2.4 Sound field in rooms

Sound waves behave differently in enclosed spaces than in open areas. In rooms, sound interacts with surfaces, creating complex patterns of reflection, absorption, and diffusion. Understanding these interactions is the foundation for designing spaces with good acoustic properties.

This topic covers direct and reverberant sound, room modes, reverberation time and its prediction equations, absorption and diffusion strategies, objective room acoustics parameters, simulation methods, and measurement techniques.

Sound propagation in rooms

Sound propagation in rooms describes how sound waves travel and interact within an enclosed space. Every surface a wave encounters changes its energy, direction, or both. Getting a handle on this behavior is what lets you design concert halls, recording studios, and lecture rooms that actually sound good.

Direct vs reverberant sound

Direct sound travels straight from the source to the listener with no reflections. It arrives first and carries the clearest information about the source.

Reverberant sound reaches the listener after one or more reflections off room surfaces. It builds up as reflections accumulate and gives a room its characteristic "liveness."

The ratio of direct to reverberant sound changes with distance from the source. Close to the source, direct sound dominates. As you move farther away, the reverberant field takes over, especially in rooms with hard, reflective surfaces.

Critical distance

The critical distance is the point where the direct sound level and the reverberant sound level are equal. Inside this distance, you mostly hear direct sound. Beyond it, the reverberant field dominates.

Critical distance depends on three things:

• Room volume and total absorption (more absorption pushes the critical distance farther out)
• Source directivity (a more directional source extends the critical distance along its main axis)

This concept matters practically: if a listener sits well beyond the critical distance, speech intelligibility drops because the reverberant sound masks the direct signal.

Room modes

Room modes are standing waves that form at specific frequencies when sound reflects between parallel surfaces and interferes constructively. They're most problematic at low frequencies, where wavelengths are comparable to room dimensions.

At a mode's resonant frequency, some locations in the room will have a strong buildup of sound pressure (antinodes) while others will have almost none (nodes). This creates uneven bass response that varies from seat to seat.

Axial vs tangential vs oblique modes

• Axial modes form between two parallel surfaces (e.g., floor-ceiling). They involve just one pair of walls and have the strongest effect on the sound field because the energy bounces back and forth along a single axis.
• Tangential modes involve four surfaces and carry less energy per mode than axial modes.
• Oblique modes involve all six surfaces of a rectangular room. Individually they're the weakest, but there are many more of them at higher frequencies.

As frequency increases, modes of all three types become more densely packed. Above a certain frequency (the Schroeder frequency), modes overlap so much that the sound field becomes essentially diffuse and individual modes stop being audible problems. Careful room dimension ratios and the use of absorptive and diffusive treatments help mitigate mode problems in the low-frequency range.

Reverberation

Reverberation is the persistence of sound in a room after the source stops. It results from the accumulation of countless reflections off room surfaces, each losing a bit of energy at every bounce. Reverberation strongly shapes how we perceive a space: too much and speech becomes muddy; too little and music sounds dry and lifeless.

Reverberation time (RT60)

RT60 is defined as the time it takes for the sound pressure level to drop by 60 dB after the source stops. It's the single most commonly cited room acoustics metric.

RT60 depends on:

• Room volume (larger rooms generally have longer RT60)
• Total absorption (more absorptive surfaces shorten RT60)
• Room shape (affects how evenly sound decays)

Optimal RT60 values vary by use. Speech-focused spaces like classrooms work best around 0.4–0.8 s. Concert halls for orchestral music typically target 1.5–2.2 s. Recording studios and control rooms often aim for very short, tightly controlled values.

Early vs late reflections

Early reflections arrive within roughly 50–80 ms after the direct sound. Your brain fuses them with the direct sound, so they reinforce clarity and contribute to the perception of spaciousness and source width.

Late reflections arrive after that initial window and blend together into the reverberant tail. They give a room its sense of envelopment and warmth but, if too strong, reduce intelligibility.

The balance between early and late energy is controlled by strategic placement of reflective surfaces (to strengthen useful early reflections) and absorptive materials (to tame excessive late energy).

Diffuse vs specular reflections

• Specular reflections behave like light bouncing off a mirror: the angle of incidence equals the angle of reflection. They occur on smooth, flat surfaces and can create strong, discrete echoes.
• Diffuse reflections scatter sound in many directions at once, produced by irregular or textured surfaces. They help fill the room with even reverberant energy without creating distinct echoes.

Most well-designed rooms use a combination of both. Specular reflections can be useful for directing early energy toward listeners, while diffuse reflections create a natural, enveloping reverberant field.

Sabine equation

The Sabine equation is the classic formula for estimating reverberation time:

$RT60 = \frac{0.161V}{A}$

where $V$ is the room volume in cubic meters and $A$ is the total absorption in sabins (square meters of equivalent open-window absorption).

