In the last few years, AI has made some incredible things possible in the world of audio.

One tool that my team and I at DeFacto Sound often use is called La La Lie.

It uses AI to split apart the sonic elements of any audio or video file, like separating voice

from music, or even splitting a song into individual instruments.

Their latest model is called Andromeda, and it's their most advanced version yet.

Here's a quick demo.

We start with this.

Run it through La La La Lie, and then get all of this.

Lookin' through my files and I'll check in the news.

Right now, until February 26th, our listeners can get 20% off a La La La LaS subscription using promo code 20K. That code only works

if you pay by card and not PayPal. To sign up, visit

la la lao-a-l-al-a-I and use promo code

20K. You'll find a link in the show description.

You're listening to 20,000 Hertz.

The stories behind the world's most iconic and fascinating sounds.

I'm Dallas Taylor.

Here in the U.S., most people have probably never heard of the BBC Radiophonic Workshop.

Even as a sound designer, I had never heard of it until well into my career.

But over in the UK, the Radiophonic Workshop was a cultural institution for four decades.

It was basically a pioneering sound design department funded by the British government,

almost like NASA, but for sound.

And it was staffed by a rag-tag group of engineers and composers whose entire job was to craft

...