Elapsed: 0m
Start measurement to see projection
At current noise level and accumulated dose
Noise dose is the fundamental metric for assessing occupational noise exposure and protecting workers' hearing. It expresses the cumulative sound energy a person receives during a work shift as a percentage of the maximum permissible exposure. A noise dose of 100% means the worker has reached the regulatory limit for the day. SonaVyx calculates noise dose in real time using your device's microphone, applying A-frequency weighting per IEC 61672-1 to match the sensitivity of human hearing.
The Occupational Safety and Health Administration (OSHA) in the United States uses a 5 dB exchange rate (also called the doubling rate or trading ratio) to calculate noise dose under 29 CFR 1910.95. The Permissible Exposure Limit (PEL) is 90 dBA as an 8-hour Time-Weighted Average (TWA). For every 5 dB increase above the criterion level, the allowable exposure time is halved: 95 dBA for 4 hours, 100 dBA for 2 hours, 105 dBA for 1 hour, and so on. OSHA also defines an Action Level at 85 dBA TWA, which triggers mandatory hearing conservation program requirements including audiometric testing, hearing protector availability, training, and recordkeeping. The 5 dB exchange rate is unique to OSHA and is based on older research; most international standards and NIOSH recommend the more protective 3 dB rate.
ISO 9612:2009 (Acoustics — Determination of occupational noise exposure) uses the equal-energy principle with a 3 dB exchange rate, which is also adopted by NIOSH, the EU Directive 2003/10/EC, and most countries outside the United States. The daily noise exposure level LEX,8h represents the A-weighted equivalent continuous sound level normalized to an 8-hour working day. Under ISO 9612, the exposure limit is 87 dBA LEX,8h with upper and lower action values at 85 dBA and 80 dBA respectively. The 3 dB exchange rate means that every 3 dB increase doubles the sound energy and halves the allowable time: 88 dBA for 4 hours, 91 dBA for 2 hours, 94 dBA for 1 hour. This is more scientifically accurate because it reflects the actual energy doubling.
SonaVyx captures audio at 48 kHz with echo cancellation and noise suppression disabled for measurement integrity. The raw audio is processed through an A-weighting filter implemented in Rust compiled to WebAssembly. Every second, the equivalent continuous level (Leq) is computed and fed into the dose accumulator. For each 1-second sample, the fractional dose is calculated as Ci/Ti, where Ci is the sample duration and Ti is the allowable time at that level using the appropriate exchange rate formula: T = 8 / 2^((L-Lc)/ER), where Lc is the criterion level and ER is the exchange rate. The total dose is the sum of all fractional doses, expressed as a percentage.
The Noise Reduction Rating (NRR) is a single-number rating on hearing protection devices that indicates the attenuation provided under laboratory conditions. However, real-world performance is significantly lower due to improper fit and wear. OSHA's derating method calculates the effective exposure as: Effective Level = LAeq - (NRR - 7) / 2. NIOSH uses even more conservative subject-fit derating factors: 75% of NRR for earmuffs, 50% for formable (foam) plugs, and 30% for other types. SonaVyx shows both the unprotected and protected exposure levels to help you select appropriate hearing protection for your noise environment.
OSHA uses a 90 dBA criterion level with a 5 dB exchange rate and an 80 dBA threshold (levels below 80 dBA are not counted). NIOSH uses an 85 dBA criterion with a 3 dB exchange rate (equal-energy principle) and no threshold. The NIOSH Recommended Exposure Limit (REL) of 85 dBA TWA is more protective and aligns with international standards. In practice, NIOSH calculations yield higher dose percentages for the same noise exposure, making it easier to exceed the limit.
Modern smartphones can achieve accuracy within 2-3 dB of professional dosimeters when properly calibrated. The main limitations are microphone frequency response (smartphone mics may roll off below 100 Hz) and placement (a true dosimeter is worn on the shoulder near the ear). SonaVyx is excellent for screening surveys, preliminary assessments, and awareness, but official OSHA compliance monitoring requires a calibrated Type 2 personal noise dosimeter with NIST-traceable calibration records.
An OSHA dose of 50% corresponds to a TWA of 85 dBA, which is the Action Level. At this point, employers must implement a Hearing Conservation Program (HCP) including: baseline and annual audiometric testing, offering hearing protection devices at no cost, providing training on noise hazards and HPD use, and maintaining exposure records. If the dose reaches 100% (TWA 90 dBA), engineering or administrative controls to reduce noise are required.
LEX,8h (daily noise exposure level) is defined in ISO 9612 as the A-weighted equivalent continuous level normalized to 8 hours: LEX,8h = LAeq,Te + 10 log10(Te/8), where Te is the effective exposure duration. TWA (Time-Weighted Average) under OSHA uses the 5 dB exchange rate formula. For identical exposure conditions, LEX,8h and TWA yield different values because of the different exchange rates. LEX,8h is used in the EU and most international regulations, while TWA is specific to US OSHA.
Long-term exposure to levels between 70-80 dBA may contribute to hearing loss over decades, especially with continuous exposure. The World Health Organization recommends limiting occupational exposure to 75 dBA LAeq,8h for lifetime protection. While OSHA's 80 dBA threshold excludes these levels from dose calculations, they are not truly "safe" for prolonged daily exposure. The EPA identified 70 dBA Leq(24h) as the level below which no hearing loss occurs over a lifetime.
SonaVyx calculates the dose accumulation rate (percent per hour) from the elapsed measurement period, then extrapolates to the full 8-hour shift. The formula is: Projected Dose = (Current Dose / Elapsed Hours) x 8. This assumes the noise environment remains similar for the rest of the shift. The projection updates every second and becomes more accurate as more data accumulates.