After a musician is finished in the studio, but before you get to hear their work, there is one crucial step in between. That step is called mastering.
Mastering is performed by a mastering engineer, and it is the process of preparing the final stereo mix for wide distribution. A final stereo mix will usually be balanced and sound great on its own, but that final stereo mix is for archival purposes, not commercial release. That final stereo mix is usually referred to as the master tape.
So, what happens during the mastering stage? Any number of the following: compression, equalization, stereo widening, added reverb, hard limiting, and if needed, additional processing effects, such as a harmonic exciter, tape saturation, etc.
Today, however, I’d like to discuss just one of these aspects, the one tool that seems to be overused on almost every release over the past two decades, and that’s hard limiting.
What is hard limiting?
Hard limiting is accomplished by a piece of gear – or more commonly, an audio plugin – called a limiter. The limiter restricts an audio signal from going above a certain volume.
The limiter was first developed for use in amplifiers and radio broadcasts to protect circuitry and speakers from overloading. But in the mid-90s, limiters were applied in an audible way to try to increase the perceived volume on CDs.
Limiters, however, don’t work the same as compressors. Compressors gradually rein in peaks in the music, and can sound very musical, because they work with the dynamics of the music. Limiters stop a signal dead, and are not gradual or musical at all.
But why should you care that limiters are being used?
Let me show you a picture, it’s a picture of the song “Inactive” by Weird Al Yankovic, released on his 2014 CD/iTunes album Mandatory Fun:
Note: all waveforms pictured in this article have been volume matched, because the above clip has such a high average volume, the file as a whole has to be turned down -6db to have the same average perceived volume as a dynamic master.
See how the song starts off with a quiet introduction and then, bam, it’s basically a solid brick the entire time? That’s a hard limiter at work. The quieter parts of the verses and chorus have been turned up as loud as digital can go (0db in digital audio is the threshold, the signal can not go above 0db).
The problem with turning the quiet parts up so loud, is that when the song gets even louder, as music naturally does during a chorus or bridge, there’s no room for the sound to go anywhere, it’s already at max volume. So instead of a dynamic rise and fall in the song, you get this fatiguing brick of music. Using a hard limiter you can technically make the entire song, loud or quiet, all-loud all-the-time.
Now just last week, Weird Al released a remastered version of his entire discography, and thankfully, the mastering engineer did not use a hard limiter on the master. Let’s take a look at the new remastered version of the exact same song:
As you can see here, we have some actual dynamics happening. This isn’t the most dynamic song in the world, it’s still fairly compressed, but remember that compression is a gradual, musical process, where limiting is a hard, cold process.
Let’s compare these two side by side to really reveal the difference:
All those peaks and valleys – the ones missing on the top, limited version – those are what make music fun to hear. The pounding kick drum, the cracking snare, a driving chorus. Without quieter sections and moments, there can’t be any loud sections.
But Weird Al is comedy music, right? Surely legendary rock albums receive more careful, delicate mastering on CD/iTunes?
Well, no. Over the last 20 years an album doesn’t get much bigger than Green Day’s American Idiot. Over 15 million copies sold, numerous hit singles, a Broadway musical, a worldwide tour…
And here’s what the opening track looks like:
Where did all the dynamics go? Where’s the pounding drums? The chunky guitars?
Well, they’re on the master tape. But not on the CD, or your iTunes download. How do I know? Because a hi-res remaster was quietly released a few years ago, and here’s what that looks like:
There’s the dynamic sound I associate with a legendary rock album! And you can buy this beautiful remaster from a number of hi-res retailers. But only if you know to look for it in the first place. But the vast, overwhelming majority of Green Day fans bought American Idiot on CD or iTunes and have no idea how painfully squished their copy truly is.
Again, here’s a side-by-side comparison, the CD/iTunes master on top, the hi-res remaster on the bottom:
And what they trimmed away with the limiter:
I want to stress here that the dynamic version is not better simply because it’s a hi-res release, it’s better because a hard limiter was not used during the mastering process. Make an mp3 of that hi-res release and it’ll still mop the floor with the original CD release.
