Introduction
In our last lesson, we established the fundamental signal flow of subtractive synthesis: an Oscillator generates a harmonically rich waveform, a Filter sculpts its timbre, and an Amplifier controls its final volume. You learned that the filter acts as the primary "sculptor," but we only touched upon its function in the broadest sense.
This lesson delves deep into the heart of the filter module. We will explore the primary filter types used in synthesis—low-pass, high-pass, and band-pass—and understand how they shape the spectrum of a sound. Crucially, we will dissect the concept of resonance, perhaps the single most important parameter for creating the dynamic, characterful sounds associated with classic synthesis. Given your background in physics and electronics, we will not only examine what these components do but also explore the underlying principles of how they work, connecting abstract filter graphs to the behavior of electronic circuits.
By the end of this lesson, you will be able to explain how filters sculpt timbre and apply these tools to transform a static oscillator tone into a vibrant, expressive sound.
The Anatomy of a Filter
At its core, a filter is a frequency-dependent amplifier. In subtractive synthesis, its primary job is to attenuate, or "subtract," specific ranges of frequencies from the oscillator's raw signal. The two most fundamental parameters of any filter are its type and its cutoff frequency.
Let's start with a clear, concise overview of the most common filter types and their essential characteristics.
This video from Mixed Signals provides an excellent animated introduction to the core concepts of filtering.
Watch the following sections: Introduction: The brief introduction explains the filter's basic purpose. Filter Types: Pay close attention to the animated graphs explaining low-pass, high-pass, and band-pass filters. Note how each affects the frequency spectrum relative to the cutoff frequency. Filter Slope: The section on filter slope introduces the concept of how sharply a filter attenuates frequencies.
As the video explains, the main filter types are:
- Low-Pass Filter (LPF): The most common filter in synthesis. It passes frequencies below the cutoff and attenuates frequencies above it. Sweeping the cutoff from high to low makes a sound progressively "darker" or "muffled."
- High-Pass Filter (HPF): The inverse of a low-pass. It passes frequencies above the cutoff and attenuates those below it. This is useful for thinning out sounds or removing low-frequency mud.
- Band-Pass Filter (BPF): A combination of a low-pass and a high-pass filter. It passes only a specific band of frequencies around a center frequency, attenuating frequencies both above and below this band.
Filter Slope and Poles
The video introduced the concept of slope, measured in decibels (dB) per octave. This defines the "steepness" or aggression of the filter. A shallow slope of 6 dB/octave creates a gentle attenuation, while a steeper 24 dB/octave slope creates a much more dramatic cutoff.
This slope is determined by the filter's internal design, specifically the number of "poles" it has. Each pole, which in a simple passive circuit is an RC (resistor-capacitor) pair, contributes approximately 6 dB/octave to the slope.
- A 1-pole filter has a 6 dB/octave slope.
- A 2-pole filter has a 12 dB/octave slope.
- A 4-pole filter has a 24 dB/octave slope.
This is why you will often see synthesizers advertise their famous "24 dB/octave Moog filter" or "12 dB/octave Oberheim filter." The 4-pole (24 dB/octave) design is known for its aggressive, distinctly "electronic" sound, while 2-pole (12 dB/octave) filters are often considered more subtle or "musical."
Resonance: The Sound of Synthesis
While cutoff frequency and slope define the basic shape of the filter, resonance gives it its unique character. Resonance, also known as Q (for "quality factor"), creates a narrow boost or peak in the frequency response right at the cutoff frequency.
This peak emphasizes any harmonics that fall within it, imparting a "whistling," "singing," or "squelchy" quality to the sound. Sweeping the cutoff frequency of a resonant filter causes this peak to sweep through the harmonics of the sound, creating the classic synth filter sweep that is a hallmark of genres from funk to acid house.
To understand how this behavior arises, we can draw an analogy from your physics background and then look at the electronics.
Of Responses & Resonance - Sound On Sound
This classic article from Sound On Sound's "Synth Secrets" series provides an intuitive physical analogy for resonance before discussing its electronic implementation.
Please read the section titled "Step 4: Resonance". The author analogizes filter resonance to a string vibrating in sympathy with an external sound source, which is a useful mental model.
