Sound Reflection in Concert Hall Design

By Sarah Okonkwo · April 29, 2026

1. Project overview: what, where, who, and why

In 2024, our team was brought into the late design phase of the Ravenport Civic Arts Hall, a new 1,180-seat concert venue in a mid-sized U.S. city. The building was already under construction: a traditional “shoebox” room (approx. 44 m long × 22 m wide × 15 m high) with a stage designed to host chamber ensembles through a full symphony orchestra. The client was the city’s performing arts authority, with an operator who had managed a multi-use theater but not a dedicated acoustic hall.

The core project goal was simple to state and harder to execute: deliver consistent acoustic clarity and warmth for unamplified music, while still supporting speech intelligibility and occasional amplified events without resorting to heavy electronic enhancement.

Our team included an acoustician (lead), a systems engineer (PA/assisted listening), the architect of record, a theater consultant, and the general contractor. I served as the audio project documentarian and technical liaison between the acoustics team, the AV integrator, and the construction trades.

2. Challenges and requirements at the outset

We entered the project during interior framing, which meant several key surfaces were already locked in: balcony fronts, concrete structural walls behind finish assemblies, and mechanical chases. Early room models used optimistic reflection assumptions, but field conditions were trending toward over-absorption due to value-engineered finishes. The largest risks were reflection-related:

Weak early lateral reflections in the rear orchestra seating, leading to reduced envelopment and perceived intimacy.
Specular “hot spots” under the first balcony due to a flat, low soffit design.
Stage-to-audience coupling at risk because the initial stage shell drawings had shallow side returns and a high percentage of absorptive masking curtains.
Flutter echo potential between long, parallel side walls if the finish remained smooth gypsum.
Conflicting requirements: the operator wanted a “warm” hall for orchestra and also demanded “no echo” for lectures. This often triggers design decisions that over-dampen the room.

Measurable requirements were agreed upon in a kickoff meeting with the owner and operator:

Reverberation time (RT60): 1.8–2.0 s mid-band (500 Hz–1 kHz) for unoccupied hall; target occupied RT60 ~1.6–1.8 s.
Clarity (C80): -1 to +3 dB in most seats for orchestral use (avoid overly “dry” values).
Speech intelligibility (STI): ≥ 0.55 with sound reinforcement engaged, with HVAC in “performance mode.”
Noise criteria: NC-15 in the hall, NC-20 on stage.

The uncomfortable truth: we had about 14 weeks until substantial completion of the room finishes, and major geometry changes would be expensive. The work needed to focus on reflection behavior using tools that were compatible with construction realities.

3. Approach and methodology chosen

We used a hybrid method: predictive modeling to evaluate reflection paths and overall decay, then mockups and in-situ measurements to validate the specific surfaces that were driving audible issues.

Tools and workflow:

3D acoustic model in ODEON (v16) for ray-tracing and reflection analysis, using updated geometry from the BIM model (Revit export).
Material absorption and scattering values sourced from manufacturer data where possible; conservative assumptions where not available, especially for wood slat assemblies and seating.
On-site impulse response measurements as soon as the room was enclosed (before seats), using a dodecahedron loudspeaker and measurement mic positions mapped to seating zones.
Construction mockups for two key reflective elements: a diffusive sidewall panel and an under-balcony reflector treatment.

We defined decision gates aligned with construction milestones: (1) finalize sidewall finish strategy, (2) finalize stage shell reflection plan, (3) commit under-balcony treatment, (4) confirm seat absorption assumptions and finalize RT targets.

4. Step-by-step execution narrative

Week 1–2: Reality-checking the model. We started by updating the acoustic model to match as-built framing. The early design had a more articulated sidewall, but field conditions showed long, mostly flat runs. The balcony soffit was flatter than the drawings suggested, with a larger under-balcony footprint. We ran baseline predictions and saw two immediate red flags: (a) reduced lateral energy in the rear third of the orchestra level, and (b) a cluster of strong early reflections under the balcony landing in a narrow band of seats around rows H–K.

