Languages:
The SCENIHR opinion states:
3.3.3.Units of noise exposure
3.3.3.1. Sound pressure level and dB SPL
One parameter of the acoustic (sound) wave which is generally used to assess sound exposure to humans is the sound pressure level expressed in μPa or Pa. Human ear’ audible sound pressure levels range from 20 μPa (hearing threshold) till 20 Pa (pain threshold), resulting in the scale 1:10,000,000. Since using such a large scale is not practical, a logarithmic scale in decibels (dB) was introduced which is also in agreement with physiological and psychological hearing sensations.
dB of sound pressure level (dB SPL) is defined as: 20 log10 p1/p0 where p1 is actually measured sound pressure level of a given sound, and p0 is a reference value of 20μPa, which corresponds to the lowest hearing threshold of the young, healthy ear. In the logarithmic scale the range of human ear’s audible sounds is from 0 dB SPL (hearing threshold) to 120-140 dB SPL (pain threshold) (see table 1 below).
Table 1: Typical sound pressure levels for daily life sounds
3.3.3.2. Loudness level and filter A [dB(A)]
The human ear is not equally sensitive to sounds (tones) of the same sound pressure levels but different frequencies. This subjective or perceived magnitude of a sound by an individual is called its loudness. The loudness of a sound is not equal with its sound pressure level and differs for different frequencies. In order to assess loudness of a sound the isophonic curves are explored. Isophonic curves relate the characteristic of a given tone expressed in dB SPL to its subjective loudness level expressed in phones (see figure 1 below). As it could be seen in the figure below, the frequencies 3-4 kHz are the most sensitive within sound frequency range from 20 Hz to 20 kHz that can be heard by human ear. For frequencies lower than 3-4 kHz and higher sound frequencies, the ear becomes less sensitive.
While sound pressure measurements should give a reading of the sound pressure in dB SPL, in the context of human hearing it is more practical to provide also a value which corresponds more closely to the hearing sensation or loudness in phones. The A, B, and C filters used currently in sound-level meters were aimed at mimicking isoloudness curves over frequency under different conditions of sound intensities, i.e. for sounds of low, medium, and high loudness levels, respectively (IEC 651, 1979). The “A” network modifies the frequency response to follow approximately the equal loudness curve of 40 phons, while the “C” network approximately follows the equal loudness curve of 100 phons. A “B” network is also mentioned in some texts but it is no longer used in noise evaluations. The popularity of the A network has grown in the course of time. In current practice, the A- weighting curve filter is used to weight sound pressure levels as a function of frequency, approximately in accordance with the frequency response characteristics of the human auditory system for pure tones. This means that energy at low and high frequencies is de-emphasized in relation to energy in the mid-frequency range.
Correlation between noise effect hearing loss and sound exposure levels measured in A, B, or C weightings would not be very different. B (or even C) weightings provide a better correspondence between loudness and moderate (or high) acoustic levels, however A weighting differs only from B and C as underweighting frequencies below about 500 Hz. Since the human ear is much more resistant to noise-induced hearing loss (NIHL) at and by low frequencies A weighting is more in correspondence with NIHL risk.
It should be noted that the A-filter has been adopted so generally that sound pressure levels frequently quoted in audiology literature simply in dB are in fact A-weighted levels. Many older general purpose sound level meters are restricted solely to A-weighted sound pressure level measurements.
3.3.3.3. Decibel measures in audiometry
Different decibel measures are used in audiometry (evaluation of hearing sensitivity) than in sound pressure measurement. They depend on the reference value.
Pure-tone audiometric thresholds are expressed in dB HL (hearing level) and are referred to hearing thresholds of normal hearing young individuals. The differences between dB HL and dB SPL arise from isophonic curves. Their corresponding values are given in the table below.
Table 2: Audiometric hearing thresholds of normal ears
Similarly to dB HL, the dB nHL (normal hearing level) values are referred to hearing thresholds of normal hearing individuals but they regard non-tonal sound stimuli (e.g. clicks).
