The Theory of Drum Programming

This article was originally published in 2004

Contents

  1. Preface
  2. Tools of the trade
  3. The basics of Beats
    1. Functional elements of the beat
    2. Primary and secondary hits
    3. Kicks
    4. Snares
    5. Rides
    6. Auxiliary and other hits
  4. Feel
    1. Dynamics
    2. Timing
  5. Achieving Realism
  6. Achieving “Fake-ism” (or, Fun things to try)
  7. Timbres
    1. Kick sounds
    2. Snare sounds
    3. HiHats
    4. Cymbals
  8. Drum machines of note
  9. Auxiliary percussion
  10. Tips and other items of note
  11. Appendix: resources on the web

Before we begin…

If you’ve found this article helpful, please make a small $2 donation via paypal by clicking the button. Donation is optional, but very appreciated. It’s not much and it takes only a moment. Thanks for your help!

Preface

This document is a collection of knowledge regarding what is commonly known as “Drum Programming”, or the art of creating drum tracks electronically. I will admit from the start that it is oriented mostly towards “groove” music, and not as much towards styles such as heavy rock or jazz. Those programming for pop, country, dance, hip-hop, or acid jazz (to name a few) will find much here for their benefit.

I would also like to state from the beginning that any sort of theory regarding art or music has numerous exceptions. The purpose of theory is not to restrict us to doing things a certain way or to circumscribe our creativity, but rather to show us the most common ways to accomplish things. The true artist will understand the balance between following the rules (to provide a familiar context for the listener) and creatively breaking them (to surprise the listener and add distinctiveness to the piece). In a nutshell, don’t take this too seriously and don’t get obsessed with finding exceptions to things I say.

Any feedback from you the reader can be sent to me at captnace@juno.com(that email is defunct, just leave a comment), for the purpose of bettering this article. Questions can also be sent, and I’ll answer as best I can. Please put something obvious like “Drum programming thing” or “Drum article you wrote” in the subject line so I don’t delete it, as I get a lot of spam at that address.

A word on notation: It would be impractical to insert standard music notation into this essay, so in order to write about rhythms I’m employing a technique musicians use to talk about them. The four beats of a standard 4/4 measure are referred to as 1, 2, 3, and 4 (wow!), and the eighth notes between them are referred to as “&”. Thus, “1&” is the eighth note between 1 and 2. The sixteenth notes are referred to as “e” and “a”, “e” being directly after the beat and “a” being directly before it. Thus the sixteenth notes in a measure are labeled 1, 1e, 1&, 1a, 2, 2e, 2&, 2a, etc. This is a common practice among musicians of all types, though mostly for drummers who deal more often with rhythm issues.

Tools of the trade

The first thing you need to program drums is naturally a bit of gear that will allow you to do so. I’ll break these down into four categories: hardware, hardware sampler, standalone software, and plugin software. Let’s take a look at each.

  • Hardware: This encompasses things like drum machines, workstation synths, or a computer sound card with built in wavetable. The advantage to using these devices is that they are much more affordable than samplers, and often readily available. Most home studios have a synth workstation of some kind, and it’s not hard to find a soundcard with a synth installed. Probably the biggest disadvantage, however, is timing. Because they rely on MIDI, hardware synths have a natural latency that can make drum tracks sound loose and sloshy. MIDI is a serial communication, which means it sends data one bit at a time. Just playing one note via MIDI requires the sending of well over 20 bits of data. This can cause a significant delay that, while it may not be noticeable on a string track or pad, can ruin a drum feel. The other significant disadvantage to this kind of equipment is that you are stuck with the sounds you have. Although you can do some editing or processing to the sounds, the sample set typically can’t be changed or updated. As we’ll discuss later, drum sounds have a tendency to go in and out of style, so the killer drum kit on your new workstation may be terribly dated in a few years.
  • Hardware sampler: A step up from the hardware synth for the drum programmer is the hardware sampler. A good sampler has features that will make your drum programming much more professional. The ability to update your samples, and to use much larger samples will obviously give you better and more appropriate timbres, and the improved sound-sculpting tools will help you with dynamics and feel. Like any other MIDI hardware, however, you will still have latency problems and timing issues. Regardless, samplers were the pro standard for drum programming for years and are still in use.
  • Standalone software: Over the last half of the 90’s, many standalone music composition programs have evolved, utilizing samples and software synthesizers. Many “old school” technophiles will remember “tracker” programs, like Octamed, Fastracker, Impulse tracker, and Mod tracker. In fact, my personal choice for drum programming is Buzz, a modular environment whose interface is based on these tracker programs. More recently programs with sophisticated and user-friendly GUI’s like Reason, Fruity Loops, and Rebirth have brought software drum programming to the masses. The advantage to these programs is their relative affordability, upgrade-ability, and rock-solid timing. The disadvantage is that they require a computer (as if you didn’t have one?), and they can take a lot of resources away from your DAW application. Sometimes you must also take the time to render the output to wav or aiff before you can use them in your DAW. This can take extra time. Nevertheless, the flexibility of these applications makes them the choice of many drum programmers.
  • Plugin software: The final category, and most recent development, is the software plugin. Unlike standalone software, these can be introduced into your computer DAW environment in a very ergonomic way and programmed from your DAW midi interface. Many standalone applications can also interface with a DAW through a plugin interface, making them quite flexible to work with. They feature the same upgrade-ability, affordability, and flexibility of standalone software, and are definitely the wave of the future.

