07Traffic: tires, engines, horns and noise cameras

The short version: Road traffic is the noise source that reaches the most people. In Europe, about 100 million people lived with road noise at or above 55 dB Lden in 2012, five times the count for railways. Above about 20 mph (32 km/h), most of what you hear from a passing car is its tires on the pavement, not its engine, which is why speed and road surface matter as much as mufflers. The loudest single vehicles - modified exhausts, motorcycles, trucks using engine brakes - are the target of federal limits, local ordinances and a new and disputed tool, the noise camera.

Is traffic really the biggest source of city noise?

What you hear: A steady wash of sound that never quite stops, louder at rush hour, with trucks standing out above the cars.

In Europe, yes, by a wide margin. The European Environment Agency (EEA) counted about 100 million people in its 33 member countries exposed to road traffic noise at or above 55 dB Lden in 2012, and 32 million above 65 dB Lden.¹¹⁷ Railways came second at 19 million people above 55 dB Lden, aircraft third at more than 4.1 million, and industry in urban areas at about 1.0 million.¹¹⁷ Those figures come from a 2017 EEA briefing with 2012 as the reference year, and the advocacy group Transport & Environment reported the same numbers in May 2017.¹¹⁸ The EEA's 2025 report, using 2022 data, put road noise at about 92 million people at or above 55 dB Lden, rail at about 18 million and aircraft at about 2.6 million.²

Lden (day-evening-night level) is a 24-hour average that adds a penalty to evening and night sound, because the same noise bothers people more after dark. It is the metric the EU uses to map road, rail and aircraft noise; sound and decibels explains how it differs from US metrics. The US does not publish a comparable national exposure count for all roads; its federal highway rules use hourly levels instead (see the noise wall section below).

The EEA briefing reports that the World Health Organization ranks road traffic noise as the second most harmful environmental stressor in Europe, after air pollution, with effects including sleep disturbance, cardiovascular disease, hypertension, cognitive impairment and annoyance.¹¹⁷ What the studies behind those effects show is covered in the health section.

The closest US figure covers Interstates only. The Bureau of Transportation Statistics' National Transportation Noise Map, released March 21, 2017 with 2014 data, estimated that 98.00% of the US population was potentially exposed to Interstate highway noise below 50 dBA (24-hour LAeq), 0.25% to 70 to 79 dBA and 0.06% to 80 dBA or more.²³ BTS describes the model as a worst case, with no shielding from buildings or terrain, so these are potential exposures, not measurements.²³

Why doesn't cutting traffic in half make it twice as quiet?

What you hear: A road with half the cars sounds only a little calmer.

Decibels are logarithmic, so sound energy and perceived loudness do not track each other the way intuition expects. According to a US concrete-pavement research center, halving traffic volume lowers the sound level by only about 3 dBA.¹¹⁹ A change of 1 to 3 dB is hard to notice, 5 dB is noticeable to most people, and a 10 dB drop is judged to be about half as loud.¹¹⁹

The same brief gives three other rules of thumb. A truck is about 8 dBA louder than a car. Roadway sound drops about 3 to 5 dB for each doubling of distance from the road. And each 10 mph increase in speed raises the level by about 2 to 3 dBA.¹¹⁹ By those rules of thumb, slowing traffic or taking trucks off a road can do more than removing a share of the cars.

Are tires or engines louder?

What you hear: At low speed, an engine note and the clatter of an exhaust. On the highway, a roar or whine that seems to come from the road itself.

For cars at steady speed, tire-pavement noise - sound made where the tread meets the road - becomes the main source above about 20 mph (about 32 km/h). For trucks, that crossover speed is higher.¹¹⁹ Below the crossover, the engine, intake and exhaust dominate. Above it, a quieter engine makes little difference, and the road surface and tire tread take over.

The most influential pavement factor is "positive texture": small bumps protruding above the surface that the tire strikes as it rolls. Quieter surfaces have a very fine, "negative" texture, with voids and grooves below the surface rather than bumps above it.¹¹⁹

Engineers measure this with microphones mounted a few inches from a test tire as it rolls. In the US the usual method is OBSI (on-board sound intensity); elsewhere a related method called CPX (close-proximity) is more common.¹¹⁹

This is also why electric vehicles are not much quieter at highway speed: above the crossover, the tires are the noise. The artificial sounds electric cars must play at low speed are covered in sounds made on purpose.

