Professional Mastering - Analog & Digital
Professional Mastering
251K+ subscribers
There’s always a degree of subjectivity in music production. For every engineer who prefers odd-order harmonic distortion, one will prefer even, and another will be indifferent. For every mixing engineer who enjoys the natural resonances of a mix, another will find them problematic and try to reduce them.
But by studying the physics of sound, the interaction between waveforms, and then observing the strong patterns that emerge, we can find borderline objective ways to make any production sound better.
In this video, let’s explore the relationship between harmonics, dissonance, and the plugins used during a session.
And in the process, we’ll find ways to use common plugins like compressors, saturators, and equalizers in a way that augments what we perceive as musical and enjoyable.
By the end, I think you’ll agree that Knowing this Will Drastically Improve Any Mix
Watch the video to learn more >
Dissonance occurs due to destructive phase interference between waveforms.
But the perception of dissonance is believed to be caused by the cochlea’s inability to differentiate multiple similar frequency signals.
On the waveform side of things, phase cancellation between 2 or more signals causes pulsing, or points at which the signals null. The further away the signals are in frequency, the quicker they null, and once they’re far enough apart, the cycle of each null is so fast that it’s hard to notice.
On the psychoacoustic side of things, the cochlea has a finite resolution. Similar to how an EQ or any FFT processor can break up frequency ranges into bands, the cochlea has distinct regions in which it can accurately translate incoming info.
But whereas an EQ can have thousands of these regions, the cochlea only has 24, meaning it’s difficult for it to distinguish closely related frequencies, since they stimulate the same critical band.
For example, if I have a 200Hz sine wave, I can listen to it and hear it without any confusion. (Demo)
If I have a 210Hz sine wave, I could do the same thing. (Demo)
But if I play these 2 waves simultaneously, I’ll hear amplitude modulation.
Depending on the listening setup, this could be caused by the physical interaction between air molecules, as we covered when discussing waveforms. However, if I play 200Hz in the left channel and 210Hz in the right, and you listen with headphones, you’ll still hear amplitude modulation.
Watch the video to learn more >
The signals aren’t physically interacting since they’re isolated to their respective channels, and yet the modulation is still present.
This indicates that the phenomenon of dissonance isn’t strictly the physical interaction between waveforms, and could be caused solely by psychoacoustic factors, in some instances.
I bring this up since later, when we discuss how processors influence dissonance, I want you to know that whether the listener is using headphones or speakers, the dissonance will still play a role.
So far, we’ve discussed sine waves, or single-frequency waveforms. But the majority of instruments, unless you’re playing a sine-wave generator, will have multiple sine waves occurring simultaneously.
If I play a note on a piano and observe the response, notice that I don’t have a single frequency. Instead, I have a fundamental and a complex series of overtones or harmonics.
Most of these overtones occupy frequencies closely related to our Western scale. For example, as I play C3 (demo), I’ll observe overtones at C4 and C5.
But, I’ll also notice overtones that stray slightly to significantly away from any note - the 7th order overtone from this particular piano, while playing C3 occupies 1823Hz, which is 39Hz flat of the closest note A#6. (Demo of 1823Hz)
Now, if I play C#3 while C3 plays on this same instrument, the fundamentals of the notes and their overtones interact. (Demo)
This occurs both to the physical waveforms and while they’re processed by the cochlea.
If we look at this interactive graph, which I’ll leave a link for in the description, we can observe a map of 3 note chords on a piano. The orange peaks indicate places of high dissonance, whereas the troughs show areas of low dissonance.
If I click on a point in which the chord has 3 notes in unison, there’s very low dissonance - this makes sense - they’re all the same note. Meanwhile, the minor third, major third, major fourth, major fifth, and perfect octave all occupy the areas in which dissonance is low.
This suggests that our scales are based on keeping dissonance to a minimum.
This large red peak is the most dissonant 3-note chord. As you imagined, it’s not an enjoyable sound, and isn’t used too often when making music.
But keep in mind that this is just 3 notes played at once.
Imagine you have a full mix with multiple instruments, each playing distinct parts.
For most compositions, these parts are written to be in key, which should, in theory, keep dissonance to a minimum. Yet when their overtones interact, dissonance occurs.
This happens in just about every recording, so it’s not something to obsess over; however, certain forms of processing can improve this or exacerbate it.
