Drum Programming (Part 5)
A Series By Warren Cann
This month we look at the importance of backing up data, and explain how to produce more authentic drum patterns.
To date in this series we've learned the value of working with a method to achieve order and consistency. I've shown you how to examine the arrangement of a song and how to construct a song map sheet. I've given you a demonstration arrangement, we've programmed a simple series of patterns into a song chain, and we've filled in a map sheet, all of which takes us to the first, most elementary level. Now we shall start to elaborate upon our initial efforts and take the song to the next stage.
However... since we actually programmed something last month, let's check something very important. What was the very last thing you did? You saved the data, right? Save your work frequently! And not just at the end of each session. Try and save every hour or so. Losing an hour's worth of work is bad, but not nearly as bad as losing a day's worth. So, save regularly; you won't regret it.
Make life easy on yourself and get into the habit of documenting your actions. Be methodical with your saving, naming, and labelling practices. You really don't need the aggravation of trying to figure out which of four files is the most recent version, and it's even worse if you can't find a song at all. I know it's not very exciting, but do the housekeeping and keep the plot tidy, I guarantee it will pay off.
If your drum machine can only save its information via cassette tape-dump, then practice with data you don't mind losing until you understand the procedure and have it absolutely down pat. You don't want to mess around with this operation on a machine full of data which you may or may not be able to reproduce. It's no fun: you can't continue working until you've saved, and if you execute a bad save... Use the cassettes which are specially formulated for saving computer data. Leave enough blank space on tape between saves. Save more than once — always have a back-up. Better yet, save the data again on an entirely separate cassette. This is not being too paranoid. Once your data is saved and verified, label the cassette shell itself, as well as the case; you don't have to duplicate all of the detailed information you're going to write on the case's inlay card, but you must be able to identify the cassette to match it with the appropriate box.
If your drum machine can save (or dump SysEx data) to floppy disk, then use only high quality disks. You won't be able to buy back your lost work with the money you 'saved' buying cheap disks. True, you can usually get away with murder on most floppies, but it only takes one hiccup on the surface of that disk to relegate your data to the Twilight Zone. And make a back-up. When you spill that cup of coffee and capillary action sucks the liquid between the halves of the disk into it's innards [been there, done that — Ed.], you can sit back and smugly reach for your second, back-up disk. No problemo.
Use the facility to name files to give your data a readily identifiable name/abbreviation. I use the software update system whereby data is given a name and a version number, such as 'Song Title V1.0'. Subsequent minor changes get logged as 'Song Title V1.1', 'V1.2', etc. Major changes increment to 'V2.0', and so on. Don't name your files 'some of the time'; name them each and every time, without fail. When your disk operations are complete, immediately label the disk. Do not wait till later just because you're on a creative buzz or you're too tired. No matter how tired or spaced out you are, pull yourself together for the few moments it takes to wrap up your work by doing the save operations. It's all too easy to make a mistake at the end of a long session, so don't let your fingers get ahead of your brain; be slow, be sure, do it right.
Alright, back to the song. I would like to take this opportunity to address one of the single most common failings of non-drummers who programme drum parts — the failure to take into account the physical logistics of playing a drum kit. After nearly 50 years of music created with the contemporary basic kit — snare, bass drum, toms, and cymbals — it's no surprise that we're attuned both to its sounds and to the way that it is played. You don't actually need to be able to play drums to notice when something impossible is being played — most people's ears immediately catch the fact that something odd has occurred, even if they can't tell you exactly what it was that grabbed their attention, hence the initial shock value and use of fast, syncopated bass drum rolls (the drum machine equivalent of 'N-n-n-n-nineteen').
One characteristic of the drum machine, much used and abused, is its inhuman ability to 'play' figures too complicated and/or rhythmically independent of each other for any mortal to actually play. And why not? Whoever said, "The only rule is... there are no rules" was my kinda guy. However, there are times when it's prudent not to follow this dictum too closely when rhythm programming, because you will immediately shatter the illusion of the music being 'played'. Whether you're trying to sound as 'real' as possible, or are creating obviously programmed beats, it's important to realise that our ears are accustomed to certain rhythmic characteristics associated with the acoustic drum kit. In my opinion, the best approach is to attempt to generally remain faithful to certain basic principles when programming drums and percussion — I'll let you argue amongst yourselves as to whether 'we like it because we're used to it', or whether 'we like it because we like it'.
A drummer with a full complement of limbs has only four 'sources' with which to achieve polyrhythms, and they obviously can't be in more than one place at any given moment. No matter how clever or dexterous the player, there are physical limitations that cannot be avoided. For example, novice programmers tend to forget the incredibly simple fact that when a drummer plays a fill he or she also stops playing the hi-hat or ride cymbal. Obvious, yes? But many people make the rudimentary faux pas of programming their fills directly on top of the hi-hat pattern, then leaving it to drone on relentlessly throughout the entire song, regardless of the fact that it's a dead giveaway that a drum machine is being used. OK, a drummer can play a fill in which the right hand continues to play the ride or hi-hat while the left hand plays accents on various drums (and it's a good technique when you don't want to really make a meal out of things, yet still want to punctuate something), but that's more a matter of using syncopation as rhythmic variation. What I'm concerned with here is your average, full-on, drum roll/fill. So, just when should the hi-hat stop?