Now I could go on and on, but you get the picture. Hard limiting is irreversibly ruining every CD/iTunes download you buy. Once this process is applied during mastering, there’s no way for listeners at home to reverse the process. Listeners would have to wait for a better remaster to come along, if it ever does.
So the big question I’m sure you’re left with right now is, “why?”
Why are labels hard limiting music?
Well, that depends on who you ask. Ten years ago, engineers would say that it was so their latest album would be perceived louder than any other album in your shuffle. The human ear perceives louder sounds as clearer or better, so if they could crank even a delicate acoustic guitar up to digital 0, why not?
However, today all streaming services and most audio software now include level-matching features that are automatic and built in, especially when shuffling. So iTunes will turn down the latest album to a much quieter volume to make sure it’s the same volume as an album from 30 years ago.
This volume leveling that’s performed automatically by Spotify, iTunes, and almost every other service available today, ironically makes the “loud” modern music sound quieter than a CD from 30 years ago.
More commonly today, engineers will say that they hard limit so the album sounds good on the tiny speakers used in phones, laptops, and other common playback devices.
However, all of those devices have the ability to include on-the-fly limiting that could squish a dynamic signal if the user wanted that, assuming the tiny speaker is all the listener had to play their music on. So that reason doesn’t hold much water either, given that a dynamic, proper master could be sold, and a limiting function could be applied when and where necessary.
So these days, I think it’s done simply because that’s how it’s been done for so long now, and always-loud, fatiguing mastering is the sound most listeners are used to today. That doesn’t make it the right decision though.
Any $20 set of headphones (so not ear buds, but anything larger than that) will reveal the difference in a hard limited master that’s stripped away 6-9db of dynamics.
Even if you happen to think low dynamic music is preferable, not only does hard limiting strip the dynamics away, when it’s pushed too hard (as it commonly has been over the last 5-10 years), the limiter also introduces audible distortion into the music.
Let’s go back to my first example above, Weird Al’s “Inactive”. Here’s an audio clip that plays 8 seconds of the CD/iTunes limited master, and then 8 seconds of the dynamic master. Listen to how much distortion is present on the first clip but completely absent from the second clip. (use headphones for better detail, best heard on the “ch” in “cheese”)
You should be able to hear the static in the first clip clearly, even on inexpensive headphones. But to really bring the point home, here’s another render of the above clips, but with all of the low end frequencies turned down, so all you’re hearing is the top end. Note: I didn’t turn up the distorted frequencies, I simply rolled off the low frequencies so you could hear the high end better:
Again, the first clip is the CD/iTunes master with hard limiting, and the second clip is the remaster without hard limiting. The distortion from hard limiting on the first clip is constant.
And this distortion is not unique to this Weird Al release, it’s present everywhere, from Adele and Sia’s latest albums, all the way back to some of the best 90s albums.
So now we have not only terribly squished dynamics, but audible digital clipping/distortion too, all in the name of making the CD/iTunes master louder than the previous song you heard, because labels think you’ll like it more if it’s louder.
I think I’d prefer my music without digital distortion, and with more natural dynamics. I know where the volume button/knob/icon is if I want to turn it up.
A few artists have compared their recent releases to older, dynamic releases and have begun reversing this process, demanding more dynamic releases. Some artists have taken on full remaster campaigns of their entire discographies, like Weird Al, Tom Petty, Rush, Paul McCartney, and many others. But the vast majority of artists and labels just don’t seem to care or acknowledge the damage they’re doing. Meanwhile, most listeners don’t know it’s even going on. However, once educated on the process, most music fans hunt down more dynamic releases when available. If you’d like more information about any of the above, start at wikipedia’s page for the “Loudness Wars” and go from there, they’ve collected a ton of good research and resources.
And of course, music fans are trying to educate as many people as possible to hopefully turn the tides: