The subject of mixing is a vast, complex topic with many differing philosophies and methods of approach. Before actually diving into any of these “hands-on” mixing techniques, tips, and tricks, I'd like to define mixing and break it down. Let's jump right in!
The following is an expert guest post by Jim Emerson. Jim holds the title “Executive Audio Engineer” at Reasonzone Audio. With a deep-rooted passion for his work, he has had the opportunity to help musicians around the world realize their creative potential.
What Exactly is Mixing?
It’s very important to understand what mixing actually is. At the very core, it can be boiled down to a simple concept: making your song sound good... However, this basic and simple explanation is easily confused with the practice of mastering - a very different topic covered in the next article in this series. To expand on this rudimentary definition of mixing, we can break the topic into five categories:
- Chapter 1: Organization of a mix session
- Chapter 2: Solo processing of individual channels in your mix
- Chapter 3: Creative processing of your mix elements
- Chapter 4: Group processing of similar aspects of your mixdown
- Chapter 5: Automation of mix elements
As a numbered list, the process of a mixdown from start to finish may not make much sense, but in the following article, I will break down and explain each category in detail to clear things up.
Chapter 1: Organization of a mix session
Easily the most overlooked aspect of a mixdown is the organization of your session. I know you as the reader are likely not here to be lectured about organization, nor do you need a lesson on it, so I’ll try to keep this brief. It’s tempting to jump right in and start mixing away, but often this can lead to chaotic projects that are hard to navigate. This state of disarray will likely stunt your workflow and cause silly mistakes - it’s very easy to lose track of your place in a session or accidentally delete/move channels and plugins. Because of this, it’s crucial to take the time at the beginning to set yourself up with an efficient workspace.
The most important part of session organization is the grouping of similar elements. This allows for much more control later on, and breaks the mixdown into manageable parts. My grouping varies from client to client, but I find that the typical mixdown will consist of five sections:
These groups should each have their own “bus”, or mixer channel that each group is sent to. This allows you to mute/solo the entire rhythm section at once, or completely change the volume of all of your vocals independently of the rest of the groups.
The rhythm section of my sessions consists of all rhythmic and percussive elements: typically a kick drum and snare, and several high-hat channels or cymbals, toms, snaps and claps, or clicks and impacts. The bass group is pretty straightforward, as well as the vocal group: all bass channels in the former, and all vocal channels, including backups and dubs, in the latter. “Body” is sometimes a bit more ambiguous, however. I often will have every single instrument (or synth channel in electronic sessions) grouped here, but included sometimes is a channel that isn’t necessarily “bass” residing in the lower register of the mix.
Finally, grouping all of the effects together allows for a lot of creative control later on. Most songs will sound alright without their effect elements, but these swooshes and sweeps and atmospheric drones and subtleties add a lot of “flavor” to a mix and can bring a song from “good” to “great”. Finding a grouping method that works for you is very important. My method illustrated here is tried and tested and works for me, and I’d recommend trying something similar as it streamlines the mixing process I have yet to explain in this article.
A logical follow-up after grouping your session is color-coding. Though not necessary, I’ve found that it drastically speeds things up and keeps things tidy. Hand-in-hand with color-coding is the arrangement of the channels in your sequencer. Keeping each group together not only in your mixer, but in your sequencer as well, speeds up the automation process later on and keeps it organized.
Chapter 2: Solo processing of individual channels in your mix
Your first steps after organization should focus on the processing of each channel independently. A common piece of advice thrown around is to always process your different song elements while the rest of the track is playing. The reason behind this suggestion is because it doesn’t matter what your kick drum or main vocal sounds like if it doesn’t sound right in context with the rest of the mix. While this is a good piece of advice to follow, I’ve found that some “solo processing” on each channel beforehand is appropriate.
As you’ll see, this section of the process isn’t strictly done with each channel soloed, but there is a lot of work that can be done on individual channels before actually mixing them together that will result in more control over your sound and ultimately a cleaner, more powerful mix. To break it down, “solo processing” consists of a few steps:
- Basic leveling
- Basic panning
- Basic compression
- Basic filtering
- Basic equalization
Once again, a numbered list is hardly apt to describe this part of the mixing process, so strap in as I go over each aspect in detail. Grab a drink.