Whatever tools you choose to use, it’s important to recognize a few key features that will enable you to get the most out of them. These are things to look for when considering what tool you will use.

  • The ability to import samples, and the ability to layer and crossfade between different samples.
  • A variety of adjustable/modulation parameters (e.g. can you change the tuning, filter cutoff, sample start time, envelope, playback direction, switch between samples, etc.)
  • The ability to make fine adjustments in timing.
  • DSP or effects processing capabilities.

Many of the techniques I’ll discuss rely on these features. If your equipment doesn’t support these and you’d like to try something that does, see the appendix at the end of this tutorial for some options.

1. The basics of beats

Back in the days when rock ‘n’ roll was just another style among many, before it and its variants dominated the popular music market, the rock beat was referred to as a “backbeat”. The reason for this is because it emphasized beats 2 and 4 of a 4/4 measure, rather than the traditional 1 and 3 (think polka). This is generally done with the snare drum.

Of course, with the backbeats being emphasized, something has to tell us when the downbeats are. This is generally the job of the bass drum (a.k.a. kick). The kick part generally has two functions: To define the downbeat of phrases or measures, and to act as a rhythmic counterpoint to the snare. Because of this, the kick rarely plays on the same beat as the snare (the most notable exception being the classic disco/house/trance beat where the kick plays a steady quarter note pulse).

With all this crazy back-and-forth going on, someone has to keep things together. This is where the third element comes in, the ride. (Do not confuse the ride element with the ride cymbal. Drummers speak of “riding” any number of instruments– including the ride cymbal– but other instruments are common as well). A ride part is typically a steady rhythm part played at a constant interval. Quarter note, eighth note, and sixteenth note ride parts are common.

Note that all these parts are functional elements of the beat, not to be confused with the role of the actual instruments in a given beat. For example, the kick and snare interplay could be done with a set of low/high bongos, the middle and side of a djembe drum, a floor tom and a cowbell, or any number of bass/treble combinations. As mentioned above, the ride part can be any number of instruments.

Each functional element has two kinds of hits: primary and secondary. The primary hits are those that define the basic feel of the beat. The secondary are those that are less important, usually quieter, and exist for extra interest. The difference will become clearer as we examine each element individually.

  • Kicks: There are generally 2 primary hits for the kick. The first one is invariably beat 1. The second is located just about anywhere between beats 2 and 4, the simplest being on beat 3. It is quite common to place the second primary kick on 2&, 3&, and 2a, however. It’s this placement that really defines the feel of the beat. Common places for the secondary hits include 1&, 1e, 1a, 2e, 3e, 3&, 3a, 4e, 4&, and 4a. Naturally, the placement of any secondary kicks between beats 2 and 3 must sit well with the placement of the primary kick in that region. Of course, the most common exception (and it is quite common) is the “4 on the floor” beats mentioned earlier, and the variations associated with them.
  • Snares: The most common and rarely altered primary beats for snares are 2 and 4. You will occasionally find beats where one or the other is moved or simply not used, but it’s most common to have both on 2 and 4 and quite rare that at least one of those beats doesn’t have a primary snare hit. Secondary snare hits are an art unto themselves, and are commonly found on 1e, 1&, 2e, 2&, 2a, 3e, 4e, 4&, and 4a. Of course, just about any beat is eligible for a secondary snare hit if it’s part of a pattern, but those listed are most common for single secondary snare hits. Remember too that the placement of these hits is dependent on the placement of primary and secondary kicks. The two instruments play off of one another.
  • Rides: For any ride part, the primary hits are simply the quarters, eighths, or sixteenths that we’ve chosen for the part. Secondary hits are common, and are typically the insertion of the next finer note type (in other words, a sixteenth note ride part may have 32nd notes inserted here and there as secondary hits, an eighth note ride part inserts some 16th note hits, etc.). One exception is when the ride is on the offbeats, not uncommon in disco, house, funk, or rock.

