Designing Voice Blips for Villains and Antagonists (Beyond "Lower Pitch")

The default advice for a villain voice is "make it lower." It is not wrong, but it is the weakest tool in the box, and it produces a result that is only marginally distinguishable from a normal character. A villain who speaks at 600 Hz instead of 900 Hz sounds like a baritone hero. A villain who speaks at 600 Hz with a modulated subharmonic, a rhythmic irregularity, and a sharp spectral edge sounds like something the player should be afraid of.

The difference between those two sounds is not a matter of degree. It is a matter of which acoustic cues the ear reads as "threat." Pitch is one of those cues, but it is the least specific, and the easiest to overuse. The cues that actually produce the threat response are lower in the list: subharmonic content, spectral instability, rhythmic irregularity, and proximity to the human vocal tract's limits. Each one can be adjusted on its own, and each one moves the perceived threat level in a consistent direction.

This article is about those cues, and the parameter ranges that produce them without crossing into cartoon territory. The friendly/hostile tone design guide covers the two-pole version of the same question; this article is about the villain specifically, which is a narrower problem with its own set of conventions.

Why does lowering the pitch alone produce a hero, not a villain?

The human voice has a well-established relationship between pitch and perceived dominance. Studies of both human and primate vocalizations have found that lower-pitched voices are rated as more dominant, more attractive, and more authoritative across cultures. That is why the instinct to lower the villain's pitch is so common: it is the same cue that makes a hero sound impressive.

The problem is that dominance and threat are not the same thing. A lower pitch reads as "this person has power," which is equally compatible with a hero, a mentor, or a villain. The listener does not hear "villain" unless a second cue pushes the interpretation toward threat. Without that cue, the low-pitched character reads as a normal character who happens to be baritone.

Cue	Produces	Villain-specific?
Lower pitch	Dominance, authority	No — also reads as hero
Subharmonic content	Roughness, growl	Yes — reads as threat
Spectral instability	Unpredictability, unease	Yes — reads as threat
Rhythmic irregularity	Unnaturalness, danger	Yes — reads as threat
Upper-register strain	Aggression, shouting	Partially — reads as anger

The cues that actually distinguish a villain from a hero are the ones that make the voice sound unnatural. A low-pitched voice is still a natural voice; it is the voice of a large person. A voice with subharmonics, unstable formants, or rhythmic irregularities is not the voice of a person, and the ear reads the unnaturalness as a threat signal.

What does subharmonic content actually do, and how do you generate it?

A subharmonic is a frequency component that is below the fundamental pitch, typically at half or a third of the fundamental frequency. In a real voice, subharmonics are produced by the irregular vibration of the vocal folds, which happens when a person is speaking at a pitch near the bottom of their range or is fatigued. In a synthesized voice, they can be added directly.

The perceptual effect is a growl or a roughness. The fundamental is still present, but the subharmonic adds a lower, rougher layer that reads as "this voice is not entirely human." Studies of creature vocalizations in film and games have found that adding a subharmonic one octave below the fundamental is one of the most reliable ways to signal "large predator" without changing the pitch itself.

Subharmonic ratio	Effect	Use
0.5× fundamental	One-octave growl	Standard villain voice, subtle threat
0.33× fundamental	Two-octave rumble	Monstrous villain, heavy threat
0.25× fundamental	Very deep rumble	Cosmic villain, approaching overkill
0.9× fundamental (near-unison)	Beating, unease	Psychological villain, uncanny

The simplest way to add a subharmonic to a synthesized blip is to layer a second oscillator at half the fundamental frequency and mix it in at a lower level. A 600 Hz fundamental with a 300 Hz subharmonic at 20 percent of the fundamental's amplitude produces a voice that is clearly low-pitched and clearly rough, without the subharmonic overwhelming the blip.

function playVillainBlip(freq, duration, subLevel) {
    const ctx = getAudioContext();
    const now = ctx.currentTime;

    const main = ctx.createOscillator();
    const sub = ctx.createOscillator();
    const mainGain = ctx.createGain();
    const subGain = ctx.createGain();
    const masterGain = ctx.createGain();

    main.type = 'sawtooth';
    main.frequency.value = freq;

    sub.type = 'sine';
    sub.frequency.value = freq * 0.5;

    mainGain.gain.value = 1.0;
    subGain.gain.value = subLevel;

