Audioshake uses A.I. to separate a mix-down into stems
Image Credit: Audioshake
With a single audio file, Audioshake uses artificial intelligence to isolate the vocals, drums, guitar, bass and other instruments.
Audioshake’s technology could make remixing, mash-ups, remasters, removing bleed from multi-tracked live recordings, grabbing samples, and instrumentals for karaoke or background music a breeze. Trained on tens of thousands of high-quality stems to learn what each instrument sounds like, Audioshake uses A.I. tech to recognize and isolate different components.
We’ve seen similar tech previously from the likes of Deezer with Spleeter. While Audioshake admit that their product isn’t perfect, it’s the next best thing when you don’t have access to the original stems. They’re clearly doing something right, as earlier this year, Audioshake won Sony’s Demixing Challenge above researchers from Facebook and ByteDance (TikTok’s parent company). Sound separation goes beyond music to improve products such as hearing aids.
Labels and publishers can sign up to a 15 minute web conference demo taking place tomorrow and over the course of the next two weeks. For musicians, audio engineers, music supervisors and producers who need multiple tracks that you own ‘stemmed’, contact Audioshake. They will create samples that you can choose whether to buy.
Audioshake are now putting the technology in the hands of those in the music industry. Audioshake Live is an on-demand stem creation platform. The demo below shows a program that’s as simple as uploading a song and choosing the stems, before being presented with waveforms that can be muted, soloed or mixed with a click. From here you can listen in the player or download the extracted tracks. Audioshake Live looks set to be a subscription platform aimed at labels and publishers who regularly need to create stems. Those interested in Audioshake Live can sign up for updates here. Musicians who need to create stems on a one-off basis can continue to contact Audioshake.