Week 3: First measurement window (shell closed, no seats). Once the roof deck and most perimeter walls were sealed, we brought in a measurement kit: a NTi DS3 dodecahedron source (battery powered), an RME interface feeding the source, and matched measurement microphones (Earthworks M30 and iSEMcon EMX-7150) into Room EQ Wizard for quick checks, then processed the data in EASERA for detailed metrics. We ran sine sweeps and extracted T20/T30, EDT, C80, and early reflection timing at 18 mic positions. Without seats, the mid-band RT was around 2.3 s, which was acceptable as a pre-seat number, but the EDT was uneven under the balcony and early reflection peaks were noticeably concentrated.

Week 4–6: Sidewall strategy—diffusion without killing energy. The architect’s finish package had shifted toward fabric-wrapped absorptive panels for aesthetics. We pushed back with data: adding too much absorption on the side walls would improve speech but reduce orchestral support and lateral energy. We landed on a hybrid sidewall:

Lower side walls (up to 3.2 m): hardwood slat system over an air cavity with intermittent absorption behind slats (approx. 20% open area). This maintained reflection strength but controlled mid/high buildup.
Upper side walls: diffusive wood paneling using varying-depth wells (30–90 mm) arranged in pseudo-random sequences to increase scattering in the 700 Hz–4 kHz range.

The fabrication detail mattered: slat thickness (18 mm), spacing (12 mm), cavity depth (75 mm), and where the absorber was placed (we specified 25 mm mineral wool only behind every third bay). This prevented the finish from becoming a hidden broadband absorber.

Week 7–9: Stage shell reflections—making the orchestra talk to the room. The stage shell was originally designed as a set of flat panels with heavy velour legs. The first orchestra rehearsal would be dead-on-arrival if early reflections back to the players were weak. We revised the shell concept without changing the rigging points:

Added angled side returns (7–10° splay) to direct energy both to the audience and back across the ensemble.
Specified a three-piece overhead cloud using 50 mm thick laminated plywood panels, each cloud tilted 4–6° to avoid pinging a single seating area.
Reduced the default masking: the legs remained available, but the “standard orchestral mode” called for minimal curtain coverage.

We coordinated with rigging and lighting because the clouds could not interfere with line sets and lighting focus. The final cloud elevation was set at 10.8 m above stage, preserving sightlines and keeping early reflections within the desired timing window (roughly 20–80 ms for supportive energy depending on seat position).

Week 10–11: Under-balcony correction—breaking the hot spot. The under-balcony issue was not simply “too reflective” or “too absorptive.” It was too specular in the wrong places. A flat gypsum soffit was producing localized reflection peaks. We avoided adding large absorptive areas (which would create a dull under-balcony compared to the rest of the hall). Instead, we designed a pattern of shallow convex reflectors:

Module size: 600 mm × 1200 mm
Maximum curvature depth: 40 mm
Material: painted MDF over ribs, mechanically fastened for service access
Layout: alternating orientation to increase scattering

A small portion near HVAC diffusers received discreet absorption (25 mm glass fiber behind micro-perf) to address localized high-frequency hiss without turning the entire zone into an acoustic sink.

Week 12–13: Seating and final tuning assumptions. Seating is where reflection and decay predictions often fail. The selected seat had moderate absorption occupied and relatively low absorption unoccupied, which is good for consistent RT between rehearsal and show. We requested manufacturer absorption data and validated with conservative inputs. We also ensured the seat backs had perforation and internal damping to avoid “drumhead” resonances.

Week 14: Commissioning measurements and operator training. After seats were installed, we re-measured at 24 positions with the same source level and mic mapping. We also ran a short operator workshop: how to deploy stage curtains, how to choose shell configurations, and how those choices change reflections and clarity more than any EQ ever will.

5. Technical decisions and trade-offs made

Diffusion vs absorption on side walls. The operator’s instinct was to add fabric absorption to “fix echo.” We demonstrated that flutter echo is solved more effectively with non-parallelism and diffusion than with blanket absorption, especially in a concert hall. The trade-off: diffusion panels cost more and required tighter carpentry tolerances. We limited the diffusive treatment to where it mattered most: upper side walls and mid-hall zones.

Under-balcony design: scatter, don’t soak. Absorbing the balcony soffit would have lowered RT under the balcony but created tonal mismatch with the rest of the hall. Scattering addressed the reflection concentration while keeping energy.

