Church Sound Transformation: RT60 2.8s → 1.2s

The Challenge: A Century of Echoes

St. Michael's Church, a 500-seat stone church built in the 1920s, had served its congregation faithfully for over a century. But as the worship style evolved from traditional liturgy to contemporary services with spoken word, music, and multimedia presentations, the acoustics became a serious problem. Congregants in the rear pews consistently complained that they could not understand the pastor. The worship team struggled with muddy sound despite investing in a modern PA system. Video recordings were unusable due to excessive reverberation.

The building presented a classic acoustic challenge: a large volume of 4,200 cubic metres with stone walls, a timber-vaulted ceiling at 12 metres height, and stone tile flooring. Every surface was hard, reflective, and historically significant, meaning any acoustic intervention had to be visually sensitive.

Measurement: What SonaVyx Revealed

Using the SonaVyx RT60 measurement tool, the engineering team captured impulse responses at 8 positions throughout the nave and transepts. The log sine sweep method provided clean IR data with excellent signal-to-noise ratio despite the ambient noise from HVAC and street traffic.

The results told a stark story:

• RT60 at 500 Hz: 2.8 seconds (target for speech: 0.8 to 1.4 seconds)
• RT60 at 1 kHz: 2.5 seconds
• RT60 at 250 Hz: 3.1 seconds (even longer at low frequencies)
• STI measured via SonaVyx STIPA: 0.42 (poor — below the 0.50 minimum for acceptable speech)
• Average Leq during a typical service: 78 dBA (the Lombard effect — people raising their voices)

The Schroeder backward integration curves showed a clean, linear decay with no anomalies, confirming that the reverberation was diffuse and evenly distributed. The impulse response showed strong early reflections from the side walls arriving within the first 20 ms.

Diagnosis: Root Cause Analysis

The SonaVyx AI diagnostic engine identified three contributing factors:

1. Excessive room volume: At 4,200 m³, the Sabine equation predicted RT60 of 2.9s with the existing surface absorption — closely matching the measured value.
2. Uniformly reflective surfaces: Stone walls with an absorption coefficient of just 0.02, stone floor at 0.01, and the timber ceiling at 0.10 provided a total absorption of only 95 Sabins. The room needed approximately 220 Sabins to achieve a 1.2-second RT60.
3. No absorptive treatment: Unlike modern worship spaces that incorporate carpet, cushioned seating, and acoustic panels, this historic building had none.

The treatment calculator quantified the absorption deficit per octave band and recommended panel locations that would maximise acoustic impact while minimising visual intrusion.

Solution: Strategic Acoustic Treatment

Working within heritage conservation guidelines, the team installed:

• 80 m² of fabric-wrapped acoustic panels (NRC 0.85) on the upper portions of both side walls, above the window line, using colours sympathetic to the stone
• 40 m² of suspended ceiling clouds positioned over the seating area at approximately 8 metres height, well below the historic vaulted ceiling
• All panels were mechanically fixed to allow future removal without damaging the heritage fabric

The placement was guided by the SonaVyx room scan tool, which identified the primary reflection paths from the stone side walls and the ceiling zones contributing most to late reverberation. The treatment targeted early reflections on the side walls and diffuse field energy via the ceiling clouds.

Results: Before and After

Post-treatment measurements using the SonaVyx before/after comparison confirmed:

Metric	Before	After	Improvement
RT60 @ 500 Hz	2.8s	1.2s	57% reduction
RT60 @ 1 kHz	2.5s	1.0s	60% reduction
STI (average)	0.42 (poor)	0.71 (good)	+0.29 points
Service Leq	78 dBA	72 dBA	6 dB quieter

The congregation immediately noticed the difference. Feedback from the first service after installation was overwhelmingly positive, with elderly members reporting they could finally understand the pastor without hearing aids. The music team found they could reduce PA levels because the direct-to-reverberant ratio had improved dramatically.

Lessons Learned

• Measure before specifying: The per-octave-band RT60 data prevented over-treatment at high frequencies where the ceiling already provided some absorption.
• Target the reflection paths: Side wall panels at ear height would have been less effective than the above-window placement that intercepted the strongest reflection angles.
• The Lombard effect matters: The 6 dBA reduction in service levels was an unexpected bonus — less reverberation meant less voice-raising, creating a virtuous cycle of reduced noise.
• Heritage sensitivity is achievable: Well-chosen panel colours and positions allowed a 57% RT60 reduction without compromising the building's historic character.