When trying to make a balanced mix, you can control the low end attenuating the bass’s fundamentals on the kick, or using bass ducking. For a balanced high-frequency range, try attenuating 6-10kHz with an MB compressor before amplifying the highs with an EQ/high shelf filter.
These tips are in no order, so apply them to your mixes as needed.
By gradual high-pass filter, I’m specifically referring to 6-12dB per octave sloped filters - I recommend these for 2 reasons. First, they sound much more natural, as if the instrument or vocal originally lacked these frequencies - second, 18dB per octave filters and higher cause 180-degree phase shifts.
Now, this may not be a problem, but if you have another very similar signal that occupies the frequency at which this phase issue occurs, it could result in a notch filter due to phase cancellation.
As you’d imagine the sharper the slope, the more aggressive and numerous these phase shifts become. Again, they may not cause issues, but when they do it may be hard to understand what exactly is sounding off.
Let’s take a listen to a demo designed to showcase the worst-case scenario of using sharp filters.
A big way a mix can become unbalanced is between the kick and bass - when these overlap in frequency, it’s hard to find a good relationship between the two. With an EQ inserts on the kick, side-chain the bass or synth, and observe where they overlap.
Then attenuate the bass’s fundamentals on the kick. Alternatively, if you know the notes of the bass, you can attenuate those on the kick.
Let’s listen to how this balances the kick and balance and overall helps the low end.
Bass ducking is pretty common since it helps balance the relationship between the kick and bass. That said, I’ll place the compressor on the bass, side-chain the kick, and then isolate the trigger to only the low frequencies, before achieving about 1-3dB of compression.
It’s an old technique, but can definitely come in handy and can be done with just about any compressor.
Let’s take a listen.
Like the low end, the high end of a mix is difficult to get right - you want clarity and brightness without an unwanted harsh sound. The best way to achieve this in my opinion is dynamically controlling 6-10kHz before amplifying the high range with an EQ.
To illustrate this idea clearly, I’m going to process a full mix with a high shelf filter amplified a few dB. Then I’ll enable a multi-band dynamics processor to attenuate 6-10kHz so you can hear the difference.
Although I’ll demonstrate this on a full mix, apply this concept to individual instruments, especially vocals.
Let’s quickly cover one of the biggest mistakes when trying to achieve a balanced mix, compressing the stereo output. I can understand the appeal - it’ll control dynamics, make the sound loud if auto-make-up gain is on, etc, but it ruins the musicality.
It does this by altering the totality of the mix’s dynamics, meaning if it’s not set correctly, or if it’s used aggressively, your mix’s balance will be completely off, at least from a dynamic perspective.
Let’s listen to what I’m talking about, and keep in mind you should avoid this.
Last chapter I brought up the idea of compression being musical, so let’s look at that a little more. Although some instruments are more related to groove and rhythm, all performances have some relation to it - meaning whenever compression is used, it will always interrupt or alter that rhythm.
We mainly control this with the compressor’s attack and release, to determine when and for how long compression occurs.
By dividing 60000/BPM, we know the time of a quarter note of our track in milliseconds. If we set our attack and release to this value, we’ve essentially told the compressor to wait 1 quarter note before compressing, and then compress for 1 quarter note.
By using multiples or divisions of this quarter note value, we could do something like waiting a 1/16th note, then compressing for a half note. It really depends on the signal, but try some variations to keep the natural rhythm of your track intact.
Let’s take a listen to this concept applied, and notice how it achieves dynamic balance while retaining the musicality and rhythm of the demo.
A huge part of balancing a mix is establishing levels with the faders, but if I can, I’d suggest not conceptualizing the process as ‘I need more kick’ or ‘I need more synth,’ etc. Instead, think of it in terms of the frequency range that the signal you’re affecting occupies.
So if I need more low mids, maybe this low register guitar would help balance the response - or if I’m lacking some highs I could bring up the drum overheads slightly.
If you think of each signal as occupying a space in the frequency spectrum, and you tailor the levels to filling that space, you’ll find it easier to have all instruments included in a way that sounds balanced overall.
Let’s take a listen to leveling and try to keep the frequency response in mind.
Ear fatigue occurs when a mix is too static - in other words, the frequencies are always the same, at the same amplitude. We keep this from happening by retaining dynamics, but introducing subtle frequency response randomization is another way this issue can be remedied.
I’ll use the shade EQ and introduce randomization filters, before attaching these to various EQ filters. Although the effect is subtle, you’ll notice that this variation helps keep things balanced.
2 tools I’d recommend checking out are the Gullfoss EQ and Soothe 2, both of which mainly affect the frequency response, but also dynamics. If you place Gullfoss on your various buses, it should help create room for each respective signal, especially if you side-chain one bus to another.
Soothe 2 on the other hand will dynamically attenuate resonances, which is especially helpful when trying to balance a vocal with surrounding instrumentation.
Let’s listen to Soothe 2 and Gullfoss equalizers being introduced to a mix.
Maybe I’m getting too literal here, but let’s cover balancing the peaks and troughs of a waveform, in this case, our full stereo mix. In RX I’ll use the phase tool and click suggest to measure if the phase has been offset - then I’ll process according to that suggestion.
As a result, the polarity of the waveform becomes balanced, resulting in a mix that sounds the exact same, but has its peaks more accurately represented when processed.