Twelve Easy Pieces
Sometimes, gimmick is good. Here are twelve easy recipes for unusual sound design - quick tips to get the studio juices flowing.
© Dan Phillips
This article first appeared in Electronic Musician's Desktop Music Production Guide magazine in 2002, under the title "Causing Effects."
Audio examples of the techniques described in this article are available on this page of Electronic Musician's website.
The last shall be first
Normally, people think of reverb as being the last effect in the chain - but things can get interesting if you swap the order, and place the reverb *before* other effects.
For instance, I've created great distorted lead "guitar" synth sounds by running a basic sawtooth patch through reverb, and then into heavy distortion; the reverb produces random-sounding overlapping between notes, reminiscent of feedback.
On a more ambient note, try running the reverb output through stereo modulation effects, such as chorus or phaser. The result: rich, swirling, and spacious.
Do it again
A friend of mine used to be fond of saying, "Nothing succeeds like excess." If you're working towards a unique sound, but you haven't found quite what you're looking for, try duplicating the processing. If you've used one chorus, use two; if you've used one reverb, add on a second.
You can place the duplicate effects either in series or in parallel; I prefer to use reverbs in parallel, but experiment to see which works best in any particular case. Vary the parameters slightly between the two, to create a more rich, complex sound.
Using effects from two different manufacturers will also generally guarantee that they'll sound at least a little different from one another, and thus (hopefully) complement each other.
Is it a bug, or a feature?
Ever wonder about how Garbage created the strange, sputtery effect on the bridge to "Stupid Girl," from their first, eponymous album? It wasn't the latest effects plug-in, or a painstakingly sculpted multieffect masterpiece, but rather the sound of a malfunctioning digital multitrack.
When one of their machines suddenly went on the fritz, they were quick to recognize an interesting timbre - and so instead of hitting the stop button and powering down, they whipped out a tape and started recording (using another machine, of course!).
Equipment malfunctions can be annoying interruptions - but they can also create happy accidents, as the errant gear makes bizarre sounds that you might never be able to create otherwise. So, the next time your favorite studio toy is acting up and spewing out mangled audio, try recording its output instead of turning it off.
Say goodbye to the dry
Singer/songwriter Glen Phillips (www.glenphillips.com), now a solo artist and formerly the lead signer of Toad The Wet Sprocket, has a favorite trick for creating ambient pads. He'll run his guitar through delays and/or reverb, and perhaps a chorus as well. So far, this sounds pretty normal - but then he kills the dry signal completely.
("Dry" means the original sound, before effects; "wet" means the output of the effect.)
Using only the wet signal creates beautiful, lush pads, which are wonderfully organic in quality but difficult to identify as a guitar. Of course, this trick can work for other signal sources, such as keyboards and backing vocals.
Back in the dawn of the sampler age, it was cool to explore the percussive possibilities of kitchen equipment, augmenting a rhythm track with spoons hitting glasses and pans crashing to the floor. While I still love a good cast-iron clank from time to time, household gear can be a source of sustained, ambient sounds as well.
Try walking around your house and turning on any piece of electrical equipment - computer fans, food processors, blenders, razors, hair dryers, and so on. If there's a pitch within the noise, set up a mic and get it into your sampler. With today's plentiful RAM, I'd recommend sampling 10-30 seconds of sound, so you can avoid (at least initially) the time-consuming task of looping.
After the sound is in the sampler, give it an envelope with slow attack and release times, run it through delay effects, reverb, and possibly a chorus, and voila - a new ambient bed.
Looking for sympathy
Have you ever been annoyed by something in the studio that rattles when the amps are turned up? Snare drums buzzing, drums ringing, glasses dancing on a table? Those are called "sympathetic resonances," and their energy can be harnessed for good as well as for evil.
For instance, try using a hand-drum (such as a Djembe) as a resonator. Using a boom stand, place a microphone deep inside the drum, and then sing into the open end. You might also try tuning the drum head to match the key of the song.
Or, use a speaker to run previously recorded audio through a drum, and then record the result. Producer Ethan Johns used this trick on an upcoming CD collaboration from Glen Phillips and folk-circuit favorites Nickel Creek, turning a kick drum into a quirky reverb.
Be your own personal sample CD
This trick is from one of my favorite bands, Geggy Tah (www.geggytah.com). On their first CD, Grand Opening, they took a single vocal phrase from one track, and used that one sample to create almost all of the elements for a completely different song - released on the same album. They've re-used other elements from that first CD in their latest release, as well, lending a certain self-referential consistency to their oeuvre.
So, the next time you're about to reach for a sample CD, consider plundering your own catalog instead.
The LFO is the sound
With a tempo-controlled LFO, you can set the frequency in note-length and beats per minute, instead of hertz. This makes it easy to program flangers, phasers, and other LFO-driven effects so that they pulse in time with the music. Many modern effects processors include tempo-controlled LFOs.
These can be very cool for processing everything from drum loops to pads - but the LFO can also become a sound in and of itself.
You can start with almost any source signal, including simple white noise. Run the signal through a tempo-controlled flanger or phaser, set to 8th or 16th notes. If you can set the LFOs for the left and right channels to be 180 degrees out of phase, so much the better. I like to use downward sawtooth shapes for the LFOs, but other waveforms will work as well.
