The Acoustic Paradox at the Heart of Church Design
Walk into Salisbury Cathedral. Stand under the crossing tower and clap your hands once. The sound will continue for 6–8 seconds, building and cascading through the stone vaulting in a wash of overlapping reflections. It is, by any sensory measure, extraordinary.
Now imagine trying to understand a sermon delivered in that space without electronic amplification. You cannot. The words become an unintelligible smear, each syllable obliterated by the reverberant tail of the previous one. For conversational speech at approximately 5 syllables per second, a room with RT60 above 2.5–3 seconds cannot support intelligible unamplified speech. A room with RT60 of 6–8 seconds cannot support intelligible amplified speech either — because the amplification just makes the reverberant field louder.
This is the central paradox of church acoustics. The acoustic properties that create the overwhelming sensory experience of a great cathedral — long reverberation, envelopment by reflections, the sense of sound emerging from and dissolving into the architectural fabric — are precisely the properties that make congregational speech communication impossible.
Every church building project at some scale faces this tension. This article will explain why, give you the numbers to work with, describe what design strategies actually resolve it, and show you how to calculate whether your design achieves something people can both sing and understand in.
What RT60 Means in Worship Acoustics
RT60 is the time in seconds for sound to decay by 60 dB after the source stops. In a worship space, this number controls three things simultaneously: how speech intelligibility is perceived, how music sounds, and how the space "feels" acoustically.
For speech, the critical relationship is between RT60 and the rate of articulation. Normal conversational speech delivers approximately 4–5 syllables per second. Each syllable lasts approximately 150–200 ms. For each syllable to be perceptible as a discrete acoustic event — a precondition for intelligibility — its reverberant decay must fall to a negligible level before the next syllable arrives.
The approximate threshold is this: if RT60 exceeds approximately 2 seconds in a small room, or approximately 3 seconds in a large room, the decay of each syllable is still clearly audible when the next syllable begins. Words blur into each other. Consonants — which carry most of the phonemic information and are inherently lower in energy than vowels — become inaudible. The speech transmission index (STI) drops below 0.45, the threshold that is typically described as "poor" intelligibility.
For music, the RT60 requirements are fundamentally different and, for some musical traditions, directly opposed to the speech requirements.
Pipe organ: The organ is designed specifically for reverberant acoustics. The legato style of organ playing — where notes are held and overlap — is designed to fill the reverberant space with a sustained harmonic texture. A pipe organ in a dry room with RT60 of 1.0 seconds sounds thin, disconnected, and flat. The same organ in a space with RT60 of 3.0–4.0 seconds sounds rich, majestic, and deeply satisfying. Organs in English cathedrals were designed and voiced for spaces with RT60 of 5–8 seconds.
Choral music: Congregational and choral singing both benefit from acoustic support — a room that gives back energy to the singers, helping them hear each other and encouraging blend. This requires RT60 above approximately 1.5 seconds for small choirs and above 2.0 seconds for large cathedral choirs. Rooms with RT60 below 1.0 seconds feel dry and effortful for singers; the congregation hears isolated voices rather than a unified choir.
Contemporary worship music with amplification: Electronic keyboards, electric guitars, drum kits, and in-ear monitor systems function best in dry acoustic environments. Excessive reverberation smears the rhythmic definition of contemporary music, blurs chord changes, and makes the sound system mix impossible to control. Contemporary worship bands typically prefer RT60 of 0.8–1.2 seconds.
The worship space serving multiple musical traditions is therefore asked to serve three fundamentally incompatible acoustic requirements simultaneously. This is not a problem that acoustic design alone can fully resolve — it requires either the worship community to make a choice about which acoustic priority comes first, or the installation of electroacoustic systems and adjustable acoustic elements that allow the space to function differently for different uses.
RT60 Targets by Tradition and Volume
Based on published research and measured data from high-performing worship spaces internationally, here are empirically derived RT60 targets:
| Tradition | Volume (m³) | Target RT60 at 500 Hz | Acceptable STI Range |
|---|---|---|---|
| Evangelical / Pentecostal | 1,000–5,000 | 1.0–1.4 s | 0.65–0.80 |
| Mainline Protestant (word-focused) | 1,000–8,000 | 1.2–1.8 s | 0.60–0.75 |
| Contemporary multi-use | 2,000–10,000 | 1.0–1.6 s | 0.60–0.75 |
| Roman Catholic (modern) | 2,000–15,000 | 1.5–2.2 s | 0.50–0.65 |
| Anglican cathedral / choir | 5,000–30,000 | 2.0–3.5 s | 0.40–0.60 |
| Historic stone cathedral | 20,000–80,000+ | 4.0–8.0 s | 0.25–0.45 |
| Small chapel | 200–800 | 0.8–1.2 s | 0.65–0.80 |
The STI ranges reflect what is achievable with sound reinforcement in a space with the specified RT60. High-reverberation spaces can achieve reasonable STI only with very tightly controlled distributed speaker systems — delay-line systems, column arrays, or individually aimed directional speakers that deliver high direct-to-reverberant ratio at the listener position.
