Abstract (summary)

TranslateAbstractListening to music is one of the most common forms of recreational noise exposure. Previous investigators have demonstrated that maximum output levelsfromheadphones can exceed safe levels. Although preferred listening levels (PLL) in quiet environments may be at acceptable levels, the addition of background noise will add to the overall noise exposure of a listener. Use of listening devices that block out some of the background noise would potentially allow listeners to select lower PLLs for their music. Although one solution is in-the-earearphones, an alternative solution is the use of earmuffs in conjunction with earbuds. There were two objectives to this experiment. The first was to determine if an alternative to inthe- earearphonesfor noise attenuation (the addition of earmuffs to earbuds) would allow for lower PLLs through a portable media player (PMP) than earbuds. The second was to determine if a surrounding background noise would yield different PLLs than a directional noise source. This was an experimental study. Twenty-four adults with normalhearing. PLLs were measured for threeearphoneconfigurations in three listening conditions. Theearphoneconfigurations included earbuds, canalearphones, and earbuds in combination withhearingprotection devices (HPDs). The listening conditions included quiet, noisefromone loudspeaker, and noisefromfour surrounding loudspeakers. Participants listened in each noise andearphonecombination for as long as they needed to determine their PLL for that condition. Once the participant determined their PLL, investigatorsmade a 5 sec recording of the music through a probe tube microphone. The average PLLs in each noise andearphonecombination were used as the dependent variable. Ear canal level PLLs were converted to free-field equivalents to compare to noise exposure standards and previously published data. The average PLL asmeasured in the ear canal was 74 dBA in the quiet conditions and 84 dBA in the noise conditions. Paired comparisons of the PLL in the presence of background noise for each pair ofearphoneconfigurations indicated significant differences for each comparison. An inverse relationship was observed between attenuation and PLL whereby the greater the attenuation, the lower the PLL. A comparison of the single noise source condition versus the surrounding noise condition did not result in a significant effect. The present work suggests thatearphonesthat take advantage of noise attenuation can reduce the level at which listeners set music in the presence of background noise. An alternative to in-theearearphonesfor noise attenuation is the addition of earmuffs to earbuds.

Full Text

TranslateFull text Turn on search term navigationHeadnoteAbstract

Background: Listening to music is one of the most common forms of recreational noise exposure. Previous investigators have demonstrated that maximum output levelsfromheadphones can exceed safe levels. Although preferred listening levels (PLL) in quiet environments may be at acceptable levels, the addition of background noise will add to the overall noise exposure of a listener. Use of listening devices that block out some of the background noise would potentially allow listeners to select lower PLLs for their music. Although one solution is in-the-earearphones, an alternative solution is the use of earmuffs in conjunction with earbuds.

Purpose: There were two objectives to this experiment. The first was to determine if an alternative to inthe- earearphonesfor noise attenuation (the addition of earmuffs to earbuds) would allow for lower PLLs through a portable media player (PMP) than earbuds. The second was to determine if a surrounding background noise would yield different PLLs than a directional noise source.

Research Design: This was an experimental study.

Study Sample: Twenty-four adults with normalhearing.

Data Collection and Analysis: PLLs were measured for threeearphoneconfigurations in three listening conditions. Theearphoneconfigurations included earbuds, canalearphones, and earbuds in combination withhearingprotection devices (HPDs). The listening conditions included quiet, noisefromone loudspeaker, and noisefromfour surrounding loudspeakers. Participants listened in each noise andearphonecombination for as long as they needed to determine their PLL for that condition. Once the participant determined their PLL, investigatorsmade a 5 sec recording of the music through a probe tube microphone. The average PLLs in each noise andearphonecombination were used as the dependent variable. Ear canal level PLLs were converted to free-field equivalents to compare to noise exposure standards and previously published data.

Results: The average PLL asmeasured in the ear canal was 74 dBA in the quiet conditions and 84 dBA in the noise conditions. Paired comparisons of the PLL in the presence of background noise for each pair ofearphoneconfigurations indicated significant differences for each comparison. An inverse relationship was observed between attenuation and PLL whereby the greater the attenuation, the lower the PLL. A comparison of the single noise source condition versus the surrounding noise condition did not result in a significant effect.