By now you are probably wondering about where hihats fit into the picture. Well, if you haven’t guessed, they are ride instruments. But many other instruments can take the place of the hihat, including the ride cymbal, floor tom, tambourine, shaker, and even the crash. Hihats are unique in that they can achieve different timbres through opening and closing. This phenomenon can be used to achieve dynamic levels (i.e. riding a closed hat in quiet sections and an open hat in loud sections), or to create a certain feel within the basic groove (opening and closing the hat on different beats, as often heard in breakbeats and disco).

You may also be wondering where toms, crashes, cowbells, and other such paraphernalia fit in. I’ve mentioned a few special cases above, but most of these work into secondary hits of some kind, or are specifically used for fills. In addition, auxiliary percussion, a classification that includes pretty much any percussion instrument not included in a standard drum kit, is kind of a classification of its own, which we may deal with later. For now we’re dealing with the basic beat.

2. Feel

“Feel” is a hard word to define clearly, especially as it is used somewhat loosely both to refer to the overall impact of a specific beat, and the attitude with which is it performed. For example, we would say that a house beat has a different feel than a jungle beat, because the kicks are quite different and one may be played swung and the other straight (more on that in a minute). We might also say that two drummers play the same beat with a different feel, here referring to the minute variations in dynamics and timing that define the overall attitude of a player’s performance.

In drum programming, we are of course concerned about the first sense of the word, in that we want to create a feel appropriate to the track, but the real challenge lay for us in the second sense of the word. One of the biggest complaints about programmed drums is the lack of “feel” in this sense, since a machine, rather than a person, is performing it. The machine makes no assumptions about feel. It simply plays back the notes as they were literally communicated.

Essentially, this leaves us with two issues to deal with: timing and dynamics.

Timing: Once we have placed the notes where we want them in a general sense (i.e.-we’ve composed the basic beat), we need to take a look at how we want the beat to feel. If you want a straight, electronica type beat you are already there. Humans tend not to be so good with time, though, so for a more human feel we’ll need to loosen things up. The first thing to understand is the concept of swing. Swing is the human tendency to play offbeats late. Many styles exaggerate this sound to different degrees, and on different offbeats. Full on swing will take you to a triplet feel (3 notes in the space usually taken by two), but a natural swing should be somewhere in the middle ground between triplets and duplets.

The most common type of swing is to swing the sixteenth notes. This means that the e’s and a’s of each measure will be late. The downbeats and the &’s will be right on time. Blues music and certain folk styles are also known to swing the eighth notes, meaning that the &’s will be a bit late. How late is up to taste, but never exceed pure triplets.

Most software or hardware that is used for sequencing or tracking has some kind of swing parameter which can be used to swing a beat quickly, but I often find it better to create the swing manually by delaying each appropriate hit the desired amount. This allows you to randomize to some extent the amount of swing for a more human feel.

And speaking of randomizing the timing, a little randomness is good on every hit in a beat if you want a more human feel. The only beat I would stray away from randomizing is beat 1. Beat one usually needs to be smack on, and it doesn’t contribute much to a feel to make it late. You should experiment with making other beats late or early (called dragging or rushing, respectively), especially the primary snare hits. You may see some tips related to this later.

Dynamics: A huge part of the feel of the track is the loudness with which each hit is played. We’ve dealt with this concept a little bit already in discussing primary and secondary hits. With few exceptions, primary hits should always be louder than secondary. This alone will add quite a bit of life to a beat.

But it doesn’t end there. Dynamics become critical when we are dealing with ride elements. One of the things that will kill a programmed drum track is a loud, stale, sixteenth note hihat part. Adding realistic dynamics to a ride part of any sort, be it hihat, ride cymbal, etc., is a must. The simplest way to do this is a tiered approach, where you make the downbeats the loudest, the &’s second loudest, and the sixteenths the quietest of the three. You can experiment with other ways, but the rule of thumb to go by with ride parts (or any part for that matter) is that no two hits within a quarter note of each other should have the same volume.

