How sound waves, glass composition, interlayers, and frames conspire to create — or destroy — acoustic comfort in buildings.
Sound is mechanical energy — a pressure disturbance that propagates as a longitudinal wave through a medium. Unlike light, sound cannot travel through a vacuum. It needs matter: air, water, glass, concrete. When a source vibrates — a truck engine, a conversation, a subwoofer — it cyclically compresses and rarefies the surrounding air molecules, creating regions of high and low pressure that travel outward at roughly 343 m/s at room temperature.
For a single homogeneous panel, sound reduction improves by approximately 6 dB for every doubling of mass, and by 6 dB for every octave increase in frequency. This is the Mass Law: TL ≈ 20·log₁₀(m·f) − 47, where m = surface mass (kg/m²) and f = frequency (Hz). Doubling glass thickness from 4 mm to 8 mm adds ~6 dB — meaningful, but expensive. Structural decoupling (DGU) and damping (lamination) defeat the mass law's limitations at high and low frequencies respectively.
Three properties characterise any sound wave relevant to window design: frequency (pitch — how many oscillations per second), amplitude (volume — how much pressure), and phase (the timing relationship between two waves). Noise control engineers exploit all three. Frequency determines which reduction mechanism works best; amplitude tells you how much reduction is needed; phase matters in special cases like active noise cancellation.
"Every 10 dB of sound reduction halves the perceived loudness — meaning a window that cuts 30 dB makes external noise seem eight times quieter to human listeners."
— Fundamental acoustic perception principle, ISO 532BAdjust frequency and amplitude to see how sound wave characteristics change. Note how low-frequency waves have longer wavelengths — these are the hardest to attenuate with glass.
Orange = incoming wave · Teal = transmitted wave (after glass reduction)
When an airborne sound wave strikes a glass pane, energy is partitioned into three components: a fraction is reflected back into the source room, a fraction is absorbed within the glass itself (converted to heat through damping), and the remainder is transmitted through to the other side. The acoustic performance of a window is a measure of how little energy reaches the transmission side.
Every glass pane has a critical frequency (the coincidence dip) at which airborne sound waves perfectly excite bending waves in the glass, causing transmission loss to drop sharply. For a standard 6 mm monolithic pane, this critical frequency falls around 1,800–2,000 Hz — squarely in the middle of the speech intelligibility range. At this frequency, the glass resonates like a drumhead, dramatically amplifying transmission.
The viscoelastic interlayer (PVB or SGP) introduces structural damping that dissipates the energy of bending waves before they can radiate sound on the transmission side. This is why an 8.76 mm laminate (4+0.76PVB+4) can outperform a monolithic 10 mm pane at mid-frequencies by 5–8 dB, despite being lighter. The interlayer converts mechanical vibration energy into heat, effectively "smearing" the coincidence dip.
Sound doesn't only travel through glass. It also flanks through window frames via structural vibration. A high-performance acoustic glazing unit set in a poor frame loses up to 8–12 dB of effective performance. This is why true acoustic window systems specify thermally broken aluminium, uPVC multi-chamber profiles, or timber frames with elastomeric seals — all chosen for their mechanical decoupling properties, not just thermal performance.
Five principal glazing strategies are used in acoustic design, each with distinct physical mechanisms, cost profiles, and performance characteristics.
STC (Sound Transmission Class) is the primary North American metric. OITC (Outdoor-Indoor Transmission Class) weights low frequencies more heavily, making it more relevant for traffic and aircraft noise. European standard Rw serves a similar purpose. The values below are representative of tested assemblies; installation quality affects real-world performance significantly.
| Glass Config. | Thickness | Weight kg/m² | STC | Rw (Europe) | Coincidence Freq. | Verdict |
|---|---|---|---|---|---|---|
| SGU Float | 4 mm | 10.0 | 25 | 24 | ~3,200 Hz | Minimal |
| SGU Float | 6 mm | 15.0 | 27–28 | 27 | ~2,000 Hz | Basic |
| SGU Float | 8 mm | 20.0 | 29–30 | 29 | ~1,600 Hz | Low |
| SGU Float | 10 mm | 25.0 | 30–31 | 30 | ~1,300 Hz | Low |
| SGU Float | 12 mm | 30.0 | 32 | 31 | ~1,100 Hz | Moderate |
| SGU Float | 15 mm | 37.5 | 33 | 33 | ~900 Hz | Moderate |
| SGU Float | 19 mm | 47.5 | 34 | 34 | ~750 Hz | Moderate |
Note: STC gains diminish rapidly beyond 10 mm for monolithic glass. The coincidence dip drops into lower frequencies with thicker glass, creating new problems even as mass law improves overall performance.
