Modern Approaches: Processing Rhythms

Whether triggered from a machine, programmed in the box, sampled from the past or recorded live, drums serve as the foundation to all forms of club music. Essentially, if an artist is unable to properly deliver on the rhythmic portions of their track – if the kicks don’t cut through, if the snares don’t snap – it doesn’t really matter what the rest of the track is doing: it won’t move a dancefloor.

Fortunately, in the modern era of production the possibilities for creating rhythmic components are seemingly endless. One can scour through an abundance of sample packs (both official and unofficial), make use of decades’ worth of drum machines and samplers, grab rhythmic pieces from a massive range of recorded material, manipulate odd bits of audio into percussive hits, or – of course – record live drummers and percussionists. With so many options, producers are able to think far beyond traditional sounds when programming their rhythms. In the right hands, a synthetic kick and snare can have as much sonic force as a full drum kit. Manipulated cleverly, field recordings of unfamiliar sounds can become propulsive rhythmic loops.

Still, programming drums for club music can at times feel restricting, as many club genres are dictated not only by their basic rhythmic structures but also by the tempos at which their patterns are realized. The question then becomes, how does one transmit their own personality amidst these (somewhat) predetermined rhythmic parameters and sonic demands of club music? Once again, the Modern Approaches series has turned to a wide range of working artists to help us gain some insight into the craft, each producer discussing how they collect, arrange and process the rhythmic components of their tracks.

Right now I am very into using the [Dave Smith Instruments] Tempest; it’s a versatile and intuitive machine. I especially like how the pads are laid out in more of a horizontal way, which makes it easier to play than an MPC or a Maschine, whose square layout can cause you to have to put your hands on top of each other to play). The Tempest’s drum pads also have a unique sensitivity that is perfect for playing live. At the moment I’ve been trying to avoid the computer screen when programming rhythms, so even if it’s a very monotonous, repetitive percussion line, I try to play it through the whole track.

I’ve also been using the Vermona DRM-1 drum synth and the MFB Tanzbär analog drum machine. I use the MFB in a little bit of a non-typical fashion, as I don’t really use the sequencer, but instead use it in combination with drum pads so that I can trigger the Tanzbär’s sounds live throughout a song, or loop the patterns using an Electro Harmonix 4500 looper/recorder or the Boss RC-505 Loop Station. I also often use what some people would maybe call “toys,” these old Yamaha DD series drum pads, which I like and are often used in Champeta music from Colombia and other kinds of music like that.

I try to process my rhythms in a way that leaves a lot of space. To help achieve this, I like to use delays – some very short delays, or delays that are so low or dry that you almost can’t perceive it’s there. Somehow, with all these little delays it can create some space. I also like to use spring reverbs on claps and elements like that. There is something warm, something beautiful about the spring reverbs – it can give the percussion a vintage sound, but I try not to have things sound too vintage. So, in addition to using the few spring reverbs that I like, I’ll use something like a Strymon Timeline pedal, which sounds very modern.

We use hardware for the drums on all of our tracks. In our studio we have a Roland TR-909, a TR-808 and a TR-606, as well as an Akai MPC2000XL sampler. Generally, when we want to program a rhythm with heavy swing, we use the the 909 and the MPC together, and then the 808 and 606 for straighter patterns (they don’t have any swing), also in combination with the MPC.

The MPC is used in absolutely everything we do as it really adds a “human element” to the tracks. Almost all of the sounds we have in the MPC are samples or recordings of live drums or percussion, which we use as extra layers on top of an existing drum machine track. For instance, if we use a lot of 808 on a track, we may add a live rim or a snare from the MPC to add some variation. The beauty of the MPC is that you drum in the rhythms in real time, and the pads pick up the velocity of your fingers, so every hit varies in volume and no loop is ever exactly the same. Trying to get the human movement in there is pretty central to all of our drum programming.