The overall volume and balance of a mix is the most important thing to gain proficiency in. Perhaps “solo processing” is somewhat of a misnomer in this case, but basic leveling definitely fits into this section of the mixing process.
First and foremost, “fader definition” is a concept you have to grasp. To illustrate, consider the following scenario: the mixer channel fader is sitting at 0dB (90%) for your kick drum and 0dB (90%) for your snare, and the kick/snare balance is perfect. They’re sitting right where you want them and it sounds great. In this case, moving the fader for the snare up to only 91% or 92% will make a noticeable difference and may throw the balance off. Now consider a different scenario: the mixer channel fader is sitting at 0dB (90%) for your kick drum and 0dB (90%) for your snare, and the snare needs to be 6dB louder for the balance to be right. Pushing the snare’s fader up to 91% or 92% won’t really make a difference at all: your kick is still going to be far too loud. In the first scenario, the faders in question have a higher definition - you only need to tweak them a little bit to get the results you want. Apply this concept to every mixer channel fader in your project, and you can see how it could become an issue: low fader definition means that your levels will be much more difficult to get right, and you may not even be able to increase some faders enough to balance everything. You’ll hit 100% and run out of room. The solution is to get your levels close to where they need to be before your channel faders.
On every one of the channels in a mixdown, I will insert a gain utility before the channel fader and do a few passes through of the song, adjusting the volume of each channel to get the levels close to where they need to be. This is the most subjective and abstract part of a mixdown, but a decent way to approximate where the levels need to be is by following the spectral balance of “pink noise”
Pink noise rolls off at a rate of 3dB per octave, with the low end of the spectrum having the most volume/power. This balance is a great starting point when considering where your levels should be along the sonic spectrum, but should not solely be relied on. Some mixes need to be more bass-dominant, and some may utilize the mid-range more than lows or highs to convey what the artist was feeling when the song was composed. That’s why this aspect of the mixdown is the most subjective: it heavily depends on personal preference. However, using gain utilities to get channels’ levels close to where they should subjectively be before delving any further into the mix is a good call and will help you out in the long-run.
Volume and pan are the most important tools you have. These basic audio functions alone can result in a good, solid mix if you have the luxury of starting with high-quality source material. The simple act of moving a channel left or right can help breathe a lot of life into an otherwise dull mix. In my opinion, the most useful way to use panning is by using it to avoid “frequency masking”. Frequency masking is when the signal from one channel in your mix affects how one or more other channels are perceived by the listener. For example, if you have a synth lead and a vocal channel playing at the same time in a similar octave range, one can easily overpower the other and make the mix less defined and clear. Or, two synthesizers playing a melody and a supporting melody can clash and cause unwanted frequency spikes or harshness in your mix. The simple solution is to pan them away from one another a little bit. Moving one of the melodies to the left speaker a tad and the other to the right can clear up a lot of frequency masking issues without heavy-handed equalization or sacrificing the relative volume of each element.
The “basic panning” step in your mixing process should seek to open your mix up and “give it some breathing room”. By using this simple tool, you can balance out your stereo field and create a three-dimensional sound-stage. Suppose you have both a piano and a guitar playing at the same time. Edging one towards the left and the other towards the right can clear out the center of your sound-stage for important rhythmic elements and basslines or vocal tracks. Something I like to do during this step is to visualize my sound-stage as a cone emanating upwards. Higher frequency content gets panned out more, keeping balance on left and right. Important elements like your kick drum or vocal channel stay near the center, and every other element in your song is somewhere between. Always seek to create this space.
Bonus tip: I’m sure a lot of readers may already do this, but it can be fun and sound very good to pan your rhythmic elements out as if they were a real drumset on a stage (or in recorded material, it might already be a real drumset). Picture yourself as someone sitting in the audience, looking at the stage. Use this visualization to move the rhythmic elements where you would expect them to be if it were a live show.
