Odd Time Signatures In Composition

“Movement 03/365″ by odolphie via flickr

Odd time signatures. I used to dread them. Their unexpectedness, the way they would challenge my musicianship, the sheer sight of those unusual figures would give me sweats. Yet, some of the odd meter based music I was listening to or playing was truly appealing to me. I decided it was time for me to explore this (then) uncharted territory, through composition first.

I quickly found out that some things worked and some others didn’t. The critical factor for me was to be able to come up a rhythmic thread that united melody, harmony, and, of course, the rhythmic frame. In this article, I will present a few concepts that I found to be particularly effective.

The “mother concept” of these tools was to find and apply claves (rhythmic patterns used as tools for temporal organization) to my odd meter based compositions. A strong clave would have enough rhythmic DNA to be easily remembered (and then performed) and to allow for a smooth flow of the music. I found that having such a strong backbone to my compositions made them work.

Although 3/4 (and its 3/8 sibling) is technically an odd meter, we generally refer to odd meters as rhythmic signatures containing any odd number of beats from 5 on, i.e., 7/8, 9/4, 11/16, 17/8 etc. Most of the examples here are in 5/4, but the concepts can be used to any odd meter!

On this subject, I would like to recommend two terrific books. The first is called “Creative Rhythmic Concepts for Jazz Improvisation”, by Irish bassist and composer Ronan Guilfoyle. A strong clave would have enough rhythmic DNA to be easily rememberedThe second is “Factorial Rhythm” by the iconic Mick Goodrick – the pages devoted to odd meters should offer you some interesting perspective.

In this article, I have deliberately chosen to not include any pitches whatsoever. The first reason for this is that the harmonic and melodic implications of these concepts would require a few dozen pages to introduce. Also, I believe that discussing rhythmic concepts exclusively will give you more room to experiment and find your own compositional ideas. At last, all of these concepts can be applied to melodies, bass lines, comping parts or drum grooves, all of which have a very different roles from one other – any hint at this would defeat the underlying self-exploration incentive of theses pages.

At the core of my early findings, it appeared to me that the smallest building blocks of odd meters (or any meter for that matter!) were two rhythmic cells, a duplet and a triplet, referred to in this article as “2 cell” and “3 cell” respectively (notice that then have two kinds of beats – a short one and a long one). In this article, I will use a quarter note as my duplet, but any rhythmic value is fine. Adding a number of these cells together can produce any type of odd meter. For instance, a bar of 13/8 can be spelled 3 2 3 3 2, or 2 2 2 3 2 2… you get the idea!

Here are our 2 cell and 3 cell with their subdivisions:

Now, having too many beats or rhythmic accents might not be what you are looking for in terms of rhythmic density. Here are a first set of manipulations you can use. Here, our 5/4 clave reads 3 3 2 2. We can tie some of those cells together to get less rhythmic activity (first staff), or just mute some cells (second staff).

Conversely, you might be looking for more rhythmic activity.

Here’s a chart of subdivision for 2 and 3 cells:

The next logical step would be to replace our basic cell with one of the more developed cells. Here’s an example of various manipulations of the 3 cells:

When using subdivided cells, I would strongly suggest that you stick with one of them (one type of subdivided 3 cell for instance) so as not to confuse the original clave.

So far we have seen these odd time claves as being one dimensional elements. But there is so much more to them! Let’s begin discussing the relationship between melody and comping. Here’s a great example, Pat Metheny’s “the First Circle”. The time signature here is 22/8 or… 11/4. The notation used here (one bar of 12/8 followed by one bar of 10/8) should help with the reading. The top staff is the melody while the bottom staff is the clapping pattern used in the intro of the song:

The clave for this song is 3 2 3 2 2 3 3 2 2. Notice the rhythmic manipulation of the 2 and 3 cells. Now, take a look at how the two interweaving lines. The melody line (top staff) is being played off of the rests of the clapping pattern (bottom staff). While the entire pattern is long and intricate, this allows for some room for the melody to “sing”.

Generally, though, I would suggest that you adopt a freer modus operandi as far as the melodies you write are concerned. Metheny’s song is a great example of melody and comping sounding at different times so as not to interfere with each other, while abiding by the same clave. I encourage you to explore different ways to superimpose over the odd meter clave, i.e., concerted melody and clave, or melody and clave independent from one another. The right mix is probably somewhere in the middle. Strive for balance, a mix of lightness AND grounding.

Augmentation and Reduction can generate new ideas from the original rhythmic pattern.

On the subject of superimposing melodies over comping patterns, here’s something worth investigating. Very common in Indian music, augmentation and reduction can generate new ideas from the original rhythmic pattern. For those of you who are unfamiliar with this, augmentation and reduction are very simple concepts – simply multiply or divide all notes and rests values by the same number! The note value and rhythmic placement of the notes is preserved relative to each others. Here, you will see how the melody sounds when it is “processed” in 1/2 reduction, 2X and 1,5X augmentation. The original pattern (5/4, 2 3 3 2 clave) is still played underneath. Note that when using a 1/2 reduction, you will have to play the melody twice to fit over the original pattern. Likewise, 2X and 1.5X augmentation will take 2 and 3 bars of the original pattern respectively in order to have a complete cycle. The resulting polyrhythms sound pretty hip! You can also use the augmentations/reductions for another section of your piece.

