“The breakthrough takes probably a minute. That can be followed by 30 years implementing it.” Jeff Bloom knows this from personal experience. The US-born, UK-based scientist and technologist conceived a signal processing algorithm for matching audio signals to each other in the early 1980s and has spent the succeeding years establishing the concept in film post-production and extending its reach into consumer electronics.
After gaining his BSc in Physics and then working as a musician and teacher in Cleveland, Bloom came to the UK in 1973 to work on his PhD in Cardiff, writing his thesis on human auditory localisation and binaural phenomena.
During the late ’70s, Bloom became interested in digital recording studios, contributing articles on the subject to Studio Sound. Through this he met a sound recordist for involved in ADR. “I’d never heard of the process but the idea struck me about lining up signals for editing,” Bloom says.
Four years’ work resulted in WordFit in 1984. “I was petrified somebody would come out with something before me,” Bloom says. “Which is why I patented the concept.” Digital Audio Research (DAR) was formed with Guy McNally, Nick Rose and Gerry Cain, to develop and produce a commercial product, which Bloom admits originally, took 20 minutes to process five-seconds of speech. DAR also produced the SoundStation DAW, which incorporated WordFit.
Despite being the size of a “small refrigerator”, WordFit was used on films of the time, including David Lynch’s Dune and Spielberg’s The Goonies, with both Universal and Warner Bros buying units. Bloom left DAR in 1994 with the intellectual rights to his invention and continued to develop the concept of synching words to lip movements under the banner of SynchroArts.
With VocAlign and ReVoice established for the pro market, SynchroArts moved into consumer, developing Singtones for mobile devices. The unassuming Bloom credits his technical director John Ellwood and DSP expert Jonathan Newland for their part in the development of the pitch processing on ReVoice, describing both as being “genius level”.
The implementation of his original idea continues, with research into the formance of words – the different sounds and shapes – with the aim of adding that to the timing and pitch algorithm. Bloom says this could have applications in language teaching, allowing students to record and mimic themselves instead of teachers.
“I’ve been really lucky to find questions and problems that were really interesting to me,” Bloom concludes. “Perhaps, more important, I was lucky to find environments in which to do the work and some great people who could support my research or project developments.”
Pictures: Top: Jeff Bloom and daughter Sophie in Croatia, July 2016. Second: WordFit system from 1984. Third: ReVoicePro3.2 main screen. Last: DAR SoundStation Sigma audio workstation, which incorporated WordFit
Launched last year, Genius!2 is the second edition celebrating those clever people whose inventions have transformed the world of professional audio. The 30-page supplement is also available to read in a handy digital-edition form.