Updated: Mar 17, 2022
You've made your first mix.
It was exciting, interesting, your learnt a lot and you're very happy.
Perhaps you even have published it on Spotify.
So after working so much on it, you're bit exhausted and just move to something else.
A week or two pass before you go and listen to the song again and... argh! It doesn't sound great! What you thought was at least an ok mix, maybe not stellar but listenable, turns out to be something you're embarrassed by, not proud of.
What happened? How could it sound so good just two weeks ago, and so bad now?
First of all, find some consolation in the fact that it happens to more or less everyone. Part of the skillset for making audio productions is to understand how to judge audio productions... so it's not surprise that, at start, most people will misjudge their own mixes.
However: where does that leave you? How do you make sure it does not happen again? How do you get better at making mixes, and get better at judging mixes - especially your own?
While these are subjects that require way more than a blog post, it's perhaps possible to give some pointers here.
The key is to learn first what makes a production good, and then to learn how to get to that "what" starting with the material you have (and even decide if it's possible at all).
Critical listening of good productions
First - what is a good production? It could be good moment to go and read my first post, about what makes a good recording. And then listen critically to some of these good productions. Better, listen to a lot of them.
How do you find them? Basically, take any hit song from the 70s, 80s and 90s and you will almost infallibly find one.
That's because these were the years were hit productions were invariably made by people really dedicated to their own field - recording, mixing, mastering and so on. These people knew what they were doing, and they were doing it day in day out, in big old studios with great recording spaces. And of course performers were really good at singing or playing, because the business was just as competitive as today, but they had none of the modern aids to support them. Technology didn't allow for anything else than full commitment. And there was money enough in the business that record companies could support that commitment.
Note that I'm not saying that there aren't more recent productions which are good: of course there are.
But in recent years, with the advent of home studios and "everyone can publish music" thing, there's a lot of published stuff that is... questionable. After all, you're reading this post and your mix is published on YouTube or Spotify, right?
But of course if you know for a fact that a production is great, feel free to pick something more recent.
Once you have a good production in your hands, maybe one you are very familiar with, you need to listen to it in a critical way.
In rough terms, this means that instead of focusing only on the result ("wow! I love this!") you explicitly look at the relationship between the individual parts. First and foremost the level relationships: before anything else, a mix is balance. A set of faders in relative positions, possibly changing a bit over the duration of the song.
Yes, I'm well aware that critical listening may mean much more, such as listening with calibrated levels into a well known and good sounding listening space thru a consistent and good playback chain.. but for our purposes, there's no point in going that far. We aren't making equipment reviews and comparisons.
Don't worry too much about the format - high quality, low quality.. you're not listening to the sounds and timbres, you're trying to listen to the arrangement and the balance - the level relationship between the various parts.
So take that metal epic that you love so much: what is the level of the guitars relative to the drums? Louder? Quieter? The same? Does it change over the course of the song? What's the level of vocals with respect to the guitars? How much do you hear the bass?
Or take that electronic/hip hop track - is the beat always the same, and does it play when the vocals are going? Does it change level? Does the bass stands out (in the sense of appearing louder), or the kick?
And so on.
If you've never done it before, you will often be very surprised of what you discover.
You will find that these thundering guitars, which you thought were super loud (and you would have mixed super loud) really aren't... they are actually pretty low, but they "pop" at all the right places because everything else gets out of the way. So they really feel loud to the listener, without really being it.
You will find that heavy guitars aren't really as distorted as you may think as a listener - too much distortion is buzzy and fizzy, as you find out every time you listen to a guitarist with too little skill and too many pedals.
You will find that certain productions are "drums-driven", other are "bass-driven" - in the sense that the main rhythmical pulse and drive is given by either instrument.
You will find that the hip hop production is very sparse - when the kick hits, there's nothing much else going on, so it feels huge even if, in level, is not that loud. And the pads behind the vocals are, well, very quiet, because it takes just a hint of harmonic content for our brains to lock on as listeners and provide a reference for a rapped or spoken vocal line.
And so on and so on. It will be an eye opener.
Understand the effect of relative balance
Once you grasp that most of a good mix is just the relative balance between the parts, you want to focus on the specifics: what effect the relative loudness of each element produces, on the total result? Use your chosen song... what if the drums were quieter? What if there were no guitars? Do the relative levels change over the course of the song?
Keep also in mind that different good productions are, well, different.
Tracks in the same genre will share some traits.For example, in rock there will be at least a part where the guitars almost overwhelm the vocals, but not quite, and the drums tend to be loud as they provide energy and excitement. But even in the same genre there may be different ways of mixing a song. That is to say, there many different faders settings in each section/part that work, but will produce a different final feeling, and it's up to you (and the artist/producer/label) to choose the flavor you want.