Total absorption $A$ is calculated by summing $\alpha_i S_i$ for each surface, where $\alpha_i$ is the absorption coefficient and $S_i$ is the surface area.

The Sabine equation assumes a diffuse sound field with evenly distributed absorption. It works well for rooms with relatively low average absorption ($\bar{\alpha}$ below about 0.3) but overestimates RT60 in highly absorptive rooms. Despite this limitation, it remains the standard first approximation.

Eyring equation

The Eyring equation improves on Sabine for rooms with higher absorption:

$RT60 = \frac{0.161V}{-S \ln(1 - \bar{\alpha})}$

where $S$ is the total surface area and $\bar{\alpha}$ is the average absorption coefficient across all surfaces.

The key difference: the Eyring equation accounts for the fact that in a highly absorptive room, each reflection removes a larger fraction of energy, so the Sabine formula's assumption of gradual, continuous decay breaks down. When $\bar{\alpha}$ is small, the two equations give nearly identical results. When $\bar{\alpha}$ exceeds about 0.3, Eyring is noticeably more accurate.

The Eyring equation still assumes a diffuse field, so it has its own limitations in rooms with very uneven absorption distribution.

Sound absorption in rooms

Sound absorption converts acoustic energy into heat (usually a tiny amount) when sound encounters a surface or object. Controlling absorption is the primary tool for managing reverberation time, reducing noise buildup, and improving speech intelligibility.

Absorption coefficients

The absorption coefficient ($\alpha$) describes how much sound energy a material absorbs at a given frequency, on a scale from 0 (perfectly reflective) to 1 (perfectly absorptive).

Absorption coefficients are frequency-dependent. A thick carpet might have $\alpha = 0.1$ at 125 Hz but $\alpha = 0.65$ at 4 kHz. Manufacturers publish these values across standard octave bands (typically 125 Hz to 4 kHz), and you'll use them directly in Sabine and Eyring calculations.

Porous vs panel absorbers

Porous absorbers (fiberglass, mineral wool, acoustic foam, heavy curtains) work by friction. As sound waves push air back and forth through the material's pores, kinetic energy converts to heat. They're most effective at mid and high frequencies. Thicker panels or panels mounted with an air gap behind them extend absorption to lower frequencies.

Panel absorbers (thin plywood panels, perforated panels, membrane absorbers) work by resonance. The panel vibrates at its resonant frequency and dissipates energy. They're most effective at low to mid frequencies, making them a good complement to porous absorbers.

For broadband absorption across the full frequency range, rooms typically use a combination of both types.

Placement of absorbers

Where you put absorbers matters as much as which absorbers you choose:

• First reflection points on walls and ceiling are high-priority locations for improving clarity and intelligibility.
• Even distribution across surfaces prevents the sound field from becoming lopsided (e.g., one end of the room much deader than the other).
• Corners are effective locations for treating low-frequency standing waves, because particle velocity is highest at boundaries for many room modes.
• Ceiling clouds and wall panels are common treatments in lecture halls and offices.

Effect on reverberation time

Adding absorption directly reduces RT60, as both the Sabine and Eyring equations show. But the effect is frequency-dependent: adding porous absorbers will shorten RT60 at mid and high frequencies more than at low frequencies, potentially creating an unbalanced decay.

To achieve a target RT60 curve across the full frequency range:

1. Calculate the existing RT60 at each octave band using measured or estimated absorption data.
2. Determine how much additional absorption is needed at each frequency.
3. Select materials whose absorption profiles fill the gaps.
4. Run the calculation again to verify the result.

Acoustic simulation software automates this process, but understanding the underlying math helps you make informed design decisions.

Diffusion and scattering

While absorption removes energy from the sound field, diffusion redistributes it. Diffusers scatter sound in many directions, breaking up strong specular reflections and creating a more uniform reverberant field. This is especially valuable in performance spaces where you want envelopment without echo.

Diffusers vs reflectors

Reflectors are smooth, flat surfaces that redirect sound specularly. They're useful for sending early reflections toward an audience (e.g., overhead reflectors above a stage), but they can also create problematic echoes or uneven coverage.

Diffusers scatter sound broadly. They preserve energy in the room (unlike absorbers) while preventing the harsh, focused reflections that flat surfaces produce. This makes them ideal for rear walls, side walls, and ceilings where you want liveliness without distinct echoes.

Scattering coefficients

The scattering coefficient ($s$) quantifies how much of the reflected energy is scattered versus specularly reflected, ranging from 0 (perfectly specular) to 1 (perfectly diffuse). Like absorption coefficients, scattering coefficients are frequency-dependent and are used as inputs in acoustic simulation software.

