What Is Reverberation? Sound Arriving After It Has Already Stopped

Acoustics divides what arrives at a listener into three stages. The direct sound travels straight from source to ear and carries the clearest information about what was produced and where it came from. Early reflections arrive within roughly the first fifty to eighty milliseconds after bouncing once or twice off nearby surfaces, and because the ear integrates sound over that window, these are not heard as separate events but fuse with the direct sound, making it louder and giving an impression of the room's size and the listener's position in it. Later arrivals, having bounced many times, come from all directions in such density that individual reflections cannot be distinguished, and this diffuse tail is reverberation proper. A single distinct late arrival is heard separately as an echo, which requires a delay long enough and a reflecting surface large and flat enough, and echoes are generally a defect in room design rather than a feature.

The field is built on a small number of quantities and a limited set of physical remedies:

• •Reverberation time, defined as how long a sound takes to decay by sixty decibels, which is the single most used descriptor of a room
• •Sabine's formula, derived around 1900, relating that time to the room's volume divided by its total absorption, which founded architectural acoustics as a quantitative discipline
• •Absorption, using porous materials that convert sound energy to heat through friction in air movement, which works well at high frequencies and poorly at low ones
• •Bass traps and resonant absorbers, which target low frequencies that porous material cannot touch because the wavelengths are metres long
• •Diffusion, using irregular surfaces to scatter reflections in many directions rather than absorbing them, which keeps a room lively while removing harsh distinct reflections
• •Room modes, standing waves at frequencies whose wavelengths fit the room dimensions, which make bass response uneven and are why small rectangular rooms are acoustically difficult

There is no good or bad reverberation time, only appropriate and inappropriate for a purpose, and the requirements conflict directly. Speech needs clarity, which means each syllable must not be masked by the tail of the previous one, so classrooms, lecture theatres and theatres aim for short times of well under a second, and a reverberant classroom measurably impairs learning, particularly for children with hearing difficulties or learning in a second language. Orchestral music wants blending and warmth, with concert halls typically aiming for around two seconds, which lets notes overlap and the hall contribute to the sound. Organ music and chant in a cathedral live comfortably with five seconds or more, which is why that repertoire was written for those spaces and sounds thin elsewhere. Recording studios want control rather than any particular value, so they combine dead recording spaces with artificially added reverberation. Multi-purpose venues attempt compromise with adjustable elements including movable panels, retractable curtains and variable ceiling heights, and generally satisfy nobody completely.

Adding reverberation to a recording has driven a long technical progression. Echo chambers were real rooms with a speaker and a microphone, and several famous ones in recording studios are as distinctive as any instrument. Plate reverbs used a large suspended sheet of metal excited by a transducer, giving a dense smooth decay in a unit weighing a couple of hundred kilograms. Spring reverbs, small and cheap, became the characteristic sound of guitar amplifiers. Digital algorithms from the 1970s built artificial decay from networks of delays and filters, and the artefacts of early units became sounds in their own right that are now deliberately emulated. Convolution reverb takes a different approach, recording the impulse response of a real space by capturing how it responds to a brief broadband sound, then mathematically applying that response to any recording, which reproduces the acoustics of a specific building convincingly. The same measurement technique is used by acousticians to characterise real halls, so the tool serves both the engineer and the producer.

Reflections arriving within the first fifty to eighty milliseconds fuse with the direct sound and tell you the room's size, while later ones pile up into a diffuse tail. Decay to sixty decibels below the source is the standard measure, and Sabine related it to volume over absorption around 1900. Speech needs well under a second and orchestral music around two, which is why multi-purpose halls satisfy nobody completely.