Let's Get Theoretical
Writing With Mixing In Mind

Let's Get Theoretical
Writing With Mixing In Mind

I like to think of writing a song like making a puzzle, and mixing a song like trying to actually put together the pieces of that puzzle. When I’m recording a composition I was not involved in creating from the start, it’s often a more challenging puzzle than if I’m making the pieces from scratch myself, and I can dictate how they fit together from the start.

In this post, which I received some requests for, I’ll walk through how to create those puzzle pieces yourself, using music theory as a guide. You’ll see that once you have the means to write parts that fit snuggly with each other, mixing and recording become much easier and sound completely natural!

Please note: This post will take a little bit of music theory to get the most out of. At some point, I would love to have a guest on here to give a quick theory primer; for now, I’ve linked to some great music theory introductions on the bottom of this post.

Building your puzzle pieces

There are plenty of little tricks you can use during your writing process to make recording and mixing a breeze, and most of them boil down to writing different instrumental parts to fill different roles.

It helps to think of things spatially on a piano. Say we are writing a song with a vocalist, a guitarist, a synth player, and a drummer. We’re going to make a shell of a song using just that information. Keep in mind that each of these instruments in this example can be substituted with any other instrument you normally use to write.

I always write new parts by putting them into one of two categories (usually with some light overlap):

  • Parts that use a new section of the keyboard
  • Parts that reinforce a section of the keyboard that is already in use

Vocals (or, any melodic instrument)

So with that in mind, let’s say I’m starting the song with a vocal melody, for example, that spans this range on the piano:

Do re...

You can see that my vocal part in this case ranges from B2 to E4, which is generally my most frequented area melodically as a baritone singer. This range works well for my voice - I’m not trying to push it to the limit and write a part that strains my voice and what it can naturally accomplish. It’s great to start with vocals for this reason, since vocal ranges are more variable than other instruments across different players (as opposed to piano, for instance, which can hit the same notes no matter who is playing it)

Vocals are also unique in the sense that every head of every vocalist is different, and thus every vocalist has notes which resonate acoustically in their mouth, throat, and nasal cavity better than others, and are thus easier and more natural to sing. This combination of range and resonances often leads to certain keys being preferred by vocalists (I lean towards F, Bb and Ab, personally). In this case, we’ll pretend we’ve tried or had our vocalist try different keys, and we’ve picked the key of E major for this song.

Guitar (or, any chordal instrument)

The easiest next step in writing the song for me, since I’m also a guitar player, would be to add a guitar part. Since the vocals are going to be the main focus of the song in this case, I’ll want the guitar to sit behind them, and reinforce them. I’ll also want to fill out the keyboard a little bit more, so I’ll write a part that fills out the high-mids a bit (knowing that I have synth to fill out the lows). Notice how I’m already using terms like “fill out the high-mids” during the writing process, rather than the mixing and recording process!

Here is the range of my guitar part in green, with my vocal range still in blue:

mi fa...

You’ll notice here, I do have some overlap (indicated by the red arrow). It’s difficult to achieve complete isolation between parts in practice. Here I’m trying to illustrate that this style of writing often has overlap; as long as it’s only a few fleeting notes, this kind of overlap is negligible - we’re concerned with the region about which most of the notes in a part are clustered. Of course, avoiding overlap entirely will almost always sound good in comparison!

Guitar players may have noticed this, but it’s worth pointing out: this range (B3 to E5) would consist of chords spanning only three or four strings. That’s right! More often than not, it sounds great to leave unused notes for others to take, especially when those notes are better suited for other instruments. I very regularly omit the bottom note of the chord from my guitar part so that the bass player can have that note for themselves. With any sort of mid-range instrument, I usually use a lot of chord inversions and root note omissions.

In contrast, for some styles, like singer-songwriter and folk, full open chords are often played on guitar, since there are usually no other instruments to fill that role, and acoustic guitar is more often than not a star player in those style recordings, and can take a more central role in the song.

Synth (or, any bass instrument)

Next, we need to fill in those bass notes! Enter: the bass synth.

so la...

I’ve noted the bass synth’s range in magenta.

You’ll notice this time that we have a gap between the lower vocal part and higher bass synth part, indicated by the red arrow. Don’t panic! When in doubt, leave more space between parts - better to have more room you can fill in later if need be (say, if the low-mids need beefing up when you listen to a rough mix of your song)

Now we’ve used all of our non-percussive instruments. But wait - there’s still a ton of empty space at the top of the piano! Do we need to fill that in?