Before you start editing velocity data, however, let’s stop and think about something. When you strike an object hard, and then strike it soft, there is a difference in the sound. The difference goes beyond volume; the two hits have a different tone, a different decay rate, and a different sound to the attack. To realistically simulate dynamics, you’re going to have to go beyond simple volume changes. The ideal method is to have multiple samples of an instrument being struck at different velocities. When you don’t have this luxury, you can simulate the effect through envelope modulation, filtering, and sample-start modulating. Of the three, the latter is my personal favorite, but you should experiment with them and decide what feels right for the beat you’re working on.

As with timing, a little randomness is good with dynamics, not only with the volume, but also with whatever other parameters you use to simulate dynamics. Other methods of creating dynamics and feel in general are discussed in the tips section below.

3. Achieving Realism

Although I’ve already said a good bit about achieving realism with respect to dynamics and timing on individual parts, there are some more general concepts to consider when simulating a real drummer as we bring all the parts together.

The first is what I call tone matching. This means that your drum tones should sound like drums that would likely be in the same kit. Don’t use a crusty old hihat with sparkly high-end drums.

The second is performer matching. This means it has to sound like the same drummer is playing all the parts. For example, big John Bonham style drums paired with a wimpy hihat will sound disjointed and decidedly fake. A drummer who is pounding his snare like an anvil will have no less mercy on his hihat, crashes, or anything else. This goes beyond a volume issue; drums that are struck hard sound different. When picking samples to use for a part, keep this in mind.

Third, keep in mind the inherent ambience of each part. They should sound like they’re being played in the same room. Sometimes the easiest way to do this is to use all dry samples and run them through an artificial reverb. Of course, this isn’t always the most pleasant sound, so if you can get a collection of drum samples recorded by the same drummer on the same kit in the same room, that’s a real plus.

You also want to consider whether a real drummer could conceivably play the part you are programming. For example, do you have a hihat, ride, crash, and snare all playing onthe same beat? Unless Shiva is your drummer, that’s not gonna happen. Remember that most drummers have only two arms (or less), so program accordingly (as I’ll point out later, however, “feel before real”. In other words, if it feels wrong to make it two-armed, just go with what feels good. Real drummers have been known to overdub now and then).

Next, consider if the drum part as a whole is too repetitive. Usually this is a dead giveaway for a programmed part. In some styles of music, repetitive drums are not really a problem. In others, such as rock, metal, or jazz, it’s death. If you’ve laid out the basic loop through the track, go through and start removing a hit, adding a hit, changing a velocity, or making some kind of random change at different points in the song especially in the secondary hits. Always check to make sure it sounds appropriate. This will really help greatly to break up the monotony.

Finally, you need to listen to the overall beat and decide if anything sounds unnatural. Does and open hihat cut off too quickly or fade in an electronic way? Do the crashes ring out long enough (this is a big issue!)? Are instruments cutting off the tails of other instruments? If something feels wrong, don’t be satisfied. Keep working on it until it feels good.

4. Achieving “fake-ism” (Fun things to try).

As I said before, “feel before real”. In other words, if it doesn’t feel right when you try to make it sound real, go ahead and let it sound fake. Most of the people who are concerned about your drums sounding real aren’t going to be fooled by even the best programming anyway. Everyone else doesn’t care if the drums are real; they just know whether or not their feet are tapping and heads bobbing to the music. With that in mind, here are some techniques to try.

Backwards sounds: The reverse snare hit was a staple of the big-hair rock of the 80’s, but reverse sounds don’t have to be used in that way. Reverse kicks can be great for trip hop and hip-hop, reverse hihats and ride instruments can add a nice texture to the beat as well. The important thing to remember with reverse sounds is that where they end is more important than where they begin. Try putting a reverse kick on the last beat of a measure so that it ends (or is cut off by) the first beat of the next. Try reversing the offbeats (the e’s and a’s) of a sixteenth note hihat part for an interesting swing effect. The strange dynamic curve of the backwards sound can add a unique feel to a standard beat.

