The Big Misunderstanding
I wanted to share this because there is so much confusion around the internet on what mastering really is. "Mastering is making your track loud." "Mastering is part of mixing; the engineer always mixes and masters." "Mastering is when you put these six plugins on your master channel."
I can see where the confusion stems from. It's really hard to describe a process that may be different every time, or may occasionally be doing nothing at all. And it's something that even the masters of mastering, while giving useful tips on what they do, don't really cover the essence of what mastering actually is in simple terms.
But before we start:
What Isn't Mastering?
Mastering isn't mixing. Mixing is combining all the tracks in a multi-track session to sound appropriate and interesting with each other and to tell a story. Read here if you want to learn more. Mastering is about working with the 2-track export of that mix to polish and finalize it, presenting the mix in its best light.
Mastering isn't just making your song louder. Mastering often includes increasing the final volume, but there is far more to mastering than just turning the volume up.
- Mastering isn't a repeatable plugin-chain. Not every song needs the same things to sound its best. Some songs need to be brighter, others darker. Some need to be more compressed, others more dynamic.
Now that we have a couple of the biggest myths out of the way,
What Is Mastering?
In the simplest way I can describe it, mastering is the quality control check for your finished mix. Mastering refines the mix in the ways it still needs refinement. Mastering compensates for mistakes the mixing engineer made or inaccuracies in the mixing engineer's playback system. Mastering smooths and sands down songs to sound similar to each other and similar to what people expect commercially made music to sound like.
Even to me, that doesn't sound very helpful. Why can't we just know exactly what a mastering engineer does?
Because a mastering engineer adapts.
A mastering engineer does something different every time. And he doesn't know what needs to be done until he hears the song.
When you tune a guitar, not every string needs to be tightened: some may need to be loosened, and some may need to be left as is. And when a string needs to be tightened, more tightening isn't always better than less. After all, the goal of tuning isn't to tighten all strings by arbitrary amounts. Of course, the goal is to have all strings in tune, tightened just enough to ring true on specific notes.
Likewise, mastering a song doesn't require the same action for all songs. But of course, mastering is more complex than tuning a guitar, because a machine can't determine the subjective qualities of what makes a good song in the same way that a machine can determine good tuning, and because mastering covers many variables instead of the one variable per string of tuning a guitar.
Let's explore a few of these variables:
Examples of What a Mastering Engineer Might Do
- A mix may arrive that the mastering engineer finds too dynamic. It's the mastering engineer's job to compress the peaks of the mix to smooth out the overall volume over time, so the song feels more contained. In this case, the mastering engineer will use a compressor, or a series of compressors.
- A mix may arrive that's too compressed. In this opposite scenario, the mastering engineer needs to provide more dynamic range to give the song life and energy. Assuming the compression wasn't too extreme, the mastering engineer can use expansion (the reverse of compression, using a compressor with a ratio of less than 1:1) to bring out peaks and provide punchiness and motion to the track.
- However, if the mix arrived with far too much compression, expansion won't be a suitable band-aid. The mastering engineer will need to request that the mixing engineer back off on compression, particularly on the mix bus, and then send a new version of the song for mastering with this revision. The mastering engineer may not actually be doing anything to the audio in this case, but he's still doing his job by acting as the quality control check for the music, making sure it doesn't get published with that major flaw.
- A song may have been mixed by a poor engineer that isn't skilled enough to realize he left a lot of 200 Hz mud in the mix, or a lot of harshness at 6 kHz. That's okay: we're all at different points in our learning curve and are learning things in a different order. The mastering engineer would EQ out the peak at 200 Hz or 6 kHz, according to what needs cutting. And boost with EQ if the song needs boosting.
- A song may have been mixed by an engineer that loves the punch and power of drums, and he mixed the drums 6 dB too high relative to the rest of the song. In this case, a mastering engineer may send the song back to the mix engineer instructing him to lower the drums by 6 dB before it can be ready for mastering.
We are beginning to see that the mastering engineer compensates for having a less-than-stellar mixing engineer. This can be a big part of a mastering engineer's job, but what if the mix engineer is talented?
- A song may have been mixed by a great engineer on a poor playback system. Say the mixing engineer's monitors are very sizzly and bright at 12 kHz, and he has his subwoofer too quiet by 3 dB, and his room has a null at 50 Hz. Even though the mixing engineer is good, his monitors and his room and his subwoofer settings aren't, and he likely compensated by EQing out too much 12 kHz, mixing all bass too loud by 3 dB, and especially boosting 50 Hz until it sounds right to his ears. The more experienced mastering engineer in a more carefully treated room with higher caliber speakers can compensate for these shortcomings by adding frequencies missing at 12 kHz, cutting all bass by 3 dB, and especially cutting at 50 Hz, bringing the song to a closer degree of perfection and balance.
- An album of songs may have been mixed by a good engineer, but the songs likely sound different from each other since they were mixed on different days. Some are brighter, others are darker. Some are louder and others are quieter. A mastering engineer would brighten the dark songs, darken the bright songs, and generally do whatever needs to be done so all songs sound similar dynamically, spectrally, and in volume.
