Who makes it
Frédéric Le Louët
Sound designer, supervising sound editor and re-recording mixer.
I have worked in sound for twenty-eight years, in film, television, animation and video games, as a sound designer, supervising sound editor and re-recording mixer. I am a member of the Academy of Motion Picture Arts and Sciences.
VoxSplit Pro started with an editing problem I kept running into, on many different films: two actors speak over each other, and the line cannot stay as it is. ADR is not possible, there is no other usable take, or you want to preserve one actor’s performance.
I also studied computer science, and that is what let me take the step: I set out to build the tool I was missing. Each version is tuned by ear, on real dialogue, the way I would work in an editing room. VoxSplit Pro is made for that precise moment: when a line is good, but another voice runs over it.
The data
What the separation is trained on
The training data includes speech corpora released under open licences. We also use our own recordings to test and improve how VoxSplit Pro sounds.
Professional Sound Alliance
Limited Edition Audio supports the Professional Sound Alliance, which campaigns for fair licensing of professional sound.
The seal shows our support for the Alliance. It is not a certification, and it is not an endorsement of VoxSplit Pro.