Conclusion: The present work suggests thatearphonesthat take advantage of noise attenuation can reduce the level at which listeners set music in the presence of background noise. An alternative to in-theearearphonesfor noise attenuation is the addition of earmuffs to earbuds.

Key Words: Music, noise, portable media players, preferred listening levels, recreational noise exposure

Abbreviations: DRC 5 damage risk criteria; HPD 5hearingprotection devices; KEMAR 5 Knowles Electronics Manikin for Acoustic Research; OSHA 5 Occupational Safety and Health Administration; PC 5 personal computer; PLL 5 preferred listening level; PMP 5 portable media player

The concept of recreational noise exposure refers to any noise an individual is exposed to outside of their occupation. Sources of recreational noise exposure can include weapons, gardening and landscaping equipment, attendance at concerts, etc. (Clark, 1991). Recreational noise exposure in the form of listening to music is one of the most common. Listening to music awayfromhome has increased dramatically over the last 20 yr due to the development of progressively smaller portable music systems.

Concerns regarding the potential forhearing lossthrough the use of headphones have existed for quite some time (Wood and Lipscomb, 1972; Catalano and Levin, 1985; Rice et al, 1987; Turunen-Rise et al, 1991; Airo et al, 1996; Fligor and Cox, 2004; Portnuff and Fligor, 2006; Farina, 2007; Kumar et al, 2009). When portable stereos were commercialized in the late 1970s, concerns regarding the potentialhearing lossrisk due to exposure to music for longer periods of time increased due to the portability of the devices. Currently, popular formats for portable media players (PMPs) are CD players or digital music players (i.e., iPod, MP3 player, etc). As of June 2008, over 150 million iPods had been sold (Gaba, 2008) with average quarterly sales currently around 22 million (Apple, 2009).

Previous investigators have documented that the maximum output of PMPs can exceed safe levels and, therefore, these devices have the potential to causehearingdamage if used at the highest levels. Two studies conducted by Fligor indicate that the maximum output of PMPs as measured through a Knowles Electronics Manikin for Acoustic Research (KEMAR) and converted to free-field equivalent values can exceed 120 dBA (Fligor and Cox, 2004; Portnuff and Fligor, 2006). These measurements were made by placing headphones on a KEMAR with the output of the devices set to maximum and a removal of the ear canal resonance. However, levels alone do not reflect risk tohearing loss. Although the devices were shown to emit potentially hazardous levels, the levels are only potentially hazardous when users set them to maximum levels.

In order to predict someone's risk ofhearingdamage, the levels at which people set their PMPs must be considered in combination with the amount of time they spend listening to them. Damage risk criteria (DRC) are a combination of duration and level based on rates of noise-inducedhearing lossfollowing noise exposure. In the United States, the Occupational Safety and Health Administration (OSHA) sets the DRC at 90 dBA for 8 hr. Exposures that exceed the DRC are considered hazardous tohearing(OSHA, 1983). It is worth noting that DRC are based on free-field measures and not on levels measured in the ear canal. In order to compare levels measured in a KEMAR or a listener's ear canal to DRC, the ear canal or KEMAR values must be converted to free-field values by subtracting out the ear canal resonance (International Standards Organization [ISO], 2002, 2004). Values measured in the ear canal will always be higher than those measured in the free field due to the ear canal resonance (Berger et al, 2009).

To investigate whether usage patterns suggest risk ofhearing loss, studies have evaluated the typical levels at which people listen to PMPs (Catalano and Levin, 1985; Rice et al, 1987; Turunen-Rise et al, 1991; Airo et al, 1996; Williams, 2005, 2009; Farino, 2007; Hodgetts et al, 2007; Torre, 2008;Hoover and Krishnamurti, 2010) and queried individuals regarding their use of these systems (Zogby, 2006; Danhauer et al, 2009). Fromthese studies, we know that most users select levels within a safe listening range, but a small percentage of the population reports routinely listening to their PMPs at a potentially hazardous level. Average usage time surveys indicate that typical PMP users wear their devices for less than 4 hr/day with only about 10% wearing them for more than 4 hr/day (Zogby, 2006; Ahmed et al, 2007; Farino, 2007; Danhauer et al, 2009).

