by Craig Anderton
Is it time for a rave revival already? Gee, it seemed like it was only...well, it was over two decades ago when L.A. Style released “James Brown is Dead,” so we’re due.
Techno music doesn’t have “leads” in the traditional sense of screaming guitars and vocals; many times they’re sampled sound snippets whose source can be anything from political speeches to old movies. Trying to find appropriate samples is only one task—the other is laying them into the tune’s rhythmic bed.
Although you can always use a digital audio editor to lay samples on a timeline and then cut and paste them to create cool effects, I much prefer putting these samples in a playable format and doing these tricks in real time—the process is more spontaneous, and more fun. Even if you’re not into samplers, many virtual drums let you load samples into the drum’s “pads” and play them from a keyboard.
However, note that samples rarely have inflections that match the music’s rhythms perfectly, which can be distracting. Some musicians who don’t take a digital audio editing approach attack this problem at the sampler itself, by breaking phrases down into individual samples and triggering different words from different keys at the desired rhythms. However, there are a lot of sequencer tricks that can produce similar effects with less effort.
FIRST, A SAMPLING SAFARI
Finding appropriate samples is the first task (getting clearance for them is another one, but that’s a whole other issue). Ideally, the samples not only stand on their own, but can work with each other to create composite effects where the whole is greater than the sum of the parts. A simple example is finding phrases which can be combined in ways to make different sentences.
A friend recently turned me on to a grade-Z sci-fi movie, “Invisible Invaders,” which you can not only catch on Netflix but is a gold mine of samples. The premise is that earth’s invaders can only be destroyed with sound waves. One sample, “Sound is the answer,” became the song’s title. Other samples that made up the collection were:
Some of these are fairly long, and at 135 BPM, I wanted to have the words line up with the rhythms as much as possible, and also mutate the samples for other and perhaps more nefarious purposes. Here are some tricks that worked for me.
SAMPLE TRUNCATION WITHIN THE SEQUENCE
You can shorten samples within a sampler or digital audio editor, but the easiest approach with MIDI sequencing is just to shorten the note’s duration. For example, I wanted to follow “The device must have used sonic rays” with “The device must have used sound vibrations.” Rather than cut and paste to replace “sonic rays” with “sound vibrations” and create another sample, I simply shortened the note for the first sample so that it ended after “...must have used,” then added a note for the “sound vibrations” sample immediately after to create the composite sentence (Fig. 1).
Fig. 1: The E note triggers the first sample up to “...must have used,” while the G# triggers the sample for “sound vibrations.” The pitch bend messages speed up the first sample slightly so that it can end before the next sample, which I wanted to have start on the beat.
Truncating notes to extremely short times gives nifty percussive effects that sound very primitive and guttural. Generally, I map a bunch of samples across the keyboard as a multisample so that each sample covers at least a fifth, making different pitches available. Playing several notes at the desired rhythm, and setting their durations to 30-50 ms, gives the desired effect. This works best with sounds that have fairly abrupt beginnings; a word such as “whether” has an attack time that lasts longer than 30-50 ms. As one example, I wanted a series of eighth-note “ohs.” Triggering “only two theories seem to make any sense” with a note just long enough to play the “o” from “only” did the job.
SETTING SAMPLE START TIME WITHIN THE SEQUENCE
What if you want to play back the last part of a sample rather than the beginning? This is a little trickier. Put a controller 7 = 127 (maximum volume) message where you want the phrase to start in the sequence, and a controller 7 = 0 message somewhere before that. Jog the note start time so that the controller 7 = 127 message occurs right before the section of the phrase you want to hear (Fig. 2).
Fig. 2: Adding a message for controller 7 = 0 (circled in red for clarity) mutes the phrase “The device must have used,” but the second message for controller 7 = 127 unmutes just before the sample says “sonic rays.”
Note that in a multisampled keyboard setup, this will affect any other samples that are sounding at the same time. To fix this, set up the different samples multitimbrally.
THE EARLY BIRD CATCHES THE EAR
It seems that many samples work best if they’re nudged forward in the track so that they start just a bit ahead of the beat. This is probably because some sounds take a while to get up to speed (like the “w” sounds mentioned earlier). Another factor might be that the ear processes data on a “first come, first served” basis. Placing the sample very slightly before the beat gives it more importance than the sounds that follow it right on the beat.
CREATING WEIRD DOUBLING EFFECTS
If a sample covers a range of the keyboard rather than just one key, you can play two samples at the same time for groovacious effects. For example, copy a note that triggers a sample and transpose it down a half-step. The lower-pitched sample takes longer to play, so move it slightly ahead of the higher-pitched sample. Depending on the start times of the two notes, you’ll hear echo, flanging, and/or chorusing effects. If they start and end with about the same amount of delay, you’ll hear a way cool flanging effect in the middle.
USING PITCH BEND TO CHANGE RHYTHM
If a sample works perfectly except that you need to shorten or lengthen a single word, no problem--apply pitch bend to just one portion of the phrase. Bend pitch down to lengthen, bend up to shorten. This can also add some fun, goofy effects if taken to an extreme. Fig. 3 shows this technique applied to several notes.
Fig. 3: The first note rises in pitch (thus shortening the sample); the fourth and fifth notes bend downward to lengthen the sample. The right-most note shortens the beginning, lengthens the middle for emphasis, and shortens the end.
Combining all these tricks means you can lay samples into the track that sound as if they were cut specifically for your tune. Yes, they do take a little work—but the tight phrasing can preserve the rhythmic integrity of the song, and really make a difference.
Craig Anderton is Editor Emeritus of Harmony Central. He has played on, mixed, or produced over 20 major label releases (as well as mastered over a hundred tracks for various musicians), and written over a thousand articles for magazines like Guitar Player, Keyboard, Sound on Sound (UK), and Sound + Recording (Germany). He has also lectured on technology and the arts in 38 states, 10 countries, and three languages.