A 24-inch round duct silencer that is 72 inches long delivers 60 dB of insertion loss at 1,000 Hz but only 10 dB at 63 Hz in certified test data (Vibro-Acoustics, CMB-LV-F1 Certified Performance Data). That gap sums up duct silencing: mid and high frequencies come out easily, rumble does not, and every extra decibel costs pressure drop and fan energy. The laboratory method behind these catalogs has a new edition, with ASTM E477-26 active since its March 10, 2026 update (ASTM International, E477-26 2026). On the receiving end, 54% of surveyed building occupants are dissatisfied with sound privacy and 34% with noise level (Graham, Parkinson and Schiavon, Buildings and Cities 2021). We aggregated data from Vibro-Acoustics certified test sheets, ASTM International, the UK Department for Education, UC Berkeley’s Center for the Built Environment, and peer-reviewed papers from Forum Acusticum 2023 and Archives of Acoustics; our companion roundup on subwoofer and low-frequency HVAC noise transmission covers the structure-borne side, while this one stays inside the duct.
TL;DR
- 1,000 Hz insertion loss climbs from 45 dB to 60 dB when a 24-inch silencer grows from 48 to 72 inches, while 63 Hz only moves from 7 to 10 dB (Vibro-Acoustics, CMB-LV-F1 Certified Performance Data).
- Silencer pressure drop including system effects should ordinarily stay at or below 0.35 in. w.g. (85 Pa) (ASHRAE HVAC Applications Handbook 2015, via Vibro-Acoustics Silencer Selection Instructions).
- Modeled silencer self-noise jumped from 59 dB(A) to 98 dB(A) sound power when the assumed face velocity rose from 6 m/s to 19 m/s (KFB Acoustics, Forum Acusticum 2023).
- ASTM E477-26 is now the active silencer test method, last updated March 10, 2026 (ASTM International, E477-26 2026).
- New-build classrooms are capped at 35 dB LAeq,30min including building services noise (UK DfE, Building Bulletin 93 2015).
- Classroom background noise rose from 29.7 dBA to 49.5 dBA between fan off and the highest fan setting (Oktav et al., Archives of Acoustics 2025).
- Modeling only duct-borne noise underestimated the average open-office HVAC level by 12.3 dB (Fiorini et al., Forum Acusticum 2023).
- 39% of surveyed occupants say acoustical quality interferes with their work, the largest of any IEQ factor (Graham, Parkinson and Schiavon, Buildings and Cities 2021).
- Over 60% of cubicle occupants say acoustics interfere with getting their job done (UC Berkeley CBE, Acoustical POE Survey).
- Mechanical (heating, cooling and ventilation) noise ranked 5th of 11 acoustic complaint sources, behind four speech and phone items (Abbaszadeh et al., Healthy Buildings 2006).
1. Insertion Loss by Octave Band and Silencer Length
The certified tables make one pattern hard to miss: added length pays off in the 250 Hz to 1,000 Hz bands and barely moves the 63 Hz band. Going from a 48-inch to a 72-inch unit on the same 24-inch duct adds 15 dB at 1,000 Hz (60 minus 45) and just 3 dB at 63 Hz (10 minus 7). Even the biggest model in the family, a 60-inch duct with a 180-inch silencer, tops out at 18 dB at 63 Hz.
This is why fan rumble complaints rarely get fixed by buying a longer silencer. Low-frequency problems usually need treatment at the source, a different silencer geometry, or a different fan selection, while mid-frequency problems can be solved with a few extra feet of absorptive baffle. The Vibro-Acoustics CMB-LV certified performance data also warns that dynamic insertion loss (DIL) above 50 dB may be limited by noise flanking around the silencer or along the duct walls. Catalog values of 55 to 60 dB therefore describe the silencer, not what the room will get.
