TLDR: What ISO 3382 Tells You and Why It Matters
ISO 3382 is the international standard for measuring room acoustic parameters. It defines how to measure reverberation time (RT60), early decay time (EDT), clarity (C50, C80), definition (D50), and lateral fraction (LF) using impulse response methods. If you design, build, or commission rooms where speech intelligibility or music quality matters — classrooms, concert halls, offices, hospitals — ISO 3382 is the measurement framework that determines whether your acoustic design actually works.
The standard comes in three parts: ISO 3382-1:2009 covers performance spaces (concert halls, theatres, opera houses), ISO 3382-2:2008 covers ordinary rooms (classrooms, offices, meeting rooms), and ISO 3382-3:2012 covers open-plan offices. Each part specifies different parameters, measurement procedures, and minimum position counts. Getting the right part matters because using concert hall methods in an open-plan office will give you meaningless numbers.
This guide walks through the standard part by part, explains each parameter with real field data, covers equipment requirements, and identifies the five mistakes that invalidate most acoustic measurements. It draws on a field case study where a £4M school hall in Manchester measured RT60 2.8 seconds — nearly three times the BB93 limit — because the architect specified decorative ceramic tiles with NRC 0.15 instead of perforated acoustic tiles.
The Problem: When Measurement Reveals Design Failure
In 2024, a £4 million multi-purpose school hall in Greater Manchester failed its acoustic commissioning test spectacularly. The architect had specified hand-made decorative ceramic tiles for the upper walls — a beautiful terracotta finish that won a regional design award. The tiles had an NRC of 0.15, meaning they reflected 85% of incident sound energy. Combined with a polished concrete floor (NRC 0.02) and full-height glazing on the south elevation (NRC 0.05), the hall had almost no absorption above the suspended acoustic ceiling, which covered only 60% of the ceiling area due to rooflight penetrations.
The post-completion ISO 3382-2 measurement revealed RT60 values of 2.8 seconds at 500 Hz and 3.4 seconds at 1 kHz — against a BB93:2015 target of 0.8–1.0 seconds for a school hall of that volume. STI measured 0.38 at the rear seats, categorised as "poor" per IEC 60268-16. Children in the back rows could not understand assembly speeches. The remediation cost £180,000 in retrofit acoustic panels and took the hall out of service for six weeks during term time.
The root cause was not ignorance of BB93 — the architect had checked the RT60 target. The failure was in material specification: the design team assumed "acoustic ceiling" was sufficient without calculating total absorption against room volume. A 10-minute calculation using the Sabine equation would have shown the shortfall before a single tile was ordered.
Part 1: Performance Spaces (ISO 3382-1:2009)
ISO 3382-1:2009 defines measurement procedures for performance spaces — rooms designed primarily for music, speech, or both. It covers concert halls, opera houses, theatres, recital halls, and multi-purpose auditoria. The standard specifies eight acoustic parameters derived from the room impulse response.
The Eight Parameters
Per ISO 3382-1:2009 §3, the standard defines the following measurable quantities:
| Parameter | Symbol | Unit | Clause | What It Measures |
|---|---|---|---|---|
| Reverberation time | T20, T30 | seconds | §3.2 | Time for 60 dB decay (extrapolated from 20 or 30 dB range) |
| Early decay time | EDT | seconds | §3.3 | Decay rate of first 10 dB — subjective reverberance |
| Clarity (music) | C80 | dB | §3.5 | Ratio of early (0–80 ms) to late energy — musical clarity |
| Clarity (speech) | C50 | dB | §3.6 | Ratio of early (0–50 ms) to late energy — speech clarity |
| Definition | D50 | ratio | §3.7 | Fraction of energy arriving within first 50 ms |
| Centre time | Ts | ms | §3.8 | Centre of gravity of impulse response — clarity balance |
| Lateral energy fraction | LF | ratio | §3.9 | Fraction of early lateral energy — spatial impression |
| Late lateral energy | LG | dB | §3.10 | Late lateral sound level — listener envelopment |
Source and Receiver Requirements
ISO 3382-1:2009 §5.1 requires an omnidirectional sound source with directivity deviation no greater than ±1 dB in any direction within the frequency range 100 Hz to 5 kHz. In practice, this means a dodecahedron loudspeaker. Balloon pops and starter pistols are acceptable for survey measurements but do not meet the directivity requirement for engineering-grade work.
Receivers must use pressure microphones for T20, T30, EDT, C50, C80, D50, and Ts measurements. LF and LG require a figure-of-eight microphone oriented with its null axis pointing at the source. Per §5.3, source height should be 1.5 m (typical seated musician head height) and receiver height 1.2 m (seated listener ear height).
Minimum Position Counts
The standard does not specify a fixed minimum but recommends enough source-receiver combinations to characterise spatial variation. In practice, 3 source positions (stage centre, stage left, stage right) and 8–12 receiver positions distributed across stalls, circle, and balcony provide meaningful spatial resolution. Each source-receiver combination produces one impulse response from which all eight parameters are extracted.