What is quiet pavement, and does it last?

What you hear: The roar drops noticeably when you drive from one stretch of highway onto a newer, darker or grooved one.

Several surfaces can cut tire noise, but the numbers vary widely from study to study. A March 2025 literature review for the California Department of Transportation (Caltrans) collected these results:¹²⁰

Open-graded friction course (OGFC) is a porous asphalt with connected air voids that absorb some of the noise made at the tire. Durability is the catch. Arizona's rubberized overlay lost about 2 dB of its benefit over four years, and an experimental section in Washington State lost nearly all of its benefit in just over two years.¹²⁰

Arizona's Department of Transportation started a Quiet Pavement Pilot Program in April 2003, working with the Federal Highway Administration (FHWA) and the Volpe Center.¹²⁰ As of the Caltrans review in March 2025, quieter pavement was still not an FHWA-approved noise abatement measure for federal projects, although a proposed change to the federal rule hinted it might become one.¹²⁰

The proposal in question is FHWA's overhaul of 23 CFR Part 772, published on October 18, 2024, with comments due December 17, 2024. Among other changes, it would lower the Table 1 impact levels by 1 dB(A) and apply Type I project requirements to federally funded abatement on existing roads, which it says include "changing the pavement surface or building noise barriers."¹²¹ As of October 2026, a search of the Federal Register found no final rule, so the 2010 rule and its table below still apply.¹²²

When does a highway get a noise wall?

What you hear: Traffic noise that suddenly softens when a concrete or wooden wall appears between the road and the houses.

In the US, federally funded highway projects follow a rule at 23 CFR Part 772. When predicted noise at a nearby property "approaches" or exceeds the Noise Abatement Criteria (NAC), the state has to consider abatement, usually a wall. The criteria are written in Leq(h), the steady level that carries the same sound energy as the actual fluctuating traffic over one hour. The table below is the federal Table 1, as reproduced in New Hampshire's 2016 noise policy:¹²³

"Approach" means within 1 dBA below the criterion, so a home predicted at 66 dBA Leq(h) counts.¹²³ States also flag a "substantial increase" over existing levels; New Hampshire defines that as 15 dBA or more.¹²³

Considering a wall is not the same as getting one. A barrier is not "feasible" unless it cuts at least 5 dBA for at least one impacted home. Then it has to be "reasonable," which each state defines within the federal rule. New Hampshire requires at least 7 dBA at one benefited home, aims for 10 dBA or more at first-row homes, and starts its cost test from a base of 1,500 square feet of wall per benefited home, adjusted by policy factors.¹²³

Insertion loss is the number of decibels a wall takes off compared with no wall. For homes directly behind a barrier, the CP Tech Center brief puts it at about 5 to 10 dBA, and says effectiveness falls off beyond about 300 to 500 feet (90 to 150 m) from the road.¹¹⁹ By the perception rule above, 10 dBA is roughly half as loud. People several blocks back often notice little change.

In Massachusetts, MassDOT says it uses the federal Type I noise abatement program under 23 CFR 772 and approves a barrier when "enough people will receive a high enough benefit for the cost." Its page lists no request process for residents; it points questions to MassDOT's online contact form.¹²⁴ Some existing walls came from the former Massachusetts Turnpike Authority's Type II (retrofit) program, and a 2026 MassDOT FAQ on new barriers in Allston says state policy included a formal vote by residents on whether to advance a barrier.¹²⁵

Is it illegal to honk for no reason?

What you hear: Short taps at a green light, long blasts in gridlock, a car alarm's cousin at 2am.

In the US, horn rules are set mainly by state law. North Carolina's is typical in shape: a vehicle must have a horn "capable of emitting sound audible under normal conditions from a distance of not less than 200 feet," the driver may use it only as a reasonable warning, and the horn may not emit an "unnecessary or unreasonable loud or harsh sound."¹²⁶ So a horn has a minimum audibility, and its use is limited to warning, but the statute sets no decibel cap.