Stage shell cloud thickness and weight. Thicker panels offer better low-mid reflection but add rigging load. We held to a panel mass that the existing rigging could support with a safety factor, accepting that the cloud’s useful reflection would be strongest above ~200–250 Hz rather than trying to brute-force 125 Hz support.

Multi-use requirement handled with variable acoustics. Rather than tuning the hall to be “speech perfect” all the time, we used variable elements: curtains and stage configuration. This avoided designing a permanently dry hall that would disappoint acoustic music.

6. Results and outcomes with specific details

Post-installation measurements (occupied simulation with seat absorption, HVAC in performance mode) showed the hall landed close to target:

RT60 (500 Hz–1 kHz): 1.78 s average across main floor and balcony; under-balcony zone averaged 1.70 s.
EDT: improved uniformity; under-balcony EDT dropped from a pre-treatment equivalent of ~2.1 s to ~1.75 s, aligning better with the room.
C80 (orchestral configuration): ranged from -0.5 to +2.8 dB across measured seats; previously modeled hot spots under the balcony exceeded +4 dB in a narrow band, which was reduced through soffit scattering.
STI with reinforcement (L/R point-source arrays aimed for speech events): 0.60–0.66 in most seats at nominal level, with HVAC at NC-15.

Subjectively, the first dress rehearsal with a 62-piece orchestra revealed two wins directly tied to reflection decisions: musicians reported better ability to hear across sections (stage shell side returns and cloud), and the rear orchestra seating gained a sense of “wrap” without adding any electronic enhancement (upper sidewall diffusion).

Timeline and budget impact were tracked carefully. The under-balcony reflector modules added roughly $68,000 in fabrication and installation and required 12 calendar days in the finish schedule (overlapping with other trades). The sidewall changes were a larger scope shift but were offset by reducing other decorative elements; net impact to the interior package was approximately +1.4% of the room finish budget.

7. Lessons learned and what could be done differently

Lock reflection-critical geometry earlier. The balcony soffit shape and sidewall articulation were high-leverage items that should have been protected from late value engineering. We recovered with scattering, but the cheapest fix would have been shaping the soffit from day one.

Don’t rely on “RT targets” alone. The early design meetings focused heavily on RT60, but the most audible problems were about early reflection timing and directionality. Modeling and measurement that include C80/EDT and reflection mapping prevented us from “fixing” the wrong thing.

Mockups save arguments. The sidewall and under-balcony mockups allowed the architect and owner to approve a finish that was both attractive and acoustically functional. If we had built those mockups a month earlier, we could have reduced redesign churn.

Operator training is part of acoustics. Variable acoustics only works if staff understand it. After training, the operator adopted a default orchestral setup (minimal curtains, clouds deployed) and a speech setup (additional curtains, modest PA level). That operational discipline protected the hall’s musical performance.

8. Takeaways applicable to other projects

Design reflections intentionally, not incidentally. Early lateral reflections and controlled scattering often matter more than another 0.1 s of RT adjustment.
Under-balcony zones need geometry or diffusion. Flat soffits create specular reflection problems; scattering treatments can fix hot spots without making the area acoustically “dead.”
Stage shell configuration is a system. Side returns, overhead clouds, and curtain policy must work together; a reflective cloud doesn’t help if the stage is wrapped in absorption.
Use measurements to validate assumptions before finishes lock. A single measurement window when the room is enclosed (even before seats) can reveal reflection issues that models miss due to real-world construction details.
Protect key acoustic surfaces from value engineering. If budget cuts are inevitable, prioritize preserving reflection geometry and scattering. Decorative changes elsewhere are often less risky than altering sidewall behavior.
Document trade-offs with numbers. Project managers respond to quantified impacts: schedule days, cost deltas, and measurable acoustic outcomes. That documentation speeds decisions and reduces late-stage rework.

In this project, sound reflection wasn’t an abstract theory—it was a series of construction decisions: angles, panel depths, spacing, and where not to place absorption. The hall’s final performance came from treating reflection as a controllable design resource, validating it with measurements, and ensuring the operator could maintain those conditions event after event.