The Pros Don't Care About -14 LUFS
Sam from Musik Hack gives the real deal with LUFS
Post Courtesy of Musikhack
The pros don’t care about -14 LUFS/-16 LUFS all, and as an indie producer, mixer, or mastering engineer, neither should you. Simply looking at a Billboard chart will show you that most music with high production values is still being mastered around -8 LUFS. Here’s how these numbers became ubiquitous:
Broadcasters (TV for example) and streaming media providers (like Spotify) were getting inconsistent volume as their number one customer complaint. As a result, standards bodies in Europe (EBU) and the US (AES) worked hard to develop guidelines to solve the issue via track metadata and volume normalization, then published them.
At first it seems simple: pick a loudness target. Push loud tracks down and quiet tracks up. But then, when you push quiet tracks up, you push any peaks in that material above 0 and need to slap on a limiter. When thinking about raising up a quiet track, the standards bodies came to realize this was a big problem for music that’s relatively quiet, but still has some loud peaks. Can you imagine how bad classical music — with a really low integrated LUFS value — would sound, normalized up and smashed into a limiter? And what about non-music material, such as film audio or podcasts? In the end, EBU decided everybody should broadcast at -23 LUFS. That gave them the ability to reduce loud material way down to match high dynamic range material like classical or non-musical broadcasts, and they wouldn’t need to push very many things up.
As it turns out, portable music players don’t have the juice to play music loud enough for consumers when material is delivered at that level, and they have too much noise in their amplifiers, which would make the signal to noise ratio bad if you cranked it. So the standards bodies said fine, we’ll make more standards to try and get manufacturers to make higher quality portable amps (good luck with that!) and said we’ll relent for music and suggest -16 LUFS instead. The AES calls out the aspirational target of -23 LUFS, but never bothered to recommend it in today’s environment.
But wait! Albums intentionally have tracks at different volumes, so how do we maintain that while normalizing? The standards bodies decided OK, we’ll take the loudest track (by LUFS) on the album and normalize it to -14 LUFS, so that it’s a little louder than our target, and use the same normalization factor on all the rest of the tracks to keep them relatively matched for album listeners. We’ve now arrived at -14 LUFS: a double compromise based on the difficulty reconciling the dynamic range of different audio content. Not only that, we’ve learned the risk of distorting your track on delivery comes from providing a track that’s too quiet, not too loud! That’s when a client might use a limiter.