We use quite a bit of processing on our drums to help them have a more unique sound. We’ve found that the Sherman Filterbank is great for adding a bit of grit to thinner or clean samples (we almost always use it with the Roland 606). We also have quite a few filters on our modular rig that are good for adding some grit while still being a little cleaner than the Sherman, which is a beast.

We record all our drums separately and give them each their own channel. When it comes to the mixing process, it’s really important to be able to come back to remove something or redo an element if it’s not working. As the years have gone on, we have spent more time in this area, getting to know how to properly EQ, side-chain and compress our drums (which is an art in itself). We are quite big into side-chaining hats and other percussive elements, as that can really add extra grooves and movement to rigid patterns. We’ve also found that the use of side-chain across the entire mix is really something that can create space but also add character to the rhythms.

The starting point for my percussive elements is my personal noise bank that I’ve built over the years. I have some great little binaural mics which I make field recordings with. For instance, a while back when I could hear some building work going on, I ventured out next to a big hammering machine, which was smashing huge metal spikes into the ground and making some great sounds in the process. From things like that, I’ve built a big collection of field recordings. I’ll also sample any nice little sounds and bits that come out of my production process, as well as clicks and noises off old records, sounds of a broken piano (which made the percussion in my track “Remnants”), and whatever else I can get my hands on. I combine these sounds with some standard drum samples when needed, as well as synthesised percussion sounds I’ve made using white noise (these are good for organic, ever-changing clicky hats when used with an LFO applied to the release).

I usually start by building some basic rhythmic patterns that fit the track, and then put a lot of work into how I can mess up the patterns, while still keeping a remnant of their original form. I do a lot of mousing around with the MIDI of the basic rhythms too, moving things a bit left or right to try to construct some sort of more natural groove, and I often apply 16-note shuffles, too (making every second 16th note a bit late), which can give the rhythms that skippy techno effect. In the end, I usually end up with more than 100 percussive elements for each track.

Randomisation and modulation on a big scale is also one of my favorite ways to manipulate rhythms. For example, I will choose interesting parameters for the percussive elements (like their tuning or envelopes) and then have these oscillate and change randomly throughout the track, which introduces more complexity. I also like to apply some randomisation to the individual MIDI notes as well, so that at any given position in the pattern, a note may or may not trigger, or it might change the sample being used.

Randomising or modulating sends is also great for occasionally sending a percussive sound to a huge reverb that itself has a randomised gate or a weird delay. In the end, a lot of rendering and fiddling with audio is necessary after using this much modulation and randomisation on the percussion. It doesn’t always sound good on the first try, so it’s a matter of rendering the chaos to audio, and then chopping and editing that to get something a bit more cohesive.

Then there’s lots of different widening approaches I use to make all the different hits interact in some interesting psychoacoustic manners – some hits popping out, hitting left, right, down the middle, behind your head, etc. Binaural widening effects are good for this, as are normal widener plug-ins, or just simple delays in Ableton offset on each side (either left or right, by a few milliseconds). You have to beware of unwanted phasing effects with all this though, as some sounds can cancel each other out and not punch through in a club because of the widening effects.

Right now, my biggest thing in building rhythms is to actually import outside tracks, and, not necessarily sample them, but try to duplicate a rhythm that the imported track has. For instance, I’ll bring in a salsa rhythm to a project, extract the groove in Ableton and use it in the groove pool, or I’ll just try to imitate the pattern at whatever BPM I’m working at.

I do also like sampling quite a bit, but I don’t go sampling the whole record or break. Instead, I like to go through old vinyl records and take one millisecond snippets of hi-hat or small pieces of drum patterns that I would not have come up with on my own. If I’m taking a bigger pattern, I cut out most of the low-end (from about 400 Hz and below). That way you hear a little of the low-end from the original sample, but you’re also really making the other kicks and low-end elements in the track punch through, and then the sample can be there to provide the grit.