Source & ©:
SCENIHR,
The SCENIHR opinion states:
3.3.4.Methodology of noise measurement
Sounds are usually identified by their frequency spectrum, which is also relevant to human perception because the ear analyses sounds in the cochlea by a spectral analysis.
The elemental component of a frequency spectrum is a sine wave or sinusoid with a specific frequency. All sound waves can be described as a linear superposition of sinusoids. Each sinusoid can be characterized by its frequency, its amplitude and the phase in relation to the zero-time mark. Sinusoids with the same frequency and amplitude superimpose either constructively by adding up to a sinusoid with double amplitude if the phase difference is zero and destructively by cancelling out if the phase difference is 180 degrees (or antiphase) resulting in no sound of that characteristic frequency at a given point.
Sound originating from speech and music can similarly be described by their spectrum. In general terms signals can be divided in signals with a tonal character and with a noisy character.
- Signals with a tonal character exhibit a spectrum made up of a basic frequency component (f0) with harmonics (components that have a frequency which is an integer multiple of the basis frequency (n*f0) and a related phase.
- Signals with a noisy character exhibit a spectrum which is more complex than a linear superposition of basic frequencies and their harmonics.
Sound measurements are done by determining the amplitude of the spectral components or by detecting the sound pressure through a physical device, e.g. a microphone. The total sound level of a signal is a root-sums-of-squares of the amplitude of all the spectral components.
Signal levels, including noisy signals and music, are measured by placing a calibrated sound meter (SPL meter) at the centre head location of a potential listener. This method is generally used to determine the risk for hearing loss in working conditions.
The method distinguishes between various possible measures:
- The averaged level, which is the average level of all frequency components over a certain time period
- The level measurement can be recorded by filtering according to the A, B or C filter; dB (A)
- The peak level indicating the highest level recorded either of the total (weighted) signal or of specific components
- The 8-hour equivalent level (Lequ, 8h) which is a measure for the risk on hearing damage based on certain criteria
The method can also be used to determine the level of music in the open field. Due to the dependence of sound waves on the exact listening situation, as detailed in 3.2, it is clear that this type of measurement is not suitable to head phone use where only a small space between the head phones and the inner ear is exposed to sound waves.
Sound levels of signals presented through headphones are usually measured by artificial ears. Most common are two types, the occluded ear simulator (OES) and the 2 cc coupler. In audiometry and hearing aid specifications all measurements are measured using one of these two couplers. The design of the couplers is based on the resonance properties of the ear canal and the impedance of the tympanic membrane.
In the link of sound transfer from the open field to the ear, there is another transfer characteristic to be included and that is the baffle effect of head and torso. The head effects are usually determined by using a manikin or as they are also called HATS, head and torso simulator. It consists of a torso and head in which artificial ears are included. The sound pressure is measure at the eardrum. If compared with the free field, this gives the head-related transfer function (HRTF).
It is obvious that HATS and the couplers are based on measurements, averaged over many torsos and ears of both genders taking a multitude of anatomic features into account. Sound levels in individual ears will always differ somewhat from these values. These have to do with the following features:
- Shape of the torso and clothing
- Hair style and head shape
- Shape and volume of the outer ear and ear canal
- Impedance of the tympanic membrane
- Distortion of the sound field caused by other listeners or objects in the room
For the purpose of estimating the risk of the use of individual music players we assume that the calculated sound levels based on the use of artificial heads and ears are good estimates of the real levels.
The risk for hearing damage depends on sound or noise level and exposure time. Criteria were originally developed using working conditions as a reference which are typically measured in the open field. If we want to assess the risk of PMPs we have to compare the levels produced by earplugs or headphones with the measurements done in free field. This implies we have to determine the HRTFs for the different PMPs.
The output level of a PMP is determined by using an artificial ear. It measures the actual sound pressure at the eardrum. To calculate the risk for hearing damage, the free field level has to be calculated by using the inverse HRTF.
Source & ©:
SCENIHR,
This summary is free and ad-free, as is all of our content. You can help us remain free and independant as well as to develop new ways to communicate science by becoming a Patron!
BECOME A PATRON!