Pitch-shifted sounds: Many styles of electronica would not be what they are without pitch-shifted drums. Usually this is a byproduct of slowing down or speeding up a breakbeat, but you can do it on individual hits as well. This is an alternative for making dynamics if you only have one sample of an instrument.

Gating: A popular drum technique of the 80’s, but it can be tastefully done yet today. Gating simply refers to cutting off the tail of the sound artificially. Working with a sequencer, you have the advantage of being able to cut the tails of drum hits off on a specific beat. Try cutting off the snare when the kick hits, or cutting short the down beats of an 8th note hihat part. You can get some interesting feels with in-tempo gating.

Filtering: Filtered drums have been around a while, and in many ways are somewhat dated. Nevertheless, creative filtering can make for some very lively drum parts. Instead of running the part through a static filter, try an envelope filter, LFO filter, or some other kind of dynamic filtering that will liven up the part.

Other processing: In the quest to take away the static, predictable quality of programmed drums, any number of effects can be employed. Ring modulation, phase shifting, flanging, chorusing, and pitch modulation are just a few. Experiment with effects that change over time to get some variation in your drum parts.

Electronic sounds: Sometimes there isn’t a sound in nature to achieve the sort of feel you want. A great number of electronic sounds have been inducted into the percussion instrument hall of fame, and many more wait to be discovered. We’ll discuss some classic drum machines below.

5. Drum Timbres

In this section, we’re going to look at the variety of timbres that we encounter for each instrument class, and what elements determine that sound. We’ll also touch on how playing velocity affects the sound.

Kicks: A real kick drum is a large membrane drum, often with two heads, that can be anywhere from a tiny 15″ to upwards of 24″ in diameter. It’s played using a felt beater played with a footpedal. Some drummers utilize a double-beater arrangement with two pedals to play faster kick passages. Other drummers make use of two kicks to play faster passages.

The kick sound consists of the beater sound (a thumpy, quick attack), the body resonance (which determines the depth of the sound) and the ring of the heads (which generally determines the length of the kick). The ambience of the room is also a factor. A harder kick hit will generally be brighter, a little less ringy (but more ambient), and have a lot of beater sound. Softer kicks will have a mushy attack and a longer, lower ring.

In general, the busier a kick part is, the tighter you want the kick to be.

Snare: A snare is another two-headed drum, usually around 14″ in diameter and anywhere from 3 ½ to 7 or more inches deep. The bottom head of a snare drum is spanned by a set of metal coils called “snares”, which rattle against the head when hit. The tension of the snares against the head is adjustable, and the snares can even be turned off completely. Snares are generally played with wooden sticks, wire brushes, or plastic or wooden “bundles”. Snare drums tend to be quite loud and piercing compared to the rest of a drum kit, and as a result tend to be more ambient than other instruments.

The snare sound is affected by the size of the instrument, the tension of the heads, the tension of the snares, and the velocity and location of each strike. Bigger snares have a deeper tone and more low-end oomph. Tighter heads cause the snare to have more of a distinct tone, and looser heads give more of a toneless “thud”. Loose snares make a snare sound more rattly and raspy, while tighter snares give more of a “snap”. Hitting the snare softly generally causes a less defined attack and more snare sound than drum sound. Conversely, striking the snare hard gives more emphasis to the body of the drum and a more distinct attack. Hitting the snare in the center of the drum gives more tone and body to the sound, while hitting near the edge (called a “rim shot”, though this term gets muddled with “rim click” in many sample labels) gives more of a “crack”.

Hihat: The hihat is a very versatile and potentially expressive instrument that comes in a variety of sizes from 12 inches in diameter to 14 or 15 inches. It consists of two matched cymbals mounted on a stand with a foot pedal that can move the top cymbal up away from the bottom one or down against it. The distance between the two is adjustable, and the looseness or tightness of the hats is a key factor in the sound. Larger hihats tend to have a deeper tone, while smaller ones tend to have more sizzle. A loose hihat has a sloshy sound, with a longer decay. Hitting the hihat harder can cause it to be sloshier as well, as can hitting it near the edge. Hitting the hat near the middle (the bell) can make a distinct ringing tone. Finally, hihats can be played without even using a stick by simply operating the footpedal. This has less defined attack than the struck hihat, and has a kind of “swallowing” effect as the air rushes out from between the two cymbals.