Your snare would be slightly off to one side, your kick in the center, hats and cymbals further panned out with some in the center, and perhaps a few floor toms are off to the side and staggered. Psychoacoustically, the drumset being panned as if it were actually on a stage is more pleasing and natural to the listener, rather than having all of the rhythmic elements sitting dead-center and stacked up.
Once channel levels are approximately where they should be and the mix is panned out, I like to move on to the compression of individual channels. Compression is a function of volume, so one might ask why it comes after the initial balancing of channel levels. Truth be told, you could go through and compress channels that need it as a first step instead of what I’ve got laid out here. However, I think it’s important to get volume and panning taken care of first because the levels will be adjusted again later on anyway, and because getting the volume and pan positions set will paint a more clear picture of what the mix actually needs. During this stage, I feel that channel compression is a tool to tame your transients, level out large peaks, and accent vocal nuances or other performance nuances in recorded material. Later on, compression can be used as a creative tool.
Taming your transients is an important step to take before moving forward, especially in the rhythm group of the mix. The “transients” I keep referencing are quick, loud bursts of sound - the initial hits of your drums and cymbals in this case. These high-energy elements with sharp attacks can stack up and cause issues down the line for compressors, limiters, or plugins in general. Not every channel will need to be compressed, but I find that in the majority of my clients’ sessions, I end up throwing at least a few compressors into the rhythm group somewhere. Typically, these will have fast attack settings and a medium release, serving to catch the transients and bring them down in volume a decibel or two. Doing this, and appropriately setting your output gain afterword can really fill out the rhythm group and tighten up a performance. It may not only be your rhythm channels that benefit from some light, fast compression, however. Sharpening up the initial hit of a synth lead by employing a slower compressor attack and release can add some “punch” to the channel, helping to define the sound and allow it to cut through a dense mix more easily.
When it comes to leveling out large peaks, I find myself primarily talking about vocal tracks. Vocals are typically very dynamic, ranging from a whisper to a passionate yell. This extreme range, while imparting a lot of emotion into a song, can be a pain to mix properly. Compression can help by lowering this dynamic range while retaining the timbre of the loud and soft phrases. Commonly, I will use a medium to long attack with a long release and low ratio with “soft knee” enabled. Something similar to these settings usually allows transparent heavy gain reduction in vocals, causing them to better sit with the rest of the mix. This practice also ties in with what I meant by “accenting vocal nuances or other performance nuances in recorded material.” It can often be desirable to leave in the breath noises from a vocal channel or the fret noises in a bass guitar or 6-string channel. Transparent compression like this can bring these features up in volume and allow them to better sit in the mix.
As a bonus tip and tangent about accenting performance nuances, sometimes you need to heavily compress your vocals but don’t want the breath noises to increase. One might ask why you couldn’t just cut out the breath noise. Well, you can, but that’s often a very tedious process and can make the performance seem unnatural. In cases like this, I will often snip the breath noises out and move them to a separate channel with no compression. Thereafter, you can heavily compress the vocals as needed and the breath noises will remain the same volume.
This part of the process is relatively straightforward, though I would like to clear up a common misconception. It is often passed around that “you should highpass everything in your mix that isn’t kick or bass”. While in theory this isn’t a bad practice, people will typically throw around a number with it (e.g. “highpass everything at 150Hz that isn’t kick or bass”). Therein lies the issue. Every mix is different, and every channel needs its own attention. Do not blindly highpass anything at any specific value just because someone tells you to. This usually results in a flat, cold mix. A ton of the low end energy is missing and the mix sounds “sterile”.
With that being said, I do find myself highpassing almost everything at some value. For example, female vocals probably won’t have a lot of low to low-mid content, but there is often microphone rumble down there. Cut it out. The same goes for cymbals, hats, snaps, claps, etc. A lot of the low end content can be cut from those without actually altering the perceived sound. It doesn’t seem like much, but it adds up. Your mixes will be more clear and thank you for it.
