Standing Waves in Concert Hall Design

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

In early 2024, Sonus Gear Flow was brought into the renovation of the Riverside Civic Concert Hall, a 720-seat, mid-century venue in Madison, Wisconsin. The city wanted the room to serve two jobs equally well: unamplified orchestral programs and amplified touring acts (jazz, folk, small-format rock) without a “one-size-fits-none” compromise. The hall had a loyal audience and a busy calendar, but complaints had become consistent: “boomy” low end, uneven bass between rows, and dead spots onstage that made musicians overplay.

The project team included the city’s capital projects office (owner), Keller & Mott Architects (architect of record), Hearthstone Acoustics (acoustical consultant), and our group handling electroacoustic design and measurement-driven commissioning. The motivation was not cosmetic. A prior sound system refresh had improved clarity at mid/high frequencies, but low-frequency issues persisted—especially during spoken-word events and when amplified bass instruments were involved. The suspicion was modal behavior: standing waves driven by room geometry and stage-house coupling.

Our goal was to deliver a documented, repeatable solution: quantify the standing-wave problem, integrate architectural and acoustic treatments that wouldn’t compromise sightlines or historic finishes, and commission the room so the results were verifiable—not just “better than before.”

2) Challenges and requirements at the outset

The venue is a classic shoebox with a deep balcony: approximately 32.4 m long (back wall to stage shell), 18.6 m wide, and an average ceiling height of 12.1 m (higher over the main floor, lower under the balcony soffit). Stage depth is 12 m with a proscenium opening 14 m wide. The floor is raked concrete with hardwood over sleepers. Side walls are plaster on masonry; the rear wall is mostly flat plaster with decorative pilasters.

The operating constraints were typical of a civic hall but tight in practice:

Schedule: 14-week construction window, with the hall dark only 9 continuous weeks. Remaining work had to be nights and Mondays.
Budget: $620k total for acoustics + audio infrastructure. The city earmarked $210k specifically for room acoustic upgrades.
Historic appearance: no “studio look.” Visible treatments had to match plaster and wood finishes.
Flexibility: the hall alternates between orchestral (no PA) and amplified concerts (PA) weekly.
Measurement proof: success criteria had to be measurable: smoother LF response across seats and fewer modal peaks with shorter decay.

Early data from the venue’s in-house team (basic RTA snapshots) suggested strong, narrow peaks around 47 Hz, 63 Hz, and 94 Hz depending on location. The most problematic symptom was spatial variance: at mix position the 63 Hz band was heavy, while in the first 10 rows it nearly vanished. That’s textbook standing-wave behavior—especially axial modes along length/width and coupled with a reflective back wall.

3) Approach and methodology chosen

We treated the project like a forensic exercise first, then a design exercise. The methodology had four pillars:

Baseline mapping: high-resolution measurements of frequency response and decay time across many seats, not just a single “money seat.”
Modal analysis + model validation: compute expected axial/tangential modes from room dimensions and validate them with measured peaks/decays.
Architectural + acoustic interventions: prioritize geometry and broadband control before relying on DSP band-aids.
Commissioning plan: define target metrics in advance and measure again after each major change.

The tools were standard but used rigorously. We measured with Room EQ Wizard and Smaart v9, using a Rational Acoustics RTA-420 and a Earthworks M30 for verification. Playback was through a temporary ground-stacked system (two QSC KS118 subs and two QSC K12.2 tops) driven by a Q-SYS Core 110f so we could do repeatable sweeps and log DSP settings. For impulse-response work we used a NTi XL2 as a secondary recorder to cross-check time alignment.

4) Step-by-step execution narrative

Week 1–2: Baseline measurement and symptom confirmation

We started with the room empty (no audience), HVAC on (as in show conditions), and stage curtains in their typical positions. We ran sweeps from a consistent source location: downstage center, 1.5 m above the deck. Measurements were taken at 26 listener positions covering stalls, under-balcony, balcony front, and balcony rear. At each seat cluster we captured magnitude, phase, waterfall, and RT60 (noting that RT60 below 125 Hz is less reliable but still indicative when compared consistently).