Next, crank up the resonance (or feedback) of the flanger/phaser until you can hear the pitch distinctly. Finally, tune the center frequency so that the LFO creates laser-like sweeps - and you have a distinctive element to add to a loop.
Have you ever been using a delay and accidentally created a positive feedback loop - meaning that the sound builds in volume with every repeat until it's a pulsing, distorted mess? If so, you probably reached quickly to pull down the volume or yank out the patch cord. If you're interested in a creative experiment, however, feedback loops don't have to be a bad thing.
First, a bit of caution: this exercise can create runaway volume levels, so before starting, make sure to turn down your speakers to a very low volume, and do not use headphones.
To create a feedback loop, begin with a delay effect. I like to use very long delays, between 8 and 40 seconds, but short delays will also produce good results.
You may be able to set the delay's internal feedback volume so high that it's equal to or greater than the input level, to create the loop entirely within a single unit. You can also patch several effects together in a chain, so that the last one feeds back into the first - filters, delays, reverbs, pitch shifters, etc.
Once you've created the loop, "seed" it with a bit of audio input - a noise burst, a vocal phrase, a keyboard or guitar lick, a loop, anything. After seeding the feedback loop, wait for a while and allow it to build on itself. Sometimes I'll even walk out of the studio for ten minutes or more, and then come back to hear what's happened.
Alternately, stick around as the loop builds, and modulate elements within the loop - change delay times, pitch shift amounts, filter cutoffs, volume levels, and so on. This is especially cool with very long delays, since it takes a while for these changes to come back around.
Once the sound has built into something cool, record the output. (I'll sometimes just press record at the very beginning, since hard disk space is cheap and plentiful.) You can then load the sound into a sampler, use it unaltered in an audio track as an ambient bed - or use it as the basis for further experimentation.
Sometimes called "modulatable" or "glide" delays, interpolated delays smooth out changes as you change the delay time, creating a temporary pitch-bend effect. (Choruses and flangers are made out of very short interpolated delays, driven by LFOs.) You can find this type of delay on various effects processors, including the Korg OASYS PCI, Lexicon PCM 81, and Eventide DSP 4000, DSP 7000, & Orville.
These algorithms generally allow access to the degree of smoothing, so that you can control the amount of time that it takes to slide from one delay time to another.
With higher degrees of smoothing, you can create very cool tape-stop and munchkin effects by switching the delay time from very short to very long, and back again. Use this in conjunction with feedback, as above, for even more interesting results.
This is currently my favorite production trick, and it's dead simple. Start with a pad - strings, vocals, analog sawtooth, or the like. Send the pad through a gate with external triggering capabilities; this could be a plug-in, an analog gate with a key input, or a MIDI-controlled gate. Next, the fun begins.
Create a one-note rhythm pattern in your sequencer, to use as the gate trigger. You could use straight 8th or 16th notes, but you'll get more interesting results by varying the rhythm and/or note duration.
Next, send the output to the gate trigger input. If the gate responds directly to MIDI, you're done. If it requires an audio key input, send the MIDI channel to a synth patch with a fast release time, to precisely control duration, and then use the audio output of the synth as the key.
At this point, your pad will have become a pulsing rhythm, controlled by the gate. If the gate allows it, work with the attack and release controls to hone the sound.
After this, you can experiment with endless variations. Add a tempo delay after the gate, to further augment the rhythm (with a tempo delay, you can set the time in note-length and tempo, instead of milliseconds); program filter sweeps or other timbral changes in the pad; vary the gate's attack and release in real time; or change the duration of the trigger pulses.
In an interview I did many years ago, Thomas Dolby also recommended programming the gate (or expander) so that the volume is not completely off when the gate is closed; this means that the pad will continue to sustain throughout, with the triggered gate providing a gentle pulse on top.
The name "vocoder" comes from what the device was originally intended to do - to encode voices for data-efficient telephones. While the technology never came into common use for those weekend calls to Grandma, it did give us the processed vocal timbres of "Oh Superman" and "Mr. Robato."
For a completely different sound - and one which blends more easily into a mix - try giving the vocoder a rhythmic twist by using drums instead of vocals for the modulator input. (The modulator input controls the rhythmic and harmonic content, while the carrier input provides the "dry" sound.)
As the carrier, use a pad with high-frequency content, such as airy vocals or strings. The result is something similar to the triggered gate trick described in this article, but with a unique, softer twist.
For a thick, unusual drum timbre, try using white or filtered noise as the carrier, and then mix the vocoder output with the dry signal. The final effect is something like a compressed room reverb run through a fluttery cassette tape - weird and ear-twisting.
An even dozen
So, that's the wrap - twelve tricks to break out of the sonic doldrums. Have fun, and after trying a few of these, go off and invent some of your own!
Dan Phillips is a singer/songwriter and producer, and Product Manager at Korg R&D. Check out his music at www.danphillips.com.
Mic and Mixer
Capture the sustained, tonal qualities of household appliances. Just remember that you can't plug a mic into this kind of mixer. (Photo by Jean Morrison)
Interpolated delays let you change the delay time smoothly, without clicking - generally creating a pitch-bend effect during the transition. This screen-shot from an OASYS PCI delay algorithm shows the Smoothing parameter, which controls how long it takes to slide from one delay time to another