Recently I was mastering a single for a client. I had asked for some references to use, and they sent me several of their previous singles to use as a starting point to gauge their expectations, I analyzed the file to get a few stats to help me know how to proceed, particularly regarding how loud the track was.
I wasn’t surprised by what I found out.
It measured really loud.
The Loudness Wars
Throughout the early 2000’s and well into the 2010’s there was a phenomenon in the music industry called the Loudness Wars. Most people familiar with music are probably familiar with the term, but in case you’re not here’s a quick recap.
The Loudness Wars were a period of time in the recorded music industry built on a false notion that if your CD was louder than one of your competitor’s you had a competitive advantage over them. As with so many bad ideas, it was built on a partial truth: that louder sounds better.
You can easily prove this by doing a quick experiment. Take a song and play it for someone. Then play them the same song that’s just a little louder. Now ask them which one sounds better. Inevitably they will say the louder one. Every time.
It’s a curious feature of our hearing system that has been exploited in so many ways. Want to make a piece of gear sound better? Just make the output a little louder. Want to make a stereo sound better than another? Just make it a little louder. Want your CD to sound better than someone else’s? Just make it louder.
Limits and Limiting
Here’s the problem: there’s an upper limit. As I hope you’ll understand by the end of this post, a more important metric for making great music isn’t loudness, it’s dynamic range.
Any playback medium has an upper limit: a point at which it can go no further. In terms of analog electronics, a signal cannot exceed its power supply voltages. With digital signals (CD, MP3, streaming, etc.), the signal cannot exceed 0 dBFS (decibels full scale).
When it comes to the “loudness” of a recorded medium, it’s all in our brains. I.e. our brains are taking the physical stimuli of air movement caused by a transducer (a speaker, a headphone) and turning it into what we “hear” in our consciousness.
One of the ways that our brains determine “loudness” is by the contrast between the loudest portions and the quietest portions. It turns out this contrast is really important.
In order to get a music file “loud” one has to use a brickwall limiter. There are other tricks and tools, but this is the primary way. It’s called a limiter because it “limits” the dynamic range.
A good analogy here is packing your suitcase. There is a physical limit to how many things you can stuff into your suitcase. When you’re packing for a trip you start putting all the items in that you’ll need, and as so often happens you start running out of room. What do you do? You start compressing things, smooshing them as small as they will go, so that you can fit everything. Then you sit on top of it and zip that puppy up. And then you hope it doesn’t explode everywhere.
This is essentially what a limiter does. There is a finite amount of audio “space” on a CD or a digital music file. In order to make the most of that space, mastering engineers employ a limiter to limit the dynamic range of the audio. As it turns out, for some genres this is really great up to a point. If genres like rock, EDM, or metal, had too much dynamic range they wouldn’t have as much impact as they do. Limiting the dynamic range is an essential part of the aesthetic of that genre.
But as with all things, there is an upper limit. As music is squeezed further and further, there is a point at which the music loses its vitality and punch.
The only way to get a track “louder” is by using a limiter to keep the music from going over a predetermined threshold, and essentially turning up the lower, or quieter, parts of the music, thereby reducing the difference between the loudest parts and the quietest parts.
When the Loudness Wars were in full swing, mastering engineers were pressured into squeezing more and more “loudness” out of the music they were given in order to hit the tenuous target of “louder”.
The result? The music sounds weak, wimpy, distorted, and farty. Metallica’s album Death Magnetic is a really great example. It’s one of the quintessential examples of loudness gone too far.
A New Era
Enter the streaming era. Players like Spotify, Apple Music, Google Music, and many more offer streaming music on demand. Since they are essentially selling you a user experience rather than music, it’s very important to them to make your user experience as pleasant as possible, so as to keep you as a customer, and thereby keep getting your money.
Imagine listening to one of your favorite tracks on shuffle. It’s a little bit too quiet, so you turn it up. The song ends and you excitedly wait for the next great song. It turns out this next song is comparatively much louder than the previous track. It completely blows your ears out with how loud it is. Annoyed, you turn the volume down. The next song is even quieter than the first, so you have to turn it up even more, but by this point you’re scared of having your ear drums assaulted again.
How likely would you be continuing to use this music delivery system? Over time you’d probably use it less and then eventually drop it.
To prevent this, streaming music services have started using loudness normalization to ensure that their customers have a smooth experience.
Using a measurement called LUFS (Loudness Units Full Scale), they will reference a track to their standard, and simply turn it up or down as needed. LUFS is basically a standardized, objective measurement of loudness.
What is happening right now is that a lot of producers, engineers and mastering engineers are taking this old paradigm into the new world of online streaming, trying to make their products as loud as possible. The problem is that when the files get delivered to the streaming services, they will just get turned down to be the same relative loudness as everything else.
If they were sacrificing clarity and punch for the sake of being louder, now the music will just sound smaller and less interesting.
They’ve sacrificed “loudness” for an appropriate dynamic range. In the end, this hurts everyone involved.
Why is this important? Because so often what happens is that in order to achieve “loudness” (which again, was a myth) we’ve inevitably had to sacrifice dynamic range to achieve a sketchy standard. We’ve had to push beyond the appropriate dynamic ranges of the genre.
So what does this mean?
We need to get away from asking “is my track loud enough” to “does my track have an appropriate dynamic range for my genre”. This is a much more useful metric than “loudness”, and one that will hold up over time so much better.
Let us help
Mastering can be an intimidating experience. It doesn’t have to be. We’d love to walk alongside you through this last stage of your project.
It can also be really hard to know exactly what you’ll be getting, so we offer a free mastering sample to give you an idea of what your final product could sound like. Send us a track today and get your free sample!
let us help
Making music on your own can be overwhelming and frustrating. Did you know we offer remote production coaching? You can schedule a free 30 minute consultation video chat. We can take a listen to your song and help you get unstuck.