“We have the most accurate and synthesized voice in the world”, says director of Android

android

In an interview with Wired, Hugo Barra, responsible for managing the development of Android, said that Google’s mobile platform, Jelly Bean, today has the most accurate and synthesized voice in the world.

Although he did not reveal who the “owner of the voice” would be, the executive took the opportunity to detail to the digital magazine some curiosities about how the product was developed, including the choice of the ideal profile among the candidates.

“It is a very, very complicated process. And it starts with the discovery of a voice talent. A group of designers, engineers and speech scientists worked to describe the person’s personality and the voice personality we were trying to create, ”said Barra.

“We wanted something ‘friendly’ and there were literally 15 different ways to describe what that friendship means. So we asked a casting agency for 10 candidates. We recorded with the 10 candidates and did blind tests with all kinds of different people, and we took a vote to reach just two people. We recorded more with these people and did some tests to finally decide: “ok, let’s go with this person.”

Barra says he doesn’t know the person’s name and that his identity is a secret. “In fact, nobody knows her name. It is not something that you need to disclose, because it needs to be the voice of Google ”.

As for the personality of the virtual assistant, the Google executive did not intend to create something that would come to play with the user but a computer faithful to the answers and always available. «There was not a single person in the company with the idea that we should have gone in another direction.»

“We don’t want to make fun of you. Google is a neutral party – it is not your friend, secretary or sister. It is not your mother. It’s not your girlfriend or boyfriend. It is an entity requesting information. You ask, we answer. And it is very important that this entity is impartial. Adding jokes and other mannerisms would take the voice away from that ”.