The article mentions that resonance in active filters can boost frequencies, unlike passive filters which can only attenuate. It also correctly states that the behavior is the solution to a second-order differential equation. This is precisely the same mathematical framework used to describe damped harmonic oscillators (like a mass on a spring), which exhibit resonant behavior when driven at their natural frequency. In a filter circuit, the resonance is created by feeding a portion of the filter's output signal back to its input.
For a deep and practical dive into the electronics of how this feedback creates resonance, the following video is unparalleled. It walks through building an active filter and implementing resonance from the ground up. The visualizations on the oscilloscope will be particularly illustrative for you.
DIY SYNTH VCF Part 2: Active Filters & Resonance
The channel "Moritz Klein" offers detailed tutorials on DIY synthesizer circuits. This video explains why simple passive filters are insufficient and how active filters using operational amplifiers (op-amps) and feedback are used to create resonance.
Please watch the following two key segments: Active Filters & Buffering: Watch from the start to understand why active filters are necessary. The explanation of op-amps as buffers with high input impedance and low output impedance to isolate filter stages is crucial. This will likely connect with concepts from your electronics labs. Creating Resonance with Feedback: This is the core of the topic. Watch from this section. Pay close attention to how he visualizes resonance on the oscilloscope as an "over and undershooting behavior." His step-by-step analysis of the circuit, explaining how feeding the output back to the input creates this ringing oscillation at the cutoff frequency, is a masterclass in the electronic "why" of resonance.
The key insight from the video is that resonance is a controlled instability. By creating a feedback loop, the filter is made to "ring" or oscillate in response to the input signal. The frequency of this ringing is determined by the component values (resistors and capacitors) that also set the cutoff frequency. The amount of resonance (Q) corresponds to the amount of signal fed back into the input.
Self-Oscillation: The Filter as a Source
What happens when you turn the resonance up to its maximum value?
As the feedback in the loop increases, the filter becomes more and more unstable. At a certain point, the gain of the feedback loop at the cutoff frequency exceeds unity. The circuit no longer needs an external input signal to sustain its ringing; it becomes a self-contained oscillator.
This is known as self-oscillation. A self-oscillating filter produces a very pure sine wave at its cutoff frequency. This means you can "play" the filter by adjusting its cutoff frequency knob or controlling it with a keyboard's control voltage. It's a powerful sound design technique for creating laser zaps, percussive pings, and melodic sine wave leads.
Making sounds with analogue electronics - Part 2: Subtractive ...
This article provides a concise explanation of how filter resonance, when pushed to its extreme, leads to self-oscillation.
Read the short section Filter oscillation. It summarizes how a filter with high resonance can be conceptualized as an amplifier with a frequency-selective element in its feedback loop, causing it to oscillate at the frequency of highest gain—the resonant peak.
Conclusion
In this lesson, we have moved beyond a simple block diagram and explored the rich, character-defining properties of the synthesizer filter. You now understand not just the what, but also some of the why behind its behavior.
Key Takeaways:
- Filter Types: Low-pass, high-pass, and band-pass filters are the primary tools for subtracting frequencies and shaping timbre.
- Cutoff Frequency: This parameter determines the point in the spectrum where the filter begins to act.
- Filter Slope: Measured in dB/octave and determined by the number of "poles" in the filter, this controls the steepness of the attenuation.
- Resonance (Q): A peak at the cutoff frequency, created by an internal feedback loop. It emphasizes harmonics and imparts a characteristic "synth" sound. At its extreme, it causes the filter to self-oscillate, producing a pure sine wave.
You now possess the core vocabulary and conceptual understanding of the filter. You know that sweeping a low-pass filter's cutoff makes a sound darker or brighter, and that adding resonance emphasizes that sweep dramatically.
Our sound is still static, however. A synthesist rarely adjusts these knobs by hand for every single note. The real power comes from automating this control. In our next lesson, we will introduce ADSR envelopes, the primary tool for shaping a sound's characteristics over time, including how they are used to dynamically control the filter's cutoff frequency to create evolving, percussive, and expressive sounds.