| Metric | Value | Source |
|---|---|---|
| DIL at 63 / 125 / 250 / 500 / 1,000 Hz, 24 in. duct, 48 in. long, no flow | 7 / 12 / 24 / 33 / 45 dB | Vibro-Acoustics, CMB-LV-F1 Certified Performance Data |
| DIL at 63 / 125 / 250 / 500 / 1,000 Hz, same duct, 72 in. long, no flow | 10 / 18 / 37 / 48 / 60 dB | Vibro-Acoustics, CMB-LV-F1 Certified Performance Data |
| Chimney silencer: minimum design target vs measured insertion loss | 16 dB target, 21 dB measured | KFB Acoustics, Forum Acusticum 2023 |
| A-weighted sound power at the chimney outlet before vs after the new silencer | 126 dB vs 105 dB | KFB Acoustics, Forum Acusticum 2023 |
Context note: the KFB case is an industrial dust-collection stack (three centrifugal fans feeding a 5 m chimney), not a building duct, but it shows the same rule in the field: the silencer met its acoustic target without degrading flow. The KFB Acoustics Forum Acusticum 2023 paper attributes the gap between predicted and measured spectra mainly to noise generated by turbulent airflow, which is the subject of section 3.
2. Pressure Drop: The Trade-Off Behind Every Decibel
Pressure drop is where acoustic ambition meets the fan curve. For the same 24-inch, 48-inch-long silencer, pressure drop goes from 0.08 in. w.g. at 1,000 fpm to 0.74 in. w.g. at 3,000 fpm. Doubling velocity from 1,000 to 2,000 fpm takes it from 0.08 to 0.33 in. w.g., roughly four times higher, which is the square-law behavior designers plan around: halve the velocity and the silencer can be longer, or more aggressive, for the same pressure budget.
The bigger surprise is the installation. Catalog pressure drops assume ideal straight duct on both sides, and a real mechanical room rarely provides it. The Vibro-Acoustics Silencer Selection Instructions list system effect multipliers that can exceed the silencer’s own loss. A silencer bolted to a centrifugal fan or discharging into a vaneless mitered elbow can double or nearly triple the pressure drop the designer budgeted.
| Metric | Value | Source |
|---|---|---|
| Pressure drop, 24 in. duct, 48 in. long, at 1,000 / 2,000 / 3,000 fpm | 0.08 / 0.33 / 0.74 in. w.g. | Vibro-Acoustics, CMB-LV-F1 Certified Performance Data |
| Recommended maximum silencer pressure drop including system effects | 0.35 in. w.g. (85 Pa) | ASHRAE HVAC Applications Handbook 2015, via Vibro-Acoustics Silencer Selection Instructions |
| System effect factor, mitered elbow without turning vanes at the discharge (D = 0) vs centrifugal fan at the discharge (D = 0) | 2.9 vs 2.0 | Vibro-Acoustics, Silencer Selection Instructions |
| Total pressure loss coefficient of the KFB chimney silencer | ζ = 4.8 | KFB Acoustics, Forum Acusticum 2023 |
| Outlet flow with the new silencer vs with the old silencer still in place | 1,417,873 vs 1,200,216 m3/h | KFB Acoustics, Forum Acusticum 2023 |
Outlier note: the same selection guide shows the effect can run the other way. Inlet cone silencers matched to the fan hub drew 99% of reference fan power, while plenums with blunt, turbulent splitters drew 110%. Good aerodynamic design is part of acoustic design, because a fan that uses less power also makes less noise.
3. Regenerated (Self-Generated) Noise vs Face Velocity
Every silencer is also a noise source. Air squeezing through the passages between baffles creates turbulence, and that regenerated noise rises quickly with velocity. In the certified data, raising forward-flow face velocity from 1,000 to 1,500 fpm lifts generated noise by 5 dB in every octave band from 63 Hz to 8,000 Hz. Running a silencer 50% faster costs 5 dB of self-noise before any change in insertion loss.
