PROGRAMMING UPBEAT PATTERNS

 I'm sure we're all familiar with variations of this rhythm:


   1.......2.......3.......4.......
 S --------0---------------0-------
 R ---------------0---0------------
 K 0---0-------0---0---0-----0---0-

 Where S is snare, R is snare rim or light snare drum, and K is kick drum. This will henceforth be known as the "hip-hop upbeat pattern" because of what the rim is doing. And what *is* the rim doing? Quite simply, the rim is adding interest to a very basic drum pattern, and it is my opinion that the addition of these "upbeats" to practically any rhythm- even a two-step- can add rhythmic interest.

 Let's look more closely at the five hits between an eighth note after beat two and and an eighth after beat three (or 2'5" and 3'5" respectively, where the first number represents the location in quarter notes and the second the location in thirty-second notes.) What we notice is that the entire group of notes is exactly one quarter-note in duration, but more importantly, that this quarter note is *staggered*,-- that is, it begins on a weak beat and ends on a weak beat. Also of note is the fact that beat three- a very strong beat- is merely continuing a pattern started in the weak beat 2'5". The "downbeat" on beat three can be eliminated alltogether, leaving the staggered quarter note between 2'5" and 3'5"- and staggered eighth between 2'8" and 3'4"- in place. While our rhythm is undoubtedly a simple one, it is syncopation embodied,-- the accent on weak beats, the staggered figures; all the elements are there. This is one of my complaints with Classical theories of rhythm,-- simply taking note of which notes fall on "strong" beats and which on "weak" beats based on an underlying motor rhythm is not enough; attention must be paid to which beats are *structurally* strong and *structurally* weak based on the rhythm itself. In our rhythm, beat three is a stong motor beat- there is even a hit on the beat- but, structurally, it is very weak. The whole point of a music theory is to analyze what gives a specific piece of music its identity, not what unifies all music everywhere.

 The remainder of this essay deals with the dirty work of drum programming as it relates to the upbeat. Our notation will also change; from now on the relative location of drum hits will be given in inter-onset-intervals- that is, a numerical indication of time- which, for the purposes of convenience, will be the same as the resolution of most modern sequencers-- 1920 ppqn. So if I say a drum hit is 960 after another hit, I mean it is an eighth note later.

 The question remains unanswered,-- what *is* the rim doing. Or, more specifically, what *should* it be doing. There are 960 possible relative locations for it in the eighth note between 2'5" and 3'1", some of which work better than others. There are three obvious choices:

 THE DUPLET:

 This one's a no-brainer. It starts with a kick at 2'5" and continues with a snare at 2'7" and then to another kick at 3'1" and a snare at 3"3 and finishing with a kick at 3'5". In inter-onset-intervals:

 2'5"                    3'1"
 K-----------S-----------K-----------S-----------K
 480         480         480         480

 The final appeal in drum programming is the "groove." Nobody can explain *why* something "grooves", but most musical people can agree when it does. Just because a rhythm is easy to notate or "on the grid" doesn't necessarily mean that it has more of a "groove" than others that don't, and this rhythm is a prime example. The snares want to "breathe" a bit more- but not too much- suggesting that the snares should be moved farther to the right. Later, we will examine an extremely off the grid rhythm that does just this. As for this rhythm, it might be appropriate for 190+ BPM tempos, but at slower tempos it just sounds sloppy.

 THE DOTTED-NOTE FIGURE:

 This one is the one first presented at the start of this essay. It is a classic dotted-note figure:

 2'5"                    3'1"
 K-----------------S-----K-----------------S-----K
 720               240   720               240

 In this instance, the snares are moved *too* far to the right. This only seems to groove at extremely slow tempos, so that the separateness of the kick and snare is articulated.

 THE TRIPLET:

 And, of course, we have the triplet:

 2'5"                    3'1"
 K---------------S-------K---------------S-------K
 640             320     640             320

 This is closer to what we want, but it's still not quite it. As well, I have a reservation about using triplets, as they want to "roll along", leaving the listener expecting further triplets. Triplets are striking, to be sure, but, in my opinion, they sound out-of-place in a drum'n'bass context.

 Well, those are our three obvious choices. If we find none of them are to our liking, we can try any of the 957 more. One useful exercise is the old "greater than" or "less than" trick of guessing a number between 1 and 1000 in ten guesses. In this instance, we would start at tick 512, and if we find the beat is "pulling towards" a beat, we move it forwards by 256 ticks, and if we find the beat is "pushing away from" a beat, then we move it backwards by 256. Continue this process eight more times until you have determined which beat you want to a resolution of one tick. If nothing else, this process is useful for learning to *listen to* your beats,- to develop a trained ear for rhythmic nuance,- and hear which modifications "want" to be made.

 I have my own theories of groove, mostly centering around nontuplets,-- that is, triplets of triplets. Alas, I know of no sequencer whose resolution is a multiple of nine, so inter-onset-intervals will be given in decimal points. The theories that yielded the following syncopation still haven't fully metastatized, but I humbly give you...

 THE TAGUTCOW FIGURE:

 2'5"                    3'1"
 K--------------S--------K--------------S--------K
 613.333        346.666  613.333        346.666

 Of course, the decimals repeat infinitely.

 In less arcane notation that's 133 ticks after 2'7" at 1920 ppqn.

 2'5"        2'7"        3'1"        3'3"
 K--------------S--------K--------------S--------K
 480         |133  347   480         |133  347


 There are "sympathetic numbers" in my theories that vary proportionally according to resolution, and at 1920 ppqn, these are 133 and 107. In my old sequencer Vision, which operated at 480 ppqn, I used the numbers 33 and 27.

 This essay began with an examination of the hip-hop upbeat pattern, in which the upbeat was accentuated in two eighth notes in the center of the bar. However, in basic on-the-grid rhythms, the upbeat can be accentuated *across the bar*. Instead of a staggered figure, we have an entire staggered *rhythm*. A whole new world is opened up with this superimposition of rhythms. Try moving tamborine rides 613 ticks to the right, and tweaking the velocity. Something I haven't tried is using this syncopation recursively, or using a delay to accentuate the upbeats and the upbeats of upbeats. The downbeat can fall on any one of these upbeats; if you want to be adventurous you can experiment with polyrhythms.

 Even if you don't believe the Tagutcow figure grooves, the moral is clear; there exists an infinitude of rhythms "off the grid" that groove as much as or more than many "on the grid." Experiment!