Part 2: Ordinary Rooms (ISO 3382-2:2008)
ISO 3382-2:2008 applies to the rooms most architects actually design: classrooms, meeting rooms, offices, hospital wards, corridors, and atria. It focuses on reverberation time as the primary parameter and defines two accuracy grades.
Engineering vs Survey Grade
Per ISO 3382-2:2008 §5, the standard defines two measurement grades:
| Requirement | Engineering Grade | Survey Grade |
|---|---|---|
| Source positions | ≥ 2 | ≥ 1 |
| Microphone positions | ≥ 3 per source | ≥ 3 |
| Source-receiver combos | ≥ 6 | ≥ 3 |
| Minimum S-R distance | > 2 × V^(1/3) / c × f_low^(-1) | Same |
| Frequency range | 100 Hz – 5 kHz | 250 Hz – 2 kHz |
| Expected precision (T30) | ±2.5% | ±10% |
Engineering-grade measurements are required for compliance verification against standards like BB93:2015, ANSI S12.60-2010, and DIN 18041:2016. Survey-grade measurements are acceptable for preliminary assessments and diagnostics.
The Sabine and Eyring Methods
ISO 3382-2:2008 Annex A describes two calculation methods for predicting RT60:
Sabine (§A.1): RT60 = 0.161 × V / A, where A = Σ(αᵢ × Sᵢ). Valid when average absorption coefficient ᾱ < 0.3 and absorption is relatively uniformly distributed. Overestimates RT60 in highly absorbent rooms.
Eyring (§A.2): RT60 = 0.161 × V / (−S × ln(1 − ᾱ)). More accurate when ᾱ > 0.3 or when absorption is concentrated on specific surfaces. Converges to Sabine for low absorption coefficients.
Ready to calculate? Use the AcousPlan RT60 calculator to run both Sabine and Eyring predictions for your room. Input your dimensions and surface materials — results in under 30 seconds.
Air Absorption Correction
Per ISO 3382-2:2008 §A.3, air absorption becomes significant above 2 kHz in rooms larger than approximately 200 m³. The correction adds 4mV to the denominator of the Sabine equation, where m is the energy attenuation coefficient of air (dependent on temperature and relative humidity per ISO 9613-1). At 20°C and 50% RH, m = 0.0069 Np/m at 4 kHz. Neglecting this correction in a 2,000 m³ auditorium can underpredict RT60 at 4 kHz by 15–20%.
Part 3: Open-Plan Offices (ISO 3382-3:2012)
ISO 3382-3:2012 addresses open-plan offices, which require fundamentally different parameters from enclosed rooms. In an open-plan space, the concern is not reverberation within a single volume but rather how sound propagates between workstations and how quickly speech becomes unintelligible with distance.
Key Parameters
| Parameter | Symbol | Unit | What It Measures |
|---|---|---|---|
| Spatial decay rate | D₂,S | dB per distance doubling | Rate of A-weighted SPL decrease |
| A-weighted SPL at 4 m | Lp,A,S,4m | dB(A) | Speech level at reference distance |
| Distraction distance | rD | m | Distance at which STI falls below 0.50 |
| Privacy distance | rP | m | Distance at which STI falls below 0.20 |
| Comfort distance | rC | m | Distance at which SPL < 45 dB(A) |
Per ISO 3382-3:2012 §4, a well-designed open-plan office should achieve D₂,S > 7 dB, Lp,A,S,4m < 48 dB(A), and rD < 5 m. These targets ensure that conversations beyond 5 metres are unintelligible (providing privacy) while maintaining acceptable ambient levels.
Measurement Method
Measurements are taken along a straight line of workstations using a loudspeaker at desk height (1.2 m) reproducing a standardised speech spectrum. SPL is measured at multiple distances from 1 m to the maximum extent of the open-plan area. The spatial decay curve is fitted to these measurements using least-squares regression.
Equipment Requirements and Calibration
Regardless of which part of ISO 3382 you are using, equipment calibration is non-negotiable. Per ISO 3382-1:2009 §5.4 and ISO 3382-2:2008 §5.5:
- Sound level meters must conform to IEC 61672-1 Class 1
- Microphones must have traceable calibration within the past 12 months
- The sound source must be verified for omnidirectional output per ISO 3382-1 Annex A
- Temperature and humidity must be recorded (they affect air absorption and sound speed)
- Background noise must be measured and recorded — at least 10 dB below the decay range used (e.g., for T30, background noise must be at least 35 dB below the initial level)
Equipment List for ISO 3382-2 Field Measurement
| Item | Specification | Approximate Cost |
|---|---|---|
| Dodecahedron speaker | 12-driver omnidirectional, 100 Hz – 10 kHz | £3,500–£8,000 |
| Power amplifier | 200 W minimum, low noise floor | £500–£1,500 |
| Measurement microphone | ½" free-field, IEC 61672-1 Class 1 | £800–£2,000 |
| Audio interface | 2-channel minimum, 48 kHz, 24-bit | £300–£800 |
| Calibrator | 94 dB / 1 kHz pistonphone, IEC 60942 Class 1 | £400–£900 |
| Software | Room acoustic analysis (DIRAC, EASERA, REW) | £0–£3,000 |
| Tripods | Heavy-duty, vibration-isolated | £200–£400 |
| Environmental meter | Temperature, humidity, barometric pressure | £100–£300 |
Total investment for a professional measurement kit: £5,800–£16,900 depending on brand and features.