At industrial scale the effect is steeper still, because local velocities in front of the baffles can be far above the average a designer assumes. For building systems, the ASHRAE screening rule quoted in the Vibro-Acoustics selection guide, summarized in the table below, explains when self-noise matters at all: in most installations it does not, but quiet rooms and silencers placed next to the occupied space change that.
| Metric | Value | Source |
|---|---|---|
| Generated noise at 63 Hz / 1,000 Hz, forward flow, 1,000 fpm vs 1,500 fpm (3 sq ft face area) | 53 / 37 dB vs 58 / 42 dB re 1 pW | Vibro-Acoustics, CMB-LV-F1 Certified Performance Data |
| Modeled silencer self-noise at 6 m/s vs 19 m/s face velocity | 59 dB(A) vs 98 dB(A) sound power | KFB Acoustics, Forum Acusticum 2023 |
| When ASHRAE says to evaluate airflow-generated noise | pressure drop above 0.35 in., criterion below RC-35(N), or silencer in or near the occupied space | ASHRAE HVAC Applications Handbook 1995, via Vibro-Acoustics Silencer Selection Instructions |
| Rough limit on self-noise for a silencer close to the occupied space | room criterion spectrum minus 5 dB per octave | Vibro-Acoustics, Silencer Selection Instructions |
| Normal low / medium / high velocity ranges: circular dissipative vs rectangular dissipative silencers | 0-1,500 / 1,500-3,000 / 3,000-5,000 fpm vs 500-750 / 750-1,250 / 1,250-2,000 fpm | Vibro-Acoustics, Silencer Selection Instructions |
Context note: the velocity row shows why rectangular and circular silencers cannot be compared on face velocity alone. The circular low-velocity band (up to 1,500 fpm) overlaps the rectangular high-velocity band (1,250 to 2,000 fpm). For the open-plan office side of the background-noise problem, our open-plan office distraction acoustics roundup covers speech and masking data.
4. ASTM E477 Test Facts
Nearly every number in sections 1 to 3 rests on one laboratory method. According to the ASTM E477-26 listing, E477-26 became the active edition, last updated on March 10, 2026, replacing E477-20 after a lineage that runs back to E477-99. The standard tests insertion loss, airflow-generated noise and pressure drop across several airflows, because both insertion loss and generated noise change with air speed.
What the standard does not promise matters as much as what it measures. The significance section of the archived E477-20 edition says installed performance “may be significantly different” from laboratory results. Catalog figures are ratings under controlled conditions, not field guarantees.
| Metric | Value | Source |
|---|---|---|
| Active edition and last update | ASTM E477-26, March 10, 2026 | ASTM International, E477-26 2026 |
| Document length and responsible subcommittee | 16 pages, Subcommittee E33.08 | ASTM International, E477-26 2026 |
| Editions listed on the ASTM store | 7 (E477-99, -06, -06a, -13, -13e1, -20, -26) | ASTM International, E477-26 2026 |
| Quantities measured, and devices excluded from scope | insertion loss, airflow-generated noise and pressure drop at several airflows; reactive mufflers and non-ventilation devices excluded | ASTM International, E477-20 2020 |
| Straight duct assumed behind catalog pressure drops | 5 diameters at the inlet, 10 at the outlet | Vibro-Acoustics, CMB-LV-F1 Certified Performance Data |
| Accreditation claimed by one manufacturer lab | world’s first NVLAP-accredited silencer test lab, Lab Code 100424-0 | Vibro-Acoustics, Our Testing Facility |
| Standard paired with ASTM E477 for system-effect pressure drop | AMCA 210 | Vibro-Acoustics, Our Testing Facility |
Context note: the Vibro-Acoustics testing facility page traces its first aero-acoustic lab to the early 1960s and refers to 4th- and 5th-generation accredited facilities, with a 6th generation planned. For how reference test rooms are built, see our anechoic chamber acoustic testing statistics.
