A one-inch hole drilled into a wall designed for STC 56 dropped its field rating to FSTC 41 near the hole, a loss of as much as 10 points, while the same wall measured FSTC 51 away from it (Acoustics Today, Fall 2016). That gap between the number on the drawing and the number in the apartment is the whole story of partition acoustics. The 2024 International Building Code asks for STC 50 in the lab but accepts NNIC 45 in the field (ICC, IBC 2024 Section 1206), and New York City logged 463,349 residential noise complaints through 311 in 2025 (NYC Open Data). Survey data from Denmark adds the human side: 36% of residents in multi-storey housing are bothered by neighbour noise (Rasmussen and Ekholm, Inter-Noise 2021). We aggregated data from the International Code Council, ASTM International, the National Research Council Canada, Aalborg University, NYC Open Data, the New York State Comptroller and a dozen peer-reviewed and conference papers; for the materials-science side of the problem, see our acoustic metamaterials and soundproofing statistics.
TL;DR
- The IBC requires STC 50 between dwelling units in the lab, or NNIC 45 if field tested (ICC, IBC 2024 Section 1206).
- Canada’s 2015 code moved to ASTC 47 between units, a rating that counts flanking paths (Hoeller et al., Construction Canada 2018).
- A single 2x4 wood stud wall with one layer of Type X per side rated STC 37; a double stud wall with the same board rated STC 61 (Byrick, Canadian Acoustics 2011).
- Adding cavity absorption raised gypsum wall STC by about 10 points in NRC’s 250+ assembly program (NRC Construction Technology Update No. 1, 1997).
- A floor rated STC-60 measured ASTC-49 in the field, an 11-point shortfall (HGC Engineering, Canadian Acoustics).
- Back-to-back metal outlet boxes cut a wall from STC 55 to STC 51, and poorly located boxes cost up to 6 points (Nightingale and Quirt, Canadian Acoustics 1995).
- A gypsum layer placed inside a double stud cavity cut its rating from STC 64 to as low as STC 57 (Mahn, Skoda and Cunha, NRC, Canadian Acoustics 2024).
- NYC 311 residential noise complaints rose 22.2%, from 379,296 in 2024 to 463,349 in 2025 (NYC Open Data).
- Residents very annoyed by neighbour noise had 2.49 times the unadjusted odds of not getting enough sleep (95% CI 1.79-3.46) (Rasmussen and Ekholm, Inter-Noise 2021).
- NRC’s survey of 600 residents found annoyance approached zero only near STC 65 (Bradley, Canadian Acoustics 1999).
- England’s Robust Details scheme lifted compliance from 35-50% to 97% within 18 months (Smith, Acoustics’08 Paris).
1. Code Minimums: What the Rules Actually Require
North American codes still lean on a single lab number. The IBC’s airborne minimum is STC 50 in the lab, relaxed to 45 in the field, which builds a 5-point allowance (50 minus 45) for the real world directly into the code (ICC, IBC 2024 Section 1206). The same section requires penetrations for piping, electrical devices, recessed cabinets, bathtubs, soffits and ducts to be “sealed, lined, insulated or otherwise treated to maintain the required ratings”, which is where many field failures start (see section 4).
Canada took a different route. Its code rose from STC 45 to STC 50 in the 1990 National Building Code (NRC Construction Technology Update No. 1, 1997), then in 2015 shifted the target to ASTC 47 for the whole building configuration, flanking included (Hoeller, Mahn and Mueller-Trapet, Construction Canada 2018). Europe is stricter still: the same NRC authors note most Scandinavian countries require the equivalent of ASTC 55. The bar a North American apartment has to clear is lower than in much of Europe.