The key findings:

Length mode cluster: strong behavior around ~53 Hz (calculated axial length mode for ~32.4 m is ~5.3 Hz for fundamental; higher orders land in the 40–70 Hz range). The measured peak most consistently appeared at ~47–55 Hz depending on seat depth.
Width mode: a persistent peak at ~63 Hz matched width-related modes for ~18.6 m.
Under-balcony resonance: a pronounced bump and long decay at ~94–105 Hz, tied to the soffit cavity and local boundary conditions.
Decay: at several positions, LF decay at 63 Hz exceeded 1.6 s while 500 Hz sat around 1.2 s, giving a perceived “hangover” in the bass.

The data aligned with audience complaints: some seats had too much 60 Hz energy with long decay, while others sat in modal nulls where bass fundamentals disappeared.

Week 3: Design workshop and constraint alignment

We held a design workshop with the architect and acoustician to decide what was feasible. Two non-negotiables emerged: no large visible bass traps in the audience chamber, and no reduction of seating count. That meant we needed to use hidden or integrated low-frequency absorption and diffusion/geometry adjustments that could be blended into existing surfaces.

Week 4–6: Prototype testing of low-frequency control

Rather than committing immediately to custom tuned devices, we prototyped. We built two temporary absorber assemblies:

Tuned membrane panel (1.2 m x 2.4 m), 18 mm MDF diaphragm over a 200 mm cavity with mineral wool fill, aiming for a center frequency near 63 Hz.
Broadband soffit absorber concept for under-balcony: 100 mm mineral wool with a 50 mm airgap behind perforated plywood (micro-perf pattern selected for aesthetic match).

We positioned the membrane prototypes at the rear wall corners (high-pressure zones for width/length modes) and re-measured. Even with only two prototypes, we observed a 2–3 dB reduction at the worst 63 Hz peaks and modest decay improvements. That validated the strategy: tuned/pressure-based absorption at boundaries would help without changing the room’s visible character.

Week 7–10: Construction and integration

Final construction included:

Rear wall intervention: a new decorative rear-wall build-out, 250 mm deep in sections between pilasters, filled with mineral wool and faced with slotted wood panels. Behind the slots, we incorporated membrane elements tuned across 55–75 Hz in different bays to broaden the effective bandwidth.
Corner LF absorbers: two hidden corner cavities (stage-left rear and stage-right rear) converted into pressure traps with limp-mass membranes. Each cavity provided roughly 3.5 m³ of trapped volume.
Under-balcony treatment: replacing a reflective gypsum soffit finish with perforated wood panels over 100 mm absorption, targeting the 100 Hz area and cleaning up the “boxy” coloration under the balcony.
Stage shell adjustments: the existing portable shell had parallel returns that reinforced flutter and modal buildup on stage. We modified two shell panels with shallow convex curvature (subtle, visually acceptable) and added absorptive backing in the shell storage alcove to reduce stage-house coupling.

Week 11–12: System alignment and operational setup

Although the project focus was room behavior, we coordinated with the venue’s PA. The hall used a left/right system with flown point-source cabinets and a cardioid sub array. The previous tuning tried to “EQ out” room peaks; it worked at FOH but made other seats worse.

Post-treatment, we retuned with a different philosophy: minimal corrective EQ, more emphasis on time alignment and coverage consistency. Subwoofer array spacing and delay were adjusted to reduce rear-wall excitation. We settled on a 2.6 ms delay offset between sub elements to maintain forward energy while softening coupling into the back wall.

Week 13–14: Post-work verification

We repeated the same 26-position measurement grid with identical source and level. We also ran a “real program” check using multitrack playback: upright bass, kick drum, male speech, and full-range orchestral content. Finally, we did a listening walk with the house engineer and a visiting FOH engineer to confirm the measured improvements translated into operational reality.

5) Technical decisions and trade-offs made

Several decisions were debated, and the outcomes are relevant to any hall dealing with standing waves:

Tuned vs. broadband LF absorption: Pure broadband traps big enough for 60 Hz were not architecturally possible. We chose a hybrid approach: tuned membranes at pressure maxima (rear wall, corners) plus moderate broadband absorption under the balcony for higher LF/low-mid control.
Diffusion vs. absorption on the rear wall: Diffusion can help mid/high frequency envelopment, but it does little for narrow LF modes. We integrated slotted faces that provide some scattering while the cavity behind does the LF work.
DSP correction limitations: We deliberately avoided aggressive narrowband cuts in the system EQ. Cutting a 63 Hz peak at FOH can deepen a null elsewhere. The room had to be fixed physically first.
Preserving RT for orchestral use: The acoustician’s target midband RT was 1.4–1.6 s occupied. We had to reduce LF decay without “drying out” the room. Treatments were concentrated at boundaries and under-balcony regions rather than across large wall areas.