5. Room Noise Criteria by Space Type
A silencer is sized backwards from the room. The selection procedure takes the predicted room spectrum, subtracts the room criterion, and the positive difference in each octave band is the insertion loss to buy. That makes the criterion the most important number in the chain, and criteria vary a lot by space type. Under Building Bulletin 93 (2015), classrooms and lecture rooms get 35 dB LAeq,30min in new UK schools, while recording studios and rooms for students with special hearing needs get 30 dB.
BB93 counts building services noise inside the limit and requires mechanical ventilation to meet the table value at normal operating duty, so the silencer has no separate allowance. For offices, the literature summarized by the Sapienza team works in NC or RC ratings, while their LEED-compliant case study applied a dBA ceiling to open offices.
| Metric | Value | Source |
|---|---|---|
| Classroom, seminar and lecture room limit (new build / refurbishment) | 35 / 40 dB LAeq,30min | UK DfE, Building Bulletin 93 2015 |
| Recording studio, ensemble room, rooms for students with special hearing needs | 30 / 35 dB LAeq,30min | UK DfE, Building Bulletin 93 2015 |
| Open-plan teaching area, libraries, school offices | 40 / 45 dB LAeq,30min | UK DfE, Building Bulletin 93 2015 |
| Allowance for natural ventilation under normal conditions | table value + 5 dB | UK DfE, Building Bulletin 93 2015 |
| Summertime natural or hybrid ventilation, hottest 200 hours of the year | 55 dB or less | UK DfE, Building Bulletin 93 2015 |
| Unoccupied US classroom background noise including HVAC | below 35 dBA | ANSI/ASA S12.60-2010, as cited in Oktav et al., Archives of Acoustics 2025 |
| Recommended criterion for offices and conference rooms | NC/RC 30 | Fiorini et al., Forum Acusticum 2023 |
| Maximum background noise applied to LEED-compliant open offices | 45 dB(A) | Fiorini et al., Forum Acusticum 2023 |
Context note: NC curves date from 1957 and cover 63 to 8,000 Hz, while RC curves, proposed by Blazier for ASHRAE in 1981, cover 16 to 4,000 Hz with a 5 dB per octave slope, so RC reaches further into the rumble range that silencers handle worst (Fiorini et al., Forum Acusticum 2023). The ASHRAE Handbook criteria table itself is paywalled. Certification schemes that reference it are covered in our LEED and WELL building acoustics roundup.
6. What HVAC Noise Does to Occupied Rooms
The criteria in section 5 assume fans at normal duty. Field data show how quickly a room leaves them when fan speed rises. In the Alanya university classroom study, every one of five rooms exceeded the 35 dBA classroom benchmark at the highest fan setting, with levels between 47.2 and 49.5 dBA. The loudest room, A203, sat 14.5 dB above the benchmark (49.5 minus 35), and its signal-to-noise ratio at 125 Hz went negative.
Speech intelligibility moved less than the decibels suggest, which is a useful warning about relying on a single metric: the STI drop in the table is real but small. The Sapienza open-office models show the prediction side of the problem. Leaving out sound radiated by fan coils and duct walls produced the largest modeling error in that study, and the error grows as rooms get smaller.
| Metric | Value | Source |
|---|---|---|
| Classroom A203 background noise, fan off vs highest fan setting (33.0 m3/min) | 29.7 vs 49.5 dBA | Oktav et al., Archives of Acoustics 2025 |
| Rooms above 35 dBA at the highest fan setting | 5 of 5 (47.2 to 49.5 dBA) | Oktav et al., Archives of Acoustics 2025 |
| Classroom A203 STI, fan off vs highest fan setting (JND 0.03) | 0.61 vs 0.57 | Oktav et al., Archives of Acoustics 2025 |
| Classroom A203 signal-to-noise ratio at 125 Hz, highest fan setting | -4.00 dB | Oktav et al., Archives of Acoustics 2025 |
| Underestimate when only duct-borne airborne noise is modeled | 12.3 dB on the average room level, about 15 dB near 150 m3 | Fiorini et al., Forum Acusticum 2023 |
| Spatial spread of HVAC level inside 150 to 250 m3 open offices | around 3 dB | Fiorini et al., Forum Acusticum 2023 |
Outlier note: the A203 STI fell by 0.04 (0.61 minus 0.57), just above the 0.03 just-noticeable difference. The Sapienza offices showed almost no intelligibility penalty, with STI falling by only 0.01 versus the system-off condition (Fiorini et al., Forum Acusticum 2023). A quiet fan-coil design and an over-driven classroom cassette sit at opposite ends of the same scale.