Drummers tend to begin a fill with their strongest hand leading; assuming the drummer is right-handed, the hi-hat should stop immediately upon the right-hand beginning the fill. You can almost always get away with the hi-hat carrying on right up to the first beat of the fill. Let's program a very basic fill to use in our song. We haven't used pattern 08 yet, so copy pattern 01 to pattern 08. Before we continue programming the fill, erase the last of the pattern's four snare beats. One of the snare beats we're going to add is going to fall on this beat, so this way you won't inadvertently have two MIDI events on the same beat (on older drum machines, you don't need to erase this beat, as the machine isn't that smart; with the polyphonic capabilities of newer models the precaution is required or you'll get two snares flanging against each other). Now, add three snare beats to the end of its second bar as shown in Figure 1. Sound familiar?
If you like, you can now slow the tempo down a little to help give your reactions an edge. There's no law against working on your material out of real time, just don't slow things down too much or you may lose the ability to feel that particular groove; instead you'll start to hear a new one. Hit the erase button for the first snare beat of the fill, and hold it down through the remaining two beats. The 2-bar pattern you're working on is, no doubt, looping round and round, so make sure you release it in time for the first downbeat at the top of the pattern. If you're late and accidentally erase the first hi-hat beat, it will no longer segue smoothly into any pattern before it... the pattern with the fill will glitch because it has lost its first hi-hat beat.
Normally, that's all you need with regard to your initial exit point, but you can experiment; depending on the type of fill you're attempting, sometimes losing the hi-hat beat immediately in front of the normal cut-off beat works well, too. Just try to visualise the movement of the hands during the fill. You'll note in Figure 1 that I also deleted the last bass drum beat from our parent pattern 01. If you leave it in, it will sound odd because a drummer would not normally think to play that one — he's in 'fill mode', it's a momentary change of gear, and that beat would muddle things. If you want more emphasis here, you can (as I did) add a tom to those three snare beats, or even tuck a bass drum underneath them, as long as it's the same figure as the snare.
By the way, while our fill was a very simple one, others you may attempt will not be so straightforward. You'll find that during the creation of complex fills, you may achieve a certain degree of success, then you screw up and have to start all over again. Or you'll get partway through it and stumble onto something that, while not being exactly what you had in mind, sounds quite interesting in its own right — but you lose it if you carry on, still seeking that first idea. So, make a copy and fool around with that. Cover yourself. If you get halfway through programming a fill and stumble upon something you think you might want later on (especially if you don't think you could deliberately program it again), copy it to another location for safekeeping. Pay attention to happy accidents, there's certainly no shame in using them.
So far, we've removed the conspicuous hi-hat beats from during the fill, but we're still not finished. You also need to think about when the hi-hat reappears after the fill. Drummers also tend to complete a fill with their strong, leading hand hitting a crash cymbal on the first beat of the following bar. Therefore, the first beat of the bar immediately 'after' the fill isn't generally played on the hi-hat. We know that the last beat of the bar that contains the fill is also considered the first beat of the following bar — they're the same point in time, remember? — so, because that's the hand hitting the punctuation of the crash, it doesn't hit the hi-hat. Naturally, the hand can't be in two places at once. Yes, the other hand might hit something; another crash cymbal, or perhaps the snare, but there's not much point in hitting the hi-hat. It's a drummer thing, trust me.
It's also very unusual for most drummers to include a hi-hat beat on the second eighth note in the subsequent bar. It's normally left out due to the length of time it takes for the player to move his stick back from the crash cymbal onto the hi-hat, the exception to this rule occurring at slow tempos where there's enough time to make the physical movement. Selectively erase the hi-hat beats that fall on the first two eighth notes in every pattern that follows a fill (see Figure 2); we've created three patterns with crashes so far (patterns 02, 04, and 06), so make the change to these. The three patterns are already marked for specific use, due to their having a crash on the first beat, so you're not interfering with anything, you are merely refining their application. "Why bother?", you might say. "I can't hear the hi-hat against the power of the crash, and the other beat is barely noticeable. Why wipe them?" Losing those two hi-hat beats is a small point, true, but it makes a difference. I would consider this operation a basic one vital to your programming — when combined with losing the hi-hat during the fill, the entire effect is well worth the effort, immediately freeing up the fill to sound punchier.
The hi-hat or ride cymbal part within a rhythm is the closest link between the drummer and his own internal metronome. All parts lock together integrally to produce the whole, but the hi-hat part is fundamental in producing the groove in which the kick and snare reside. Therefore, it's often the weakest link in a programmer's efforts at achieving a really effective rhythm track. The drum track in our song will definitely suffer if the hi-hat isn't paid the attention it's due, so time spent refining the hi-hat parts is definitely time well spent.
Feature by Warren Cann
Previous article in this issue:
Next article in this issue:
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!