Types of diffusers

• Schroeder diffusers use mathematically derived well-depth sequences to scatter sound predictably across a designed bandwidth. The two main types are quadratic residue diffusers (QRD), which scatter sound in one plane, and primitive root diffusers (PRD), which scatter in a hemispherical pattern.
• Geometric diffusers rely on physical shape (pyramids, hemispheres, convex curves) to redirect sound in multiple directions.
• Volumetric diffusers such as skyline diffusers use varying block heights arranged in a grid pattern. Binary amplitude diffusers (BAD) alternate between reflective and absorptive patches.

Each type has a useful frequency range determined by its physical dimensions. Deeper wells or larger features scatter lower frequencies.

Placement of diffusers

• Rear walls are the most common location, reducing strong back-wall reflections that can cause echoes for performers and listeners near the front.
• Side walls help reduce flutter echoes (the rapid repetitive reflection between parallel surfaces) and improve lateral energy.
• Ceilings can use diffusers to distribute overhead reflections more evenly.
• Diffusers are often combined with absorbers: for example, absorbers at first reflection points for clarity, and diffusers on the rear wall for spaciousness.

Room acoustics parameters

Room acoustics parameters are objective, measurable quantities derived from a room's impulse response. They let you put numbers on subjective qualities like clarity, spaciousness, and loudness, making it possible to compare designs and verify that a built space meets its targets.

Clarity (C50, C80)

Clarity indices compare early sound energy to late sound energy.

• C50 (speech clarity): ratio of energy in the first 50 ms to energy after 50 ms, expressed in dB. Values above about 2 dB generally indicate good speech intelligibility.
• C80 (music clarity): same concept but with an 80 ms boundary. Values around 0 to +3 dB are typical for good orchestral music halls; higher values suit amplified or rhythmically complex music.

Higher clarity values mean more of the sound energy arrives early, which aids intelligibility and definition. Lower values indicate a more reverberant, blended sound.

Definition (D50)

D50 is the ratio of early energy (first 50 ms) to total energy, expressed as a percentage. It's closely related to C50 but presented on a linear scale rather than in decibels.

A D50 above 50% generally corresponds to good speech intelligibility. The metric is straightforward: the higher the percentage, the more dominant the early, useful sound energy is.

Interaural cross-correlation coefficient (IACC)

IACC measures how similar the sound signals are at your left and right ears. It ranges from 0 (completely different signals) to 1 (identical signals).

Lower IACC values mean the sound arriving at each ear is more dissimilar, which your brain interprets as greater spaciousness and envelopment. Concert halls with strong lateral reflections tend to have lower IACC values and are generally rated as sounding more immersive.

Lateral energy fraction (LF)

LF is the ratio of sound energy arriving from lateral (side) directions to the total sound energy, measured using a figure-of-eight microphone oriented to capture lateral sound.

Higher LF values indicate more sound arriving from the sides, which correlates with a stronger sense of spatial impression and envelopment. Well-regarded concert halls typically have LF values in the range of 0.1 to 0.35.

Strength (G)

Strength ($G$) compares the sound level at a measurement point in the room to the level the same source would produce at 10 m in a free field (anechoic conditions). It's expressed in dB.

Higher $G$ values indicate the room is amplifying the source relative to the free field, creating a sense of intimacy and power. Lower values suggest the room provides less acoustic support. $G$ is particularly important in concert hall design, where sufficient acoustic strength lets unamplified instruments project to every seat.

Room acoustics simulation

Simulation tools let you predict how a room will sound before it's built. By modeling the geometry, surface materials, and source/receiver positions, you can evaluate design options, optimize treatments, and avoid costly post-construction fixes.

Ray tracing vs image source methods

Ray tracing sends thousands of virtual sound rays from the source and tracks each ray as it reflects off surfaces, losing energy according to absorption coefficients at each bounce. It handles complex geometries well and naturally models late reflections and diffuse fields, but its accuracy depends on using enough rays.

Image source methods calculate the positions of virtual "mirror" sources created by each reflecting surface. For a first-order reflection off a wall, the image source is located behind that wall at the mirror position. This method gives exact results for early specular reflections in simple geometries but becomes computationally expensive as the reflection order increases.

Hybrid methods

Most modern simulation software (ODEON, CATT-Acoustic, etc.) uses hybrid methods that combine both approaches:

1. Image sources handle early reflections (typically the first 2–3 orders), where accuracy of individual reflection paths matters most.
2. Ray tracing or cone tracing handles late reflections and the diffuse tail, where statistical accuracy is more important than tracking individual paths.

This combination gives both the precision of image sources for early sound and the efficiency of ray tracing for the reverberant field.