In Massachusetts, General Laws chapter 90, section 7 requires a motor vehicle to carry "a suitable bell, horn or other means of signalling."¹²⁷ Section 16 provides that "No person operating a motor vehicle shall sound a bell, horn or other device, nor in any manner operate such motor vehicle so as to make a harsh, objectionable or unreasonable noise."¹²⁸ Neither section sets a decibel limit for horns.¹²⁸

Elsewhere in the world: horns in India

What you hear: A near-continuous layer of honking at every intersection.

India is a frequently cited case study for horn culture. NPR reported in September 2025 that a 2019 study found average noise in Mumbai often above 80 decibels, and counted 27 honks in one minute at a single intersection.¹²⁹ Sumaira Abdulali of the Awaaz Foundation, a Mumbai activist, told NPR that horns reach about 120 decibels; that is her estimate, and no weighting or distance was given.¹²⁹ Fines run up to about $25.¹²⁹ Road minister Nitin Gadkari proposed replacing horn sounds with Indian classical instruments; Abdulali called the idea a disaster.¹²⁹

In February 2020, Mumbai Police ran a "punishing signal": decibel meters at traffic lights reset the red-light countdown whenever honking at the intersection exceeded 85 dB, so the drivers who honked waited longer.¹³⁰

What are the rules on loud motorcycles and exhausts?

What you hear: A crackling, popping exhaust that sets off car alarms, or a motorcycle you can hear for blocks.

In the US, the Environmental Protection Agency (EPA) sets noise limits for new street motorcycles under 40 CFR 205.166: 83 dBA for model year 1983 and 80 dBA for 1986 and later, with moped-type motorcycles at 70 dBA. The exhaust must keep meeting the standard for one year or 6,000 km.¹³¹ Those numbers come from the federal test procedure and do not translate directly to a roadside reading.

Motorcycles built after December 31, 1982 carry a label on the chassis and a label on the muffler with matching codes. According to the advocacy site NoiseOFF, an unlabeled or non-matching exhaust on a regulated motorcycle violates section 4909(a)(2) of the federal Noise Control Act.¹³² Federal law makes tampering a violation, but enforcement in practice falls mostly to states and cities (see who makes the noise rules).

California's exhaust limits, as listed by the California Highway Patrol (CHP), are among the most specific: 95 dBA for passenger vehicles under 6,000 lb under an SAE stationary test (Vehicle Code 27151), 80 dBA for motorcycles (VC 27202), and 80 dBA for pickups from 6,001 to 10,000 lb and commercial vehicles (VC 27204).¹³³

How do noise cameras work, and do they hold up?

What you hear: Nothing different. A noise camera is a roadside sound meter wired to a license plate camera; when a passing vehicle crosses a set level, it records the plate and the sound.

The idea is contested. Supporters point to deterrence and relief for residents on streets used for late-night revving. Skeptics, including the California Highway Patrol, question whether the devices can reliably pick out the right vehicle. Where each program stands:

New York City. The Department of Environmental Protection (DEP) piloted noise cameras from June 2021. The meter triggers at 85 decibels or more, measured from at least 50 feet; the DEP letter does not state the weighting. As of August 2023, fines were $800 for a first offense, $1,700 for a second and $2,500 after that.¹³⁴ In December 2023 the City Council approved the "Stop Spreading the Noise Act," allowing the devices in all five boroughs.¹³³

California. SB 1079, by Senator Anthony Portantino, was signed on September 19, 2022 (Chapter 449). It added Vehicle Code 27150.4 and told the CHP to evaluate sound-activated enforcement devices and report by January 1, 2025.¹³³ In the CHP's trial, one vendor's system logged 328 events between February 9 and March 27, 2024, with exhaust noise from 72.7 to 115.3 dBA; a second vendor's interface was unsuitable, and a third produced no data. The CHP does not recommend the devices as a standalone basis for citations.¹³³

The same report notes that Knoxville, Tennessee tested a device from 2022 without issuing citations, and that Edmonton, Alberta ran a pilot from 2016 to 2020 before returning to traditional enforcement.¹³³

