I tend to work on music while wearing headphones, because the kids are asleep and don’t seem to appreciate when I pump up the jams at 1 in the morning. I finish the song, and it sounds excellent in my headphones — every note is crystal clear, the bass is powerful, the vocals are shimmery. Every sound is where it should be. However, when I load it onto my phone and play it in the car — it sounds terrible. What happened!?!
Mastering is hard. Or, at least, I’ve found it to be so across the last few decades, when every few years I’m ready to release something. Turns out, not everyone listens to music in a perfect acoustically-balanced room, or through thousand-dollar headphones (including me). This means that our music is subject to the imperfections of their audio environment. How dare they! Mastering solves this, but is hard. Below, I’ve collected some mastering tricks that I use to improve how my music sounds.
The basic stuff
Quick searching online will yield some of the following mixing/mastering tips. Definitely abide by these, as they will go a long way:
- Make sure your different tracks don’t collide in the EQ. Different instruments have different key frequencies, so make sure there is not too much overlap. Bass guitar is 100–500 Hz (and, really, much wider in a lot of cases), vocals are 1–3 kHz, piano is somewhere around vocals, etc. Go through your tracks and get a sense for where the key frequencies are — use a band-pass or parametric filter to see which freqs make it sound good. For any ranges that don’t matter for an instrument, kill them! Otherwise, the transient sounds in the irrelevant ranges will add to your overall amplitudes, without actually contributing to the sound.
- Compress each track individually. Apply a light compression (maybe 2:1 or 4:1) to your tracks. Particularly drums — which hit hard and can redline your track overall — and vocals, which can fluctuate based on your singers closeness to the mic or their inflection. Don’t apply so much compression that you hear buffeting (pulsing in and out); just apply enough to lower the highest peaks the track hits. You should be able to do so without changing the sound of the track.
- Add loudness to your track while applying overall compression. But only compress slightly! Otherwise you’ll get bad buffeting when you hit your peaks. You can clip your waveform(increase the volume past redlining) or use a limiter, but again — don’t do this too much.
In addition to these important actions, here are some other things I do to try to optimize my songs.
Compare to the pros
This is critical. While mastering, always have some professionally-mastered music to compare to. The more popular the better, since they have large budgets to pay expensive mastering engineers, who try to optimize the volume/clarity for maximal radio plays. That is what you should be targeting, even if you are worlds away from pop music. These are the pros, so follow their lead.
Find something similar in feel, with similar instrumentation and dynamics. If your song is heavy electronic with lots of bass/drums, find the latest IDM bass droppery and work with that. If you do piano or acoustic-driven work, go with some modern well-produced pseudo-folk artist. Find songs that have similar feels to what you are going for, and have those queued up so that you can go back and forth.
Using a pro song will also help offset for any imperfections in your setup. If your speakers err too treble, for example, then the pro music will sound that way as well, and your ears will adjust. Tune your system’s acoustics so that the pro music sounds good to you, and then make your music sound like that.
Muddy music ⇒ Kill 250Hz — 500Hz
Occasionally, one of my songs will drive me crazy during mastering — I just can’t get it to sound right. It will sound underwater or cloudy or throaty or chunky in a bad way. Often this happens on songs with heavy bass, as well as other lower instruments, or when my instruments cover a wide EQ range due to distortion or other effects. I love distorted bass, because it is freaking awesome, but this can wreak havoc when it comes time to master. I’ll go round and round, trying to isolate which frequencies are killing the overall mix.
The answer? 250–500 Hz. Always. Every single time. I’ve found that you can drop this range pretty substantially (maybe -8dB or so) and it has very little affect on the overall sound, other than to clean it up amazingly. This seems to be because this range is the north end of the bass drum sound (when you really want your bass drum low), and in the middle of the bass guitar range. But the good deep bass guitar is south of 250Hz, and the nice tonal quality of the bass is up in 500–1000 Hz. So that middle 250–500Hz can be reduced without negatively affecting the sound. I often remove that range from both the bass guitar track, as well as the overall mix.
Vox that hurt your head ⇒ Use a notch filter to locate
Since most listeners ears will focus on the vocal track, it is important to get this one right. When your vocalist hits their loud notes, make sure it is not sounding too screechy. Often it will, particularly when colliding with other instruments. Use a notch filter or a narrow parametric (Q=15, +15dB) and slowly go up and down the 1–3.5kHz range, and listen for when the painful screeching matches what you are hearing in the track. Once you’ve found it, stay at that frequency but turn the gain to a -5dB or so, to cancel out the badness.
This technique can really be applied to any sound that bugs you — find it and kill it.
Careful with your snare!
Snare drums are awesome, but are hard to remove from the mix once you get to the mastering stage. A good snare hits across a very wide dynamic range (say, 500Hz-4kHz), so if you try to remove it in mastering you end up killing all your other instruments. So, lower the snare in the mix.
Revise on your worst system
Once you think you have a cut that sounds good, find your worst system — maybe your car, your bluetooth speaker, or whatever. Play the reference professional song on that system, then play your song, and then go back and forth to try to figure out why yours doesn’t sound as good. If possible, use your laptop to play it, so you can be adjusting the EQ on this bad system. Keep track of what changes you want to make, and then try them out on your good system and through headphones. Don’t overcorrect for the bad system, but do take its deficiencies into account. You want people to like your music even if they are on a bad system!
Since bad speakers aren’t good at either powerful bass or clear treble frequencies, these bad speakers will over-accentuate the 500Hz-2kHz range, making your music sound nasal and hollow. This is a good reason to reduce that range, which will make your song sound more open and wide on all systems.
Play it loud; play it soft
It is critical to play your music loud — professional music can be turned up loud, and it still sounds good; nothing stands out disproportionately as painful or ugly. So you need to turn up the volume on your song to see what stands out and needs fixing.
You also need to play it quietly. Ugly things happen to music when played in suboptimal listening environments (moving cars, rooms that aren’t coated in acoustic foam, etc), and when it is quiet those ugly things increase in effect. Also, since the human ear is tuned to hear other human voices, when music is quieter, the vocals seem to stand out more than when music is loud. This is okay; the same happens with pro music. But, you want to make sure that the non-vocal part of your music is represented well.
Do all of the above about twenty times
Iterate over and over on your song. As you get something you think sounds good, listen to it in many scenarios — in your car, on your home stereo, through headphones. Listen in the morning when your ears are fresh, in the evening when you are exhausted. Create a playlist interspersing your music with your reference songs. Have some other folks listen and see what they don’t like.
Take notes as you listen, about what doesn’t sound perfect. Then when you have time, go try to fix those problems. Wash, rinse, repeat. Keep doing this until you are fatigued and finally ready to call it done.
Then, do it a few more times. As you get towards the end, you will find that you are pushing the EQ up or down 0.5 or 1 dB, versus the +/-5dB at the beginning. That means you are honing in on the sound you seek! Keep going until you can convince yourself that there are no more improvements to make.
These are some of the things that have worked for me — would love to hear what works for you!
Visit veloureo.com to explore some of my work.
Click on the plus or the heart or the thumbs-up or the clapping hands to send me good vibes.
Screenshots (other than the Ozone one) taken from Reason by Propellerhead, pretty much the best durn music software out there.