Modern practices have shifted since the rise of streaming, but where’s the place for loudness now its wars are supposedly over? And what new ways are emerging to game the system next?
Loudness can be a perceptual concept and a measurement on a computer monitor. Mastering engineers used to be limited (pun intended) by an audio file’s peak level, but now the predominant medium for music listening is streaming, and the audio community now master for loudness normalization, the loudness wars are (to a certain extent), over, and loudness is no longer about ‘winning’. Instead, it’s about control, translation and intent.
This article won’t focus particularly on the history of the loudness wars, but we’ll cover it in due course. If you haven’t seen our documentary about the subject, you can get some war stories from the people involved right here.
Loudness Normalisation Changed the Rules
Back in the days when CDs ruled the roost, there was a standardized bit depth of 16, and CD players would play exactly what they were given. This meant there was a definite ceiling for how loud a piece of music or a whole album could be. And as it turns out, this ceiling created a race to the top.
Thanks to a combination of listener demand, record label direction, artist buy-in, brickwall limiter technology and career interests, producers and mastering engineers made their music louder, and ended up pushing closer and closer to that peak ceiling, to the detriment of the music’s dynamic range. This was the fabled Loudness Wars, and a period in which louder often appeared to mean better, even when the musical cost was obvious. When every new release was fighting for impact on radio, in playlists, or on a CD changer, squeezing out more average level could feel like a competitive necessity. But once streaming platforms began turning tracks up or down to meet loudness targets, that advantage started to disappear.
The End of the Loudness Wars
After the rise of streaming, and the introduction of loudness targets and converging broadcast standards, loudness no longer creates an advantage on most platforms. Loudness decisions are now there to shape feel, with playback level dictated more by the genre itself and its norms.
Our new guiding lights are Integrated LUFS to measure loudness, True Peak measurements that represent playback better, and various other measurements derived from the new standards.
“Gaming” Loudness: What Actually Works (and What Doesn’t)
We may now operate in a world of loudness normalization, but that doesn’t mean that increasing loudness isn’t still a desirable skill – it depends more on the circumstance.
Integrated vs Short-Term Loudness Measurements
Loudness can also be measured in different ways: Short-term LUFS over a three-second window; and Integrated LUFS over the course of a whole track. Since loudness normalization is mostly measured and applied using Integrated value, there’s space to play with loudness as a momentary effect without suffering too much of a penalty.
Strategies may be to keep introductions very quiet while pushing choruses and drops very hard. Pop tracks with whispered introductions and EDM tracks with long breakdowns could be the result of Integrated and Short-Term Loudness Measurements being gamed.
Making Compression a More Upward Experience
There’s a way to keep the peaks at the ceiling but bring up the floor in order to reduce dynamic range. Using a limiter with upward compression might feel exactly the same as the usual downward process of reducing peaks then lifting the total gain, but there are some differences to going the opposite way.
With upward compression, you can ‘fill the gaps’ where the music falls below a certain point. This may be barely noticeable by your eventual listeners, but increases average level without even touching the average level itself – let alone the peaks.
Our new L4 Ultramaximizer can work in this upward mode, bringing more loudness without touching peaks in any way.
Keeping LRA Low but Integrated Loudness Down
LRA is a measurement that hasn’t escaped very far beyond the mastering room. It measures loudness on a more ‘macro’ level, ignoring the peak transients and providing a figure for how the loudness of a particular song changes through time. It reflects musical contrast across a whole piece of music.
A techno tune may typically have a low LRA, keeping the energy (and loudness) mostly consistent from section to section, with maybe a few dB changing up or down each time, if at all. Classical music, on the other hand, tends to have a larger LRA, as dynamics shift throughout the duration of a piece.
Since Integrated LUFS is the key measurement for loudness normalization, two tracks with different LRA could measure the same Integrated LUFS, and become normalized by the same amount. However, the track with lowest LRA (Loudness Range) can still be perceived as louder due to its consistency, and unrelenting loudness. With no break for a listener’s ear to reset, the persistence and incessence can be akin to extra loudness.
Going for High LRA so your Listener Goes for the Gain
There’s another cunning way to make your song louder: make your listener do it. By shaping the introduction to a song to begin at a low volume, you may cause such a drop that the listener is more likely to turn their playback system’s volume up. At this point, the trick is revealed: as your song gets louder, ending up even louder than it would have been.
This results in a track with a high LRA but that may not be turned down if the difference in loudnesses cancel each other out.
It’s not necessarily about starting ‘quiet’, per se, but starting with a higher dynamic range to ease the song in. Of course, this isn’t solely a matter for a mastering engineer, and would have to be facilitated at the arrangement stage as well.
Loudness isn’t Dead – it’s just Genre-Specific Again
The idea that the loudness wars ended, exactly, isn’t necessarily a good way to think about the current situation. Club music, for example, still prioritizes density and impact in its presentation, and the rules of Spotify and Apple Music don’t apply in the DJ booth or on the nightclub sound system.
If you’re exporting a single master for all possible use cases, then you can trade the fact that it will get turned down by a streaming service in order to get the extra peak loudness afforded by a live venue.
While we can keep in mind the loudness targets of different platforms, genre expectations ultimately play a bigger role in the loudness we’ll be dialling in for any given piece of music. Loudness between different tracks is more of a stylistic language than a competitive weapon.