Of course, we could fill that in with another instrument if we wanted. If I do this, I usually save it for the chorus sections of the song, to add some impact.

However, I often find that I leave this region empty. The reason that this isn’t as noticeable as, say, a missing bass region part, is that all of the instruments we’ve added have higher frequency harmonics. These harmonics will fill in our higher frequencies when we mix - I’ll even usually find that when I’m mixing, there is too much competition up there, and I’ll have to chop out some high frequency harmonics of, say, the bass and vocal parts, to leave room for the guitar harmonics.

Sometimes, I’ll even just add a non-pitched, higher frequency part, such as shaker, white noise, or cymbal crashes, to the chorus of songs. This adds some excitement and differentiates these sections from the rest of the song.

Drums (or, any percussion instrument)

Now, so far, our three instruments, vocals, guitar, and bass synth, have fallen into the “new section of the keyboard” category, to fill certain roles (from bullet 1 above: “Parts that use a new section of the keyboard”). Now that most of our keyboard is filled up, we can add percussion as a supporting player to beef up our song with a backbone (which fits bullet 2 from above: “Parts that reinforce a section of the keyboard that is already in use”)

Here is a depiction of our percussion instruments, in orange, layered underneath all our other parts:

ti do!

I’ve noted the percussion as orange blocks because every drum, whether you care to think about it or not, has a fundamental frequency (so long as it’s been tuned!). I try to tune my drums to the key of the song when I can, as this really locks the backbone of the song into place, and sounds much more cohesive. Here are some tips for tuning acoustic drums - if you’re using electronic drums with a pitch control built in, you really have no excuse to leave them untuned!

In this case, since our song is in E, I’ve tuned the kick drum (the lowest orange block) to a low E1, the fundamental of the track. Then, I’ve tuned two toms (the next two blocks, moving right) to a B1 and an E2, since B is the fifth in the key of E. My snare (the fourth block) is also a B, B2. Roots and fifths are great for drums, since you can use both for minor and major keys, and they are part of the arpeggio of the key. Remember, though: this is just an example - you can place your drums anywhere you’d like!

I’ve also included an arbitrarily placed fifth and sixth orange block to denote a high hat and a crash cymbal. These may not sound pitched exactly, but they are up there to indicate that these will both be adding higher frequency content, and supporting some of the harmonics from our other instruments.

This drum part is overlapping almost all of our other instrument regions! But that’s okay, because as most drum parts are, this will be a supporting part, rhythmically accenting different notes of the vocal melody, guitar chords, or bass synth, to glue the different regions together with a rhythmic backbone.

Pick up the pieces

"Sorry Grandma, rain check on the puzzles... I have to write music"

Now we have a song that will practically mix itself! We’ve created our own puzzle pieces by partitioning our keyboard to different players, and it’s fairly straightforward to see how we’ll approach recording and mixing this song. Here are some examples of things I could do now (key word could!):

  • Record vocals with a dynamic microphone, since we want to emphasize the low-mids rather than compete with the guitar by using a condenser microphone
  • Record the guitar with our amp EQ’d with low amounts of bass, and perhaps a little more than average amount of treble
  • Add some high-passed, bright reverb to the track to fill in the high end of the song

Etc, etc, etc. Once you’ve written a song this way, mixing and recording become easier, because you have a pre-defined frequency region for each instrument to use as a jumping off point. You can also easily keep writing and recording new parts if the song needs more beefing up, by adding more supporting players, such as some subtle piano chords, guitar doubling, or harmony vocals where appropriate.

This is a rather extreme example of partitioning the keyboard while writing; there’s nothing to stop you from writing three bass parts in a song, for example, but it’s certainly rare! Even in these extreme examples, however, you find that the two overlapping instruments work together and don’t fight each other.

Remember, as with all things musical, these are just helpful guidelines. Try writing some new songs like this, and then break some rules and see how it sounds!

Great music theory references

Here are some great music theory references:

Great primer from Sound on Sound that pairs well with this post

Free online class from Berklee (can’t beat that!)

A more thorough and pedagogical guide

I’d love to hear your experiences writing with this technique in the comments. Next, I’m planning a post about how to use your music theory knowledge until it becomes second nature, so stay tuned!

-Keenan Hye
Founder, engineer, and voice behind most of these blog posts

Get Social

contact - about - blog - faq