There’s a trick I’ve found to make any record sound like it glues together with the other rhythmic elements (say, the sounds from a drum machine). If I sample something from a record that has a snare, I will take the beginning of the snare hit and add it to the kicks that I’m using in the track, so that the kicks will have their same presence, but also the tonality of the sample. So if you’re using an 808 kick and just have a sample going over it by itself, usually it’s not going to sound very glued together. But if you take the attack of the kick or the snare or the hi-hat from the sample and you add it to the beginning of an 808 or 606 kick, it can really help glue their sounds together.

My go-to software drum machine is MOTU’s BPM. I like having a convenient software drum machine to get down an initial sketch of a beat instead of wasting time hooking up hardware or fine-tuning every sound. Later in the process, when I have a better idea of how the song should sound, I’ll get more detailed, replacing a lot of the initial drum sounds, though I do usually end up keeping at least the kick drums from the BPM.

I’m not afraid of quantizing, but it’s nice to create elements that don’t line up exactly with the grid, so sometimes I’ll add a delay to some part of the beat, unsync it from the song’s BPM, and slowly change the delay time, listening for interesting polyrhythmic moments that are on the edge of being off. Another way to add some feel to programmed rhythms is to integrate bits of sampled breaks. I usually don’t use the whole break, but more something like from the first snare hit to the third kick drum, for example. I’ll then try dropping it in at different points in my beat to see what happens.

It’s also nice to set things up so the drum sounds are interacting with each other, so they feel like a single organism in a room instead of a bunch of separate drum sounds occurring in vacuums. With that in mind, I will often create a buss channel just for the beat and route all the drum sounds through that buss, that way I can add effects to that channel and process the beat as a whole. I also like to make duplicate tracks of some of the drums (usually, at least the kick and sometimes the snare) and assign no output to them so they can become triggers that I can use for side-chaining or rhythmic noise-gating. I almost always gently side-chain my kick to my hi-hat patterns to make each hit’s velocity a little different.

Messing with the velocity and envelope of each drum sound can really help make a beat feel less rigid as well, even when a pattern is totally quantized and swingless. I have my DAW set so if I place one audio region overlapping another one, they crossfade. So when I’m moving a drum sound, or cutting and pasting multiples for a fill, a lot of times I’ll leave a tail of silence at the end of a region so it overlaps with the beginning of the next region, causing a crossfade. This creates variations in the attacks of each sound.

Originally, I was a sample head and went through a lot of funk and jazz records. I have a library that’s been morphing for years, because you can always do up a sound that you recorded when you were 17, mix it different when you’re 25 and pitch it down in your 30s, and it’s always a completely different sound. Next to that drum library that I use regularly, the first-and-only drum machine I will sometimes use is the Elektron Machinedrum (I think I got it ten years ago). I also often just use synthesizers to create drum sounds. For instance, I love the filter on the MS-10 because just combing that with the synth’s white noise I can get my own bass drums, snare drums or pitchy whatever sounds.

For me, it’s always about processing drums in the right way. I’ve never really liked using dry drum sounds – I love pitching them and then distorting and equalizing, layering. My bass drum sound is always several kicks on top of each other. I most always process a beat’s elements individually, and sometimes if it’s too many layers, I’ll just bounce some of them out as one sound to make it less complicated. I don’t really use too much compression, and more just focus on trying to get the volume right. For EQ, I’ll use the Waves EQ6 or the Q8, or sometimes just the onboard equalizer in Reason. I do a lot of my drum programming in Reason still because I’m just used to the onboard drum computer and it’s the quickest way to get my ideas down.

I’ve built up a big collection of samples from drum machines that I’ve found along the way, and breaks and breakbeats are a big part of my music as well. I’ve always loved finding breaks and drum loops to chop up and put in my music because it just gives you such a unique part of your song that you may not have thought of yourself.