- An album of songs may have been mixed by a great engineer and all sound consistent with each other, but they may not spectrally fit with what is appropriate for commercial music. For example, most modern pop music has abundant high frequencies on the drums, synths, and particularly on the vocals. A mastering engineer may spectrally shape each song with EQ to achieve the balance appropriate for the genre of the album. This is especially important when the mixing engineer is unaccustomed to mixing in that specific genre.
- A song may not need EQ or compression, but it might sound a little flat or sterile. A mastering engineer may reach for saturation or an exciter or tape emulation or stereo enhancement to subtly enhance the texture and character of a song.
- A superb engineer mixing on a great playback system may have introduced all kinds of inaccuracies because he's simply heard the song for too many hours. The current mix is emotional to him. The song is connected to him. He no longer has fresh ears for this music, and he may not have fresh ears for it for months. A mastering engineer may have all kinds of problems to fix not because the mix engineer is bad, but because the mix engineer is too close to the song. A mastering engineer will begin working on the song with fresh ears and quickly assess what it needs, then do it.
- A song may have been mixed by a superb engineer on a great playback system. It may not need more or less dynamics, or boosts or cuts with EQ. It may have life and character and texture already that really make it sing. If it's a quiet mix, the mastering engineer may merely raise the volume after careful compression and limiting. Or if the mix is already loud and has already been carefully compressed and limited, the mastering engineer may merely say, "This song doesn't need anything." And that is still worth his services: getting the stamp of approval from an expert with fresh ears listening on high-caliber speakers in a carefully treated room.
But all mastered songs are loud, right? Isn't loudness a big component of mastering? Let's explore that:
- A louder song may catch your attention more than a quiet song if you leave your volume control in the same place. That is, assuming you're not listening on the radio, YouTube, Spotify, Apple Music, or any other service or layer of software that manages the final volume level for you. It can be tempting to make music as loud as possible for listeners not using those services. But making a song loud past a certain point introduces distortion and audible unpleasantries, making it uncomfortable to listen to and stripping the dynamics and fullness out of the music. You can achieve volume only by trading the integrity of the music: more volume comes with more distortion, and at an exponential rate. A mixing engineer may not be the best person to establish the intended final volume of a song, and likely isn't familiar with the methods to gain the most volume for the least amount of audible distortion. The mastering engineer is accustomed to how loud music is expected to be for each genre to be considered commercially viable, and how to achieve the desired loudness by adding the least amount of distortion possible. Also, the mastering engineer has likely had a discussion with the artist and producer on how loud a given song or album should be pushed.
- A song may have been mixed too loud, and has already been heavily limited or possibly even clipped. In this case, the mastering engineer would ask the mix engineer to export a new mix without limiting, potentially with all the tracks turned down. After all, a loud mix is not a necessary ingredient in creating a competitively loud master. And as we covered, the mastering engineer is likely more familiar with the genre's expectation for loudness, the artist's expectations for loudness, and which tools are best and how to use them to achieve this loudness. Understanding this, it makes perfect sense for the mixing engineer to a deliver a quiet, dynamic, full-sounding mix for the mastering engineer to master and worry about final loudness.
- When making music loud, a mastering engineer will almost certainly employ a limiter. He may also use several layers of broadband compression to decrease the crest-factor of the mix, likely with very different settings than used for generally aiming to make the mix sound less dynamic. Multi-band compressors are powerful and dangerous tools used to increase volume while maintaining balance among the different frequency spectrums. Also, a mastering engineer could alter the frequency balance of a song to bring up the perceived volume, often by reducing fullness by lowering the bass relative to the high-mid frequencies. Other tools used to increase perceived volume are saturation, deliberate subtle distortion, exciters, and sometimes even intentionally clipping audio in the analog domain before converting back to digital. A mastering engineer will find the best mix of these many tools and how they are used to deliver volume appropriate to the song and album according to the genre and the wishes of the artist and producer. But as a lover of music, I sincerely hope mastering engineers will stop pushing loudness past this point.
Making Sense of This
Can you master your own music? Well, yes and no.
"No" in that it's really hard for the person who mixed the song to know how to overcome his own biased preferences on frequency balance and mix balance. Also, it's really hard for the person who mixed the song to give the song what it needs to sound good on every stereo and set of headphones, not just his own playback system in his own room.
And "yes" in that an engineer can do what needs to be done to a song to make it loud and clean and problem free on his own. But he may need to compensate by abandoning the song for a while to gain fresh ears for it. Or to use many reference tracks to keep the balance and dynamics and mix of the song in perspective. Or to listen to the song on many different playback systems in different spaces to be sure the song effectively translates to other systems.
None of these workarounds are as ideal as sending your song to a trusted master of mastering with a superb system in a superb room, but it can get you a lot closer. And this is important for people who can't afford to hire a mastering engineer for their music.
Clear As Mud?
I wish I could provide a clear, step-by-step outline of what a mastering engineer does. Instead, all I can do is make clear why that isn't possible.
My hope is that my examples offered some perspective on why mastering is so important despite it being something that one can't specifically explain. And it goes without saying that I hope the examples above can give you, my reader, some perspective on what kinds of problems a mastering engineer may fix. So that you can learn to avoid these problems before sending your music off to mastering or deciding to master your music yourself.
If you're interested in learning more about mastering your audio yourself, be sure to check out my guide on how to effectively master your music at home, for when hiring a mastering engineer isn't an option.
Now go make some music!