As for lowpassing, this is something that is overlooked a lot. If you have several synth channels playing, and they all have a lot of wispy, airy high frequency content, your high-end can lose definition or become harsh. Furthermore, lowpassing guitars and basses up near the top of their frequency range can clear up a lot of room for vocals and effects. There really are no hard and fast rules when it comes to this section in my opinion - you have to use your ears and do what sounds good.
Finally, we get to the last stage of solo processing, basic EQ. You can pretty much categorize these changes as “removing bad aspects of individual channels” or “accenting good aspects of individual channels”. You may see or hear words being thrown around such as “boxy”, “tinny”, or “muddy” when someone is describing an undesirable tone, or “warm” when describing something that sounds good. The truth is, nobody knows what these mean. We can generally agree on something sounding “boxy” or “warm”, but really, these words are all very subjective. What sounds “muddy” to you may sound “warm” to me. All we can do as engineers and producers is take this step and use it to make our mixes sound good to ourselves.
I commonly find myself going through every channel independently and tuning the EQ curve on each. Even slight changes of only a single decibel can alter the way a sound is perceived and how it sits in the mix. I can’t give you any catch-all values to go by here, but I find that it’s helpful to make a reasonably sized boost (3-4dB) and sweep around the spectrum before doing anything else. If there are any problem frequencies or harsh tones, this will definitely illuminate them. If I find something that sounds bad when boosted, I will cut it by a couple dB and see if the channel sounds better. Sometimes it does, and sometimes the “harshness” was simply because I had an EQ boost in that spot. Go figure. The same process can be used to accent the good in your individual channels: a small boost swept around will bring to light the golden areas of a channel that really shine. As with filtering, there really are no solid rules to follow during this process, but it is an important aspect of a mixdown.
As an aside, I would like to address the notion that you should not boost with an EQ and that you should only cut. People will say it’s counterproductive to boost your signal to make it sound good because cutting removes the stuff you don’t want, saving you headroom. Truth be told, it usually does not matter. In most equalizers, a boost and then reduction in output gain is mathematically identical to a cut aimed at accenting the same area. You can try it yourself: get two identical equalizers on two identical drum samples. Increase the high shelf on one of them at frequency x by a certain dB amount and lower the output gain by that same amount. On the second equalizer, decrease the low shelf by the same amount at the same frequency. Inverting the polarity of one of these channels will cause them to “null”, meaning that the signals have cancelled one another out and are identical. Boost if you want to.
Chapter 3: Creative processing of your mix elements
After going through all of that (the bulk of a mixdown), your mix should be sounding pretty alright. Your channels are relatively balanced and spread out, giving you a nice full and wide canvas with which to work. Your channels are tamed and cleaned up with your compression and filtering, and you’ve hopefully made everything sound a bit better with minor EQ changes here and there. What’s next is to really dig in and enhance what you’ve got through the use of parallel channels and basic audio tools such as reverb and delay.
Parallel channels are one of the most useful tools at your disposal when it comes to creative processing. Creating a parallel channel is a CPU-friendly way to make a copy of an existing channel, which can then be processed. You can completely mangle the signal coming through a parallel channel with effects while maintaining the original signal’s integrity, leading to some interesting experimentation.
Something I like to do is use a source channel to create a huge atmosphere in the back of a track by employing several successive parallel channels. Simply taking a background guitar riff and creating several parallel channels with multiple differing types of effects can really fill out a track if mixed in properly - heavy reverbs and long delays that would otherwise muddy up your mix or ruin a particular channel can be carefully EQ’ed, filtered, or panned around in the background.
Bonus tip: Using parallel channels to create movement can add a nice subtle flair to your mix. I like to create a parallel channel of the main vocal and copy/paste my vocal channel effects over. Thereafter, slightly altering the original vocal settings (i.e. changing reverb and compression balance or EQ curve) and then crossfading between the two channels over time creates a slowly evolving vocal channel that is sure to keep the attention of your listener. Experiment with this method for other channels - don’t be shy to play around!