| Configuration | Total Width | Gap Gas | STC | Rw | OITC | Best For |
|---|---|---|---|---|---|---|
| 4/12/4 | 20 mm | Air | 33 | 32 | 29 | Budget IGU |
| 6/12/6 (symmetric) | 24 mm | Air | 34 | 33 | 30 | Standard |
| 6/16/6 Argon | 28 mm | Argon | 35–36 | 35 | 31 | Improved |
| 6/20/6 Argon | 32 mm | Argon | 37 | 36 | 32 | Good |
| 4/16/6 Asymmetric | 26 mm | Air | 38 | 37 | 33 | Recommended |
| 6/16/8 Asymmetric Ar | 30 mm | Argon | 39 | 38 | 34 | Good Acoustic |
| 6/20Ar/10 Asymmetric | 36 mm | Argon | 40 | 39 | 35 | Excellent |
Using two panes of different thickness staggers their coincidence frequencies, preventing both dips from falling at the same frequency. A 4 mm pane (coincidence ~3,200 Hz) paired with a 6 mm pane (coincidence ~2,000 Hz) means the dips don't reinforce each other. This asymmetric approach can add 3–5 dB of real-world performance at no extra material cost — only smarter selection. Standard practice in acoustic engineering since the 1970s (Fahy, 1985).
| Configuration | Total Thickness | Interlayer | STC | Rw | Loss Factor η | Assessment |
|---|---|---|---|---|---|---|
| 4+0.38PVB+4 | 8.38 mm | PVB std | 34 | 33 | 0.008 | Min. benefit |
| 4+0.76PVB+4 | 8.76 mm | PVB std | 36 | 35 | 0.010 | Entry Acoustic |
| 6+0.76PVB+6 | 12.76 mm | PVB std | 38 | 37 | 0.010 | Standard |
| 6+1.52PVB+6 | 13.52 mm | PVB acoustic | 40 | 39 | 0.015 | Good |
| 8+2.28PVB+8 | 18.28 mm | PVB acoustic | 42 | 41 | 0.015 | Very Good |
| 6+0.76SGP+6 | 12.76 mm | SGP (SentryGlas) | 39 | 38 | 0.025 | Better damping |
| 6+2.28SGP+6 | 14.28 mm | SGP thick | 42–43 | 42 | 0.028 | Excellent |
| 8+2.28SGP+8 | 18.28 mm | SGP thick | 45–46 | 45 | 0.030 | Outstanding |
| 10+2.28SGP+10 | 22.28 mm | SGP thick | 48 | 47 | 0.030 | Premium |
SentryGlas® Plus (SGP), developed by DuPont (now Kuraray), has a shear modulus roughly 100× greater than standard PVB at room temperature. Research by Belis et al. (2011) and Haldimann et al. (2008) confirms SGP achieves superior high-frequency damping. However, standard PVB actually outperforms SGP at very low frequencies (below 200 Hz) because its lower stiffness allows better viscoelastic dissipation. Acoustic PVB formulations (e.g. Saflex® Acoustic) close much of this gap at mid-to-high frequencies while retaining flexibility.
| System | Build-up | Total Width | STC | Rw | OITC | Cost Index |
|---|---|---|---|---|---|---|
| Standard DGU | 6/16Ar/6 | 28 mm | 35 | 35 | 31 | 1× |
| Asymmetric DGU | 4/16Ar/6 | 26 mm | 38 | 37 | 33 | 1.1× |
| Acoustic DGU (PVB inner) | 6+1.52PVB+6 / 16Ar / 6 | 45 mm | 44 | 43 | 38 | 2.2× |
| Acoustic DGU (SGP inner) | 6+2.28SGP+6 / 20Ar / 6 | 50 mm | 47 | 46 | 41 | 3.0× |
| True Triple Glaze | 6/16Ar/6/16Ar/6 | 66 mm | 42 | 41 | 36 | 2.8× |
| Triple Acoustic | 4+PVB+6 / 20Ar / 6+PVB+6 | 74 mm | 50 | 49 | 44 | 4.5× |
| Secondary Glazing | Existing + 100mm gap + 6mm | 100+ mm | 45–55 | 44–54 | 40–50 | 1.5–2× |
Secondary glazing (an independent inner window) exploits a large air cavity. At 100–200 mm separation, the air gap resonance frequency falls below most traffic noise, delivering exceptional performance at moderate cost — often better value than complex IGU systems.