6) Results and outcomes with specific details

The improvements were measurable and operationally meaningful:

Seat-to-seat LF variance reduced: At 63 Hz, the spread across the 26 positions tightened from roughly ±8–10 dB variation to ±4–6 dB. It wasn’t perfectly uniform (no real hall is), but the worst nulls became less punishing.
Decay improved at problem modes: Average decay around 63 Hz dropped from about 1.5–1.7 s to 1.1–1.3 s in the main floor. Under the balcony at ~100 Hz, decay dropped by roughly 0.3–0.4 s.
Rear-wall reflection control: The back rows previously suffered from a low-frequency “push” and smeared clarity. Post-treatment, impulse responses showed reduced late energy, and speech intelligibility improved subjectively without over-EQ.
Operational gain before feedback: For spoken word using a handheld cardioid dynamic microphone, the in-house engineer reported ~3 dB more usable gain before low-mid buildup became objectionable, primarily because the room stopped amplifying certain notes.
Mix translation improved: Visiting engineers noted they no longer had to overcompensate for bass at FOH to make it audible in the front third of the room—an outcome consistent with reduced null severity.

Timeline-wise, the acoustic interventions took 6 weeks of the 14-week schedule including fabrication and installation. Measurement and commissioning accounted for 9 on-site days spread across the project. The room acoustic scope stayed within its $210k target by using integrated carpentry rather than bespoke visible products; the largest cost drivers were the rear-wall build-out and under-balcony soffit replacement.

7) Lessons learned and what could be done differently

Three lessons stood out:

Start measurement earlier than feels necessary: Our baseline mapping took time, but it prevented expensive guesswork. If we had joined the project during schematic design (instead of after bid set), we could have influenced geometry earlier and reduced the amount of tuned treatment required.
Prototype before committing: The temporary membrane panels were not glamorous, but they de-risked the approach. Without that proof, it would have been tempting to over-invest in diffusion or cosmetic updates that wouldn’t touch the core LF problem.
Under-balcony zones behave like separate rooms: Treating the under-balcony soffit as an acoustic system—not just a finish surface—was crucial. If we could redo one element, we would add a bit more depth (another 50–75 mm) in select under-balcony cavities to push absorption effectiveness slightly lower in frequency.

We also learned a project-management lesson: coordinated trades matter. The rear-wall cavity depended on airtight construction for predictable membrane behavior. Small air leaks around access panels measurably reduced effectiveness until sealed. On the next project, we would specify leak-check steps and require photos of membrane assemblies before closure.

8) Takeaways applicable to other projects

Standing waves are a geometry + boundary problem first. DSP can refine, but it can’t fix spatial variance caused by modes.
Measure in a grid, not a point. If you only tune to FOH, you’ll optimize one seat and potentially worsen others. A 20–30 position map is not overkill for a hall of this size.
Use pressure zones intelligently. Corners and rear-wall boundaries are where low-frequency pressure builds. That’s where tuned absorption buys the most improvement per cubic meter.
Hybrid solutions work best under real constraints. A mix of tuned membranes (for 50–80 Hz), broadband absorption (for 80–200 Hz), and subtle geometry adjustments often beats any single strategy.
Define success metrics upfront. Commit to targets like reduced seat-to-seat variance and shorter decay at modal frequencies. It keeps stakeholders aligned and protects the project from “it feels better” ambiguity.
Commission in phases. Measure after each major intervention. If a rear-wall solution moves a problem frequency rather than reduces it, you want to know before finishes are locked in.

The Riverside hall renovation reinforced a simple truth: standing waves are not an abstract textbook issue in performance spaces. They show up as inconsistent bass, unclear speech, and engineers fighting a mix that won’t translate. With disciplined measurements, targeted low-frequency absorption integrated into architecture, and restrained DSP, the room became more predictable—making every event easier to run and more consistent for the audience.