Drum Programming (Part 10)
A Series By Warren Cann.
There's more to how fast 'n' frantic a track feels than merely the bpm setting on your drum machine — this month we look at aural illusions...
A topic not commonly discussed in relation to tempo is that of aural illusions. If you're using a drum machine in a live situation you've probably experienced, at one time or another, the rest of the band frantically shouting at you, "Speed up! It's dragging!" or desperate looks accompanied by, "Slow Down! It's too fast!". Thing is, you know full well that 127bpm is exactly what you'd all decided upon in rehearsals. The song felt great at 127bpm at the gig the night before, it's set at exactly 127bpm — the drum machine's LCD isn't lying — so what's happening?
Unlike people, a drum machine isn't susceptible to fatigue or excitement. If the band is vibed up, the gig is going well and the audience is projecting a lot of energy back onto the stage, then it's natural to feed off of that energy and (deliberately or not) play harder. This usually equates as 'faster'. Or, if touring frazzle is setting in due to late nights, early morning calls, and too much travelling, it's understandable if people aren't always feeling 100%. They're tired and their perceptions are off; they interpret the drum machine's tempo as being too fast. Well, just grit your teeth and stick to the correct, rationally predetermined tempo in all but the most blatant of emergencies.
How many times have you seen a band blast through their songs a hundred miles an hour faster than their recorded versions? I mean, not just 'somewhat faster', but fast to the point where it detracted from the music? Again, it's the human factor. They may have played the material so much that they're over-familiar with it; it's not that they're bored exactly, they're just unconsciously succumbing to the understandable urge to simply play it faster in order to get the same buzz out of it.
If you've ever wondered why so many songs seem to be in the area of 120bpm, I can't offer you a definitive explanation. There are various theories: one says it's a tempo that somehow relates to the human heart-beat; another says this is the speed that most of us walk most easily, and that the ideas for countless pieces of music have been generated while someone was walking along, wrapped in their thoughts, humming snatches of impromptu melodies. I'm inclined to agree with this one. Pragmatists will say it's because we've heard it so often, for so long, that we're brainwashed... but that doesn't explain why we've heard it so much in the first place.
Whatever your tempo is, the initial selection of it seldom proves to be a problem — it suggests itself naturally when you're writing. Confusion tends to arise when it's time to take that demo into the studio and do it for real. Sometimes there's no problem; the click track is set at the same tempo used for the demo's drum machine tracks, and everyone is happy. Sometimes it's different... paranoia sets in for one reason or another, perhaps on grounds as simple as the tempo never being written down (and the programmer has forgotten it). How do you determine exactly what the tempo originally used was? You could play your demo tape and try to match tempos on your drum machine via the Tap Tempo function, or just mess around adjusting the tempo till the tape and drum machine appear to run 'in sync' for more than a few bars — but that kind of trial & error method can take too much valuable time. The easy way is to determine the tempo is via the stopwatch function on your watch (or any other such accurate timepiece). The formula for determining bpm is: 60 + (duration of measure in seconds) x (number of beats). Here, in plain English, is an example...
Listen to your tape and start the timer on a downbeat which begins a measure — try and be as accurate as you can, hit that button dead on the beat — count off eight beats, and stop the timer on the ninth beat (beginning of the third bar). Again, be as accurate as you can. Check the reading on the timer; it might show something like 3.94 seconds. Make a note of the time. Try it again — you'll probably get a slightly different reading this time, and once again write down the time. If the tempo of the track was originally set by a drum machine, you might expect these two times to be exactly the same, but your reactions aren't perfect, so try it again. When you've timed it often enough for the numbers to settle down, you'll have eliminated the errors incurred by your own imprecision. The numbers don't have to be exactly the same to the hundredth of a second, just close enough so that you see the pattern.
Let's say that you were happy with a reading of 3.99 seconds. Get a calculator, key in 60, divide it by 3.99, multiply that by 8 (the number of beats you counted), and you'll get your tempo, in this case 120.3bpm, or just a hair over 120 beats per minute. If you got 3.76 seconds, the formula will tell you that the tempo is 127.65bpm, 4.11s gives you 116.78bpm, 3.48s gives you 137.93bpm, and so on. If you're really a stickler for accuracy, time it 10 times or more, and find the average of the time before you make the calculation.