I like to think that any reverb you use can be separated into either “utilitarian reverb” or “creative reverb”. Utilitarian reverb can be used to put your song elements “into the same space” and add subtle depth to your mix. For example, sending each element in your mix to the same very quiet and small reverb can make a mix sound more natural, as if all of your song elements were recorded together in the same studio. This, by extension, should add subtle depth to your mix as well.
When using reverb creatively, the most important settings on your plugin will be the decay time and pre-delay amounts. Adjusting pre-delay plays a huge role in mix clarity - a longer pre-delay means that there is more time after the initial dry signal before the impact of the reverb. If adjusted skillfully, you can move your reverb tail out of the way of other elements without making your reverb sound unnatural.
It’s important to ensure that you “tame” your reverb before calling a mix finished. It’s often a good idea to highpass your reverb channels or otherwise EQ them to make sure they don’t clash. Just like any other channel, and even more so in some cases, reverb can cause frequency masking and needs to be mixed in to suit the rest of the channels. Furthermore, using multiple reverb units can become CPU-intensive, so it’s wise to set up a couple reverbs on send/return channels instead of using a new instance of the reverb plugin on your individual channels.
As with reverb, delay can be broken into “utilitarian delay” and “creative delay”. Utilitarian delay can be used to correct issues caused by plugin delay, and creative delay aids in the creation of a lively atmosphere and can be cleaner than reverb. Every digital audio workstation is not built the same. They’re all pretty similar, but some are better at accounting for the delay caused by your plugins. What this means is that some plugins will need more time on your processor to do their job (think sub-millisecond in most cases), and this can add up over several units. It’s typically pretty subtle, but you may want to manually correct this by shifting channels forward or backward if needed.
Creatively, delay is an often overlooked tool. Vocal channels in particular can sound very nice with delay in lieu of reverb. Throwing back to the subject of parallel channels, delay is also a powerful tool to create atmospheric effects. Entire channels can be sent to delay units; delay your delays, delay reverbs, delay your entire effect group. It’s easy to overdo it, but if done carefully and, as with anything, mixed carefully, it becomes simple to subtly fill your entire spectrum and stereo field with this effect.
Bonus tip: an interesting effect can be created with a delay unit and a distortion unit. Distortion is very sensitive to volume: a loud signal going into the distortion unit will be affected more heavily than a quieter counterpart. This means that if you send a delay effect into a distortion unit, the first delay repeats will be more heavily affected than the trailing ones. This can create some nice movement and break the monotony of a mix.
Chapter 4: Group processing of similar aspects of your mixdown
Nearing the end of a mix session at this point, group processing is a way to finely tune the balance and character of your track. As you will see, the grouping from the organization of your session plays an important role here. However, assuming you’ve done a good job during the earlier stages of the mix, this section should be a breeze.
Your group bus faders are critical. Here you can change the relative balance of your rhythm section in relation to the rest of the track, or make your vocals more prominent for a more pop-like mix. Especially after all of the creative processing, it may be necessary to spend a little time here not only balancing the relation of your groups, but altering your individual channel volumes to make everything sit where it should.
Again, assuming you’ve done a good job prior to this step, group compression should really only serve to “tighten” up your groups or add some character. I often find myself using an LA-2A compressor at this point, whether hardware or emulation. This compressor catches any stray peaks that have arisen since the beginning of the session and imparts a very warm tone to the mix in general. Group compression should typically be subtle and tasteful, serving to accent your mix and “beef it up” a bit.
After creative processing, even carefully filtered channels can start to see some unwanted frequencies building up in the low end. I find it helpful to go back through and check with a spectrum analyzer for any buildup and take care of it at this point. Vocals probably don’t need anything below 70-100Hz, so why not throw a highpass filter on your entire vocal group? You may not need it, but your vocal reverb may have introduced some mud down there and this would catch it. The same goes for your “effect” or “body” group. They likely don’t need the lowest frequency content and a group filter is a quick and easy way to address the situation.
