Vocoder

Don Lewis Takes on the Vocoder

By Susan Hayes


It is one thing to be called a pioneer, yet quite another to actually be one.

For Don Lewis, his decades-long contributions to the development of the synthesizer and related technologies like the vocoder qualify him as the real deal.  

Starting in the late 1960s, Lewis consulted behind the scenes to drive synthesizer innovation for commercial use by Hammond and Acetone. In coming decades, he contributed to development efforts at ARP, Yamaha and Roland, working directly with electronics engineers in the lab and as a demo artist.

When the synthesizer movement evolved from a desire to replicate musical instruments to replicating voice in the early 1980s, it was only natural for Roland to bring Lewis in on the ground floor as they developed their first vocoder.  

The idea really resonated with him. “I was excited. I hold the human voice as the most compelling sound most people can make. As a former church choir director, I knew the powerful impact of hearing multiple voices simultaneously.”

Driving innovation of the vocoder was the commercial success of electronic-enhanced voice box music by high profile performers like Stevie Wonder.

“Stevie Wonder used a voice box tube that you put in your mouth. The other end of the tube went down to a transducer microphone to produce a weird robotic sound. Everyone wanted to figure out how it was done, driving manufacturers to come up with their own versions.” An added benefit was that if you didn’t have enough room for backup singers when performing, you could get the same result with the Vocoder.

Under the Hood, Technologically Speaking

Roland was developing the first polyphonic instrument, the VP330. Once complete, it would play any number of notes you wanted. As chief beta tester for this and all subsequent versions, Lewis tested, critiqued and performed with these instruments at the “breadboard stage,” when engineers were putting together the first prototypes in the lab.

“We were plugging in components, just like building with Legos,” he recalled. 

At first, developers didn’t seek to emulate the human voice, but Don kept thinking about the sounds of a choir and encouraged the engineers to think bigger, going so far as to encourage them to include choir-like sounds.

His choral background also qualified him as a subject matter expert when it came to understanding how the human voice physiologically works.

The human singing voice operates by coordinating three elements: vocal chords, the mouth (as filter), and the lungs, which control amplification. An envelope generator in the throat dictates when the mouth/filter opens and closes, allowing for variations in sound.

“People can manipulate their voices, mouths and lungs on demand so the most complicated part of making the vocoder was enabling it to do the same.”

The vocoder operates using three similar yet electronic elements: oscillators (instead of vocal chords), filters and amplification, to create the sound of not just one voice, but a synthetic choir. 

Transmitting the sound to the vocoder takes place through a headset mic to 10 – 20 ‘bandpass filters’ which measure the amplitude of the harmonics generated by the user’s voice. A filter analyzes and evaluates the series of frequencies of the human voice and measures amplitude frequency at a rate of 10 – 50 Hz (herz) cycles per second.

The VP330 takes those amplitudes and instead of using the actual pitch of the original voice, it generates selected waveforms over a 49-note keyboard system with 49 oscillators, extending the range four octaves. Pressing an ‘ensemble sound’ button doubles that sound. Sound can also be controlled with a damper pedal to sustain and hold notes like a piano.

In a 1982 article published in the San Francisco Chronicle reviewing one of Lewis’s performances, columnist Joel Selvin said, “Wearing a tiny headset microphone, when Lewis opens his mouth to sing, out come the voices of a dozen young girls. He throws his head and opens his mouth in a giant grin, laughing at the comical sounds emitted by his keyboards.”

“The VP330 was a great start to add choir and string sounds to LEO,” says Lewis. “But when I got the VP550, I’d never been enthralled with anything like it! I could sing on top of it and it made me feel like it extended the power of the human voice.” 

The VP550 added strings, reverberation and ‘musical voices’ which use keys to manipulate vocal sounds instead of transposing the voice of the singer. It allowed for octaves out of the range of the human voice too - even as low as a sub-human level to create an unearthly atmosphere.

The VP770 introduced the ‘looper,’ originally called ‘sound check’ on the VP550, in the early 2000s, as well as a female choir, children’s choir, Gregorian choir, jazz scat, duets, trios.

Newer versions also integrated a dimensional beam sensory system called “Dbeam,” which uses an infrared LED sensor to change pitch and expression by raising and lowering your hand over it like an orchestra conductor.

Though research has slowed on the technology, Lewis still uses a vocoder when performing and brought one with LEO for his performance at NAMM’s 30th anniversary of MIDI in 2013. He also still sees potential for development. “The VP550 is still not where I’d like it to be. It’s no longer in development but I want to find some engineers to keep it going.”

         

Susan Hayes is a professional writer, editor and documentary film enthusiast based in Northern California.  Her Sundance Film Festival record stands at 13 films in four days.