The usage patterns for PMPs in quiet versus noisy environments may differ. Although many listeners report listening to PMPs primarily in quiet environments, commuters and frequent travelers often listen to PMPs to pass the time and tune out background noisefromtrains, buses, subways, and airplanes. Individuals using PMPs are known to increase the volume when in the presence of background noise (Fligor and Ives, 2006; Hodgetts et al, 2007; Torre, 2008). Hodgetts et al (2007) demonstrated that listeners increased the volume on PMP players by 6 to 11 dB once multitalker babble or street noise was added to the environment. Similarly, Torre (2008) demonstrated that listeners increased the levels by 6 to 10 dB once background noise was added to the environment and Fligor and Ives (2006) demonstrated an increase of 3 to 20 dB with background noise levels above 50 dBA.

Studies have shown some significant differences in the usage patterns of PMPs between males and females. Several studies have found that males select significantly higher preferred listening levels (PLLs) (around 5 dB more) than females (Williams, 2005; Fligor and Ives, 2006; Torre, 2008), but not all have done so (Levey et al, 2011). As stated previously, determination of risk is a combination of level and duration. Some authors have found significant differences in listening durations such that males use their PMPs longer than females, potentially increasing their risk forhearing loss(Catalano and Levin, 1985; Shah et al, 2009).

Although there are contradicting opinions regarding whether these devices have the potential to causehearing lossbased on average usage levels and durations,hearingdamage is certainly possible at higher levels. Most studies have found that the majority of users have usage durations lower than that needed to cause damage, and only a small percentage of the population use their PMPs at levels thatwould be concerning (Rice et al, 1987; Williams, 2005). Rice et al indicated that approximately 5% of users who had PMP levels measured would meet or exceed the DRC set by OSHA (OSHA, 1983). Williams (2005) found that 25% of those he tested were at or above the level considered "at risk." Levey et al (2011) is the one study that found higher usage durations and levels stating that among urban college students, over 50% were deemed to be "at risk."

The use of a PMP is only one source of noise exposure and should be considered in conjunction with all other noise exposure. It is reasonable to assume that use of a PMP in the presence of background noise will add to a listener's overall noise exposure. Thus, the addition of recreational noise exposure to occupational noise exposure has the potential to put someone at greater risk forhearingdamage than someone who experiences only one kind of noise exposure.

A potential solution to the increase in PLL in the presence of noise is the use ofearphonesthat decrease the amount of background noise arriving at the listener's ears. The use of isolating canalearphoneshas been shown to allow the listener to decrease their PLL (Fligor and Ives, 2006; Hodgetts et al, 2007). Most PMP users utilize the earbuds that are supplied in the original packaging of the device, but this does not mean that theearphonescannot be replaced. Fligor and Ives (2006) compared PLLs for 100 doctoral students with four differentearphones, two that provided essentially no attenuation (Koss KSC11 and Apple iPod earbuds) and two designed to block background noise (Sony MDR-EX51LP in-the-earearphonesand Etymotic ER-6i in-the-earearphones). The authors found that listeners selected lower PLLs with the in-the-earearphonesthan the others and therefore concluded that the attenuation provided by in-the-earearphonesallowed for lowered listening levels; however, other differences between theearphonessuch as frequency response could not be ruled out as contributing factors. Hodgetts et al (2007) evaluated PLLs forearphoneswith and without noise reduction and found that listeners set PLLs highest when wearing earbuds and lowest when wearing over-the-earearphonesthat incorporated noise reduction. Intermediate PLL settings were found for over-the-earearphonesthat provided isolationfrombackground noise.

