Listen! The blobs are singing!
Or at least they do in Google’s latest machine learning experiment, the awe-inspiring Blob Opera, in which a chorus of four adorable, colorful blobs serenades you with chilling opera music. Drag a blob up or down and you change the pitch in which they sing; drag them from left to right and you change the vowel sound. Each blob will also harmonize with the others, in what can only be described as magical.
The Blob Opera just sounds beautiful, with rising harmonies echoing from every blob. Four real opera singers – Christian Joel (tenor), Frederick Tong (bass), Joanna Gamble (mezzo-soprano) and Olivia Doutney (soprano) – recorded 16 hours of vocals (Ingunn Gyda Hrafnkelsdottir and John Holland-Avery also contributed), but they are not their actual voices that you hear when the blobs sing.
The team previously trained a machine learning model on those voice recordings. The blobs sing how the algorithm “thinks” opera sounds, based on what it learned in training. An additional model works to enable harmonization.
Created by David Li in collaboration with Google’s Arts and Culture team, the Blob Opera isn’t just an adorable toy – it’s a great example of how machine learning can be used to create something new and unexpected from existing data.
The machine learning based nature of the opera is why the blobs are limited to vowels rather than actual words, but the random sounds still manage to approximate the essence of a real opera, in the mind if not in the text.
But the best part of the Blob Opera isn’t just the pretty harmonies or cute characters – it’s the “Christmas surprise” triggered by clicking the Christmas tree icons, giving you the option to have the blobs sing any number of popular baubles . Christmas carols.
Joy to the world, indeed.