies out of the sound while getting the fingers to bounce away just in time to let the fundamental frequency ring.

The reason that closing your fingers helps get a tone is because it helps you control your fingers and stops them bouncing off too quickly before the dampening effect has occurred. The reason getting a really good slap is so hard is because often too much of the fingers contact the skin, and any tension will stop them bouncing sharply away.

This information provided the last piece of the puzzle for me, and really helped me to visualize what my fingers needed to do. It also helped me to understand how the technique I'd been studying for years actually worked. I recorded this video shortly after incorporating this into my technique:

Now all I have to do is fine-tune my slaps and improve my consistency. Sounds easy doesn't it? Don't hold your breath.