Overall STC and Rw ratings can be misleading because they compress frequency-specific performance into a single number. A window rated STC 40 might perform beautifully against speech but fail badly against low-frequency traffic rumble. Understanding which glass strategy targets which frequency band is essential for specification.
| Frequency Band | Noise Source | SGU 6mm | DGU 6/16/6 | Lam PVB 6+1.52+6 | Lam SGP 6+2.28+6 | Best Strategy |
|---|---|---|---|---|---|---|
| 63 Hz | HVAC, bass music | 15 dB | 18 dB | 20 dB | 22 dB | Mass only helps |
| 125 Hz | Traffic, LFN | 20 dB | 24 dB | 26 dB | 28 dB | Large gap DGU |
| 250 Hz | Trucks, trains | 25 dB | 30 dB | 33 dB | 35 dB | DGU + lam |
| 500 Hz | Voices, engines | 28 dB | 34 dB | 38 dB | 40 dB | Laminated |
| 1,000 Hz | Speech, horns | 30 dB | 36 dB | 40 dB | 42 dB | Laminated |
| 2,000 Hz | Speech sibilants | 22 dB* | 35 dB | 40 dB | 43 dB | SGP lam critical |
| 4,000 Hz | Aircraft, sirens | 36 dB | 40 dB | 44 dB | 46 dB | Any system fine |
*SGU 6mm at 2,000 Hz shows the coincidence dip — transmission loss collapses by ~8 dB. Values are approximate, based on published test data (Scholl, 2001; Hopkins, 2007; Pilkington Acoustic Guide, 2020).
"No single glazing system performs equally across all frequencies. Acoustic specification is always a frequency-weighted compromise — the art is matching the glazing strategy to the dominant noise spectrum of the site."
— Hopkins, C., Sound Insulation, Elsevier, 2007A glazing unit performs only as well as its weakest acoustic link. The frame, seals, and installation quality collectively determine whether lab-tested STC values are achieved in practice. Research by Quirt & Nightingale (2008) found that flanking transmission through frames accounted for up to 10 dB of performance loss in installed assemblies compared to laboratory tests.
Even a perfect glass-frame combination loses its acoustic rating through gaps. A gap of just 1 mm can reduce effective STC by 5–15 dB depending on its position and the frequencies involved. This is why acoustic window specifications always include: continuous compressible perimeter seals (EPDM or silicone), acoustic-grade backer rods with sealant at frame-to-wall junctions, and compressed seals or mechanical gaskets on openable sections. In controlled studies, identical window units installed with and without proper acoustic sealing have differed by up to 12 dB in effective performance.
Window acoustics research spans building physics, materials science, and psychoacoustics. Below are select findings from published literature that directly inform window specification practice.
Selecting the right acoustic glazing begins with characterising the noise source. Different sources have different spectral profiles, and each glazing type has different frequency-specific strengths.
| Noise Source | Dominant Freq. | Target Rw | Recommended Glass | Frame Requirement | Priority |
|---|---|---|---|---|---|
| Quiet residential street | 500–2k Hz | 30–35 | DGU 6/16Ar/6 | Standard uPVC | Low |
| Urban traffic (moderate) | 125–500 Hz | 35–40 | Asym. DGU or Lam PVB DGU | Thermally broken Al | Medium |
| Heavy urban / ring road | 80–500 Hz | 40–45 | Acoustic DGU (6+1.52PVB+6/20Ar/6) | uPVC or timber, acoustic seals | High |
| Railway / metro | 63–2k Hz wide | 42–48 | Lam SGP inner + DGU / Triple Acoustic | Timber or composite, full acoustic seal | Critical |
| Airport flight path | 250–4k Hz | 45–52 | Triple Acoustic IGU or Secondary Glazing | Specialist acoustic frame system | Critical |
| Industrial / HVAC | 63–250 Hz LF | 38–45 | Heavy lam (SGP) + large gap DGU | Heavyweight frame, mass loaded | Critical LF |