Cymbals: There is a huge variety of cymbals a drummer can choose from, the two most common being ride and crash. They come in a number of sizes, some well over 20 inches in diameter. Ride cymbals are generally ridden, and tend to have a mellow tone. They can be struck near the bell for a more bell-like tone (makes sense), or struck near the edge for a thinner, brighter tone.

Crashes are much brighter than ride cymbals and are often struck on the downbeat of the phrase to emphasize the change in a section. Crashes are also struck to emphasize important kicks in a rock context. One technique often used in this case is called a “choked” crash, where the drummer hits the crash with a stick and quickly grabs it with the other hand to choke off the ring. This is very useful for staccato hits. The crash can also be ridden in high-energy situations. Like all cymbals, crashes have different tones depending on where and how hard they are struck. The properties are similar to hihat and ride.

6. Drum Machines of note

Drum machines have been with us for decades, first seen on the underside of theater organs. In the beginning they were crude boxes with preset beats like “march” and “rhumba”, and their drum sounds were often made with simple white noise and sine waves. Throughout the 80’s, however, drum machines became an important part of popular music, and a few models made a name for themselves. This is by no means an exhaustive list, but they are some machines with distinctive tones that any drum programmer should be familiar with. For more information on classic drum machines visit the drum machine museum online, and for classic drum machine samples visit the music machines website.

Roland TR-808: If you have heard much classic hip-hop you’ve heard this machine. Overlooked for years, it became an underground sensation when dj’s began buying them for peanuts at the pawnshop. It was analog, employing no samples, and is most famous for its low sine wave kick. This is the kick that rattles many a subwoofer in teenage cars everywhere. The thin, snappy snare, brilliant handclap, sizzly synthesized hihats, and distinctive cowbell are also well used. Think r&b, classic hip-hop, and soul.

Roland TR-909: The brother of the 808, this machine used a combination of samples and analog synthesis to make its sounds. It’s most commonly used in trance and house music. Just about any house tune you hear will feature the thumpy 909 kick doing 4 on the floor, 909 handclaps on 2 and 4, the 909 open hihat on the offbeats, and the famous 909 snare doing 4 measure long fills.

Linn Linndrum: The Linndrum, as well as the Linn 9000, was the de facto drum sound of the 80’s. It had a raw, beefy sound, and many a drummer found his tracks replaced with the Linndrum back in the heyday of hair. Tune in to any classic 80’s radio station and you are sure to hear some linndrum sounds. It’s known for its pounding kick, thuddy snare, raspy hihat, and gritty toms. The Linndrum was so overused, in fact, that it has yet to really come back into vogue.

Roland CR8000: A precursor to the 808, it has a very similar sound. It’s different enough to be of mention, however, and featured some auxiliary sounds the 808 lacked.

Emu SP12: Although actually a sampling drum machine (and therefore not possessing a distinct set of sounds), the sp12 nevertheless is important due to the gritty sound it imparted to the samples because of its 12-bit sampling. Low bitrate sampling became a key part of the hip-hop sound thanks to this and other low-end samplers (like the Ensoniq mirage).

As drum machine technology improved, the sounds became more varied and realistic. As a result, the machines themselves became less and less distinctive. It only shows us that better technology isn’t necessarily a good thing. Still, we can recapture the days of gritty, weird drum tones with a little sampling and an appreciation for the past.

7. Auxiliary Percussion

An entire book could probably be written regarding how to program auxiliary percussion. I don’t really have that kind of time on my hands, so I’ll just discuss a few basics. Auxiliary percussion refers to any percussion instrument not part of the standard drum kit. This includes shakers, tambourines, hand drums, blocks, timbales, bell trees, triangles, gongs, and a variety of ethnic percussion instruments from around the world.

In general, you need to remember the basics of dynamics and timing that you use in programming drum kits. Remember also that auxiliary is played by a separate musician or musicians, meaning that you ought to write parts that sound as though a whole person is actually playing them (rather than half a person, such as putting two conga hits in a measure. A real conga player would be much busier). Keep in mind the concepts of tone matching, player matching, and room matching as well.

Bear in mind that a good auxiliary part will complement the kit drums. This means finding a balance between following the feel of the kit and acting as a counterpoint against it. A good basic plan is to reinforce the primary hits, and play against the secondary ones. Listen to the parts both soloed and with the kit, and make sure it sounds natural. Auxiliary is often a subtle effect in live music, so don’t feel bad if it doesn’t jump out in your programming.

