For a one minute video covering my processes, scroll down to the bottom!
This project started from a thought: it’s been a long time since I’ve put my voice in use.
As an ex-opera singer who originally went to the University of the Pacific to study music, singing was something on the forefront of my mind since I was a youth. I’d always planned to do something related to music, and that was what I focused on when searching for colleges. The only other requisite I had was a decent Japanese department. The University of the Pacific had both.
I would come to realize that my love for music wasn’t as great as others’, and eventually I would change my degree track from music to Japanese. Still, my love for music remained throughout my entire time in school. After graduating and moving to Japan for half a decade, my voice became something of a closet hobby. There was no reason and no way to use it, and so I let my voice die.
When I had the opportunity to dub this song in Japanese, I was ecstatic to have a chance to use the voice I’d long left sitting in the corner.
The Song
I went through a number of songs before arriving on the one I decided to use: Love Like You. A nice song from the cartoon Steven Universe. You can hear the original English version here:
The Lyrics
As I am a Japanese speaker, I originally intended to translate the song myself, though I knew it wouldn’t do the piece justice. I had a few lines down before I realized that nothing I wrote captured the nuance of the song well enough. So I turned to the internet.
Surely someone in Japan has translated the piece in Japanese. Surely?
As luck would have it, someone had. I found the lyrics on another youtube video:
Official translations by the Japanese Steven Universe team.
It was time to move onto the hard part.
Singing
Between memorization, warmups, and a lot of sharp and flat notes, the singing process took around 7 hours total. I probably drank around 8 large cups of water in an effort to keep my throat wet. I was very rusty. At first, I sang for around 5 hours straight, then I took some breaks between each take after that point. I ended up with hours upon hours of audio to go thorugh. Eventually, I had enough clips that were good enough for me to move onto the next step.
In order to get a version of the song that was good enough for me to not want to cringe in a corner. I spent another several hours listening to my takes, cutting pieces I hated out, and pasting together mashed up frankenstein versions of the song. Some of the takes were so bad I immediately tossed them. Others had one good note among a pile of garbage renditions. The worst part was realizing that after all that work, I still wasn’t happy with it.
The thing is, as a professionally trained opera singer, I lacked the more breathy style of singing that most songs use. I know how to use my diaphram, I sing loud, and my vowels are really round. For a song like this, I just… didn’t sound good. Think choir girl trying to sing Mariah Carey’s “All I Want for Christmas is You.”
After all that work… I had to scrap it and start over again. This time, however, I worked smarter, not harder. I sang phrase by phrase. And at the end of it… I had a piece that I approved of.
Post-Editing
Post-editing was the next part that took forever to do. Singing phrase by phrase is great, but now you’ve gotta make sure the breathwork works togeteher and that the piece sounds like a smooth song taken in one go. I spent another several hours on this step, playing with things like echo and delay. After a while, I did get it to a point where the song was stylized without being over-the-top with effects.
Finally, in Premiere Pro, I put the video and audio together and then subtitled it in both Japanese and English. This process was not too terrible, though I ended up having to use captions alongside graphics to get it to align well.