Now, here’s something surprisingly fun to do that has yielded interesting results for me. As a listener, I have always liked music that had sophisticated forms. As a composer, unusual forms have helped me generate fresher music, as I was forced to get away from more common time signature formatting.

Here’s the (very) basic outline of my tune “Pendulum” (that you can listen to on my website http://www.edouardbrenneisen.com):

What you see here are the two main phrases of the song, 5 bars of 4/4 for the A section, while the B section consists of 4 bars of 5/4. Of course, the melodies of my piece are NOT mere series of quarter notes. Notice, however, how many beats comprise the A and B phrases? 20 each! The grouping of the underlying rhythmic patterns, as well as the harmonic rhythm (not shown here) make the two sections sound entirely different, but they end up having the same overall “length”. You can use this device to balance the overall shape of your compositions, and yet add a distinctive rhythmic flavor.

And now, the grand finale. While this article’s primary focus was on odd time signatures, here are a few things that touch on polyrythms.

It has been my experience in composing music that the outcome of your writing is not so much about how much music you put into it, but rather about how much you can stretch, modify and develop few ideas (now that I think about it, good improvisation is also largely dependant on that factor). Since coming up with a solid odd meter backbone for the piece you are working on can yield substantial material already (if you go through all the “tricks” I’ve discussed here, chances are you found quite a few variations of the original idea). Here’s how you can make more with less.

Case scenario: you have come up with material in 4/4 and 5/4. I have chosen two very simple rhythmic ideas, a son clave (named after the famous Cuban style that uses this rhythm a lot) in 1/2 reduction, and a 5/4 pattern (3 2 3 2, with the 3 cells subdivided as 2 + 1):

Now, I could use these two rhythmic patterns in two different sections. Totally fine! But what would happen if I played them together? What would happen if I stacked the 5/4 pattern over the 4/4 one and then did the opposite? Something that would go like this:

In order to be effective, you must be careful to assign these patterns to two different instrument groups, say the top staff is the melody and the bottom staff is a bass line. Or you could be playing the top line on a snare drum and the bottom line on a bass drum. You could experiment with placing weaker harmonies on stronger rhythmic elementsThe two systems, 5/4 over 4/4 and 4/4 over 5/4, have the exact same length (20 beats each, as shown earlier), and employ the exact same material. But through careful arranging, we are able to drastically change the dominating context. Same polyrhythm, but different sound. Look at the brackets in the top staves, they highlight the superimposed rhythm. To complete a “cycle” (where beat 1 of the 4/4 rhythm is being played at the same time as beat 1 of the 5/4 pattern), you will need to play the 5/4 thing four times over five bars of 4/4, whereas it will take five instances of son clave over four bars of 5/4.

This technique will greatly expand your rhythmic and form palette when writing music that demands for odd time signatures. Oh, and don’t forget, at this stage, it is still possible to manipulate those rhythms as we’ve seen earlier in this article! Harmonic implications in odd-time signature based compositions.

After all, I couldn’t not touch on the subject of the relationship between harmony and odd meters. While a thorough examination of this rapport is beyond the scope of this article, I would like to mention a few things.

As far as harmonic rhythm is concerned, we are faced with a few different options here. I encourage you to try and work with different pairings as far as strength of the harmony and strength of the rhythmic placement are concerned. In other words, since the distribution of strong and weak harmonic points are disturbed in our odd meters world, you need to first identify when those strong and weak spots could occur. Assigning the corresponding harmonic sounds and functions (strong harmony with strong rhythmic points/weak harmony with weak rhythmic points) is one option, and gives a very strong forward motion to the music. But you could experiment with placing weaker harmonies on stronger rhythmic elements (and the opposite, too).

Likewise, when working with bass lines, you could experiment with pairing stronger or weaker harmonic functions with stronger or weaker rhythmic points. On the subject of bass lines, I found particularly effective to assign a category of pitches (low or high – relative to the range of your line) to a specific category of beat length (for example, “low” notes are primarily being played on long beats, “high” notes are primarily being played on short beats).

Here’s an arranging trick that I found to be quite effective. You can distribute the different elements of your odd meter pattern to different instruments (drums, bass and harmonic instruments). Pat Metheny knows how to get down with some odd time ideas, even if he needs to think hard about it, as seen here. Photo by badosa via flickrFor instance, the stronger beats of your patterns, the “anchor” beats” if you will, could be assigned to bass and drums, while the weaker beats can be played by the harmonic instrument(s). This works particularly well when the harmonic activity is not too high. For even more sophistication, your weaker beats can be subdivided (as shown earlier) into new rhythmic groupings – you would retain the feel of the chosen time signature while obtaining a lighter, more subdued sound.

Oh… should I mention that ALL of these ideas work perfectly well in more common meters contexts? The metronome The “oh-so-important” tool for all musicians! Since we are dealing with short and long beats, I would also advise you to work with a sequencer, a drum machine or a programmable metronome.

More advice

Don’t overlook those important points:

• As much as odd meters are fun, more common time signatures are perfectly fine;
• Any vocabulary you develop in your writing can (and should!) be used in your playing/improvising;
• Odd meters are nothing without a strong sense of pulse;
• Be cautious in your palette of time signatures within a single composition – too much rhythmic “spice” can seriously blur the result.

I hope you find as much interest in exploring odd-time based compositions as I did. Until then, happy composing! Edouard Brenneisen, March 2009