Limitations and assumptions

Simulations are powerful but not perfect. Common limitations include:

• Geometric acoustics assumption: Both ray tracing and image source methods treat sound as rays, which breaks down at low frequencies where wavelengths are comparable to surface dimensions (diffraction effects are missed or approximated).
• Simplified material data: Simulations rely on absorption and scattering coefficients that may not fully capture real material behavior, especially at oblique angles of incidence.
• Diffuse field assumptions: Some algorithms assume more uniform conditions than actually exist.
• Air absorption: Often approximated rather than modeled in detail, which matters most at high frequencies in large rooms.

The accuracy of any simulation is only as good as the input data. Garbage in, garbage out.

Auralization techniques

Auralization makes simulation results audible. Instead of just reading numbers, you can listen to how the room will sound.

The process works by convolving (mathematically combining) a dry, anechoic recording with the simulated impulse response of the room. The result is an audio file that sounds as if the recording were played in the modeled space.

Advanced auralization uses binaural rendering (for headphone playback) or ambisonics (for loudspeaker arrays) to reproduce spatial cues, letting the listener perceive direction and envelopment. This is a valuable tool for communicating design intent to clients who can't interpret RT60 graphs but can immediately hear the difference between two design options.

Measurement techniques

Measurements verify how a real room actually performs. They're used to validate simulation predictions, diagnose acoustic problems, and confirm that treatments are working as intended.

Impulse response measurements

The impulse response is the room's acoustic fingerprint. It captures everything: direct sound arrival, early reflections, reverberant decay, and background noise.

To measure it, you excite the room with a known broadband signal and record the result. Common excitation signals include:

• Swept sine waves (also called exponential sine sweeps): sweep from low to high frequency over several seconds. They offer excellent signal-to-noise ratio and rejection of harmonic distortion.
• Maximum length sequences (MLS): pseudo-random binary signals (covered in more detail below).

The impulse response is extracted by deconvolving the recorded signal with the original excitation signal. From this single measurement, you can derive RT60, clarity, definition, and nearly every other room acoustics parameter.

ISO 3382 standards

ISO 3382 defines standardized procedures for measuring room acoustic parameters. It ensures that measurements taken by different people in different labs are comparable.

Key provisions include:

• Equipment requirements (source omnidirectionality, microphone specifications)
• Minimum number and placement of source and receiver positions
• Data analysis methods for deriving parameters like RT60, C80, D50, and others
• Separate parts for performance spaces (ISO 3382-1) and ordinary rooms (ISO 3382-2)

Following ISO 3382 is important for regulatory compliance and for producing results that other acousticians can trust and reproduce.

Omnidirectional vs directional sources

• Omnidirectional sources (typically dodecahedron loudspeakers with 12 drivers pointing in all directions) radiate sound equally in every direction. They're the standard for measuring spatially averaged parameters like RT60 and strength ($G$), because they excite the room uniformly.
• Directional sources (studio monitors, line arrays) have a focused radiation pattern. They're used when you need to measure direction-dependent parameters or simulate a realistic source like a talker or musical instrument.

The choice depends on what you're measuring and what ISO 3382 specifies for that parameter.

Schroeder integration

Schroeder integration (also called backward integration) is the standard method for extracting reverberation time from an impulse response.

The steps are:

1. Square the impulse response to get the energy-time curve.
2. Integrate backward from the end of the recording to each time point. This produces a smooth, monotonically decreasing decay curve.
3. Fit a straight line to the decay curve over the desired evaluation range.
4. Extrapolate to find the time for a 60 dB drop.

Common evaluation ranges:

• T20: line fitted from -5 dB to -25 dB, then extrapolated to 60 dB
• T30: line fitted from -5 dB to -35 dB, then extrapolated to 60 dB

Backward integration produces much smoother decay curves than trying to read the raw impulse response directly, and it reduces sensitivity to background noise at the tail end of the measurement.

Maximum length sequence (MLS) measurements

MLS is a pseudo-random binary sequence with special mathematical properties that make it useful for impulse response measurement.

Key characteristics:

• Flat frequency spectrum: MLS signals contain energy at all frequencies equally, ensuring broadband excitation.
• Low crest factor: the ratio of peak to RMS level is low, so you can drive the loudspeaker efficiently without clipping.
• Deterministic: the sequence is repeatable, allowing averaging of multiple measurements to improve signal-to-noise ratio.

The impulse response is obtained by cross-correlating the recorded room response with the original MLS signal. This process effectively "unscrambles" the room's contribution from the known excitation.

MLS measurements offer good noise immunity, but they have a notable weakness: they assume the system is time-invariant. If anything changes during the measurement (air movement, someone walking, clock drift between playback and recording hardware), the results degrade. Swept sine methods are generally more robust against these issues, which is why they've become the more popular choice in recent practice.