Nowadays I use Ableton to chop up most of my breaks, The warp function in Ableton is really helpful as long as you use the beats algorithm (which is then not timestretching the audio but is chopping up the sounds and adding reverse loops if there are any extra gaps). Once I have the rhythm chopped the way I want it, I might separate the break onto three different audio tracks – kicks, hi-hats and snares – do any appropriate EQ (mainly cutting, quite a high low-cut on the hi-hats) and then put it all back in a group and run it through my Mackie mixing desk to EQ it as a whole.

The break you use really defines the overall sound and palette of the piece of music you are working on. If you use a break with a lot of sonic character, you have to keep in mind that there are only so many other sounds that will work with it. It’s just trial and error really, but it’s not worth the time to force things to work. If I choose a snare sound or a kick sound to layer on top and it doesn’t sit well with the break without too much work, then I’ll choose another sound because it will just be much cleaner and much more effective. There are a million snare drums, so you don’t need to force that particular one you’ve chosen to work in the track. I’ve been caught up in hours of buggering with some drum sound wondering “Why doesn’t it work?” just to realize I can simply change it and that fixes the problem.

Recently, we’ve been buying up some cheap ’80s digital drum machines. We don’t have a 909 or an 808 – we just couldn’t afford them – but we’ve been able to borrow them. The reason a 909 and all these drum machines from the ’80s are so funky when compared to Ableton-driven tracks is that the tempo is not steady on these drum machines. The electric current is not steady, so the tempo is not steady. When you start a track on a laptop, it’s 128 BPM and then it stays exactly the same, but if you record a drum track from a 909 or a 606 into your laptop and then analyze the BPM, you’ll see it fluctuates like crazy.

We’ll use these hardware drum machines in the course of our jams, multitrack them into Logic, and then we’ll do a lot of mixing on the drums. We’ll mix a track for three weeks, doing things like running a hi-hat through a speaker and recording it back in with a microphone, or using FX processors like the Eventide H3000 and the Lexicon PCM 70, or running sounds (or even samples from Logic) through our Echolette tube preamp unit to give them more character and meaning. We don’t really do much in the box. It’s really not that there isn’t a plug-in that can do what the Eventide or the Lexicon do – nowadays there are plug-ins that, in theory, can do anything – but the analog-to-digital and the digital-to-analog converters on these machines from the ’80s just sound so sweet to us. It doesn’t matter if you have an Eventide plug-in – it just won’t sound like the machine, it just won’t cut through the mix like the machine. It’s like a print of Picasso, compared to a painting from Picasso. It does the same thing, but it also somehow doesn’t really do anything close to it.

In general, we think anything you can do to take sounds out of the box and then put them back into the box is helpful. If your hi-hat is sounding boring and lame, blast it on some speakers and record it with a cheap microphone, then put it back in and see what it sounds like. Anything like that can help your track and give it more character than all the stuff you hear on Beatport.

I like for my drums to fit my track in terms of their sound and feeling. When I have my music going and I’m trying to find the right drums, I’ll actually play the pattern on the keyboard and audition different kicks and snares in combination with each other. I can’t really do that if I just have a folder full of kicks and a folder full of snares that I just click through. I need to play it almost like a live drummer would, so I can say, “OK, Kick 1 and Snare 20 work together,” or “Oh, wait, Kick 6 and Snare 30 work even better together.”

I’ve been making my own drum kits and amassing drum kits for a very long time; we’re talking since 1999. Anything that I go back to constantly is something that I custom made. That’s how I’ve always worked. There were times where I would spend hours and days and weeks just building drum libraries. I would go find a kick and a snare that I like, and then I would process them, and then save that into a folder (though you don’t always want the processed thing, so I’d save the raw sound too).

Once I determine what drums work best for the track, then I go to what compressor works well with this sound. A lot of times, when you compress too much, what you might gain in a little bit of punch you lose in the ability to control your attacks, and most of the time you will lose a lot of low-end if you over-compress. Another thing that I take into consideration is, if I put reverb on a snare, am I compressing it before or after the reverb? You’re going to get two totally different sounds each way. If I want it to sound like I sampled a live drum kit, or like there’s a live drum kit actually playing on the record, I will probably apply the reverb first, and then compress that. If you’re sampling a drum loop from a record, or a snare hit and a kick hit from a record, what you’re likely sampling is something that came from a live performance in a studio, and its natural reverb is already being compressed by the nature of the mastering and mixing [on the original record]. If you’re trying to recreate that sound, you almost have to apply the reverb to shape the sound first, then compress it, so that you get that same feel of the sample.

