Standardizing Perceived Loudness
LUFS, which stands for Loudness Units relative to Full Scale, is an international standard for audio normalization (formally specified in ITU BS.1770). It was developed to solve a long-standing problem in audio: the fact that two pieces of audio could have the same peak level but wildly different perceived loudness. For example, a highly compressed rock song would sound much louder than a dynamic classical piece, even if their loudest peaks hit the exact same level on a meter. This led to viewers constantly adjusting the volume between different TV shows, commercials, and streaming platforms.
How LUFS Works
Unlike older peak meters that only measure the loudest instantaneous signal, the LUFS algorithm measures loudness in a way that closely mimics how the human ear works. It takes into account:
- Frequency Response: It gives more weight to the mid-range frequencies that our ears are most sensitive to.
- Duration: It measures the average loudness over a period of time, not just the peaks.
This provides a single ‘Integrated LUFS’ number that represents the overall perceived loudness of an entire program. By adhering to a LUFS target, broadcasters and streaming services can ensure a consistent volume experience for their audience.
Target Levels in the Industry
Most major platforms now have a strict target LUFS level for their content. For example:
- Netflix: -27 LUFS (for dialogue)
- Spotify: -14 LUFS
- YouTube: -14 LUFS
A re-recording mixer must mix their film or show to hit this target precisely. If their mix is too loud, the platform will automatically turn it down, often with a simple limiter that can ruin the dynamics of the original mix.