| Metric | Value | Source |
|---|---|---|
| IBC lab minimum, walls between dwelling or sleeping units | STC 50 (ASTM E90) | ICC, IBC 2024 Section 1206.2 |
| IBC field-test alternative | NNIC 45 (ASTM E336) | ICC, IBC 2024 Section 1206.2 |
| IBC floor-ceiling impact minimum | IIC 50 lab, NISR 45 field | ICC, IBC 2024 Section 1206.3 |
| Canada NBC minimum, raised in 1990 | STC 45 to STC 50 | NRC Construction Technology Update No. 1, 1997 |
| Canada NBC 2015 minimum between dwelling units | ASTC 47 (includes flanking) | Hoeller et al., Construction Canada 2018 |
| Scandinavian requirement, ASTC equivalent | ASTC 55 (other European countries at least ASTC 50) | Hoeller et al., Construction Canada 2018 |
| European countries with airborne requirements (of 35 surveyed) | 29 | Rasmussen, Forum Acusticum 2011 |
| Strictest and most lenient European airborne requirements | Scotland DnT,w 56 dB; England and Wales DnT,w + Ctr 45 dB | Rasmussen, Forum Acusticum 2011 |
Context: Rasmussen counted 8 distinct airborne descriptors in European regulations, and stresses there is no general conversion between them, so these values are not directly comparable to STC (Rasmussen, Forum Acusticum 2011). Most recent available data: Rasmussen 2011. In the US, ASTM E336-25 only allows ATL and ASTC to be measured in spaces of at least 25 m3, which can exclude small bedrooms from the most rigorous field metric.
2. Lab STC of Common Wall Assemblies
The lab data are consistent on one point: decoupling beats mass. Doubling the drywall on a single stud wall added only 4 dB, while splitting it into a double stud wall added 24 dB (Byrick, Pliteq, Canadian Acoustics 2011). A standard single stud wall sits well below the code line, and only isolated assemblies (resilient channel, double studs) clear STC 50 with room to spare for field losses. Isolation is also fragile: the same paper reports that a resilient channel short-circuited by screws into the studs can lose up to 10 STC points.
NRC Canada’s consortium program tested more than 250 gypsum board assemblies and found that a cavity absorber adds about 10 STC points, while doubling cavity depth adds about 10 points with absorber but only about 4 without (NRC Construction Technology Update No. 1, 1997). Its regression model across 360 walls explained STC with r² = 0.903 and a standard error of 2.0 dB (Warnock and Quirt, Canadian Acoustics 1995), which means lab STC is highly predictable. The field is where it stops being predictable.
| Metric | Value | Source |
|---|---|---|
| Single 2x4 wood stud, 1 layer 5/8 in. Type X per side (2 layers) | STC 37 (STC 41) | Byrick, Canadian Acoustics 2011 |
| Single stud with resilient channel, 1 layer (2 layers) | STC 50 (STC 60) | Byrick, Canadian Acoustics 2011 |
| Staggered stud on 2x6 plate, 1 layer (2 layers) | STC 41 (STC 47) | Byrick, Canadian Acoustics 2011 |
| Double 2x4 stud, 25 mm gap, 1 layer (2 layers) | STC 61 (STC 65) | Byrick, Canadian Acoustics 2011 |
| Double stud, two layers each side (12.7 mm Type C / 15.8 mm Type X) | STC 64 / STC 66 | Mahn, Skoda and Cunha, NRC, Canadian Acoustics 2024 |
| Staggered 2x4 studs on an 8-inch nominal plate | STC 60 | WoodWorks, Thorburn 2021 |
| Effect of adding cavity absorption | About +10 STC | NRC Construction Technology Update No. 1, 1997 |
| Stud spacing 400 mm to 600 mm | +1 to 2 STC | NRC Construction Technology Update No. 1, 1997 |
Context note: the two staggered stud values (STC 47 and STC 60) are not in conflict; the WoodWorks assembly uses a wider 8-inch plate and a different layering, which shows how much detailing matters within one “assembly type” (WoodWorks, Acoustical Considerations for Mixed-Use Wood-Frame Buildings). Most recent available data for the like-for-like comparison: Byrick 2011.
3. The Lab-to-Field Gap and Flanking
A lab test isolates one partition between two heavy, decoupled rooms. A building connects that partition to continuous floors, ceilings and side walls. In one wood-frame project measured by HGC Engineering, a floor assembly rated STC-60 measured ASTC-49 above a small room and ASTC-53 above a larger one, shortfalls of 11 points (60 minus 49) and 7 points (60 minus 53), with interior walls attached to the joists accounting for a 4-point difference between the two rooms (Edwards, Niinivaara and Lorimer, HGC Engineering, Canadian Acoustics).
