“Should we have a jingle for Studio Snack?” We’ve turned that question into a year-long creative experiment running alongside our episodes. There will be a jingle by Christmas. Or a collection of terrible sound clips for our out-takes episode. Either way, we will learn things.
Narcis has started with a verbal brief: a shopping list of snacks for an AI-generated jingle. The result is – well – interestingly like K-pop (It’s in our Less is more episode).
I’m Lydia and I do the nerdery for our episodes. So I’ve gone down a rabbit hole of research to find out if we might generate sound from images.
In a music education tool, I’ve made a bit of staccato string music from a pattern of sprinkles and painted a string and voice cake.
Trying an AI tool that reads an image and interprets it as music, I’ve shown it some images from our episode experiments. There’s space noise from our dayglo sewn carrots. Our double decker ‘buscuit’, read as a brick, is percussion. There’s East Asian style music from our pretzel ampersand, read as an S made from a snake. The image it understood best was an AI-generated coffee and cinnamon bun emoji. And for the second time* I have broken AI with an Eccles cake, generating feedback so loud that my file name includes “BEWARE – TURN SOUND DOWN LOW”.
*Listen to our Coffee and cake episode to hear the first story.
And I’ve experimented on Soundbow, a visual sound instrument, drawing almonds and a rock bun and tracing them with a sample of my own voice. You can hear the results at the end of our Less is more episode.
Where to next on our odyssey of sound? There’ll be more in our next episode and on the next Studio Snack soundbite.