Finally, your group equalization should serve as one of the last ways to fine-tune your mix. Wide, sweeping cuts and boosts of subtle amounts will help your groups sit together well and make the mix cohesive. An important step is the process of inverse equalization: suppose you find that a boost of 2dB at 2,000Hz on your vocal group sounds really good. You will likely now want to go to your “body” group and make a 2dB cut at this same spot. The figures don’t have to be exact, but this process will lock the groups into one another and give the mix a “finalized” sound. Rinse and repeat for all of your groups as needed, creating “pockets” for each group to sit in.
Chapter 5: Automation of mix elements
Automation, or telling your audio software how and when to change parameters over time, should be reserved for last. This step signals that you are essentially finished. Every mix won’t require that you automate your channels and plugins, but it definitely helps to relieve monotony or dullness.
Channel volume automation can be used to fade in new sections or shift the listeners focal point from element to element, and the volume of entire groups can be automated to ensure that nothing gets lost during climactic sections of the song. In the same vein, channel pan automation can be used creatively or functionally to create movement or fix frequency masking introduced by new sections respectively. For example, If you have a lead playing, and then a counter-melody comes in on another lead, try having the first one panned center, volume fade the second one in at 25% left pan and pan the first one 25% right simultaneously. These values are arbitrary, but should serve to illustrate the concept.
Channel effect automation can be used to alter the wet/dry balance of plugins as needed: automating your reverbs from dry to wet during swells or transitions is a great way to create tension and release. However, effect automation isn’t relegated to wet/dry levels. You can automate essentially any parameter on any plugin for very neat and novel effects. Use a healthy amount of experimentation during this stage.
Furthermore, don’t shy away from EQ automation. Suppose you have an instrument that comes in only during choruses. It's the focal point while playing, but it clashes with your pads a little bit in the upper mids. During the verses, the upper mids sound weak if you cut that part out of the pads. Solution? Automate an EQ on your pads to cut that region only during the chorus.
Bonus tip: “effect throws” are a compelling and dynamic way to accent certain parts of your mix. For example, you can fill the space between vocal phrases by automating the send value of a delay to increase at the end of vocal sections or on certain words. If done right, the delayed vocals will only come in when the vocalist isn’t singing, or the words you automated the delay up on will be repeated, but nothing else.
Automation is hugely important to a mix, but as you can see, it’s very easy to start automating things and lose track of what you’re trying to accomplish. I think experimentation is great and encourage it, but sometimes you have to go into this step with a goal in mind or you’ll get sidetracked.
Finalizing the mixdown
If you’re here, you’re basically saying “there’s nothing more I want to change about my mix. I like how it sounds and I’m ready to call it done.” Don’t celebrate yet - there’s still a few things to take care of before exporting your song and sending it to all of your Facebook friends.
Checking against a reference track
Find a song that’s similar to yours. Lower the volume of the reference track until it matches the volume of your mix and add a filter to your master channel. With either a highpass or a lowpass, sweep through the spectrum of both your mix and the reference mix. Listen to relative balance and the overall timbre of the reference track and make sure your mix is sonically similar. This is difficult to do objectively, but pays off in the long run. A tip I’ve found to make it easier to listen to your mix objectively is to transpose the entire mix up or down a semitone or two. Your audio software should make this as easy as a couple clicks. Just make sure you transpose it back to the original tuning after you’re done (or not, maybe it sounds better this way).
Set export settings
You’ll want to master your mix after you’re done (next article!), so there are a couple things to keep in mind when exporting. It’s important to master a lossless file, so you will want to export either an .aif file or a .wav file. Furthermore, it’s important to turn “dithering” off. This is a super quiet layer of noise that your software adds in during export, but the mastering process can bring it up in volume. We don’t want that. If you’re happy with the mix, and have made it this far, export it and prepare for mastering!
I know this is a lot of content and I’ve made a lot of suggestions and given a lot of advice. It’s crucial to keep in mind that every mix is different and you cannot apply the same settings or techniques to every mix. Take what’s written here as more of a mix outline and fill it with your own settings and methods. And most importantly, have fun mixing and check out the guide on mastering!