So if you analyze different productions in the same genres, you will find similarities (in terms of balance) but also differences. And that's fine, and it's the reason why certain songs have certain vibes and others have some other.
Do that kind of analysis long enough, and you will build a vocabulary that will allow you to set the balance more or less in the right position right away, and not make the beginner's mistake of mixing something very loud because as a listener you perceive it loud.
Arrangement is king
You may have noticed that so far I haven't mentioned effects, sounds, and timbres. That's because, in the order of things that make a good mix, they are far less important than the arrangement and the balance.
An arrangement is simply about deciding when each instrument/part which makes up your song is playing and hearable; and how important it is to define the song. Moving instruments and parts in and out over the course of the track and changing their relative level, is what makes the track.
The only exception is usually the vocal part, because when there's a vocal, it usually needs to be up and front all the time - since that's what people generally focuses on.
Effects are not as important as you think
So sure, a snare or a kick sound must fit overall, but what really makes it stand out or not is not its timbre, equalization or compression - it's how it is arranged, and therefore its balance relative to other parts in any given moment - how much space there is for it to be heard.
People obsess and spend hours (days!) shaping a snare sound, learning about compression, equalization and whatnot, while they dedicate no thought at all to the overall arrangement - which is what really make things happen in your track.
It's not that compression, equalization, saturation, excitement and merry company aren't important or useful. They are, but they are tools to achieve an effect, reinforcing the overall structure which is the arrangement.
Old school and new school
Now, in old times the arrangement was often set well before mixing took place.
Either there was a composer, or the band rehearsed the songs over and over, naturally finding a good arrangement ("don't play here! Go quiet there! Increase your level for the solo!") then they hit the recording studio, set up with a bunch of mix, and that was it. Your track were already recorded with the desired arrangement "baked in".
Nowadays, however, it's very common to build songs piecemeal. And with electronic instruments and DAWs it's incredibly easy to move things about and around. This means that arrangement decisions are often made later. Especially when it's a one-man shop - and you are at the same time the composer, the performer, the arranger, the recording engineer and of course the mixing engineer!
So these days arrangement often doesn't happen as part of composition or emerging from rehearsal, but as an integral part of an ongoing mixing.
You have two choices here:
either you mimic the old school way of doing things, figuring out the arrangement first, then recording/creating the part and then going about mixing the lot.
Or - if you find that boring - you need to be very disciplined in focusing on the arrangement much more than you do on individual sounds. Possibly do different sessions for these different tasks and keep them separate in your head.
A couple of good tricks here are to first make a mix with no effects (no compression, no eq, nada) just to determine what takes the spotlight at any given time. And to realize that you can't have more than three or four things taking the spotlight at the same time.
When vocals are there, they must be one of them. The rhythm section (drums or bass) is generally another so that usually leaves you with one or max two other elements to be present. All the rest has to go - either quieted down to provide a little background, or away entirely.
The mute button is your best friend.
Practice, practice practice
Once you've learned how to analyze and dissect a good production, it's all about practicing. Try to use balances and arrangement tricks in your own productions that you've picked up along the way. Don't be afraid to be unorginal. You cannot really "copy" an arrangement, but you can use similar ideas in your productions. More cowbells is fine, but the key is where.
A key element here is to keep it simple. Start with small things - mainly muting stuff which doesn't need to be there when it doesn't need to be there, and making sure that what is there makes sense. Meaning that only one or few parts are under a clear spotlight. Get parts in an out. If you need, use automation to reproduce the level changes you would have if good musicians were playing together - building a dynamic balance, which means a balance that change over the course of the song. Remember, if your faders aren't dancing during the song, then the playing must!
If you decide to insert an effect, be very clear with yourself on what you want to achieve and how adding that effect contributes to achieve it. If you cannot explain it in minute detail, don't add it. There is no problem at all in experimenting, and there's space for serendipity and surprises... but mostly there is space for knowing why you are doing something, what you want to achieve and why that specific effect helps you achieving that.
Also remember that the great productions you are getting inspired by were made when none of the technical facilities we have now existed. Recording time was very, very expensive and effects were very limited, so you had to get a sound right and mostly work with that, with limited options to change it.
If you find yourself obsessing for hours on a sound because you feel unsatisfied, it's almost certain that it is not about the sound, but about your arrangement choices (or lack thereof).
Remember also that practice is not doing a lot of random things, but trying out the same things in a sistematic and deliberate manner. If you're just mixing lots of songs without particularly paying attention to what you're doing, you're not practicing, and you will likely produce a lot of bad mixes anyways.
Keep things simple, do stuff because you're trying to achieve something specific, then check if you have achieved it. Don't be afraid of rolling back or restarting from scratch. It's all part of the practice.
Know (your ears') limits
One of the main reasons for bad mixes is ear fatigue, and one of the biggest skills in mixing is to know where to give up - at least for a while.