7. Occupant Complaints About HVAC and Office Noise
Occupant surveys rarely isolate HVAC noise, but they show where acoustics sits among workplace complaints: at the bottom. Across 897 buildings and 93,662 responses in the CBE occupant survey database, sound privacy drew the highest dissatisfaction of any item at 54%, and acoustical quality had the largest negative effect on self-reported productivity.
The older CBE drill-down data show that HVAC noise is a real but secondary complaint. Among dissatisfied occupants, mechanical noise trailed four speech and telephone items in the CBE green building analysis. Acoustics was also the only category with a negative mean score in LEED-rated and green buildings, which suggests green design strategies that improved air quality and thermal comfort did not carry over to sound.
| Metric | Value | Source |
|---|---|---|
| Dissatisfied with sound privacy / temperature / noise level | 54% / 39% / 34% | Graham, Parkinson and Schiavon, Buildings and Cities 2021 |
| CBE database after cleaning (offices share) | 897 buildings, 93,662 responses (692 offices, 77%) | Graham, Parkinson and Schiavon, Buildings and Cities 2021 |
| Cubicle occupants saying acoustics interfere with getting their job done | over 60% | UC Berkeley CBE, Acoustical POE Survey |
| Acoustical POE analysis sample | 4,096 respondents in 15 buildings | UC Berkeley CBE, Acoustical POE Survey |
| Mean acoustics satisfaction (scale -3 to +3): non-green vs LEED/green buildings | -0.20 vs -0.27 | Abbaszadeh et al., Healthy Buildings 2006 |
| Rank of mechanical (HVAC) noise among listed acoustic complaint sources | 5th of 11 | Abbaszadeh et al., Healthy Buildings 2006 |
| Survey database behind the 2006 analysis | 181 buildings, 33,285 respondents, 46% average response rate | Abbaszadeh et al., Healthy Buildings 2006 |
Context note: the CBE acoustical POE survey page also reports that people are significantly more dissatisfied with sound privacy than with noise level (P < 0.01). Most recent available data on HVAC-specific complaint ranking: Abbaszadeh et al., CBE, 2006. Manufacturers describe rooftop units as one of the most common causes of HVAC noise problems and a major reason for tenant complaints (Vibro-Acoustics, Rooftop Unit Noise Control Solutions), but that claim comes without a survey figure.
Summary: HVAC Duct Silencer Statistics by the Numbers
| Metric | Value | Source |
|---|---|---|
| DIL at 63 Hz / 1,000 Hz, 24 in. duct, 48 in. vs 72 in. long | 7 vs 10 dB / 45 vs 60 dB | Vibro-Acoustics, CMB-LV-F1 Certified Performance Data |
| Chimney silencer measured insertion loss vs 16 dB target | 21 dB | KFB Acoustics, Forum Acusticum 2023 |
| Recommended maximum silencer pressure drop incl. system effects | 0.35 in. w.g. (85 Pa) | ASHRAE HVAC Applications Handbook 2015, via Vibro-Acoustics Silencer Selection Instructions |
| Modeled self-noise at 6 vs 19 m/s face velocity | 59 vs 98 dB(A) | KFB Acoustics, Forum Acusticum 2023 |
| Active silencer test method | ASTM E477-26, updated March 10, 2026 | ASTM International, E477-26 2026 |
| Manufacturer lab claiming first NVLAP silencer accreditation | Lab Code 100424-0 | Vibro-Acoustics, Our Testing Facility |
| New-build classroom limit incl. building services | 35 dB LAeq,30min | UK DfE, Building Bulletin 93 2015 |
| Natural ventilation tolerance above table value | +5 dB | UK DfE, Building Bulletin 93 2015 |
| Classroom noise, fan off vs highest setting | 29.7 vs 49.5 dBA | Oktav et al., Archives of Acoustics 2025 |
