I was fiddling around in vocaloid mobile and found a new way to emulate Japanese pronunciation with an English voice. A lot of people say to use the [l0] phoneme at a high consonant velocity to simulate a Japanese r, but even then, an American accent shines through. I’ve found that doing this can give an even more percussive simulation of the 4 phoneme (this was done in the Vocaloid Mobile Editor, so the interface may look a bit foreign haha).
in the short phrase “から”, I was able to get the ら sound by using a 16th [-] note before it, shortening the [d V] (or ra) note to a 32nd, and filling the rest of the space the note would fill with another [-] note.
But I’m not just going to be talking about emulating the 4 phoneme
Traditionally, Japanese t’s are much softer than English t’s, and we’d use the same method of shortening the [d] phoneme as pictured above to soften it up
Note: I’ve also found the @ phoneme to be a good replacement for a Japanese “o” (occasionally using @U based on the context of the @ note in use)
With a bit of finagling, your English voice could be singing mildly accurate Japanese in no time!
Another note: this method is recommended for Voicebanks with shallow u: phonemes (I.e. sonika, cyber diva, daina). Voices with a more heavyhanded u: (like Fukase English) may sound clunky
Here’s a cover I made using this method with cyber diva! (Sorry the mixing isn’t too great haha)
Have fun!
in the short phrase “から”, I was able to get the ら sound by using a 16th [-] note before it, shortening the [d V] (or ra) note to a 32nd, and filling the rest of the space the note would fill with another [-] note.But I’m not just going to be talking about emulating the 4 phoneme
Traditionally, Japanese t’s are much softer than English t’s, and we’d use the same method of shortening the [d] phoneme as pictured above to soften it up
Note: I’ve also found the @ phoneme to be a good replacement for a Japanese “o” (occasionally using @U based on the context of the @ note in use)
With a bit of finagling, your English voice could be singing mildly accurate Japanese in no time!
Another note: this method is recommended for Voicebanks with shallow u: phonemes (I.e. sonika, cyber diva, daina). Voices with a more heavyhanded u: (like Fukase English) may sound clunky
Here’s a cover I made using this method with cyber diva! (Sorry the mixing isn’t too great haha)
Have fun!