Our hearing tends to get used to high frequency. Listen to something enough times, and you will start to find it dull. You will be tempted to reach for that EQ knob and add high frequency. You will go "wow! This sounds great"... until you let it stay and you hear it with fresh ears a couple days after, where it will be a fizzy mess.
Pause often. Pause long enough. Listen to completley unrelated stuff when you're pausing, in order to reset your ears.
Also, keep your listening level down, when you're mixing. You should be able to have a conversation over the running mix without having to shout or raise your voice. This will both allow you to keep your hearing perspective longer, and produce mixes that sound great at low volume and sound awesome when played loud.
How do I know if I am improving?
First of all, it will be a matter of confidence and consistency of results. A big factor is speed.
The speed with which you get from zero to a decent mix, which translates reasonably well between your main speakers, headphones, the phone and a PA system. The speed witch which, when you hear something, you can go in the mix and immediately fix it.
The speed with which you get to that point where further tweaking gives only (slightly) different mixes, not better mixes. And of course then amount of times when you leave a mix in the evening, and you still like it the morning after.
Internet and YouTube - here be dragons
Well, of course, the elephant in the box. There are millions web sites and YouTube channels promising to teach you whatever you need to learn. Heck, this blog is one of them.
Some are even good.
Be wary of a couple things when exploring all that material: everybody with a few dollars can put up something on Spotify, without really any regard to the overall quality. Even more, everybody can put up something YouTube, without even the few dollars. Much, if not all, of YouTube material is made with the idea of capturing an audience, and material needs to be produced every week of every month to keep and grow that audience.
The result is that while most of the material is doing with enthusiastic intentions, there's a lot of questionable value, where the excitement of half-understood ideas (or the urgency to publish some content) prompts the author to put stuff out before it's ready or well thought. Not too different from the people who know that a song or a mix is not yet ready for prime time but they have to put it out somewhere... or even worse, when they think it is ready for prime time, but it is not. Also, often the focus is on something perhaps interesting and eye-catching, but absolutely trivial.
In too many YouTube channels, a big deal is made of stuff that.. well, it's not. The result is a waste of yout time and consumes your energy for little gain. You risk to spend a lot of time focusing on stuff that may be vaguely useful, but adds quite little to the Really Important Things - which are the quality of the material, the quality of the performance, the quality of the recordings and the quality of the arrangement.
How do you distinguish the chuff from the wheat? Well, it's relatively easy:
anything that focuses on gear is usually fun, but not very useful. It's _really_ not about the gear.
anything that focuses on sounds and timbres - it's not that they are not important, but for natural sounds, you will know how it must sound; for non-natural sounds (from electric guitars to synts) you can just do as you please. There's no rules.
anything where "magic" appears - that is to say, you don't get a clear idea on how and why a particular effect works. It can be fun, but it will leave you confused and unable to reapply it, and it will be a waste of time
anything that has to do with compressors. Just kidding, but there's such an obsession with compression in audio engineering channels/sites that it seems like they're the only thing that matters. Fact is, you can make a fantastic mix without using even a single compressor. Just take an awesome singer, an awesome song and record him or her doing his or her thing.
anything that asks you to listen to sound examples.. on YouTube. YouTube compresses sound so that it's fast to transmit on busy data networks. It's as a far from a proper listening chain as it's possible to get. It does not take in account at all the listening room, location and chain, which have a profound effect on timbres and sounds. In the wrong room, you can listen to a Fazioli or a U47 and it'll sound horrible.
anything with bombastic statements about this or that. The reality of mixing is that there's an enormous palette of acceptable options - balances, arrangements, sounds, lyrical content, song lengths and so on.
anything which presents an "infallible" recipe for success. Again, the reality is that success (popularity, commercial success, numbers of likes, the lot) depends on a lot of factors which are both out of anyone's control, and ultimately random. No technique is infallible (or, flipping it, always wrong), even when it comes down to a good mix.
A final word
There are, of course, a gazillion more things to say than I wrote here and I haven't even scratched the surface!
It's worth remarking once more that mixing is just the last step of a chain. You can't make a good mix if the material, the performance of the recordings are bad. If you are in that condition, and you can - consider rewriting, rehearsing and re-recording before attempting a mix.
Also, like for any complex skill, one never stops learning. The beauty of audio engineering is that you can always improve, no matter how skilled and experienced you are. The key is always the same: evaluate your results with a dispassionate eye, and compare them mercilessly with the best you know. If there are differences, find what they are, and - one a a time - figure out what they are due to and how to control them.
In matter of time - practice takes time so you can't expect quick results. People and intensity differ but consider that some three to five years at it is a reasonable amount of time to become good at mixing.
Take it easy, and have fun. In the meantime, happy mixing!