Watch the video to learn more >
These are the 3 main types of processing that affect dissonance: either they can be used to control the amplitude of overtones, or they introduce new overtones.
EQ is probably the most immediate example.
Filters can either amplify or attenuate certain frequencies - meaning they can be used to directly affect the relationship between overtones.
One mistake I often see is using a sharp bell filter, increasing the gain while sweeping to find the supposed issue, then attenuating. But the issue usually isn’t a single frequency - rather, it’s the interaction between that frequency and others. Again, dissonance requires this interaction either physically or psychologically.
If you have the ear to find the frequency or frequencies causing dissonance, that’s incredible - it’s the hearing equivalent of finding a needle in a haystack.
If not, the best place to start is finding overtones of various instruments that don’t correlate to a note in the Western scale.
As we covered with the interactive graph, when frequencies are in unison or perfect octaves, dissonance is at its lowest, with the second least dissonant combination being the perfect 5th.
So, if you’re hearing something unmusical in a mix, and an instrument has overtones that don’t correspond to a note, that’s a quick and easy way to reduce dissonance. With your EQ, dip the unmusical overtone and perhaps boost the nearest in-key note to improve the ratio between the 2.
Watch the video to learn more >
Compressors often shape the waveform of the signal they’re compressing. Any reshaping of a waveform is actually the addition of new lower amplitude waveforms - this is why, if I compress a sine wave, I can observe overtones in addition to the original frequency.
Various parameters change how the wave is shaped, resulting in overtones of differing amplitudes and frequencies.
But when looking at a compressor alone, it’s incredibly difficult to tell how it will distort the waveform, let alone determine if the overtones caused by the distortion will create dissonance with other signals.
So, if you’re noticing unpleasant levels of dissonance after compression, you have 2 options. 1, keep adjusting until you retain the timbre you want while making the issue less noticeable.
A slightly longer release and a couple of milliseconds of lookahead, combined with a softer knee setting, will greatly reduce waveshaping, which will lower the amplitude of overtones, causing less dissonance if these overtones are the culprit.
Option 2 is to use a compressor designed for mastering. These introduce incredibly low levels of distortion - some notable examples are FabFilter’s Pro-C2 with its mastering algorithm, and ToneBooster’s compressor V4 with its Punch algorithm.
I’m not saying using these is necessary, just that they’ll likely cause more musical compression by reducing dissonance in the long run.
But, by far the most influential processing type, as it relates to dissonance, is Saturation.
Saturation is the purposeful reshaping of waveforms for the sake of generating overtones.
Depending on the component used or emulated, be it tape, transistor, tube, and so on, different harmonics form.
Tape saturation will typically generate odd-order harmonics or overtones; the same is true for most transistors.
Tube saturation, on the other hand, is often asymmetrical, meaning it’ll create even-order harmonics, which are much more likely to be perfect octaves. This is why people often report a tube sounding more musical and warm.
The highest amplitude overtone is often the 2nd order, which is a perfect octave. It’ll likely occupy the low mids, hence the description of its warmth, and since it’s a perfect octave, it will result in lower levels of dissonance, hence the interpretation of it sounding musical.
But the accuracy of these emulations depends heavily on the developer of the plugin - for example, a developer could claim it’s emulating a tube, but just get the harmonic formation completely wrong when they code it.
So, do some research and look behind the scenes of your go-to saturators to figure out how they’re shaping the waveform.
Additionally, experiment with an emphasis de-emphasis technique - that is, insert an EQ before and after the saturator.
With the EQ before the saturator, emphasize a particular frequency; with the second EQ, attenuate to compensate for the first EQ.
This way, you can drive particular frequencies into the saturator, giving you more control over what is saturated. By controlling what gets saturated and knowing how a particular saturator generates overtones, you’ll be in a much better position to ensure your overtones introduce as little dissonance as possible.
Watch the video to learn more >
We covered a lot, so let’s just recap some of the main ideas. Whether it occurs between waveforms or during the cochlea’s stimulation, dissonance is a byproduct of the interaction between 2 waveforms that are close in frequency.
Low levels of dissonance are generally considered favorable in music, which explains why our scales and tunings of instruments emphasize areas in which we encounter low dissonance, such as a major fourth, perfect fifth, and so on.Lastly, various processors directly or indirectly affect the level of dissonance. Through intentional processing and knowledge of how the processor is working, we can minimize the amount of dissonance to cause a more musical experience for the listener.