While other singing competition series feature contestants hoping to become household names, Netflix’s “Sing On” has created a star in one of its technical elements: the vocal analyzer.
The new series from Hello Dolly puts six purposefully “eclectic” singers together on the same stage, notes executive producer and Hello Dolly co-founder Stuart Shawcross. Standing equidistant from each other, as well as from a speaker that is aimed directly at their face (because the further away one is from a speaker, the longer it takes sound to travel, Shawcross points out), the contestants on “Sing On” take turns singing parts of a popular song, with the vocal analyzer assessing their abilities to hit the exact pitch and timing of the original recording. But, because the art of karaoke is not just about mimicking, the analyzer doesn’t judge the tone of their voice.
“Our first call was for people who loved to sing because we really wanted to get that joyous party atmosphere to come across. And then in the casting process, everyone that we auditioned sang the same three songs so we were able to listen to their voices and compare [them]. A really important piece for us was that their voices all sounded different so you’re getting such a rich medley. That in itself can make the songs sound different from how you’ve heard them before,” Victoria Ashbourne, executive producer and Hello Dolly co-founder, tells Variety.
The conceit of the show came when Ashbourne and Shawcross attended a Massaoke event in London and wanted to carry the same energy they felt as part of the live crowded into a television show, Ashbourne continues. They came up with the format and then brought in Ionoco for the technical aspects of the vocal analyzer. The result is three series so far: “Sing On! Spain,” “Sing On! Germany” and now the U.S. version of “Sing On” (hosted by Tituss Burgess), which were shot back-to-back in a three week period but rolled out with a few weeks in between them starting in July of this year.
The process to perfecting “Sing On” rested heavily on the technology involved in the vocal analyzer. Shawcross reveals that said technology is a combination of an “algorithm that already existed and had been developed over the last 10 years [but] had never been used for this purpose” and a “game engine,” put together by Ionoco’s Stuart Ingram.
For “the better part of 12 months,” Shawcross explains, there was a “team working on the software itself and a team of musicologists” working on the vocal analyzer and the songs the shows would feature. “Our musicologists would trace the original vocal line of the artist [in each track] — and by that I mean they would identify every single note in the song, what pitch that note was, when that note started and when that note stopped in terms of timing,” he says.
After “four or five months” of developing the technology, he continues, they began to bring in groups of singers to test it. “We’d have musicologists or our music director, David Tench, come down and between us all we would constantly keep tweaking to make sure what the singers were doing was what the game engine was hearing.”
The analyzer was looking for an “exact replica” of the original track’s pitch and timing when someone new sang. “In terms of accuracy the system basically takes a measurement every 46 milliseconds,” Shawcross says.
None of the contestants cast on the franchise were part of the testing of the vocal analyzer during their audition process. It was only when they got to the studio to film that they were “fully briefed” on how it worked, Ashbourne says.
This was, in part, because they had to stay confined to their specific area of the stage, even if they were moved so much by the music they wanted to run around. “In the floor, basically in front of each of their light boxes, it looks black but that’s actually a grate and underneath the grate are the speakers, and they’re quite big, sizable speakers,” says Shawcross. “Ionoco measured how long it would take the sound to arrive and they would go on set every day to make sure that a) the band is always playing in time and b) that the singers all received sound at the same time so they were all in sync. Everything was controlled for every individual contestant so it was fair for all.”
Adding to the equal playing field was the fact that the contestants were told what songs they would be singing ahead of time — although “what was crucial was that they absolutely never knew who was going to sing which part, so that’s when they had to be on their toes,” Ashbourne stresses. Although she says “we were looking for those sing-along anthems that people love,” there was always the possibility that a contestant wasn’t too familiar with a particular song, and rather than get tripped up live, on stage, they were encouraged to practice. (They also had traditional karaoke screens with the color-coded lyrics to aid their performances.)
“On the day of record our vocal coach also gave them some tips on that song. It’s a competition but at the same time it’s also entertainment,” says Shawcross.
The idea was to create a “party atmosphere” in the studio, says Ashbourne. This not only included a live band and a live audience but also asking the voted-off contestants to stay and sing backup as the episode went on. The contestants couldn’t see the results of the vocal analyzer during filming; that lower-third graphic was just for the at-home audience to follow along not only with the lyrics, but also how each person was doing, technically. It became such an important feature the shot style of the show was built around it.
“A lot of people love the graphic — they can’t take their eyes off the graphic,” says Shawcross. “In our experience, more than half of the people love watching the lower-third graphic. So we wanted to give it to you big enough that if you were watching the show on a mobile device, you could see the information; we didn’t want you to be restricted by your device. Once we had what we thought was the sweet spot for that size, we then looked at, ‘How much do we need to see of the singer?'”