Walls behave the same way. The same consultancy measured as-built triple-cavity shear walls averaging ASTC-49, below their STC-50 target, even though the unsheathed version of the assembly is listed at STC-57 in the National Building Code tables; adding resilient channel on both faces brought three final measurements to an average of ASTC-53, ranging from ASTC-49 to ASTC-55 (Lorimer, Niinivaara and Edwards, HGC Engineering, Canadian Acoustics). NRC researchers note that in lightweight wood- or steel-framed buildings with continuous floor joists, floor flanking often dominates the ASTC rating, so upgrading the wall alone does not help (Hoeller et al., Construction Canada 2018). Mass timber faces the same junction problem, covered in our CLT and mass timber flanking statistics. The same structure-borne mechanism drives bass and mechanical noise, covered in our subwoofer and HVAC low-frequency noise statistics.
| Metric | Value | Source |
|---|---|---|
| Field allowance built into the IBC | 5 points (STC 50 lab vs NNIC 45 field) | ICC, IBC 2024 Section 1206.2 |
| Floor rated STC-60, measured in two rooms | ASTC-49 and ASTC-53 | Edwards et al., HGC Engineering, Canadian Acoustics |
| Penalty from interior walls tied into the floor | 4 points | Edwards et al., HGC Engineering, Canadian Acoustics |
| As-built triple-cavity shear walls, unmitigated | Average ASTC-49 vs STC-50 target | Lorimer et al., HGC Engineering, Canadian Acoustics |
| Same walls after resilient channel on both faces (3 tests) | Average ASTC-53, range 49 to 55 | Lorimer et al., HGC Engineering, Canadian Acoustics |
| Brazilian ceramic block walls, field DnT,w (3 buildings) | 42, 40 and 38 dB vs 45, 45 and 40 dB required | Buildings (MDPI) 2024 |
| Italian existing partition, measured vs required | R’w 28 dB vs more than 50 dB | Buildings (MDPI) 2025 |
Outlier note: all three Brazilian buildings failed their national NBR 15575 thresholds by 2 to 5 dB (Buildings 2024), and in the Italian case the prediction model overestimated the field result by 4.1 dB, predicting R’w 32.1 dB against 28 dB measured (Buildings 2025). In the Brazilian study, by contrast, most models underestimated the measured values, so the direction of prediction error depends on the model and its inputs; only measurement settles it.
4. Leaks, Penetrations and Workmanship Penalties
Many field failures trace to installation details rather than the wall design. NRC’s outlet box study set the scale: back-to-back metal boxes dropped a double wood stud wall from STC 55 to STC 51, and poorly located boxes cost up to 6 points (Nightingale and Quirt, Canadian Acoustics 1995). The authors also note that a 1-point STC difference is not significant, while 3 or more points is a clear difference, so a single pair of boxes can produce a measurable failure on its own.
The cheapest fixes were the most effective. Plastic air-barrier boxes left the wall at STC 55 in every placement, and with 90 mm of absorption behind the boxes the penalty fell to 1 point in every configuration. Small holes matter less than people fear: NRC found up to 48 small 6.4 mm holes per side had little effect, and it took 96 holes per side to move a single stud wall from STC 43 to STC 38 (Mahn, Skoda and Cunha, NRC, Canadian Acoustics 2024). A single larger opening can matter much more locally, as the one-inch hole in the opening of this article shows.
