Introduction to Sound Synthesis Timbre
Waveforms
An oscillator is the part of a synthesiser that generates sound. It works by repeating one short pattern over and over, hundreds or thousands of times a second. That repeating pattern is the waveform.
Repeating a pattern gives you two things to control independently. How often the pattern repeats sets the pitch: repeat it 440 times a second and you hear the A above middle C, whatever the pattern looks like. The shape of the pattern sets the tone colour of the sound. The four shapes below are the ones almost every oscillator offers, and played at the same frequency they give the same note with four clearly different tones.
Those four shapes are built the same way. Any repeating pattern can be made by adding sine waves together: one at the repeat rate itself, which is called the fundamental, then quieter ones at two times that rate, three times, four times and onward. Each waveform has its own set of levels for those sines, and the curve on screen is their sum. A sine wave is the simplest case, with the fundamental alone and nothing stacked above it. For the rest, the sound depends on two things: which of the higher sines are present, and how quickly they get quieter going up.
The four waveforms
Each of the four shapes plays the same note. The waveform is drawn above and its harmonic content below.
Listen for The pitch stays the same for all four. The sawtooth is the brightest, the square sounds hollow because its even harmonics are missing, and the triangle sounds close to a sine but slightly brighter.
What each waveform contains
The spectrum underneath plots the same sound against frequency rather than against time. Each bar is one harmonic: a sine wave at a whole-number multiple of the note you are playing, and the height of the bar is how loud that sine is in the mix. The two rows underneath give the harmonic number and the frequency it falls on, so at 220 Hz the fourth harmonic sits at 880. Move the frequency slider and the whole row of frequencies moves with it, because every harmonic is tied to the fundamental.
The sawtooth has a bar at every multiple, each one falling off as one over its number, so the second harmonic is half the height of the first, the third a third, and so on. Because every harmonic is present, the sawtooth sounds the brightest of the four. It is also why subtractive synthesis usually starts from a sawtooth, since a filter can only reduce harmonics that the oscillator produces. The square has bars only on the odd multiples, and the missing even harmonics give it a hollow tone. The low register of a clarinet sounds similar, because a cylindrical tube closed at one end also favours the odd harmonics, although the reason in the tube is acoustic rather than the symmetry of a waveform.
The triangle also contains only odd harmonics, but they fall off as one over the square of the number rather than one over the number, so the third harmonic is a ninth of the fundamental rather than a third. Its upper harmonics are quiet enough that a triangle sounds close to a sine. The sine shows a single bar, because a sine wave contains only its fundamental and has no harmonics above it.
Build a sawtooth from sines
Add harmonics to a sine wave one at a time until the sum becomes a sawtooth.
Listen for The pitch does not change as you add harmonics, because the fundamental is already there from the start. Each harmonic adds brightness, and the first four or five make the largest difference.
Adding harmonics one at a time is additive synthesis, and it shows directly that a waveform is the sum of its harmonics. By about sixteen harmonics the drawing is recognisably a sawtooth, although every part of it was built from smooth sine waves. Switch to odd harmonics only and the same stack becomes a square wave. The harmonic series guide explains where those whole-number multiples come from and how they line up with musical intervals.
A subtractive synthesiser does not build sounds this way, because it would need a separate oscillator for every harmonic. It starts from a sawtooth, which already contains them all, and reduces the upper ones with a filter. The chapter on harmonics and overtones looks more closely at the harmonics themselves.
Common questions
What are the four basic synthesiser waveforms?
The four are sine, sawtooth, square and triangle. A sine is a single frequency with no overtones at all. A sawtooth contains every harmonic, each one quieter than the last. A square contains only the odd harmonics, which is what makes it sound hollow. A triangle also contains only odd harmonics, but they get quieter much faster with harmonic number, so it sounds much closer to a sine than a square does.
Why do a sawtooth and a square sound different at the same pitch?
Pitch is set by the fundamental, and tone colour is set by the harmonics above it. Both waves repeat at the same rate, so both play the same note, but a sawtooth includes the even harmonics and a square contains none of them. A tone made of only the odd harmonics sounds hollow, similar to the low register of a clarinet.
Which synthesiser waveform is brightest?
The sawtooth is the brightest. It contains every harmonic of the fundamental, with each one falling off gently as 1 over its number, so a larger share of its energy lies above the fundamental than in any of the other basic shapes. That is also why it is the usual starting point for subtractive synthesis, since a filter can only reduce harmonics that are present.
What does a sine wave sound like?
A sine wave sounds like a single pure tone, close to a tuning fork or a softly played flute. It is the only waveform that contains only its fundamental, and every other waveform can be built by adding sine waves together.
Published Updated Glossary
In this part: Timbre
- 3 Waveforms
- 4 Harmonics and overtones
- 5 Filters