I am actually playing the drums [on a drum kit] on my new album. I’ve found it’s just much easier to develop the idea in my mind if I just play the kit, rather than working from a sampler or something like that. I record everything with five microphones that are basically just always set up, so I can jam when I want. It’s not necessarily as easy to get the right sound this way, but it does make it easier to get the idea down and have it recorded with natural dynamics. I can worry all night long about dialing in the sound as long as I have the idea down.

Once I have recorded the drum sounds into Cubase, that is where I can do all the arranging and mixing. I don’t really add any extra samples to the drum recordings, unless I want something exotic like animal sounds or rain or something like that, but Cubase is where I’ll compress all the drums and rhythmic samples as a group, and add things like extra reverb if it’s needed.

I like to use a mixture of standard sample pack hits, recordings of percussive sounds I find on freesound.org, and any breaks I come across as the basis of the drums in my tracks. From there, I use Native Instruments Battery 4 to build up a kit of sounds I like, and then once I have a beat laid down, I’ll separate each sound into dedicated channels so I can add FX to each individual element. I’ll always compress the kick, hats, snares and claps as I find it helps give more character and punch to the samples I’m using.

In addition, the Soundtoys Decapitator is a plug-in I use a lot on my drums. It’s a great distortion plug-in that adds loads of character to standard sample hits and can give you that dusty sound. Adding delays to hi-hat patterns is something I always do as well, as it adds some complexity to the patterns and really helps them to sit in the mix.

All of the drum sounds in my productions come from machines (drum machines and samplers), and I do my best to multitrack each individual sound as much as possible. Even if a piece of equipment only has a limited number of outputs, I would prefer to record multiple passes if it meant that I had a nice, clean multitrack with which to mix later. I prefer to have this control both on a mixing and an EQ/compression level, as well as an editing level.

Now that we’re experiencing, for a couple of years now, people really getting back into using hardware drum machines and samplers again, I think it’s important to note that part of making these machines sound good is recording them well. Plugging a good drum machine into a $200 interface might not be better than using a software sampler properly. But, when you track a hardware drum machine through any type of decent preamp at the right level, you’re going to be in a much better place. The fact is that you might not be getting the most out of your hardware if it’s not going through at least a decent preamp; even just a midlevel mixing board would go so much further in tracking those things properly.

I have lots and lots of sample folders that I pull my drum sounds from. Usually I’ll collect a few I like into an Ableton drum rack to start a beat (I’m also not shy about cutting up loops, either). In addition, I’ll sometimes also record live percussion on top of the programmed drums, using fun things like spray cans, sticks and weird objects to create unusual sounds.

I’m really into drums and usually use too many elements. I find that adding loads of rhythmic elements to the top-end of a track doesn’t have any negative consequences, so I like to collect a lot of shakers, hi-hats and things of that nature in the upper part of the rhythm. I try to never solo any of my drum sounds while I’m working, instead focusing on making it all work as a whole. A singled-out drum element may sound amazing on its own, but then not work well in the larger composition.

Working in Tiger & Woods, we are sample-based, so after we go find or buy some records, we’ll listen and as soon as we hear a tiny part, even just a hit in a record, we’ll sample it. If it’s really interesting, then we might grab a one-bar loop and then start messing around, changing it, slicing it and moving the pitch up and down. Whenever, after tweaking the sound, something is not solid anymore, we replace it with either a drum machine or another sample. We have a library of Casio drum machines in the studio as well as things like the Roland 606, 707 and 909, the Tempest, and some smaller musical toys like the Volca Beats.