Surveys conducted by Shah et al (2009) and Hoover and Krishnamurti (2010) found that 60-75% of college students were willing to turn down the volume of their PMP to preventhearingdamage. In the Hoover and Krishnamurti study, over 90% were interested in buying specially designedearphonesforhearingprotection, depending on the cost. It is reassuring to know that college-aged individuals are becoming more aware of the causes ofhearing lossand ways in which to avoid it. Attenuation can be obtained through the user changingfromthe standard earbudearphonesto in-the-earearphonesor through the addition ofhearingprotection devices (HPDs) (specifically earmuffs) to existing earbuds. In cases where listeners do not have access to in-the-canalearphones, the addition of earmuffs to earbuds may be a viable short-term solution.

To study the degree towhich individuals increase their PLL in the presence of background noise, the test environment needs to be similar to real-world environments. In the real world, noise is typically diffuse, comingfrommultiple directions surrounding the listener. Some studies of PMPs have used a single noise source placed directly in front of the listener (Hodgetts et al, 2007, 2009) while others have used two loudspeakers (Ahmed et al, 2007), each located at 45 angles and directed toward the listener. Differences in signal-to-noise ratios (SNRs) between the two ears may be present in situations where a single noise source is located at a listener's side. A difference in SNRs between ears would not be present in either a surrounding noise environment or when a single loudspeaker is placed at 0. A comparison of two loudspeaker configurations was undertaken as a secondary question in the present study to investigate the impacts of background noise configurations on the selection of a PLL.

There were two objectives to this experiment:

1. Demonstrate that the use of earmuff HPDs over earbuds is a viable option for decreasing PLL for PMPs.

2. Determine if a surrounding background noise yields different PLLs than a single directional noise source.

METHOD

Participants

Twenty-four adults (13 male, 11 female) ages 18 to 38 yr (mean 5 27 yr) with normalhearingcompleted the study. Normalhearingwas defined as air conduction thresholds #20 dBhearinglevel (HL) in both ears at octave intervals between 250 and 8000 Hz and interoctaves of 3000 and 6000 Hz. The participants'hearingwas screened using a calibrated Maico MA 41 audiometer and supra-auralearphones(American National Standards Institute (ANSI), 1996).

All data collection was in compliance with regulationsfromthe Institutional Review Board at the U.S. Army Research Laboratory. Informed consent was obtainedfromeach participant prior to their participation in the research study. Nongovernment civilians were compensated for participation at a rate of $20 per hour.

Data Collection

Each participant was seated in the center of a 3.6m3 3.4 m sound-treated booth facing one of the walls, designated as 0. The reverberation time (RT60) measured in the room was 490 msec (Scharine et al, 2004), similar to that of a household living room. Four Infinity Studio Monitor 150 loudspeakers were positioned in the corners of the room and were directed toward the center, facing the participant. A personal computer (PC) and two Crown XS900 amplifiers were located outside of the room for the presentation of background noise. A laptop computer with 01dB-Metravib sound recording software (www.01dB-metravib.com) was used for data collection. Background noise consisted of a 10 sec segment of white noise that was loop played during each trial. The level of the background noise was 80 dBA as measuredfromthe center of the room. The level of the background noise was chosen to be high enough to warrant a listener to increase their PLL and below hazardous levels given the duration of the study.

Two Etymotic ER-7C probe tube microphones were connected to the laptop computer for recording in the participants' ear canals. Probe tubes were placed in the participants' ear canals at a depth of 25 mmfromthe intertragal notch and secured to the participant's earlobe with tape. The depth of 25 mm was selected to be beyond the length of the canalearphonesbut at a safe distance (at least 5mm)fromthe eardrum (Zemplenyi et al, 1985).

Participants were presented 18.5 sec of chorus (50msec rise/fall times)fromthe song "All Star" by Smash Mouth played through an Apple iPod 2nd Generation Nano. Typical stereo recordings, particularly music, have differences in the sounds provided to the two ears. Therefore, the two channels of the recording were mixed together into a single channel and then split to two channels (split monaural) to ensure that the same signal was provided to both ears.

Two types ofearphoneswere used for the study: SONY MDR-E10LP earbuds and SONY MDR-EX32 canalearphones. Figure 1 shows photos of each of theearphones. The canalearphonesincluded three sizes of eartips. The experimenter selected the appropriate eartip size according to the observed fit in the participant's ear canal. Theearphoneswere selectedfromthe same manufacturer in an effort to minimize differences in frequency response acrossearphones. Theearphoneswere purchasedfroma mass merchandiser and cost approximately $10 per pair.

