Production Central: Beat of a Different Drummer (Part 2)

Read Ming tips and techniques on getting the best drum programming results
Image placeholder title
Image placeholder title

In part 1 in the last issue, I covered some of the basics of drum programming: volume, quantization, and swing and grace notes. Part 2 will cover some more advanced beat-making techniques, starting with drum-sample loops.

I started chopping loops and samples on early samplers such as the Akai S950 and MPC-60, and the E-mu 4xt. These devices were the predecessors to today''s audio workstations, but they lacked the modern beat-slicing software such as Propellerhead ReCycle or the Beat Detective feature in Avid Pro Tools that we''ve come to depend on. However, these samplers were powerful enough to help create all types of sample-based music, such as hip-hop and electronic dance. The following beat-chopping technique, though somewhat old school, is still a very powerful method for editing samples and programming beats, and you can also apply it to any sampled material such as melodic passages and vocals.

Image placeholder title

The Amen Break, chopped up in Propellerhead ReCycle

The basic concept behind beat chopping is to have multiple copies of the sample, each with a start point that is later in time relative to the previous copy of that sample. For drums, the time variation is usually placed at the nearest transient. If the drum pattern comprises the transients kick, hi-hat (hh), snare, hh, and kick, then the start point would be moved to each successive transient (see Fig. 1). Once I''ve chopped it up this way, I''ll import the files into a sampler, assign each sample to a different MIDI note, and then play in my part using these pieces. The same can be achieved through software such as Propellerhead ReCycle, which can give you multiple sample start positions within a single sample.

FIG. 1: Here''s an example of beat chopping. The top line shows the order of events in the full sample, and the lines below show what the sample would comprise if each successive transient was cut off.
Sample Chop 1 (original full sample): kick, hh, snare, hh, kick
Sample Chop 2: hh, snare, hh, kick
Sample Chop 3: snare, hh, kick
Sample Chop 4: hh, kick
Sample Chop 5: kick

There are two important reasons for chopping up drum breaks (aka breakbeats) in this fashion: First, the chopped audio can be played and sequenced in myriad interesting patterns that would otherwise take a very long time to program if you did them one note at a time. Second, the sample retains its sonic footprint, which is what gives a break its characteristic movement and tone, and is often what makes it a desirable sample in the first place.

Being able to manipulate a break or sample allows me much more freedom when using a loop. Instead of being stuck with the same drum loop or sample playing over and over, I''m able to manipulate the sample—with different parts of the loop accentuated—to create new, interesting moving passages.

A common example of this beat-chopping technique is the reprogramming of the Amen Break (referring to the break in the song “Amen Brother”; see Web Clips 1, 2, and 3) for drum ''n'' bass or the Apache break in hip-hop.

A sample or loop taken from a nonprogrammed source has a feel that I refer to as its transient DNA (which is similar to what''s contained in a groove template). This refers to the timing of the transients within the sample. The push and pull of time between the transients is what creates movement within a beat. Beats programmed to a grid often lack this push and pull, resulting in a robotic sound. So to put some life back into programmed music, I often take the feel of a sample that wasn''t created on a grid and apply it to the gridded material, or play a live passage and use its transient DNA to do the same.

For example, I like to play a live bass line against programmed drums. I often find that I like how the live-played notes work alongside the programmed beat. If I want the drums to more closely mirror the feel of the baseline, I can either remove the programmed drums from the grid and align key elements such as kicks and snares with the corresponding notes of the bass physically, or I can copy the transient DNA from the baseline with tools such as ReCycle and apply that groove in Reason''s ReGroove mixer, or by creating an audio-to-MIDI groove template in Apple Logic. I can then apply that groove as quantization to the drums or other elements of that sequence.

You can extract not only the timing, but also volume differences from one transient to another within a sample. This is particularly useful when pulling the transient DNA from a live drum progression and applying it to a programmed beat. For instance, if the one in the sampled break is heavier than the three, then the programmed material to which you apply that will follow suit. The accents give a beat bounce, and can be applied to single elements such as the kick, snare, or the whole programmed break.

Programming realistic-sounding live drums is difficult and requires care and time. The most important part of this technique is to listen to the nuances of live drums with and without supporting music. Listen to the placement of the snare and hi-hats relative to the vocals, and try to hear the room in which the drums were recorded. Are the drums recorded to a click or do they drift in time? What is the genre? Are the live drums going to be supported with programmed beats or breaks? All of these will affect how I quantize, mix, and EQ the drums, as well as how I re-create the sound of a nice drum room.

My live drum-programming method is to physically play on a keyboard, a scratch pass of the hi-hats, followed by a second pass with the kick and snare. Once I get patterns for the verse, pre-chorus, chorus, and bridge, I''ll go back and play each element in separately: kick first, then snare, then hi-hats and ride, cymbals, fills (including toms), and finally ghosts notes on the snare and toms if necessary. I''ll quantize the first note of each measure to the downbeat and then soft-quantize the two, three, and four of the beat, depending on how locked to the grid the drums need to feel. I try not to quantize the hi-hats, and I spend a lot of time varying the volume of each hit. To create excitement before a chorus or during a long verse, I use moving levels of open and closed hats.

For sound sources, I use a number of different multisample layered kits from FXpansion BFD, Apple Logic, or Propellerhead Reason Drums. They all have good volume-dependent multisamples, and some have room and microphone emulation settings.

For mixing, I have a drum sub-mix that often has a multiband EQ, an analog-sounding compressor like the H Series in Waves, and some kind of distortion or overdrive to emulate tape saturation. I use reverb on snares carefully and as sparingly as possible so that I have more freedom with the overhead microphone mix. I create the room by busing a stereo mix of the drums to a separate overhead mix bus with EQ, overdrive, and compression and reverb.

Take your time when programming “live” drums. They''re supposed to sound real, so if they don''t, keep massaging them until they''re golden.

Ming is a New York City-based artist, producer, and DJ. He owns Hood Famous Music and co-owns Habitat Music.