Interpreting Results: What the Numbers Mean
Raw measurement data requires interpretation. A single RT60 number is meaningless without context — the room type, volume, and applicable standard determine whether a result is acceptable.
RT60 Reference Values by Room Type
| Room Type | Target RT60 (500–1000 Hz) | Standard |
|---|---|---|
| Classroom (< 280 m³) | 0.4–0.6 s | ANSI S12.60, BB93 |
| Meeting room (< 100 m³) | 0.4–0.7 s | DIN 18041, BS 8233 |
| Lecture hall (200–800 m³) | 0.7–1.0 s | ISO 3382-2 informative |
| Concert hall (10,000–25,000 m³) | 1.8–2.2 s | ISO 3382-1 informative |
| Recording studio (control room) | 0.2–0.4 s | EBU Tech 3276 |
| Open-plan office | 0.5–0.8 s | ISO 3382-3, BS 8233 |
| Hospital ward | 0.5–1.0 s | HTM 08-01 |
Frequency-Dependent Analysis
RT60 must be analysed across octave bands, not as a single broadband number. A room can have acceptable RT60 at 500 Hz but excessive reverberation at 125 Hz (common with lightweight suspended ceilings that absorb mid-frequencies but are transparent to bass) or at 4 kHz (rare but indicates unusual geometry or flutter echo paths).
The Manchester school hall case illustrates this: RT60 at 125 Hz was 1.9 seconds (manageable), but at 1 kHz it was 3.4 seconds because the ceramic tiles were highly reflective at mid-high frequencies while the concrete slab provided some bass absorption. The frequency imbalance made speech sound muddy and boomy simultaneously.
Common Mistakes That Invalidate ISO 3382 Measurements
1. Insufficient Source-Receiver Distance
Placing the microphone too close to the source captures direct sound that dominates the impulse response, compressing the measured decay curve and underestimating RT60. ISO 3382-2 §5.2 specifies a minimum distance based on room volume. For a 200 m³ classroom, this is approximately 1.5 m. For a 15,000 m³ concert hall, it is approximately 8 m.
2. Not Checking Background Noise
If background noise (HVAC, traffic, equipment) is within 10 dB of the noise floor during decay measurement, the T30 extrapolation will be contaminated. Always measure background noise with the source off and verify at least 35 dB of dynamic range between initial level and background at each octave band.
3. Measuring Furnished vs Unfurnished
BB93:2015 requires measurement of the furnished, unoccupied room. DIN 18041:2016 specifies the furnished condition. Measuring an unfurnished room during construction and comparing to furnished targets is meaningless — furniture and soft furnishings can reduce RT60 by 20–40% depending on room volume and furnishing density.
4. Ignoring Temperature and Humidity
Air absorption at 4 kHz varies by a factor of 3 between 20% and 80% relative humidity. A measurement taken in a cold, dry building during winter will show significantly different high-frequency RT60 than one taken in summer with HVAC running. Record conditions and apply corrections per ISO 9613-1.
5. Single-Number Reporting
Reporting "RT60 = 1.2 seconds" without specifying the frequency bands, measurement grade, number of positions, room conditions, and applicable standard is professionally inadequate. A compliant measurement report per ISO 3382-2 §6 must include all of these elements plus the measurement uncertainty.
Summary
ISO 3382 is the foundational measurement standard for room acoustics. Part 1 covers performance spaces with eight parameters including EDT, clarity, and lateral energy. Part 2 covers ordinary rooms with engineering and survey measurement grades. Part 3 addresses open-plan offices with spatial decay and privacy distance metrics. The standard requires calibrated equipment, sufficient source-receiver positions, and proper environmental documentation. The most common failures in practice are insufficient measurement positions, uncontrolled background noise, and single-number reporting without frequency-band detail.
Whether you are commissioning a school hall, verifying a concert hall design, or diagnosing complaints in an open-plan office, ISO 3382 provides the measurement framework — but only if applied correctly. The Manchester school hall case demonstrates what happens when design decisions are made without measurement-informed calculation: a £4M building that fails its acoustic brief and costs an additional £180,000 to remediate.
Model your room before you build it. Use the AcousPlan room acoustic calculator to predict RT60, EDT, and C80 for your design — and catch problems before they become £180,000 mistakes.