There have been various attempts to create music software that isolates vocals and instruments from a mixed stereo audio file, but if you’re looking for the ultimate in convenience and accessibility, lalal.ai could be the answer.
Said to be based on the world’s number one AI-powered technology, this promises to extract vocal and instrumental tracks from any audio… and all you need to do is drag and drop your track into your web browser.
Other than choosing the level of your processing - Mild, Normal or Aggressive - it really is that simple. Once the track has been uploaded and processed, you’re presented with a preview of both the isolated vocal track and the instrumental with the vocal removed.
If you’re happy with the results and want to download them, you can sign up to one of the payment options: $2 gets you one input track of up to 15 minutes; $10 gets you 10 input tracks or a total of 90 minutes of audio; and $30 gets you 30 input tracks or a total of 500 minutes of audio.
There are various reasons why you might want to use this kind of technology. Perhaps you’re a producer who wants to sample an instrumental loop, or maybe you’re a remixer who needs a specific vocal a cappella. And then there’s good old karaoke - maybe you just want to create an instrumental backing track to sing along to.
Compared to other audio splitting technology, lalal.ai is the best solution on the market. Its machine-learning algorithm is based on 200GB of training data, though this was ‘artificially increased’ to 20TB. The digital signal processing is said to have 45 million neural network parameters, which certainly sounds impressive.
For users, though, it’s the results that matter, and the good news is that everyone can try lalal.ai for free to find out if it delivers what they need it to. All major audio formats are supported and you’ll get your results within a few minutes.
Head on over to lalal.ai now and find out what it can do for you.