The English Test (aka the “Kit Kat Challenge”)

With all the existing English-Language Vocal-Synthesis Voicebanks, and new ones on the horizon, it’s sometimes hard draw a line between which are “Engloids” and which are “Engroids”.

Luckily for us, some people got together and decided to create an English Test for Voicebanks and figure out whcih ones can ‘passably’ sing English.

We like to call it the “Kit Kat Challenge”.

The test is designed to contain a number of phoneme combinations that Japanese voicebanks (and speakers) find difficult to properly emulate and is similar in purpose to the famous English Alphabet test-sentence, “The quick, brown fox jumps over the lazy dog”.

The English Test sentence reads as follows:

“I always love a good Kit Kat bar.”

While simple in meaning, a native Japanese speaker would need to take a few tries just to wrap their head around these phoneme combinations.

I’ve taken the time to sequence some example tests for Sweet Ann, Big Al, and Oliver.

There’s a bit of hit-or-miss with the default phonemes, but if you can get the phrase to be understood by someone completely unused to listening to Engloids, then it’s a pass.

Sweet Ann

“I always love a good Kit Kat bar.”

Big Al

“I always love a good Kit Kat bar.”

Oliver

“I always love a good Kit Kat bar.”

With these examples I only had to modify the default phonemes slightly, which goes to show that these PowerFX “Engloids” can definitely hold their title as real English Voicebanks.

Go ahead and try it with your favorite VOCALOID, UTAU, or any other synthesized voice; doesn’t matter what language it’s in!

If you think you’re successful, show it to some friends at school or work, or even post it here as a comment. If it sounds good, we’ll upload it right alongside the existing ones here.