Finally, remember to make use of the concept of beat elements in your parts. Interplay between the deeper and tighter sounds of an auxiliary instrument can be just like the interplay of kick and snare, while other sounds lend themselves to ride parts. Less can be more in auxiliary parts too; many modern drum programmers get a little aux-happy and end up with a massive wall of rhythm without distinction or character.

8. Tips

We’ll end this tutorial with a collection of random tips regarding different drum programming issues, in no particular order.

  • For ride cymbals, start by getting 2 or 3 samples of the same ride. One should be “pingy” (on the bell). With the first sample, use two tracks so the attacks don’t cut off the ringing of the old one. Next, on the offbeats, change the sample start to cut of just a hair of the attack. Do this a different amount on each offbeat. You can substitute your second (non-bell) sample for every so many, at random or however it sounds natural. Finally, layer in the bell hit sample on every beat or only on the downbeats, giving each hit a random velocity. Mix it very subtly with the other two. I’ve found this gives great ride parts. Works well with crashes and open hi hats too.
  • For greasy funky beats: It’s all about timing. Experiment with playing the snare a hair late (like a fraction of a 32nd note). Swing amount is good to play with. Never use straight up triplets unless you like cheese. A very slight randomizing doesn’t hurt either.
  • Snare ghost notes: It’s best if you can get samples of ghost notes from the same snare that you are doing full hits on. If not, you can simulate them by modulating sample start time (chop off the first 5-20% of the wave), filtering (try filtering with a hpf), or just a different snare sample. Another cool trick is to take only the first attack portion and reverse it. It’s always good to have samples of snare “diddling” as well (where the drummer lets the stick bounce against the head rapidly). In certain styles drummers do a lot of this, and it’s hard to do a convincing simulation using individual samples.
  • Think logically about how a real drum kit sounds. Think about the fact that it all sits in the same room, that drums sound different when played at different volumes (not just louder and softer, but the tones are different). Think also about how a kit is typically mic’d. There are typically two overheads, close mics on the snare, kick, and toms, and two or more room mics several feet away to record the room ambience. Think about how leakage (one instrument being heard by another’s mic) will affect the overall tone, especially when you process one instrument and not another. (for example, adding reverb to a snare will add a bit to the hihat, the high frequencies of the kick, and the toms and cymbals to a degree).
  • I’ll say it again: dynamics are crucial, especially in “ride” type parts (i.e. like the ride cymbal, hihat, or floor tom). Don’t just use velocity either (unless of course you’ve got a multisample with velocity crossfades). Modulating sample start is a great way to make softer notes or ghost notes.
  • Timing is essential too; if you can introduce a little randomness or feel to it, it helps a lot. Don’t make fills too perfect.
  • If you can get several samples of the same instrument played at different volumes, this is great. I often set up a random crossfade between samples to simulate the effect of a drummer hitting a drum at slightly different places and at slightly different volumes.
  • Feel before real. And as with any music production, if it sounds good…it is good.

Appendix: resources on the web

Keep in mind, this article is from 2004. Many of these resources may be defunct or outdated.

Below is a list of things you might want to check out online if you’re new to drum programming and don’t have all the tools you need, or if after reading this you feel your programming tools are inadequate. I’ve stayed away from commercial products for the most part, because they are not hard to find information about.

  • Buzz: a full-featured standalone composing environment, complete with samplers, softsynths, and more. It’s completely free of charge and available for download. http://www.buzzmachines.com/
  • Hammerhead rhythm station: A very simple freeware standalone drum machine with a 909 like interface. Comes with some great built-in sounds and is capable of loading new ones (albeit by a convoluted method). Great for beginners. www.threechords.com/hammerhead.
  • Tuareg: By the author of hammerhead, a more complicated composing environment than hammerhead, complete with built-in synths, a drum sample sequencer, loop mixing, effects, and much more. Comes in a free version and a shareware version. http://www.brambos.com/
  • Drumsyn: a freeware analog-style drum synthesizer. Great for generating electronic drum samples. http://www.maxim.abel.co.uk/
  • VST Central: a great site for finding plugins on the web. Features a flexible search engine for finding that perfect plugin. Not just VST plugins, either. http://www.vstcentral.com/
  • Shareware music machine: A great place to find new music applications and resources. Features reviews, links, and a search engine. www.hitsquad.com/smm
  • KVR-VST: A site dedicated to software synth / sampler plugins of many formats. Great place to find news about upgrades, patches, and new products. This is becoming the premier site for news and reviews of audio plugins and plugin hosts.
  • Goldwave: A shareware audio editor with very light trial restrictions. Very useful. http://www.goldwave.com/
  • Google: great way to find drum samples. Type in “free drum samples” and you’ll be amazed at how many sites come up. Hope you have time to download a few megs (or gigs!). http://www.google.com/
  • Lovelabormusic.com: A review site for home recorded CD’s. Great place to find good indie music or send your own CD for review. http://www.lovelabormusic.com
  • NSkit: free sampled drumkit with multiple velocity layers. Pretty decent drums. www.natural-studio.co.uk
  • Kingston drums: More free drum samples. Features multiple velocity strikes of the same instruments. http://kingstondrums.bombsquad.org

