Over the last forty years, the vocoder has helped to create some of the most enduringly popular sounds in Electronic music. Singing pads, talking basslines and countless other effects can be achieved using subtly different techniques based around the same principles.

The story of the vocoder starts in the 1920s and includes everything from military voice scrambling techniques to more musical '80s Electro-Funk stylings. In this article, we'll look at the history of the technology, see how the sounds are created and look at ways they can be used in your productions. We'll also take a quick look at related vocal effects such as the talkbox and the Sonovox before rounding up some of the hardware and software vocoder options currently available.

The vocoder's invention took place CNer the course of the 1930s, a decade of intense interest in voice transmission and synthesis. By 1939, the American telecommunications company AT&T publicly demonstrated their Voder (Voice Operating Demonstrator) voice synthesizer, a robotic device that could replicate the sound of a human speaking voice with some success.

Around the same time, work was being carried out on reduced bandwidth transmission of speech, which would allow more information to be sent over a given number of phone lines. Early research was also being undertaken on pulse code modulation and digital transmission of sound. The story of the vocoder itself starts in 1928, when Bell Labs engineer Homer W Dudley began working on techniques that allowed sounds to be reproduced using basic forms of spectral resynthesis.

The first major use of vocoder technology came during the 1940s, as the need to transmit speech securely became a top priority for telecommunications experts worldwide. The SIGSALY 'speech encipherment system', developed for the Allies by Bell Labs in 1942, used vocoding techniques to encrypt important speech prior to transmission and decrypt it at the other end. Unlike previous voice scrambling systems, SIGSALY was practically impossible to crack.

The end results may be very different to what we're looking for as musicians, but these projects directly led to the vocoders we know today. Modulation time The operating principles of the vocoder aren't particularly difficult to understand when broken down into sections - the basic concepts and technology should be familiar from our studies of subtractive and FM synthesis. Like frequency modulation synthesis, vocoding works using two signals: the modulator and the carrier.

First, the modulator signal is fed into a series of bandpass filters in order to split it into bands. Each of these bands is then fed into an envelope follower that generates a control voltage based upon the amplitude of the input signal. The carrier signal is then broken down into frequency bands to match those of the modulator signal.

Each of these bands is fed into a voltage controlled amplifier (VCA) with its level controlled by the output of the corresponding envelope follower. Each band of the modulator input is thus resynthesized using the equivalent band of the carrier signal. Because the envelope followers only respond to changes in amplitude, the frequency content of the modulator input is effectively discarded.

The carrier isn't modulated by the pitch of the modulator signal but solely by its amplitude and spectral characteristics (the relative levels of each of the frequency bands). The result is a modulated version of the carrier signal which mirrors the amplitude and spectral characteristics of the modulator signal but takes its frequency content from the carrier signal.

For a common example of the effect in action, consider the use of a voice as the modulator signal and a harmonically rich synth pad as the carrier. The synth will only be heard when a voice signal is present, and its harmonic content will mirror that of the voice. This combination is probably the most common use of the vocoder, but there are a great number of alternative methods which affect the sound of the output, from the type of carrier signal to the number of bands and the cutoff frequencies of the filters. Some of the most intelligible recreations of human voices are achieved using a noise source as the carrier, but intelligibility is not necessarily the main objective for musicians.

Likewise, higher numbers of frequency bands will provide a more accurate recreation of the modulator's spectral content, but lower numbers of bands are often considered to offer a more musical sound. Synth-based vocoders offer convenient built-in carrier signal generators and keyboards in order to control the carrier's pitch, but the use of external inputs for both modulator and carrier is perhaps even more versatile.

Given that the transistor wasn't invented until the late 1940s, the first experimental vocoders and military voice encoders were constructed using complex valve circuitry. Hence, these early forebears of the synth-style vocoder or rackmount effect were significantly bigger than the typical piece of music gear. It would be a matter of decades before transistor-based equivalents arrived on the market early vocoders designed specifically for musicians including Robert Moog's 10-band prototype unit developed in conjunction with synth pioneer Walter Carlos (check it out on the soundtrack to A Clockwork Orange) and various other custom· made units from American, German and Japanese companies.

By the time the vocoder reached the mass market, a number of manufacturers each offered their own take on the concept. Most notably, these included EMS's staggeringly powerful Vocoder 5000 (1976, followed in 1977 by the cheaper and more widely available Vocoder 2000}, Sennheiser's VSM201 0978), Korg's VG-10 (1978), Roland's VP-330 (1979) and the Moog Vocoder ( 1980). The best way to get a feel for the characteristic vocoder sound is to listen to some of the classics. '80s Electro is an obvious first port of call - Afrika Bambaataa's Planet Rock and the Jonzun Crew's Pack Jam (Look Out For The OVC), both released in 1982, helped to define the classic vocoder sound. More recently, check out Daft Punk's Around The World and Harder Better Faster Stronger, Air's Sexy Boy and Kelly Watch The Stars, and MSTRKRFT's Easy Love.

A razor head A variety of effects use broadly similar principles as the vocoder or create similar sounds. The Sonovox, invented by Gilbert H Wright in 1939, is less famous than the vocoder but equally as important. The Sonovox, works by feeding signals through small speakers pressed against the neck, allowing the user to manipulate the sound as if it were their own voice. A similar method is employed by the talkbox. Rather than electronically manipulating sounds, the talkbox operates by feed ing an amplified signal through a tube held in the user's mouth. As the sound emerges from the end of the tube, mouth movements allow it to be shaped and articulated.

Speech can be thought of as a form of subtractive synthesis, with the larynx operating as the oscillator, the lungs and diaphragm controlling amplitude, and the mouth, lips and tongue filtering the sound. The talk box first gained notoriety in the early 1970s when it took off as a special effect for guitarists. Hits such as Joe Walsh's Life's Been Good (1973) and Peter Frampton's Show Me The Way (1975) prominently featured the device, and Frampton's 1976 album Frampton Comes Alive! would remain the most well known example of the talkbox sound until Roger Troutman of Ohio Funk legends Zapp arrived on the scene in the late '70s. The group's debut single, More Bounce To The Ounce (1980) laid the groundwork for the definitive talkbox sound, with Roger's synth being fed into a talkbox and expertly manipulated into a lead vocal.

Most people's idea of a classic vocoder is an analogue unit, ideally with a built-in synth and keyboard to act as the carrier signal. You can take your pick from a variety of vintage options, but if you're willing to consider digital vocoders you'll have a lot more options, from the humble microKorg upwards. Hardware vocoders might be the connoisseur's choice, but software provides a cheaper, easier alternative for beginners.