In most cases, we’ll group the beats, samples and drum machine sounds together and have them all run through one compressor and one EQ, and even then we won’t do anything too extreme; just trying to keep the signal clean and leave the rest for mastering. On the snare, we’ll use some reverb, but strictly with a gate at the end so that we don’t have that full room sound going on after the snare. We’ll also use short-tempo delays on some of the hi-hats to get them a bit dirtier. Another trick is that sometimes we’ll add a bit of reverb on the first kick drum after some filter section or some breakdown, so that we have some kind of booming effect. That’s a treatment I learned from Mark Pritchard.

Often, more important than how you compress or EQ is how you play the rhythm or what patterns you program. If the parts that you wrote are clashing with each other, it won’t feel natural and it will be much harder to mix. Some of the elements will stick out or won’t stick out because their frequency range is taken by another element’s decay or something like that. Using the space and the silence in the beat as a rhythmic element can really help when it comes time to mix.

I have a giant collection of sample packs for just about every drum machine that I’ve gotten from trading with friends and other producers. When you get them from a friend there’ll be something like “So-and-so I know recorded these drum machines through a specific preamp,” and those ones can sound a bit cooler, or they have mistakes in them, which can be really nice too. For my 1080p record [I Want to Believe], I used mostly Boss DR/Doctor Rhythm samples and SP-12 samples because I think they sound pretty unique, especially if they are overdriven a bit.

Rather than just dropping the drum samples directly onto a track, I use Simpler to control the sounds. That way, I can really screw around with the actual sample. I play a lot with decay times, which can be very helpful for creating space and gaps. The Simpler is also helpful for tuning the sounds. A lot of times, snares have a resonant point that distinguishes a key or gives it a note, so being able to maneuver that so it can sound better is often a good idea.

I usually end up with maybe ten individual sounds/layers of drums on my tracks. Before, I would focus so much on the individual sounds, but I realized maybe that is not as important as processing the group together and considering how they all sound collectively. In the mixing stages now, I’ll go inside the Simpler of each respective sound and mess with decibel level (instead of compressing each element individually), and then I’ll run it all through a couple of saturation plug-ins and compress the drums all as a group. Sometimes I’ll even bounce the whole group of drums as one stereo file, and then treat it as a break, saturating and compressing them all together, and running them through a tape emulator or something like that.

I source rhythmic sounds from everywhere. On a new track, I might start with five YouTube videos and then just take sounds of someone mowing the lawn or something. I like recording sounds from different places and using them even if they’re not perfect as can be. If I hear something on the television that’s cool, I’ll just grab my iPhone and hold it up to the speaker on the TV and send the recording to myself. Just getting some individual elements like that can make the track sound more unique than if I was just pulling from a library of drums.

For me, a lot of the time, once I get a main sound to a place I like, it’s about letting that certain element be and then shaping the rhythm around that element. I am always working with frequency spectrum in mind, so if I have sounds that are making up the rhythmic part of the track that are sounding good in the mid-high range with a little bit of lows, from there I’ll look for some other rhythmic element to fill out the top of the high-end and whatever other spaces are open in the frequency spectrum.

On my tracks, it’s one extreme or the other: either I’m recording a drummer live throughout the whole track or I’m just messing with basic drum packs that anyone can get hold off, and then just tweaking them. I’ve got this little Tascam tape deck, and if you just get any old drum sounds from Logic or wherever and you record them through the tape player and then run them back into the program, it immediately gives a character to your drums.

Besides that, I always try to make my drums groove a little bit more by either not quantizing them at all, or by using the swing from my MPC 3000 on certain elements (sometimes the MPC swing just doesn’t work, though). Usually, I’ll hear the beat in my head, vocalize it to myself, and then try to play it on the MPC, but then the swing might not be the same as the one in my head. In those cases, I just have to try to play it without quantizing the patterns. There’s definitely something that happens when you don’t quantize it. It gives it all a different feel.