So, having done this, you'll have your tempo and can carry on knowing that (cassette transport motor speeds aside!) you got it right. Just bear in mind that no two drum machines' version of a given tempo will ever coincide exactly — 128bpm on one drum machine will be very slightly slower or faster than 128bpm on another. Also, if you should do all of your drum/percussion programming on a drum machine rather than a computer sequencer, because you like the editing facilities, and use that as the master to clock your computer's software sequencer as a slave, don't be surprised to see the drum machine display 128bpm while your sequencer shows the tempo constantly flickering between 127 point-something and 128 point-something.
If you're timing something that was recorded by a real drummer, try timing different parts of the song; it's bound to have sped up or slowed down in places. This brings me to my next point: a very effective technique for eliminating the rigidity of songs using drum machines is the subtle speeding up and slowing down of tempo.
People can't keep strict tempo all the way through a song unless they use a click track, and when a band plays live they naturally tend to push when they get to choruses, or lay back slightly during moodier parts of songs. The astute pushing around of tempo is the natural result of experienced musicians playing material with which they're confident. Sometimes it's noticeable, sometimes it's not, but it's almost always there. The music has a more natural ebb and flow to it, and it's one of the reasons why some people refuse to use a click track when recording.
You can be shrewd and take advantage of this effect in your programming by incorporating such tempo changes during your song. If you start off with a chorus or a riff that a chorus is sung over, then drop the tempo back a fraction when you get to the first verse. Don't go mad or it will drag terribly and sound like the track has just fallen apart. Alternatively, when it gets to the first chorus, speed it up a fraction. But take it easy — we're talking about tiny amounts here. It's amazing how the ear can discern such small changes. Without any reference points, it's very difficult to say whether a song is 2 or 3bpm faster or slower, but with a direct A/B comparison — like when you leave one tempo immediately for another — you can notice the slightest change, especially on a track you are playing along to. Some people might think an increase or decrease of even 1 bpm too much — if that's the case, make your tempo changes in tenths of a bpm. It's very effective; the listener may not really notice it on a conscious level, but if you played him/her the exact same song, one with the changes and one without, it's likely they'd prefer the version with the changes even though they couldn't tell you why.
A little experimenting will soon show you what degree of change you're comfortable with, and make sure you're happy with it before you continue with the track. Unless your drum machine is syncing to tape via SMPTE, you probably won't be able to go back and re-record the drums if you later decide you went overboard with the tempos. Drum machines which sync to tape via FSK carry the tempo-change information as an integral part of the FSK code, and once you've laid the code down to tape you can't go back and make alterations to the tempos; you have to re-record the code itself.
The feeling of a song can be changed without actually changing the tempo at all. Many drum machines (and any worthwhile sequencer) now have a feature known as Shift or Offset, which enables you to move a part forwards or backwards against the internal clock. By shifting the beats a few clock ticks ahead you liven things up and add tension; the feel is more energetic, insistent, more 'up'. This is known as pushing the beat.
If you shift the beats backwards, the reverse happens: it's lazy, mellower, more laid back. This playing behind the beat is usually referred to as 'laying back'. Be careful with this one, however, because the difference between relaxed and 'dragging' is very narrow. You can shift just the snare or the whole kit, or whatever you want. A single clock pulse is such a small increment of time that you won't be able to hear the difference as such, it's more a matter of feeling the difference. Experimenting with this effect is easier if you already have a few other instruments playing their parts, as it helps you judge what you're doing with the drums. Have fun, but don't go berserk, or you'll lose the strict standard quantised groove by trading it for a clumsy one which never sits right.
Quantising corrects small timing errors by moving your data to the nearest beat division, according to the level of quantisation selected. This means that imperfections in your performance are tidied up automatically. All of the beats that were a fraction late have been shifted forwards in time to be 'correct', and all of the notes that were a fraction early have been moved back. It's a very powerful tool which can be an invaluable help, but it can't perform miracles — if your playing is so desperately out of time that quantising isn't helping, then you're better off programming in step time until it improves.
Many software sequencers allow you to set percentage values for quantising; instead of everything being moved exactly on the beat, you can specify a range to either side of the quantisation intervals where the notes remain as you played them — only notes outside the designated zone are pulled into time. Prudent use of this feature can make a huge difference to improving the feel of your rhythms.
Quantising gets seriously interesting when you start to create your own custom quantising templates. If you've shifted the snare ahead slightly and layered some percussion (recorded with several different plus/minus percentages), you can then command the sequencer to utilise the aggregate groove of the entire rhythm to act as the template for any future quantising — especially interesting when you apply it to synth parts, like bass lines and so on. It's fascinating, exotic stuff, but very difficult to master, as everything becomes so inter-related. Drum machines would be better off by far with even a fraction of the non-destructive quantising power of a package like Cubase.
Feature by Warren Cann
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!