Have you ever found yourself surfing across the internet for finding the karaoke for a song you would love to perform and wondered if there were any easier ways to do this on your own? Well, your wishes have come true, thanks to Deezer’s new audio separation library Spleeter.
With Spleeter, you can seamlessly separate vocals and instrumentals for any song you feed into it without having any in-depth knowledge of sound mixing. This could be lifesaving for DJs who spend a lot of time separating source tracks.
Deezer says its audio separation library is fast for getting the job done. They claim Spleeter separates audio 100x faster on GPU than in real-time. It contains three pre-trained models for separating vocals and others; vocals, drums, bass, and others; vocals, drums, bass, piano, and others.
If you’re interested to know how the results are, check out this sample generated by Andy Baio from Waxy below.
Led Zeppelin – “Whole Lotta Love”
The AI is not completely perfect but it does a decent job of isolating and separating tracks. It is quite usable in my opinion. “Spleeter is a neat tool, but in no way do we claim to have “solved” source separation. Hundreds of researchers and engineers working for decades have made the advances and built the tools on which Spleeter is based.”, wrote Deezer in a blog post.
Deezer has open-sourced Spleeter on GitHub. You can download Spleeter on GitHub from here. All you have to do is provide an input audio file and wait for Spleeter to do its magic. However, make sure you follow the installation guide here and user manual here to use Spleeter with no hiccups.
So, what do you think of Spleeter? Tell us in the comments.