    // Envelope
    masterGain.gain.setValueAtTime(0, now);
    masterGain.gain.linearRampToValueAtTime(0.7, now + 0.005);
    masterGain.gain.exponentialRampToValueAtTime(0.001, now + duration);

    main.connect(mainGain);
    sub.connect(subGain);
    mainGain.connect(masterGain);
    subGain.connect(masterGain);
    masterGain.connect(ctx.destination);

    main.start(now);
    sub.start(now);
    main.stop(now + duration + 0.01);
    sub.stop(now + duration + 0.01);
}

// Villain blip: 600 Hz fundamental, 300 Hz subharmonic at 25%
playVillainBlip(600, 0.05, 0.25);

The 0.25 subharmonic level is a starting point. At 0.15 the effect is barely audible; at 0.4 the voice becomes muddy and the fundamental is lost. The range that produces a clear villain character without losing the blip's identity is 0.2 to 0.3.

The subharmonic is most effective when the main oscillator is a sawtooth rather than a sine. The sawtooth has a bright, buzzy spectrum that contrasts with the smooth subharmonic, and the contrast is what makes the roughness audible. A sine main with a sine subharmonic produces a clean low tone rather than a growl.

How does spectral instability signal danger?

A stable voice has consistent formants and a consistent pitch. An unstable voice has formants that shift, a pitch that wavers, or a spectrum that changes in a way that is not rhythmic. The instability is what the ear reads as "this voice is not under control," and the lack of control is the threat signal.

The two techniques that produce spectral instability in a synthesized blip are frequency modulation and filter modulation. Frequency modulation adds sidebands to the spectrum, which thickens the voice and makes the pitch harder to pin down. Filter modulation sweeps a resonant filter across the spectrum, which produces a wah-like movement that reads as instability.

Technique	Parameter	Range for villain
FM depth	Modulation index	2–5
FM rate	Modulator frequency	5–15 Hz
Filter sweep	Frequency range	±200–400 Hz around the formant
Filter sweep rate	LFO frequency	3–8 Hz
Pitch waver	Vibrato depth	±20–40 cents
Pitch waver	Vibrato rate	4–7 Hz

The FM rate is the parameter that distinguishes a menacing instability from a musical vibrato. A vibrato at 5 to 7 Hz reads as expressive singing; an FM at 10 to 15 Hz reads as a machine that is malfunctioning. The 5 to 15 Hz range covers both, and the specific value determines whether the villain sounds like a person with an unusual voice or something that is not a person at all.

The filter sweep is the more accessible technique for a simple blip, because it does not require FM synthesis. A resonant band-pass filter that sweeps across a 200 to 400 Hz range at 3 to 8 Hz produces a voice that sounds like it is coming through a damaged speaker, which the ear reads as threatening in a way that a clean voice does not.

The combination of the two is where the strongest villain voices come from. An FM layer at moderate depth plus a filter sweep at a slightly different rate produces a voice that is unstable in two independent ways, and the independence is what makes it feel organic rather than mechanical. The synthesizer sound design guide covers the FM and filter techniques in more detail; the villain-specific part is the parameter ranges.

Why does rhythmic irregularity matter more than pitch for a recurring villain?

A villain who appears in a single cutscene can get away with a single dramatic voice. A villain who recurs across the game needs a voice that the player can recognize before they see the character. The recognition is what makes the voice a threat signal, and the cue that carries the recognition is not the pitch, it is the rhythm.

The rhythm of a normal character's dialogue blips follows the text reveal at a fairly consistent rate. The blips are separated by roughly even intervals, which the player reads as a natural speaking cadence. A villain's blips can be deliberately irregular: a burst of fast blips, a long pause, another burst. The irregularity is what makes the voice feel "off," and the offness is what the player learns to associate with the character.

Rhythm pattern	Character reads as	Use case
Even spacing	Normal, trustworthy	Hero, mentor, NPCs
Bursts with pauses	Unpredictable, unsettling	Unstable villain, psychological threat
Accelerating	Escalating, losing control	Villain during a monologue climax
Decelerating	Weakening, or threateningly slow	Dying villain, or slow menacing
Repeating	Mechanical, inhuman	Robot villain, corrupted AI

The rhythm is implemented at the dialogue system level, not at the blip level. The blip itself is a short sound; the pattern of when it plays is what the player hears. A dialogue system that supports per-line blip timing can produce a burst pattern by playing the blip on every character for the first five characters, then pausing for 300 milliseconds, then resuming. The blip is the same; the rhythm is different.