| Metric | Value | Source |
|---|---|---|
| Back-to-back metal outlet boxes, double stud wall | STC 55 to STC 51 | Nightingale and Quirt, Canadian Acoustics 1995 |
| Metal boxes offset 400 mm in adjacent cavities | STC 53 | Nightingale and Quirt, Canadian Acoustics 1995 |
| Worst case for poorly located boxes | Up to 6 STC points lost | Nightingale and Quirt, Canadian Acoustics 1995 |
| Plastic air-barrier boxes, all three placements | STC 55 (no loss) | Nightingale and Quirt, Canadian Acoustics 1995 |
| 1-inch hole, wall with design STC 56 | FSTC 51 away from hole, FSTC 41 near it | Acoustics Today, Fall 2016 |
| Single stud wall with up to 48 small (6.4 mm) holes per side | About 1 dB change, at 125 and 160 Hz only | Mahn, Skoda and Cunha, NRC, Canadian Acoustics 2024 |
| Gypsum layer placed inside a double stud cavity | STC 64 base wall down to STC 57; 14 to 17 dB loss below 200 Hz | Mahn, Skoda and Cunha, NRC, Canadian Acoustics 2024 |
| Single stud wall with 96 / 120 small holes per side | STC 43 to STC 38 / STC 37 | Mahn, Skoda and Cunha, NRC, Canadian Acoustics 2024 |
Context note: the internal gypsum problem is not theoretical. A 1985 NRC field case cited by Mahn et al. measured a double wood stud wall with an internal gypsum layer at FSTC 40, a wall that was predicted to reach about 60 had the gypsum been relocated. Most recent available data for outlet boxes: Nightingale and Quirt 1995.
5. What Tenants Report: Complaint and Annoyance Data
Complaint data are the field test nobody commissions. New York City logged 463,349 Noise - Residential complaints through 311 in 2025, 55.6% of the 833,000 noise complaints across all noise categories (463,349 divided by 833,000), and up 22.2% from 379,296 in 2024 (NYC Open Data, 311 Service Requests from 2020 to Present). Loud music and parties made up 64.8% of residential complaints, but banging and pounding (126,388) is the category most tied to impact and structure-borne paths that no wall upgrade alone solves.
The survey evidence explains why the complaints matter. In Denmark’s 2017 national health survey, 36% of residents in multi-storey housing were very or slightly bothered by neighbour noise, against 22% for traffic noise, and those very annoyed by neighbours had 2.49 times the unadjusted odds of not getting enough sleep (95% CI 1.79-3.46) (Rasmussen and Ekholm, Inter-Noise 2021). Some residents mask the sound rather than fix it, a market tracked in our white noise sound conditioner statistics.
| Metric | Value | Source |
|---|---|---|
| NYC Noise - Residential complaints, 2025 | 463,349 | NYC Open Data, 311 Service Requests |
| Loud Music/Party / Banging/Pounding, 2025 | 300,120 / 126,388 | NYC Open Data, 311 Service Requests |
| Loud Talking / Loud Television, 2025 | 30,836 / 6,005 | NYC Open Data, 311 Service Requests |
| Change in residential noise complaints, 2024 to 2025 | +22.2% (379,296 to 463,349) | NYC Open Data, 311 Service Requests |
| NYC noise complaints, 2024 (Comptroller count) | Over 610,000, +19% from 2023 | New York State Comptroller, May 2025 |
| Danish multi-storey residents bothered by neighbour noise (traffic noise) | 36% (22%) | Rasmussen and Ekholm, Inter-Noise 2021 |
| Never or not often enough sleep: very annoyed vs not annoyed by neighbours | 25.1% vs 11.6% | Rasmussen and Ekholm, Inter-Noise 2021 |
| Adjusted odds of being very bothered by fatigue, very annoyed vs not annoyed | OR 2.91 (2.14-3.98) | Rasmussen and Ekholm, Inter-Noise 2021 |
Context note: 311 counts measure complaints, not buildings, and include noise that does not travel through a party wall. The 2025 counts were pulled directly from the city’s dataset with the query complaint_type = Noise - Residential and grouped by descriptor; the 2024 total includes 2 stray records in other descriptors.
6. Ratings That Satisfy Occupants, and What Enforcement Achieves
The code minimum and the comfort threshold are not the same number. NRC Canada’s survey of 600 residents beside 300 measured party walls found annoyance only approached “not at all annoyed” at about STC 65, and the walls in the sample averaged STC 49.7 (range 38 to 60), with annoyance tracking measured STC at R² = 0.960 (Bradley, Canadian Acoustics 1999). Bradley recommended STC 55 as a realistic goal and STC 60 as essentially eliminating disturbance; nearly two decades later NRC researchers proposed the same ladder in field terms, ASTC 55 for satisfactory and ASTC 60 for ideal (Hoeller et al., Construction Canada 2018).