The Sabine Calculation for Worship Spaces: A Worked Example
Let us work through a realistic medium-sized church: a new evangelical congregation building with target RT60 = 1.2 seconds.
Room dimensions: 18 m long × 14 m wide × 10 m high at ridge (approximated as 8 m average height) Volume: 18 × 14 × 8 = 2,016 m³ Seating: 400 persons, upholstered pews Surface areas: Floor = 252 m², Ceiling = 252 m², Walls = approximately 480 m² total
Target absorption at 500 Hz (Sabine equation, RT60 = 1.2 s): A = 0.161 × V / RT60 = 0.161 × 2016 / 1.2 = 270.7 m² sabin
Baseline absorption (unoccupied):
| Surface | Area (m²) | Material | α (500 Hz) | Sabin |
|---|---|---|---|---|
| Ceiling | 252 | Timber boarding (10 mm, cavity) | 0.12 | 30.2 |
| Floor | 252 | Carpet on concrete | 0.25 | 63.0 |
| Walls | 360 | Brick/masonry (painted) | 0.03 | 10.8 |
| Windows | 80 | 6 mm float glass | 0.04 | 3.2 |
| Doors | 12 | Solid timber | 0.10 | 1.2 |
| Timber furnishings | 35 | Timber pews (unoccupied) | 0.07 | 2.5 |
| Total (unoccupied) | 110.9 |
Unoccupied RT60: 0.161 × 2016 / 110.9 = 2.93 seconds — far too long.
Adding 400 seated persons on upholstered pews: Absorption per person at 500 Hz: approximately 0.45 m² sabin (ISO 354 / Beranek reference data) Total audience absorption: 400 × 0.45 = 180 m² sabin
Occupied total absorption: 110.9 + 180 = 290.9 m² sabin Occupied RT60: 0.161 × 2016 / 290.9 = 1.12 seconds — within target!
But now the problem is visible: the empty RT60 is 2.93 seconds and the full RT60 is 1.12 seconds. This is a 2.6:1 ratio. An empty rehearsal feels completely different from a full service. The sound engineer who sets up the system during a Thursday evening rehearsal to an empty room is calibrating for a completely different acoustic than the Sunday morning service will provide.
Adding treatment to reduce empty RT60:
To achieve RT60 ≤ 1.8 seconds empty (a more manageable variation), we need total unoccupied absorption of: A_required = 0.161 × 2016 / 1.8 = 180.5 m² sabin Additional absorption needed: 180.5 − 110.9 = 69.6 m² sabin
This could be provided by:
- 55 m² of fabric-wrapped fibreglass panels (100 mm thick, NRC 0.95), α = 0.95 at 500 Hz → 52.3 m² sabin
- 25 m² of heavy draped fabric (velour, 3× fullness) at rear wall, α = 0.55 → 13.8 m² sabin
- Total added: 66.1 m² sabin — close to target
The key insight from this calculation: fabric treatment added to reduce empty reverberation shifts the occupied RT60 downward by the same absolute amount. In a room where the audience is the primary acoustic absorber, you cannot tune the empty condition without also changing the occupied condition. Plan for both.
Low-Frequency Behaviour: Where Stone Churches Go Wrong
The most common acoustic problem in heritage church buildings is not excessive mid-frequency reverberation — it is the frequency-dependent nature of that reverberation, particularly the bass reverberation.
Stone walls, concrete block, and brick masonry have extremely low mid and high-frequency absorption (α = 0.02–0.05 at 500–2000 Hz) but moderate absorption in the 125–250 Hz range due to vibrational resonance of the thick masonry. This creates a characteristic frequency response in many stone churches: RT60 peaks at 250–500 Hz and is somewhat lower at 125 Hz and much lower at 2000–4000 Hz.
This is actually preferable to the alternative found in many post-war concrete-block churches with thin added absorptive treatment: good mid-frequency absorption from acoustic ceiling tiles, very little low-frequency absorption (tiles are ineffective below 250 Hz), resulting in a distinctive booming bass reverberation at 125 Hz that is typically 2–3 times the 500 Hz value.
Here is measured RT60 data from three actual worship spaces to illustrate the problem:
| Space | 125 Hz | 250 Hz | 500 Hz | 1000 Hz | 2000 Hz | Character |
|---|---|---|---|---|---|---|
| Gothic stone cathedral, 45,000 m³ | 6.8 s | 7.2 s | 7.4 s | 7.1 s | 6.2 s | Flat decay, magnificent |
| 1960s concrete block church, 3,000 m³ | 3.1 s | 1.8 s | 1.2 s | 0.9 s | 0.7 s | Boomy bass, sharp decay |
| New timber frame chapel, 800 m³ | 0.9 s | 1.1 s | 1.3 s | 1.1 s | 0.9 s | Warm, speech-intelligible |
The 1960s concrete block church is the problem case. The bass reverberation at 125 Hz (3.1 seconds) is nearly three times the mid-frequency target (1.2 seconds). Adding more ceiling tiles will reduce 500 Hz reverberation further, making the frequency imbalance worse. The only way to reduce 125 Hz reverberation is with thick, low-density absorbers or resonant panel systems tuned to the bass frequencies.