The rhythm that works best for a recurring villain is the burst pattern, because it combines the recognizable (the blip's pitch and timbre) with the memorable (the irregular timing). A player who has heard the villain's voice once will recognize the next appearance before the character speaks, because the rhythm of the blips is unusual enough to stand out.

The rhythm also interacts with the pitch in a way that is worth understanding. A low-pitched voice with even spacing reads as authoritative. A low-pitched voice with burst spacing reads as dangerous. The pitch provides the power, and the rhythm provides the threat. Together, they produce a voice that is unmistakably the villain's.

What are the specific parameter ranges for different villain archetypes?

Not every villain is the same kind of villain. A scheming manipulator and a mindless monster need different voice designs, and the difference is in which cues are emphasized and which are suppressed. The four archetypes below cover most villain designs, and the parameter ranges are the ones that produce each effect.

Archetype	Fundamental	Subharmonic	Instability	Rhythm
Scheming manipulator	500–700 Hz	None or very subtle	Moderate FM (index 2–3)	Slow, deliberate
Brutal enforcer	400–500 Hz	0.5× at 0.25 level	Low FM, strong filter sweep	Bursty, aggressive
Mindless monster	300–400 Hz	0.33× at 0.3 level	Heavy FM (index 4–5)	Repeating or stuttering
Cosmic / inhuman	250–350 Hz	0.25× at 0.35 level	Dual FM at different rates	Non-repeating, drifting

The scheming manipulator is the subtlest of the four, because the voice needs to sound human enough to be trusted. The subharmonic is absent or very subtle, the FM is moderate, and the rhythm is slow and deliberate. The threat comes from the contrast between the calm voice and the player's knowledge of the character's actions, not from the voice itself.

The brutal enforcer is the most straightforward. The low fundamental and the subharmonic produce the weight, the filter sweep produces the aggression, and the bursty rhythm produces the sense of a character who is about to lose control. This is the archetype that the default "lower the pitch" advice is actually aiming for, and the added cues are what complete it.

The mindless monster is where the FM depth becomes the dominant cue. The voice needs to sound like it is being forced through a throat that is not designed for speech, and the heavy FM is what produces that effect. The subharmonic reinforces the size, and the repeating rhythm reinforces the lack of intelligence.

The cosmic or inhuman villain is the extreme case, where the voice is barely recognizable as a voice at all. The fundamental is low enough that it is on the edge of the human vocal range, the subharmonics are strong enough that the voice sounds like it has multiple layers, and the instability is multi-directional. This is the archetype that the horror sound design guide covers from a different angle; the villain-specific version is a voice that the player can still understand, barely.

What does a working villain blip set sound like in context?

The test for a villain blip is not whether it sounds scary in isolation. It is whether the player reacts to it the way the scene intends, and whether the reaction persists across repeated appearances. A voice that is scary the first time and tiresome the tenth time is a voice that has been over-designed.

The first test is the recognition test. Play the villain's blip and a hero's blip, and ask a player who has not seen the design document which one is the villain. If the answer is immediate and confident, the cues are working. If the answer requires thought, the cues are too subtle. If the answer is wrong, the cues are pointing at the wrong archetype.

The second test is the repetition test. Play the villain's blip twenty times in a row, as if the player were reading a long monologue. If the voice becomes fatiguing before the twentieth playback, the instability or the subharmonic is too strong. The cues that produce threat also produce fatigue, and the balance is the design problem.

The third test is the context test. Play the villain's blip in a scene with the game's music and sound effects present. A voice that is menacing in isolation can be masked by a loud combat scene, and a voice that is subtle in isolation can be the only thing the player hears in a quiet scene. The frequency separation described in the voice-blip mixing guide applies here as well; the villain's blip needs to sit in a frequency range that the game's other audio does not occupy.

The fourth test is the archetype test. Play the blips for all four villain archetypes in sequence and ask whether they are clearly distinguishable from each other. If the scheming manipulator and the brutal enforcer sound like variations of the same voice, the archetypes have not been separated enough. The difference between them is not just the pitch; it is the subharmonic, the instability, and the rhythm, and the combination is what makes each one distinct.

The villain voice is not a single sound. It is a family of sounds that share the same threat cues in different proportions. A designer who understands the four cues — pitch, subharmonic, instability, rhythm — can produce the full family from a small number of parameters, and can adjust any individual voice within the family without rebuilding the set. The silent protagonist guide covers the opposite case, where the voice is deliberately absent; the villain is the case where the voice is deliberately present, and the presence is the threat.