France. Bruitparif, the Paris-region noise observatory, built the Hydre noise radar from two of its Méduse sensors, each with four microphones; Méduse units had been deployed since summer 2019. The legal trial began on January 4, 2022, under the 2019 mobility law (LOM, article 92) and Decree 2022-1. Readings are normalized to 7.6 m, and one scooter measured 92.7 dB(A). The planned fine was 135 euros, reduced to 90 euros if paid within 15 days.¹³⁵

Pilot sites included Paris, Nice, Toulouse, Bron, Rueil-Malmaison, Villeneuve-le-Roi and the Haute Vallée de Chevreuse. A second phase with fines, initially planned for spring 2025, depends on type approval from LNE, the national metrology lab; a parliamentary question dated June 10, 2025 said no fines had yet been issued.¹³⁶

United Kingdom. The Department for Transport (DfT) ran a roadside trial from October 18, 2022 to February 1, 2023 in Keighley, Bristol and South Gloucestershire, Great Yarmouth and Rubery in Birmingham, with 300,000 pounds of funding. The reports were published on March 17, 2024.¹³⁷ According to the Institute of Acoustics, the cameras activated 1,777 times, and 4% of activations were attributed to excessively noisy vehicles. The proposed threshold is 95 dB LAmax at 7.5 m on roads with limits of 50 mph or less; a deterrent effect was observed, and the DfT planned to issue guidance to councils.¹³⁸ LAmax is the highest A-weighted level during the event.

The trigger levels are not comparable across programs: New York's is measured from at least 50 feet (about 15 m) with no weighting stated, the UK proposal is an LAmax at 7.5 m, and France normalizes to 7.6 m.

As of July 2026, the French motoring site Motorsactu reported that no noise radar fines had been issued; metrological testing for type approval began in late 2025, and approval was expected during 2026.¹³⁹ In New York City, DEP's annual report for calendar year 2025 says its cameras recorded 15,994 events and led to 1,691 summonses, with $1,472,265 in penalties imposed and $308,229 collected. The report does not state the trigger level, weighting or distance.¹⁴⁰

What does a "no engine brake" sign mean?

What you hear: A loud, rapid machine-gun rattle from a big truck slowing on a hill or approaching a town.

An engine compression brake, often called a "Jake brake" after one manufacturer, changes the exhaust valve timing so the engine absorbs power and slows the truck without wearing the wheel brakes. A legal note published by the Maine Department of Transportation says the brake is "quite loud" on trucks with poorly muffled or straight-pipe exhausts, and argues that the real problem is often defective or modified mufflers rather than the brake itself.¹⁴¹

In the US, according to the same note, federal rules require trucks to emit less than 80 dBA at 50 feet on a drive-by.¹⁴¹ Many towns post "no engine brake" signs, but the Maine note points out that a sign without an enforceable ordinance behind it is advisory only. It adds that federal law (42 U.S.C. 4917) can preempt local noise rules for interstate carriers unless the local rule matches federal standards, and that bans raise safety concerns on steep downgrades, where the engine brake keeps wheel brakes from overheating.¹⁴¹

Are electric garbage trucks quieter?

What you hear: The early-morning sequence of diesel idle, hydraulic whine, crash of a bin and the grind of the compactor.

Reports say yes for the driving part, less so for the compacting part. The reports below give no decibel figures. New York City's Department of Sanitation piloted a Mack LR Electric refuse truck from November 2020; it was described as "near-zero" noise, with added white noise so people know it is nearby.¹⁴² Copenhagen's Amager Resource Center received 86 electric Scania refuse trucks, and drivers reported much lower noise in the cab.¹⁴³

The most specific study comes from Finland. Tests by the research center VTT in autumn 2022 found an electric refuse truck "notably quieter" than a diesel one, but the noise of compacting the trash was unchanged. Adding a rubber mat in the hopper and an enclosure around the hydraulic pump cut handling noise.¹⁴⁴ An electric drivetrain removes the engine idle; the bang of bins and the crush of the load are a separate engineering problem.

Report an error in this section