New and Noteworthy: Issue 31

June 15, 2023 Greg

👋 Welcome to the 31st issue of New and Noteworthy, a collection of interesting ideas and projects from the world of music and audio.

Research

MusicGen: Simple and Controllable Music Generation - Meta has open-sourced their MusicGen language learning model (LLM), developed interally by their Audiocraft team. The model will generate 12 seconds of audio based on the description provided. You can optionally provide a reference audio from which a broad melody will be extracted. The model will then try to follow both the description and melody provided.

Around the web

A new era of music fluidity is here - “So fluid is music consumption becoming, that we are even seeing the emergence of music that adapts in real time to the user and their environment”

A new report from Luminate has found 120,000 new tracks released to streaming services every day

What Fortnite Creative Mode’s payout model means for music

Tracklib claim 17% of Billboard Hot 100 chart hits in the US for 2022 sampled previously-released songs. Up from 14% in 2021.

From Finland, standards for acoustic environments in buildings

Apple audio ray tracing explained. Related: Good thread on the complexity of design for immersive environments.

Laced Records CEO, Danny Kelleher, is on a mission to legitimise game music as an art form

Creating a music festival in Minecraft

iZotope, Plugin Alliance, and Brainworx join Native Instruments

Video: 1000 oscillators!

Video: What is meant when hip-hop is called “alternative”