Audio DSP: Haas Effect

---

---

Overview

A C++ implementation of the Haas effect (also known as the precedence effect), intended as a learning resource for understanding how inter-channel delay creates perceived stereo width from a mono source.

The Haas effect is a psychoacoustic phenomenon: when the same sound reaches both ears with a delay of 1–40ms, the brain fuses them into a single perceived sound and localises it toward the ear that heard it first. This is exploited in audio production to widen a mono signal into a convincing stereo image, without creating a perceptible echo.

For a deeper dive into delay in general, see the Audio Delay repository.

---

Files

File	Description
Haas.h / .cpp	Haas effect implementation — stereo delay with configurable delayed channel
main.cpp	Example: applies the effect to a 440Hz mono sine wave, demonstrates both channel configurations and two delay times

---

How It Works

Both channels receive the same mono input signal. One channel (the leading channel) passes through unmodified. The other (the delayed channel) is held in a circular buffer and output after the configured delay time:

monoInput ──►── Left  (leading) ──────────────────────────────────► leftOut
            │
            └──► Right (delayed) ──► [ circular buffer, N samples ] ──► rightOut

The brain perceives this timing difference as spatial information and localises the sound toward the leading channel. Swapping which channel is delayed flips the perceived position.

The delay is implemented as a fixed-size circular buffer — identical in structure to the reverb implementation — so no memory is allocated or freed during audio processing.

---

Delay Time Reference

Delay	Perceptual Effect
< 1ms	Phase cancellation / comb filtering — not a Haas effect
1 – 10ms	Subtle width, tight image
10 – 30ms	Clear stereo spread, natural sounding
30 – 40ms	Wide image, approaching the limit of fusion
> 40ms	Perceived as a distinct echo rather than spatial width

---

Parameters

Parameter	Type	Range	Description
delayInMilliseconds	int	1 – 40ms	Delay applied to the delayed channel.
delayedChannel	DelayedChannel	Left or Right	Which channel is delayed. The other channel leads.
sampleRate	float	> 0.0	Audio sample rate in Hz. Used to convert ms to samples.

DelayedChannel enum:

Value	Leading Channel	Perceived Image
DelayedChannel::Right	Left	Sound perceived from the left
DelayedChannel::Left	Right	Sound perceived from the right

---

Usage

Basic

// 20ms delay, right channel delayed, at 44100Hz
Sherbert::Haas haas(20, Sherbert::Haas::DelayedChannel::Right, 44100.0f);
 
float leftOut  = 0.0f;
float rightOut = 0.0f;
 
// Pass the same mono signal to both inputs
haas.ProcessSample(monoIn, monoIn, leftOut, rightOut);

Flipping the Perceived Position

// Flip which channel is delayed at runtime — no reset required
haas.setDelayedChannel(Sherbert::Haas::DelayedChannel::Left);

Changing Delay Time

// setDelayInMilliseconds calls reset() internally to avoid buffer artefacts
haas.setDelayInMilliseconds(5);

Resetting State

haas.reset();

---

API Reference

Method	Description
Haas(delayInMs, delayedChannel, sampleRate)	Construct with initial parameters.
ProcessSample(leftIn, rightIn, leftOut, rightOut)	Process one stereo sample pair. Leading channel passes through; delayed channel is buffered.
reset()	Clears the delay buffer and resets the write index.
setDelayInMilliseconds(value)	Update delay time. Calls reset() internally.
setDelayedChannel(value)	Swap which channel is delayed. Does not call reset().
getDelayInMilliseconds()	Returns current delay time in ms.
getDelayedChannel()	Returns current DelayedChannel value.
getSampleRate()	Returns the sample rate set at construction.

---

Limitations & Next Steps

No wet/dry mix — this implementation outputs the delayed channel fully wet. In a production context you would typically blend the delayed channel with the dry signal to control the intensity of the effect.

Mono input assumed — the ProcessSample API accepts separate leftIn and rightIn values, but the Haas effect only makes sense when both inputs carry the same mono signal. Passing a true stereo source will produce unpredictable results.

No delay time modulation — some production Haas implementations slowly drift the delay time to avoid phase cancellation artefacts when the effect is mixed to mono. This would require an LFO similar to the phaser implementation.

If you want to explore further, the natural next steps from here are:

• A wet/dry mix parameter
• Slow LFO modulation of the delay time to reduce mono compatibility issues
• Mid/Side processing — applying the Haas effect only to the Side signal to preserve mono compatibility

---