Enforcement is what closes the gap between those numbers and the built result. When England and Wales tied compliance to inspected, pre-approved Robust Details, compliance rose from between 35% and 50% to 97% within 18 months, against a government forecast of at least 10 years to reach 95% (Smith, Napier University, Acoustics’08 Paris). Norway shows the residual problem even in new stock: in a survey of newly built homes, sound insulation had the lowest satisfaction of the aspects rated, with 17% dissatisfied (Rasmussen, Forum Acusticum 2011).
| Metric | Value | Source |
|---|---|---|
| Residents surveyed / party walls measured | 600 / 300 | Bradley, Canadian Acoustics 1999 |
| Measured party wall STC, average and range | 49.7 (38 to 60) | Bradley, Canadian Acoustics 1999 |
| STC where annoyance approaches none | About STC 65 | Bradley, Canadian Acoustics 1999 |
| Market tiers in US multifamily (STC/IIC) | Entry 50/50, market rate 55/55, luxury 60/60 | WoodWorks, Thorburn 2021 |
| Robust Details compliance, before vs first 18 months | 35-50% vs 97% | Smith, Acoustics’08 Paris |
| Share of new homes using Robust Details / sample tests per year | 70% / over 2,000 | Smith, Acoustics’08 Paris |
| Norway new homes dissatisfied with sound insulation (two-storey housing) | 17% (27%) | Rasmussen, Forum Acusticum 2011 |
| Europeans exposed to neighbour noise that harms quality of life | More than 50 million | Rasmussen, Forum Acusticum 2011 |
Most recent available data: Bradley 1999 and Smith 2008. No US equivalent of the Robust Details compliance figures was found; the IBC makes field testing an option, not a requirement.
Summary: STC Wall Partition Performance by the Numbers
| Metric | Value | Source |
|---|---|---|
| IBC minimum between dwelling units, lab / field | STC 50 / NNIC 45 | ICC, IBC 2024 Section 1206 |
| Canada NBC 2015 minimum between units | ASTC 47 | Hoeller et al., Construction Canada 2018 |
| Strictest European airborne requirement | Scotland DnT,w 56 dB | Rasmussen, Forum Acusticum 2011 |
| Single vs double 2x4 stud wall, 1 layer Type X per side | STC 37 vs STC 61 | Byrick, Canadian Acoustics 2011 |
| Gain from cavity absorption | About +10 STC | NRC Construction Technology Update No. 1, 1997 |
| Floor rated STC-60, field result | ASTC-49 and ASTC-53 | Edwards et al., HGC Engineering |
| As-built shear walls, unmitigated | Average ASTC-49 | Lorimer et al., HGC Engineering |
| Back-to-back metal outlet boxes | STC 55 to 51 | Nightingale and Quirt, Canadian Acoustics 1995 |
| Plastic air-barrier outlet boxes | No STC loss (STC 55) | Nightingale and Quirt, Canadian Acoustics 1995 |
| 1-inch hole in STC 56 design wall | FSTC 41 near hole | Acoustics Today, Fall 2016 |
| Internal gypsum in double stud cavity | STC 64 down to 57 | Mahn et al., NRC, Canadian Acoustics 2024 |
| NYC residential noise complaints, 2025 | 463,349 (+22.2% vs 2024) | NYC Open Data, 311 Service Requests |
| Danish multi-storey residents bothered by neighbour noise | 36% | Rasmussen and Ekholm, Inter-Noise 2021 |
| Odds of insufficient sleep, very annoyed by neighbours | OR 2.49 (95% CI 1.79-3.46) | Rasmussen and Ekholm, Inter-Noise 2021 |
| STC where annoyance approaches none | About STC 65 | Bradley, Canadian Acoustics 1999 |
| Norway new homes dissatisfied with sound insulation | 17% | Rasmussen, Forum Acusticum 2011 |
| Robust Details compliance after 18 months | 97% (from 35-50%) | Smith, Acoustics’08 Paris |
Methodology and Sources
- International Code Council, 2024 International Building Code, Chapter 12, Section 1206 Sound Transmission
