Professional Analog Mastering.
Professional Analog Mastering.
226K+ subscribers
When we mix, one of the first things we do is set relative levels - we’re changing the balance between various instruments by adjusting the amplitude with channel faders.
An EQ does the same thing but in a much more granular way.
For example, say the kick’s fundamental is 60Hz. It has overtones at 180Hz, 360Hz, and some additional transients around 1kHz.
The bass fundamental is at 110Hz. It has overtones at 220Hz, 440Hz, and the percussive pluck is around 2kHz.
Meanwhile snare starts at 140Hz, has an overtone at 420Hz, one at 840Hz, and some high end around 3kHz.
These are simplifications, but you get the idea.
Each instrument is fitting into the frequency response like a puzzle - as we amplify or attenuate frequencies, it’s the same as adjusting a channel fader.
But instead of affecting the full instrument, we can adjust a fundamental, and overtone, and percussive element of the instrument and so on.
The Q value is like have multiple faders, each creating incremental change around the fundamental, overtone, etc that we want to affect.
So, when you use an EQ on an individual track, think of it as an extension of your channel fader, but instead of amplifying or attenuating everything at once, you have multiple sub-channel faders with which you can make amplitude changes.
It’s not just something to alter frequencies; it’s a complex mixer.
A 6dB change in amplitude doubles the loudness when amplifying and cuts the loudness in half when attenuating.
So when we equalize say 2kHz, and amplify it by 3dB, we’re increasing its loudness by 50%. If the Q value causes 1kHz and 3kHz to be amplified by 1.5dB, we’ve increased the loudness of those ranges by 25%.
Similarly, if we attenuate a resonance by 6dB with a notice filter, we’ve cut its loudness in half.
This gets more complex when we consider how we perceive music and how loudness variations affect the overall track’s loudness.
Our ears have a resonance at 3.5kHz, causing us to hear that range more easily. Additionally, the vocal’s 3rd formant is right around that frequency - so adjusting 3.5kHz greatly impacts perceived loudness.
Say on the vocal, I boost 3.5kHz by 3dB. That range is increased by about 50% in terms of its relative loudness.
As for the overall track, it depends on the instrumentation present - but this boost will likely have a big impact on our perceived loudness of the track if the vocal is a main instrument in the mix.
Similarly, if I increase 500Hz on the bass by 3dB, it’ll increase that range’s relative loudness by 50%, but for the overall mix, it’ll have a smaller impact on the loudness than our 3.5kHz filter on the vocal.
To show this, let’s listen to these 2 examples; I’ll boost 3.5kHz on the vocal, then bypass that filter as I enable a bell at 500Hz on the bass - again, both amplify by 3dB.
Notice how the perceived loudness of the overall mix increases more when the 3.5kHz filter is enabled than when the 500Hz filter is enabled.
I’ll also use LUFS meters on both to show how the 3.5kHz filter increases the measurement more than the 500Hz filter.
Watch the video to learn more >
This concept is a little harder to understand, or maybe just for me to explain, but some visuals should help.
Let’s say we just have a bass guitar recorded for now - we have the fundamental and overtones, like we discussed earlier. The fundamental peaks at -3dB, the 2 overtones peak at -8dB, and the percussive scrap peaks at -12dB.
Let’s say quietest part of the performance is -53dB - so right now the dynamic range is 50dB from this quietest part to the fundamental’s peak.
I’ll use a bell filter on the fundamental and boost it by 2dB. Now that peak is hitting almost -1dB - meanwhile the quietest part of the affected range is also amplified, but, only in the range that has been amplified. Other parts of the signal aren’t affected - so let’s say it’s level is now -52.5dB.
This bell filter has increased the bass’s overall dynamic range by about 1.5 dB.
Now, imagine this on a much larger scale - in which we’re using multiple filters on multiple instruments.
It goes without saying that the an EQ’s effect on the dynamic range is incredibly complex and plays a much bigger role in a mix than one would think.
For example, what if the fundamental and an overtone at 220Hz collectively cause the aforementioned -3dB peak? What would happen if I boost the fundamental by 3dB but dip the overtone by 3dB? The new peak level would be contingent upon the relative amplitude of each frequency.
If the overtone plays a bigger role in that peak than the fundamental, than the 3dB dip at 220Hz should cause the overall peak to be lower after the filters are introduced.
If the fundamental plays a bigger role in causing that peak, than the 3dB boost at 110Hz should increase the peak’s level.
There’s a lot of interconnectivity here.
To show this, let’s equalize an instrument. I’ll introduce a bell filter on a high-amplitude range and notice how the peak level increases. Then, I’ll introduce some filters to other ranges and attenuate. Notice how the peak level drops in a way that isn’t equivalent to the overall amount boosted or cut.
Watch the video to learn more >
The effect is minimal, but an EQ introduces harmonic distortion - even if it’s not supposed to.
Small variations between how the EQ measures the signal’s amplitude and the actual amplitude cause harmonic distortion.
But EQ affects distortion in a much more significant way.
For example, say I distort a 100Hz. sine wave. Depending on the wave shaping amount and symmetry, I’ll achieve various harmonics at varying amplitudes.
Typically, harmonic distortion decreases in amplitude the higher the order or frequency.
But, I could equalize one or more of these harmonics and alter the distortion’s amplitude.
Let’s say I have even and odd-ordered harmonics, and I boost everything above 200Hz with a shelf.
The THD is increased relative to the original signal.
If this EQ is after the saturator, I could even use a HP filter to get rid of the original 100Hz. sine wave. In this instance, I’ve isolated the harmonic distortion, significantly changing the relationship between clean and distorted.
What I’m about to show in the demo might be too inconvenient to do in a normal mixing situation, but hopefully, it’s interesting nonetheless.
I’ll send a bass guitar to a parallel or auxiliary track. With a linear phase LP filter, I’ll isolate the bass to just the fundamental notes of the progression.
Then, I’ll insert a saturation plugin and distort the isolated range to ensure the harmonics only generate from the fundamentals.
Then, I’ll use one more linear phase EQ to amplify or attenuate the harmonics the saturator generated.
With this set up, I can use EQ to control the harmonic distortion with a lot more precision than if I just used a distortion plugin by itself.
That said, keep in mind that EQ will control distortion even without this setup, so long as distortion has been introduced before the EQ is inserted.
Watch the video to learn more >
Admittedly, these effects aren’t super common, but, in my opinion, they’re really interesting and worth talking about.
Amplitude increases of specific frequencies can cause 2 psychoacoustic effects. First, it can cause the listener to perceive harmonics that aren’t in the original signal.
For example, if I use an EQ to significantly boost 2kHz, harmonics at 4kHz, 8kHz, and so on will be perceived, even though they’re not occurring in the original signal, the processed signal, or even the signal coming out the speakers - it’s entirely due to perception and not the physical sound.
The same could be said for the pitch. If I use an EQ to boost 100Hz significantly, say by 30dB, the pitch will be perceived as lower by almost 5%! This is nearly a semitone or half-step. In other words, the 100Hz sine wave would sound like a 95Hz—sine wave.
Something similar happens to the highs - if I boost these significantly, we perceive them as higher in pitch, although I couldn’t find exact numbers for how much is needed to cause x amount of perceived change.
So, controlling the amplitude of specific frequencies can drastically change the timbre through psychoacoustic harmonic distortion and the perceived pitch by as much as a semitone with 30dB of amplification.
Let’s take a listen to a 100Hz. Tone as its amplitude is gradually increased. Let me know in the comments if you perceive any variation in its pitch.