News

Music: an AI program from John Hopkins University to sing accurately

An exciting initiative has emerged within the prestigious American university of John Hopkins, consisting of developing a remarkable program dedicated to artificial intelligence, with the aim of addressing a common difficulty encountered by many singers: singing off-key. Indeed, researchers at this prestigious institution have succeeded in designing an innovative solution, combining technology and music, to allow artists to express themselves with exemplary vocal accuracy. Thanks to this revolutionary initiative, it will now be possible to correct errors and inaccuracies in musical interpretation. John Hopkins University is thus opening up a new dimension in the artistic field, harnessing the potential of artificial intelligence to achieve vocal perfection. This bold advancement is certainly a springboard for singers looking to improve their skills, by delivering musical notes flawlessly. In short, thanks to this innovative program, artists can reach an unsuspected vocal level, avoiding off-key notes that can tarnish their performances. The innovative union of artificial intelligence and music thus opens up undeniable perspectives and heralds promising future developments for the entire music scene.

Diff-Pitcher is a free and open-source software, which is actually a vocal tuning tool. It corrects off-key notes and transforms them into on-key notes. This method, known and widely used in the music field, is called Auto-Tune.

However, in general, this method gives an artificial and robotic voice. But thanks to advances in artificial intelligence, this new program is capable of correcting the pitch of notes without altering the voice and preserving its original timbre.

The researchers’ long-term goal is to make Diff-Pitcher operational in real-time, for example on a karaoke machine, so that everyone can sing on-key with their own voice, without disturbing those around them.

A lire aussi  Karim Dridi celebrates marginality in a moving road movie, a tribute to the "invisible."

A very useful tool for professionals

For now, this technology only works on recorded music, but it could already significantly simplify the work of producers. Indeed, they currently have to manually adjust each note of the voices of pop singers.

With Diff-Pitcher, it may finally be possible to hear the true voices of rappers, as they too will be able to sing on-key without having to all possess the same robotic voice.

Information source: francetvinfo.fr

Leave a Reply

Your email address will not be published. Required fields are marked *