Basics of Synthesizers: Creating Sound from Scratch
Synthesizer Basics: Building a Sound From Nothing
Open a synthesizer for the first time and you are met with dozens of knobs, all of them looking equally important. Here is the reassuring part: almost every synth ever made is built from four components. Oscillator, Filter, Envelope, LFO. Learn what each one does and you will never be lost in front of a synth again.
We debated whether this article belonged in the series at all. In an age where AI can generate sound, does anyone still need to learn how to turn knobs? We decided yes, though not for the reason you might expect. The point is not to use a synth. The point is to understand what a sound is made of. That instinct is what lets you hear a generated result and know what to change. Our own model breaks sound down into timbre, energy and time structure. The underlying idea is no different.
The Synthesizer Signal Flow
The oscillator produces the raw material, the filter carves it into shape, and the amp and envelope decide how it behaves over time. An LFO sits underneath the whole chain, turning knobs for you automatically. This arrangement is called subtractive synthesis.
1. Oscillator: The Raw Material
The oscillator generates the waveform. This is where you pick one of the shapes covered in Part 2: sine, sawtooth, square or triangle.
- Wave: sets the basic character of the sound. Sawtooth for bright and rich, sine for pure and heavy.
- Pitch: how high or low the note sits.
- Detune: stack two oscillators and offset one slightly and the sound opens up and thickens. The effect comes from beating between the two.
2. Filter: The Knife That Carves Harmonics
A filter removes specific frequency bands. Starting from something harmonically rich like a sawtooth, you carve away what you do not need until the timbre is what you want.
The one you will use most is the low-pass filter. It lets low frequencies through and cuts everything above.
- Cutoff: where the cutting begins. This will be your most-used knob by a wide margin.
- Resonance: emphasises the area around the cutoff, adding a vocal, whistling ring. Push it up gradually, since too much turns harsh.
- Q (Quality Factor): how sharply the curve falls away either side of the cutoff frequency.
In game audio, a low-pass filter is what you reach for when a sound is heard through a wall. You are imitating the physics of high frequencies being absorbed before they get to the listener.
3. Envelope and ADSR: Shaping Sound in Time
An envelope is the blueprint for how a sound's level changes from the moment it starts. You control it with four stages, known as ADSR.
- Attack: the time taken to reach full level. Keep it short and you get a sharp, percussive hit.
- Decay: the time taken to fall from the peak down to the sustain level.
- Sustain: the level held for as long as the key is held. Note that this is a level, not a time.
- Release: the time taken for the sound to fade to silence once the key is let go.
Two recipes will carry you a long way. Impact sounds such as a sword strike or a button click want a minimal attack and a short decay. Take the same waveform and change nothing but the ADSR, and you end up with a completely different sound.
This instinct carries over to generative tools. When a sound feels like it has no punch, the culprit is usually a blunt attack. Knowing that lets you replace a vague request like "make it punchier" with something precise: "shorter attack."
4. LFO: Putting a Sound in Motion
An LFO (Low Frequency Oscillator) is a control signal that modulates other parameters for you.
- Route an LFO to pitch and you get vibrato, a gentle wavering in the note.
- Route it to volume and you get tremolo, a steady pulsing in level.
- Route it to filter cutoff and you get a wah effect.
It is the tool that brings a static sound to life.
Synthesis or Generation: Knowing Which to Reach For
Here is how the split tends to fall in our own work.
Synthesis wins when a sound is short, simple and needs to be exactly right. A UI notification is the obvious case. One knob gets you fine control, and you always know why the sound came out the way it did.
Generation wins when a scene has many moving parts, such as a battle in a rainy forest. Building that from scratch on a synth takes days. Generating candidates takes seconds.
In practice we mix the two. Use VARCO Sound to pull a rough version of the whole scene and settle on a direction, then, if one element in the multitrack result falls short, rebuild just that piece on a synth and layer it back in. Prototype fast, finish carefully. Working that way means knowing both sides.
A Practice Exercise
Install Vital or Surge XT, both free, and work through this in order.
- Select a single sawtooth oscillator
- Sweep the low-pass cutoff slowly down and back up
- Push attack and release to both extremes
- Route an LFO to the cutoff
Fifteen minutes of this will teach you more than ten articles.
Frequently Asked Questions(FAQ)
Q. Do I need to buy a hardware synthesizer?
No. Software is plenty for learning, and free synths like Vital and Surge XT hold their own against paid products.
Q. I can't play keyboards. Is that a problem?
Not at all. Sound effect work is closer to design than performance, and any notes you need can be drawn into a piano roll.
Q. I'm a game developer. Do I really need all of this?
If you have no plans to touch a synth yourself, the ADSR section alone will serve you well. Shaping impact and decay is a skill you will need whatever tool you end up using.
Start With a Generated Skeleton for Complex Scenes
When a scene is too layered to build one piece at a time, use VARCO Sound to establish the direction first. It keeps the texture of an uploaded sample while producing multiple variations, which makes "like this, but a little different" quick to iterate on.