10 Thoughts on “The Theory of Drum Programming

  1. Walter says:

    Sorry but I don’t understand this classification: “the e’s, the a’s, the &’s, 1&, 1e, 1a, 2e, 3e, 3&, 3a, 4e, 4&, and 4a”. Thankyou in advance. Greetings.

    “The most common type of swing is to swing the sixteenth notes. This means that the e’s and a’s of each measure will be late. The downbeats and the &’s will be right on time. Blues music and certain folk styles are also known to swing the eighth notes, meaning that the &’s will be a bit late. How late is up to taste, but never exceed pure triplets”.

    “It is quite common to place the second primary kick on 2&, 3&, and 2a, however. It’s this placement that really defines the feel of the beat. Common places for the secondary hits include 1&, 1e, 1a, 2e, 3e, 3&, 3a, 4e, 4&, and 4a.”

    1. Alan says:

      “1-e-&-a-2-e-&-a” is a common way to count sixteenth notes. The “e’s” and “a’s” refers to the sixteenth notes that are not on the beat or on the eighth note between beats. In other words, if you break a 4/4 measure into 16 16th notes, they are:

      1 1e 1& 1a
      2 2e 2& 2a
      3 3e 3& 3a
      4 4e 4& 4a

      Let me know if that explains it.

  2. Martin says:

    Hey Alan,
    this is pretty awesome! I remember reading this a year or two ago, when I started out. I wish it was a little more extensive though. Things like fills, different genres, MIDI, … would also be helpful to beginners. But I guess that’s not the intend of this article and it’s more aimed towards electronic/urban music, right? The other day I came across this book called #HitIt – The Ultimate Guide to Programming Drums which covers the topics I missed in this article. Thought you might be interested as well. 🙂
    Cheers,
    -Martin

  3. Kwoods says:

    thanks for this page, i’ve already been programming drums for quite a few years now, but you never stop learning new techniques and perspectives on creating tracks. i realize this is an older page, and it’s possible no one will read this, but i just wanted to say thanks, i enjoyed reading! been thinking of switching from a pattern-based step sequencer, FL Studio, to tracker software, probably OpenMPT since it’s free and it appears to be the most popular program in that category besides Renoise, which of course is not free. i’ve asked others what the benefits are to trackers, but haven’t got a straight answer really, perhaps there is no benefit, but i’ve noticed that taking yrself out of yr comfort zone to create new music can be challenging and fun, and possibly lead to new and exciting avenues of expression and experimentation! cheers!

  4. Brian says:

    Hey, Alan thanks heaps for this info on drum programming i really found it helpful n so useful!!
    you went to great detail in explanation and i can’t thank you enough.
    i was wondering if you had any info on making melodies match basslines with chords.ect? i spent a bit of money on gear and want to explore it to the fullest. any suggestions?
    hope to hear back from ya, or at least see some new posts on your website.
    thanks again, Brian

  5. Gerry says:

    Far more useful than the $2 I chipped in to help out. I have some of the best software for my DAW, including a killer, top of the line drum sample program, alas I am a singer/songwriter/guitarist. These hints helped me more than I could ever say. Thank you so very, very much.

  6. Felix says:

    Hey man, very detailed article.
    My head is going to explode with all this info.

  7. Great & informative read. Thank you & much LUV.

  8. sam says:

    really nice words used

  9. TJ says:

    Dude, awesome write up ! Exactly what I needed to read. I came here to better understand down/up beats but I can’t wait to apply all the extra stuff I’ve learned!

Leave a Reply

Your email address will not be published. Required fields are marked *