- ASTM E336-25, Standard Test Method for Measurement of Airborne Sound Insulation in Buildings and ASTM E413-22, Classification for Rating Sound Insulation
- Warnock and Quirt, Control of Sound Transmission through Gypsum Board Walls, NRC Construction Technology Update No. 1, 1997
- Warnock and Quirt, Sound Transmission Through Gypsum Board Walls, Canadian Acoustics Vol. 23 No. 3, 1995
- Nightingale and Quirt, Degradation of Wood Stud Wall Sound Isolation by Electrical Boxes, Canadian Acoustics Vol. 23 No. 3, 1995
- Mahn, Skoda and Cunha (NRC), The Transmission Loss of Double Stud Walls with Layers of Gypsum Board Installed inside the Wall Cavity, Canadian Acoustics Vol. 52 No. 3, 2024
- Byrick (Pliteq), Examining Airborne Sound Transmission Loss in Various Wall Constructions, Canadian Acoustics Vol. 39 No. 3, 2011
- Thorburn, Acoustical Considerations for Mixed-Use Wood-Frame Buildings, WoodWorks 2021
- Hoeller, Mahn and Mueller-Trapet (NRC), Paradigm shift in acoustic regulations, Construction Canada 2018
- Edwards, Niinivaara and Lorimer (HGC Engineering), Contribution of Internal Assemblies to Wood-Frame Floor/Ceiling ASTC Performance, Canadian Acoustics
- Lorimer, Niinivaara and Edwards (HGC Engineering), Improving ASTC Performance of As-Built Wood-Frame Double-Shear Panel Wall Assemblies, Canadian Acoustics
- Evaluating Airborne Sound Insulation in Dwellings Constructed with Hollow Ceramic Blocks under Brazilian Housing Policies, Buildings 14(3), 2024
- Measurement and Prediction of Airborne Sound Insulation Performance of Different Vertical Partition Walls in Indoor Environments, Buildings 15(20), 2025
- Acoustics Today, Fall 2016 issue, page 53 (hole and field FSTC test)
- NYC Open Data, 311 Service Requests from 2020 to Present (queried October 2026)
- Office of the New York State Comptroller, DiNapoli Releases New Tool Monitoring NYC 311 Complaints, May 15, 2025
- Rasmussen and Ekholm, Neighbour noise in multi-storey housing: Annoyance and potential health effects, Inter-Noise 2021
- Rasmussen, Sound insulation between dwellings: Overview of the variety of descriptors and requirements in Europe, Forum Acusticum 2011
- Bradley (NRC), Aggregate Subjective Ratings of Airborne Sound Insulation, Canadian Acoustics 1999
- Smith (Napier University), Sound Insulation for Homes, Acoustical Society of America lay-language paper, Acoustics’08 Paris
- Data watch: much of the best lab and survey evidence is old (NRC 1995 to 1999, Byrick 2011, Rasmussen 2011, Smith 2008) and is flagged as most recent available data; newer NRC flanking data sit in the RR-331 to RR-337 report series. The Byrick 2011 tests were run at an NVLAP-accredited independent lab but published by an employee of an isolation-product manufacturer. The two HGC Engineering papers do not state a publication year in their text. NYC complaint totals differ by definition: our direct query of the 311 dataset gives 833,000 noise complaints across all noise categories in 2025, while the Comptroller reported over 610,000 for 2024; we cite each with its own scope and do not combine them. The Danish paper reports 3,893 multi-storey respondents in its abstract and 3,966 in its body. No national US statistic on the share of multifamily buildings that fail field tests was found, so none is cited. The Gypsum Association GA-600 design manual was suggested as a source but its assembly tables are not publicly readable, so no GA-600 values are cited.
- Last updated: October 18, 2026. We update this roundup quarterly, and the next refresh is expected when the NYC 311 dataset closes out calendar year 2026 and the next edition of the International Building Code is published.