| Classroom STI, fan off vs highest setting | 0.61 vs 0.57 | Oktav et al., Archives of Acoustics 2025 |
| Underestimate from modeling duct-borne noise only | 12.3 dB | Fiorini et al., Forum Acusticum 2023 |
| Dissatisfied with sound privacy / noise level | 54% / 34% | Graham, Parkinson and Schiavon, Buildings and Cities 2021 |
| Say acoustical quality interferes with their work | 39% | Graham, Parkinson and Schiavon, Buildings and Cities 2021 |
| Cubicle occupants reporting acoustic interference | over 60% | UC Berkeley CBE, Acoustical POE Survey |
| Mean acoustics satisfaction, non-green vs LEED/green buildings | -0.20 vs -0.27 | Abbaszadeh et al., Healthy Buildings 2006 |
| Rank of mechanical noise among acoustic complaint sources | 5th of 11 | Abbaszadeh et al., Healthy Buildings 2006 |
Methodology and Sources
We searched for certified silencer test data, standards documents, peer-reviewed field studies and occupant survey datasets, then read each primary document directly. Catalog values are laboratory figures under ASTM E477 ideal-flow conditions; field values come from named measurement studies. Where a manufacturer document quotes ASHRAE, we name both.
- Vibro-Acoustics, CMB-LV-F1 and CMB-LV-F2 Certified Performance Data (insertion loss, pressure drop and generated noise tables; document dated 2019, most recent available data)
- Vibro-Acoustics, Silencer Selection Instructions and Silencer System Effects (cites ASHRAE HVAC Applications Handbook 2015 and 1995)
- Vibro-Acoustics, Our Testing Facility
- Vibro-Acoustics, Rooftop Unit Noise Control Solutions
- ASTM International, E477-26 Standard Test Method and E477-20 (historical)
- Chmielewski, Nieradka and Szoka (KFB Acoustics), Forum Acusticum 2023
- Fiorini, Di Matteo, Tarsitano and Vallati (Sapienza University of Rome), Forum Acusticum 2023
- Oktav, Ener, Ozenc and Sari, The Impact of HVAC Systems on Speech Intelligibility in University Classrooms, Archives of Acoustics 2025
- UK Department for Education, Building Bulletin 93: Acoustic Design of Schools, Performance Standards (2015)
- Graham, Parkinson and Schiavon, Lessons Learned from 20 Years of CBE’s Occupant Surveys, Buildings and Cities 2021
- UC Berkeley Center for the Built Environment, Acoustical POE Survey
- Abbaszadeh, Zagreus, Lehrer and Huizenga, Occupant Satisfaction with IEQ in Green Buildings, Healthy Buildings 2006
- Data watch: the ASHRAE HVAC Applications Handbook (Noise and Vibration Control chapter) and ASHRAE Journal articles are paywalled, so ASHRAE limits appear here only as quoted in a manufacturer guide that cites the 2015 and 1995 editions; check the current edition before design use. Most recent available data for silencer catalogs is a 2019 Vibro-Acoustics data sheet, for UK school criteria is BB93 2015, for the CBE benchmarks is Graham et al. 2021, and for HVAC-specific complaint ranking is CBE 2006. Price Industries catalog and lab pages could not be retrieved during research and are not used. Catalog DIL above 50 dB may be flanking-limited, and the KFB, Sapienza and Alanya results come from single sites, so they illustrate mechanisms rather than market-wide averages. The changes between ASTM E477-20 and E477-26 are not described on the public listing.
Last updated: October 18, 2026. We update this roundup quarterly, and the next refresh is expected when the 2027 ASHRAE Handbook, HVAC Applications volume is released.