Low-frequency absorbers for worship spaces:
- Fabric-wrapped 100 mm fibreglass panels (minimum 48 kg/m³ density), mounted 150–200 mm from a rigid wall: α ≈ 0.55 at 125 Hz, 0.90+ at 500 Hz
- 200 mm fibreglass panels or rockwool blocks (same density), directly against wall: α ≈ 0.80 at 125 Hz
- Membrane (panel) absorbers — thin plywood or hardboard mounted over air cavity — tuned to 80–150 Hz: α ≈ 0.50–0.70 at resonant frequency
- Helmholtz resonators (perforated concrete block or purpose-built units): narrowband absorption tuned to 63–125 Hz
At α = 0.70 for 200 mm panels mounted off a wall, you need approximately 160 m² of panel area. In a church with a total wall area of perhaps 800 m², that is 20% wall coverage — achievable as draped fabric or decorative banners without fundamentally altering the character of the space.
Sound Reinforcement in High-RT60 Spaces
For any worship space with RT60 above approximately 2 seconds, achieving intelligible amplified speech requires careful sound system design. The objective is to maintain a high direct-to-reverberant sound ratio (D/R ratio) at listener positions, which means delivering high direct sound level from sources very close to listeners, rather than high total sound level from distant sources.
Distributed delay speaker systems: Rather than two large column speakers at the front of the church (which throw sound energy into the reverberant field to reach listeners at the rear), a distributed system uses many small speakers suspended from the ceiling at 3–5 m height throughout the space, each serving only the listeners directly beneath it. Each speaker operates at low level, creating high local direct-to-reverberant ratio. A digital delay system aligns the timing of each speaker cluster so that the direct sound from the nearest speaker arrives at each listener's ear first, maintaining correct temporal precedence (the Haas effect).
The direct-to-reverberant improvement from a distributed system over a conventional system can be 8–15 dB, which can improve STI from 0.35 (poor) to 0.55–0.65 (fair to good) in a space with RT60 of 3–4 seconds. This is the technology that makes Anglican choral services in medieval cathedrals comprehensible with modern clergy speaking.
Cardioid subwoofer arrays: Low-frequency energy from contemporary worship music is omnidirectional by nature. In a highly reverberant space, bass energy builds up in the reverberant field and creates the mud that makes contemporary music in reverberant churches sound indistinct. Cardioid subwoofer configurations — where multiple subwoofers are arranged with phase delay to create directional low-frequency output — reduce the bass energy injected into the reverberant field by up to 15 dB at the rear hemisphere, concentrating bass energy toward the audience and reducing the reverberant bass buildup.
The Adjustable Acoustic Approach
For worship buildings that serve genuinely different acoustic needs — a contemporary evangelical congregation that rents a historic stone building, a cathedral that hosts both Evensong and contemporary worship concerts — the adjustable acoustic approach provides the best long-term flexibility.
Retractable curtains and banners: Heavy draped fabric (velour or similar, 400–500 g/m², 3× fullness) deployed in alcoves and apse spaces can add 100–300 m² sabin of absorption at 500 Hz when deployed, reducing RT60 by 0.5–1.5 seconds. When retracted, the original acoustic is restored. This is reversible, relatively inexpensive ($80–$150 per m² installed), and can be designed as historically appropriate fabric hangings in heritage buildings.
Movable upholstered seating: Replacing fixed pews with upholstered stacking chairs is a significant acoustic variable. Chairs in use provide approximately 0.40–0.50 m² sabin per seat at 500 Hz when occupied. Empty upholstered chairs provide approximately 0.22–0.30 m² sabin — approximately half the occupied value, which is much better than empty wooden pews at 0.02–0.06 m². For rooms that are often empty for rehearsals, the acoustic benefit of retaining upholstered seats rather than removing chairs is substantial.
Variable acoustic panels: Mechanically rotated panels — absorptive on one face, reflective on the other — can change the RT60 of a space by 0.5–1.5 seconds depending on coverage area. These are sophisticated and expensive ($500–$1,500 per m² for commercial systems), and require careful calibration, but they are used in purpose-built multi-use auditoria and in a few high-budget church projects.
Using the RT60 Calculator for Worship Spaces
The AcousPlan RT60 calculator lets you model worship spaces with any combination of surface materials and audience absorption. For worship space design, the key inputs are:
- Enter room dimensions and select the dominant surface material for each face
- Add audience absorption as a separate surface (select "Audience — upholstered seating" material at 400 Hz centre and enter the seating area, not the number of persons)
- Calculate both occupied and unoccupied conditions to understand the variation range
- Check the frequency-band display to identify bass reverberation problems at 125 Hz
The core principle in worship space acoustics is not "more reverberant = more sacred" or "less reverberant = more intelligible." It is matching the reverberation characteristic of the space to the acoustic requirements of the specific worship tradition, and designing sound reinforcement systems that compensate for the gap between what the architecture wants to do acoustically and what the congregation needs to hear.
The cathedral is right for what it was designed for. The question is whether the new community centre church being designed this year is right for the people who will use it every week.