Music

Taking vocal software beyond Miku

by Patrick ST. Michel

Special To The Japan Times

Tokyo-based electronic musician mus.hiba has worked with a handful of vocalists during his short stint as a producer. However, many of them haven’t been human.

His go-to voice belongs to Sekka Yufu, who boasts a light delivery that fits with mus.hiba‘s slowly unwinding swirl of synthesizers. She’s a computer-generated voice powered by software called Utau, a free-to-download application resembling the popular singing voice synthesizer program Vocaloid.

“Yufu isn’t like Hatsune Miku,” mus.hiba says, comparing her to Vocaloid’s main megastar. “She’s kind of a minor character. She has a whispery voice and nothing like that exists in any of the Vocaloid voice banks. I thought it was very cute.”

Yufu’s voice appears on almost every track on the artist’s full-length debut, “White Girl,” which comes out Dec. 10 via the Noble label. It’s a strong showcase of mus.hiba‘s slow-motion style, where songs play out patiently but tend to climax in loud bursts. The music also marks a sharp departure from most Vocaloid songs, which often move at a faster tempo and treat the computerized voice as the center of the track.

For his productions, mus.hiba treats the Utau vocal bank like any other instrument. This sets him apart from other producers using vocal software, giving his work a more experimental feel and a hint at what Miku might be doing if creators didn’t dream about hearing their songs opening a Lady Gaga concert.

The Tokyo creator comes across as private and quite shy. Even some of the people he collaborates with say that they don’t know his real name. He does reveal that he grew up playing drums in various hardcore bands, but eventually grew tired of it. Then, four years ago, Vocaloid caught his attention.

“After work I would play around online and watch Vocaloid videos,” he says. “People were starting to do crazy things with the software, and it caught my attention. I thought it was something I could do.”

It prompted him to buy the software for Vocaloid character Megurine Luka, even though, he says, it was kind of expensive.

Vocaloid software has been touted as a tool that allows users to create music featuring computerized vocals that are meant to sound as human as possible — a popular Google Chrome ad starring Hatsune Miku emphasizes this everyone-can-create-and-share ethos. However, there’s a monetary barrier. Crypton Future Media, which created and distributes Miku, Luka and other famous character software packs, lists those character’s prices at ¥17,280. While mus.hiba paid up at the time, he ended up finding an alternative thanks to system issues.

“I couldn’t really use the Vocaloid library on a Mac, at least not easily, but Utau freeware had a Mac version,” he says, adding that this is how he met Yufu. “I really liked Sekka Yufu’s voice, and I had heard other creators use her in the more electronica-leaning artist community.”

Utau appeared online in 2008, created by someone who identified themselves only as Ameya/Ayame, and was free to download.

“I saw characters I had never seen before on Nico Nico one day,” says Ceriph, a Nico Nico user who has compiled an unofficial monthly ranking of Utau songs on the video-sharing website since 2010. “It turned out they were fake Vocaloid characters made by some 2channel users to make fun of Vocaloid fans who believed they were real. But somebody made actual vocal libraries from those fake characters by using Utau.”

While Vocaloid trumpets its ability to allow users the freedom to do what they want with the sounds and the characters’ images, there are restrictions. The main one is that there are only a handful of avatars to play around with, and they all cost money.

“Utau is a community created entirely by users,” says Ceriph, who asked that his real name not be used for fear of getting in trouble at his job. Those using Utau can create characters all their own with a unique voice, and share them with the world for others to use free of charge.

“Lots of Utau creators don’t really follow Nico Nico trends,” Ceriph says. “At one point, high-speed Vocaloid songs were popular, and everyone just made the same sort of song over and over again. But Utau users don’t really make fad songs.”

Yufu, mus.hiba‘s preferred Utau character, represents this trend well. Created in late 2008 by Rokka Fuyou, her appearance differs drastically from most Vocaloid or Utau characters. Whereas most feature an array of bright colors, Yufu is monochrome, reflecting her wintery image. As for Yufu’s voice, Ceriph says, “Her whispery vocals don’t really fit” with the “energetic” vocals of Miku and the other Vocaloid options.

“I always try to include the character’s theme in the song,” mus.hiba says about writing music. “I’m always thinking about her image within the song. With Sekka Yufu, that includes snow, the winter, being cold.”

The upcoming album’s title, “White Girl,” refers to Yufu. One of the tracks is titled “Slow Snow,” and much like the other songs on the album, it has an icy atmosphere that moves slowly but then surprises with moments of release. Yufu’s vocals serve as another layer in a blizzard of electronic sound.

“I think the winter is only so-so,” mus.hiba says with a laugh. “I like spring much more.”

Even within the Utau community, mus.hiba is a lone wolf of sorts.

“I was influenced by chillwave artists such as Slow Magic, Sun Glitters and Taquwami, and the Los Angeles beat scene,” he says. “When I started out, there wasn’t really anyone using Vocaloid as an instrument in the way I was.”

And unlike many of his peers, mus.hiba has gravitated away from Nico Nico to gain more widespread attention.

Mus.hiba put his songs up on Nico Nico initially, but didn’t get much of a response,” Ceriph says. “Instead, he moved to SoundCloud and similar sites that reach people who liked that type of music. That’s really rare for someone in the Utau community to do.”

That migration exposed mus.hiba to a wider range of artists, such as English producer bo en and America’s Meishi Smile (mus.hiba has collaborated with and remixed with both). Mus.hiba played on the same bill as Meishi Smile this past May at live-streaming venue 2.5D, and that performance attracted Noble’s attention.

It also earned mus.hiba the chance to work with human singers. Online imprint Bunkai-Kei Records released mus.hiba‘s “Lycoris” EP this past October. Every song features a human voice, including Japan’s Smany and New York’s Abigail Press.

“It was actually easy, because I didn’t have to program any voices,” mus.hiba says. “It was really fun.”

Moving forward, mus.hiba says a remix album of “White Girl” will come out next year, while his next full-length has started taking shape. Whether it will feature singing from people or programs remains to be seen, but if it’s the latter, mus.hiba says he won’t be breaking up with Yufu.

“If I keep making music like this, I’m going to stick with her,” he says. “She is the voice for me.”

“White Girl” by mus.hiba goes on sale Dec. 10 via the Noble label. For more information, visit www.noble-label.net.