You're ready to publish your music and you can't afford a mastering engineer. You're familiarized with what a mastering engineer does, and you're ready to tackle this yourself. But one problem: you don't know where to begin. What's the best way to approach mastering when you've never done it before? Fortunately, that's what we're covering today.
Preparing Your Files
Before you begin, you need to gather all the files: whether you're mastering an 18 song album or a 3 song EP, you need to have all your music in the same place.
For each song, open up your mix and check that everything is sounding good. You should never save problems to fix in mastering when they can be fixed more easily in the mix. If you need to make changes, now's the time.
For each song, remove any limiting or aggressive bus compression in the mix. These are things better saved for mastering. Then, export each song as a high resolution wav file using the sample rate of the song and the highest bit depth option you have available.
Arranging Your Files
When all your files are assembled, open up a new session in your DAW and place each song on a different track. It can help to first decide on song order, then space the tracks out so no one track is playing on top of another. When you do this, give some thought to how much silence sounds best after a song ends before the following one begins. For each track, add a marker on the timeline for when it begins. This will help you export each song consistently with the same song length.
Also add a track or two by an artist in your genre that sounds really good. Music you love to listen to. This will come in handy for reference checks.
Pro Tip: Turn the volume of the reference tracks down so their volume is equal with your tracks.
This will make comparisons much easier.
Play a bit of the professionally mixed song, then a bit of one of your songs. Listen to the pro mix again, and then your mix again. Is your mix sounding brighter? If so, gently use EQ to darken your song. Does your bass need to be brought up? Fix it with EQ. The goal is to tonally match your song to the reference track. When your frequency spectrum is right, your music will sound a lot better on other stereos. When you've finished EQing the first of your tracks, do the same for the rest of the tracks in your album. Now you see why it's important to have each song on a separate track: each song requires different EQ to get it to match the spectrum of your reference track.
Listening straight through can be tedious, and your ears will adjust too quickly. Don't be afraid to quickly jump around in your track and the reference track. You're looking for differences that are most apparent in the first second you start playing back each song, not hidden details revealed after ten minutes of intense listening.
Focusing on the Dynamics
Remember my post on micro-dynamics and macro-dynamics? We're going to take it from theory to practice.
In each of your songs, intently listen to the volume change within a measure. If your song is energetic and has solid percussion, you can hear the volume jump up and down within each second of music. This volume change is what we're paying attention to. If there's too much difference in intensity from the loudest moment to the softest moment, now's the time to bring in a little compression with a fast attack. A light touch is best: aim to compress just 1-2 dB off the peaks. If you feel your song needs more, add a second compressor and set that one to compress the peaks no more than 2 dB. Subtlety is the name of the game.
If you feel your song needs more dynamic diversity instead of less in each measure of music, you can try using an expander (a compressor with a ratio of less than 1:1). But the best place to fix this is back in the mix. If this is the case, go back to your mix, remove the extra compression there, export the song, and start again.
If your goal in adding compression is to achieve a little bit of flavor, aim for a colored, vintage compressor. My favorites are TuCo by Sonimus and Puigchild by Waves. With character in mind, set your compressor with a slower attack, for the compressor to actually begin working after the "threat" of loudness has momentarily passed. And aim for a slow release that about matches the amount of time before the next drum beat comes in. Compressing in this way lets the power of the drums poke through the compressor, but everything between beats gets a little squished and starts to gently pump in time with the music. This effect may be best when used in parallel, so aim for a compressor that has a wet/dry control.
Now that the micro-dynamics are in-check, focus on the macro-dynamics as you listen to the transitions in your song. When the song goes from verse to chorus, does the chorus jump out at you and feel special? If it doesn't, it should. A great way to achieve this is to use volume automation to dim the verses by a decibel or two, which then lets the chorus pop out at the listener. If you want each song section to sound special, you can use automation to gradually dip the volume from the beginning of a verse to the end of the verse, where it then pops up for the chorus. If you're subtle, the listener will never notice the drop in volume, but each new song section will feel energetic and big.
Once you've tweaked the dynamics of one song, repeat for the rest. The album's starting to come together.
Pro Tip: If you plan on pushing your song extra hard into the limiter, you may get better results
using volume automation for macro-dynamics after the limiter, not before.
For a comprehensive look at how to achieve the perfect loudness for your music, check out my blog post on the topic. But for now, we're just going to skim over the functional parts of that post.
To find the ideal loudness for your music, you'll need a tool that can measure the integrated LUFS of an entire song; bonus points if it does this by scanning an exported file. I use MAAT Digital's DROffline MkII, though other brands make tools that are equally functional. The integrated LUFS is the value we're looking for.
Next, choose one of your songs to focus on first: one that's upbeat, full, and energetic. Put a limiter on your track with a ceiling set at -1.0 dB true-peak, and lower the threshold until the limiter starts nipping at your song's peaks. Then measure the integrated LUFS value over the entire length of your track.
A decently loud, decently dynamic song should score about -14 integrated LUFS. If you want a little more volume at the cost of a little dynamic range, you can push it to -13 LUFS. If you want to go really loud, aim for -12 LUFS. But I don't recommend pushing your loudness any further than -12: you'd be throwing away the integrity and vitality of your music for outdated reasons. Volume normalization on the playback side is now the norm, marking the end of the loudness war.
If your song measures quieter than your target LUFS, lower the threshold in your limiter a bit, export, and measure again. Repeat until your song matches your target integrated LUFS value.
Once your chosen song is ringing in at the integrated LUFS value of your choice, put a limiter on the second song, and the third, and compare the loudness of each to the loudness of your chosen song by ear. At this point, the numbers matter less, and matching the qualitative loudness of the first song is more important. Again, clicking randomly from this part of this song to that part of that song will help you quickly get a feel for how loud the songs sound in relation to each other.
Pro Tip: if you don't have a plugin that measures integrated LUFS, upload your song to LoudnessPenalty.com -
a Tidal score of -1.0 dB means your song is probably close to -13 LUFS.
You've already been comparing your tracks to commercially engineered tracks in your genre. But now's the time for the other type of reference check: listening on other speakers.
Export each of your songs, and start listening to them on good speakers and bad speakers, good headphones and bad headphones. Anything that's not the speakers or headphones you made the music on. Your music won't sound amazing on every playback device, but you want it to sound decently good on all of them. And if you hear issues, you may want to make some tweaks in your mastering session and then check again. Usually, these tweaks come in the form of EQ changes.
Pro Tip: you may also want to listen to your reference tracks on the other speakers, to give you a sense
for how they sound when playing a great sounding mix.
The Final Export
When you've finished all of your changes based on the reference checks, the only thing left to do is export each song. Pick a final sample rate: usually 44.1 kHz for music listeners and 48 kHz for video work, and set your dither options to dither the sound down to 16-bits. Then export.
Pro Tip: if your DAW doesn't include dither in the export settings, your limiter probably has options for dithering.
And that's about it! If you followed the steps above, your songs are tonally balanced, have solid micro- and macro-dynamics, and are at a great loudness that allows them to sound good and loud on streaming services without paying for unused loudness with distortion. Your music is mastered and ready for distribution.
Your first master won't sound as good as your twentieth, and it's always better to have an outsider perform the master, listening with fresh ears on fresh speakers, to achieve the perspective you can't as the creator of the song. But if you've made it this far, you know how to get your music sounding pretty close all on your own.
Do you have any favorite plugins for mastering? I'd love to hear about them in the comments below.