The frequency response of theearphoneswas measured directly through use of a KEMAR, and white noise played out of a PC. The output of the KEMAR microphones and preamplifiers was routed to a laptop computer, and recordingsfromtheearphoneswere made through 01dB-Metravib software. Frequency analyses of the recordings revealed only small differences between the twoearphones(Fig. 2).

Earmuffs were used in one of theearphoneconfigurations in order to provide attenuationfromthe background noise separatefromthat provided by the earbuds. The E.A.R Model 3000 Earmuffs shown in Figure 3 were selected for this purpose as they are commonly used forhearingprotection in industrial settings.

Experimental Procedures

The participant was allotted time to become familiar with adjusting the volume setting on the PMP prior to data collection. The visual icon for the volume control on the PMP was obscured so that the listener could only see the volume icon and the first tick of the volume bar. Visualization of the first tick allowed the listener to determine if the volume was at "0" (no ticks shown) but did not allow them to rely on a visual reference for repeated PLL adjustments. The participants were instructed to turn the volume to "0" after each recording and not to adjust it again until instructed.

Participants selected PLLs for threeearphoneconfigurations: earbuds alone (Earbuds), canalearphones(CanalPhones), and earbuds in combination with the earmuffs (Earbuds1HPDs). Eachearphoneconfiguration was tested in each of three listening conditions: quiet (Q), noise through a single loudspeaker located at a 45 angle (N-ONE) to the listener's left side, and a surrounding noise emitted through all four loudspeakers (N-ALL). The experimental procedurewas the same for all test conditions, and the order of conditions was counterbalanced across participants.

Participants adjusted the PMP to their PLL for eachearphoneand noise combination. Five sec recordings of themusic weremade through the probemicrophones connected to the laptop computer. Earbuds or canalearphoneswere placed over top of the probe tubes, and in the case of the Earbuds1HPD condition, the earmuffs were placed over top of the earbuds. Once the participant listened in the noise condition, determined a PLL, and signaled to the investigator that they had made a decision, the background noise was stopped (if applicable). The process of establishing the PLL was repeated three times for each condition for a total of 27 trials. Each trial took approximately 1 min, and the total testing time was 1 hr.

Following recording of the music at the PLL, 5 sec recordings of the background noise in the listener's ear canals were made. Recordings of the music and noise were analyzed using 01dB-Metravib analysis software. Overall levels in dBA were calculated in the software through application of an A-weighted digital filter. The overall level of the music averaged over the three trials served as the dependent measure of PLL per condition. Fast Fourier transform (FFT) analyses were carried out for the noise recordings. For the N-ALL condition, the differences in levels in octave bands (250-8000 Hz) between the background noise recorded in the open ear configuration and that recorded in the noise condition served as the measure of attenuation. Recordings made of the noise also allowed for accurate calculations of overall SNR for each condition.

RESULTS

Table 1 shows the average PLL (in dBA) measured in the right ear canal in the various noise andearphonecombinations for male and female participants. The values labeled free field are the free-field equivalents calculated by applying the generic free-field to eardrum transfer function in ISO 11904-1 (ISO, 2004) to the values in 1/3-octave bands obtained through FFT analyses. As seen in Table 1, PLLs varied acrossearphoneand noise combinations. The average ear canal PLL was 74 dBA (SD 5 13.6) in the quiet conditions and 84 dBA (SD 5 12.6) averaged across the two noise conditions. Listeners had higher PLLs in noise compared to the PLLs in quiet. The male listeners selected PLLs that were higher on average than those selected by the female listeners. On average, listeners did not select PLLs that would be considered hazardous to theirhearingas evidenced by average free-field equivalent levels of 85 dBA or less (OSHA, 1983). The selection of PLLs equivalent to a free-field value of 90 dBA or greater are considered potentially hazardous to a listener'shearing, depending on the duration of exposure.

