Technology helps you copy the voice of others in just one note

What would the world be like if someone could imitate the voices of heads of state like President Trump to broadcast fake news?

Lyrebird - a Montreal-based company has just released an API that allows you to synthesize the voice of the pretty c. The special feature of this technology is that you only need at hand the data is a recording of about 1 minute from the object that you want to imitate the voice.

To prove this, Lyrebird posted some sound clips for example on the company's official website. New technology does not require the subject to say the words you want to use in the recording you want to create. In addition, you can also customize the person's intonation.

Lyrebird's API only needs one minute of recording and will use cloud technology to handle audio synthesis.

Last November, Adobe also tested a similar technology called the VoCo project . However, Adobe's technology requires a recording period of about 20 minutes and uses system resources to synthesize speech. Lyrebird's API only needs one minute of recording and will use cloud technology to handle audio synthesis.

Many people worry that these technologies will be used to imitate other people's voices and transmit bad information. What if someone uses this technology to transmit fake information from the head of state? Or how would you respond if someone imitated a family member and sent you an emergency message?

However, Lyrebird says they publicly disclose the technology and make sure there is no risk for use. The company hopes people will soon know the new technology and that copying a person's voice will become legal.

But, if for normal use, you need to copy voice to use in any case?