Recall that OSHA established damage risk criteria (DRC) as exposure to noise of 90 dBA for 8 hr or more. To determine the rate at which listeners selected potentially hazardous PLLs, we tallied the number of listeners who met or exceeded the DRC. One male exceeded the DRC in all conditions; and this was the only participant to reach it in quiet conditions. The number of listeners exceeding the DRC varied withearphonelistening condition. For the Earbuds, there were five participants (four male, one female) in the N-ALL condition and six (five male, one female) participants in the N-ONE condition. The use ofearphonesthat incorporated some type of noise attenuation resulted in fewer listeners exceeding the DRC. With the CanalPhones, there were three male participants in the N-ALL condition and two male participants in the N-ONE condition who exceeded the DRC. In the Earbud1HPD condition, only one male exceeded theDRCin theN-ALLnoise condition and fourmale participants in the N-ONE condition.

A two-factor, repeated-measures analysis of variance (ANOVA) was conducted on the PLLs measured in the right ear with noise condition andearphoneconfiguration as independent variables. All statistical tests were considered significant ata 5 .05. The analysis revealed a significant effect of listening condition, F(2, 46) 5 29.36, p , .001, with larger PLLs measured in noise than in quiet. There was also a significant effect ofearphoneconfiguration, F(2, 46) 5 13.65, p , .001, with the highest PLLs in the Earbud configuration and the lowest PLLs in the Earbud1HPD and CanalPhone configurations, and a significant interaction between listening condition andearphoneconfiguration, F(4, 92) 5 4.50, p , .01. The interaction can be explained by the significant differences between each of the noise conditions when compared to the quiet condition but no significant differences in PLL between the two noise conditions.

Increases in PLL in Noise

Increases in PLL were calculated for the right ear for each of theearphoneand noise combinations by subtracting the PLL selected in quietfromthat selected in noise for eachearphoneconfiguration. The average increase in PLL was 9.4 dB (SD 5 9.6) with the largest average increases observed for the Earbuds configuration (12.5 dB;SD59.6), the smallest increases observed for the Earbuds1HPDs configuration (5.4 dB; SD58.4), and theCanalPhones falling between the two (10.4 dB; SD 5 9.4).

Surrounding versus Directional Noise

To address the second question regarding differences between the two background noise conditions, a twofactor, repeated-measures ANOVA was conducted on the PLLs in the two noise conditions across the threeearphoneconfigurations. Results indicated only a significant effect ofearphoneconfiguration, F(2, 46) 5 12.94, p , .001. The lack of a significant effect of noise condition or interaction indicates that there was no significant difference in the PLLs between the N-ONE (directional) and N-ALL (surrounding) noise conditions. The effect size for the comparison of noise conditions was calculated as Cohen's d and was determined to be a small effect at .09.

Effects ofEarphoneConfiguration

Post hoc comparisons using Tukey's Honestly Significant Difference (HSD) test indicated that the average PLL selected for the Earbuds (M 5 84.18, SD 5 13.4) was significantly higher than the PLL selected for the Earbuds1HPDs (M 5 78.69, SD 5 12.7) and the Canal- Phones (M577.47, SD514.3) but that there was no significant difference between the PLLs selected for the Earbuds1HPDs and the CanalPhones. These findings indicate that individuals preferred a significantly higher listening level when using the Earbuds than when they used either the Earbuds1HPDs or the CanalPhones.

Signal-to-Noise Ratios

Table 2 shows the average noise levels measured in each listening condition andearphonecombination. The differences between the music levels and the overall noisemeasured in the ear canals were calculated for each noise andearphonecombination and represent the best estimate of SNR. The SNRs were calculated for each ear separately to document the differences between the ears for the two noise conditions. In all of theN-ALL noise conditions, the noise levels measured in the two ear canals were nearly identical, reflecting a diffuse noise environment. In all of the N-ONE noise conditions where the loudspeaker was located at a 45 angle to the listener's left side, there were differences of 5 to 10 dB in the noise levels measured in the two ears. This is consistent with the anticipated head shadow effect where the noise in the left (near) ear was higher than the noise in the right (far) ear. In all noise configurations, the levels of the music were the same in each ear due to themixing and splitting of the .wav file in preparation for the study. As seen in Table 3, the average SNR for the N-ALL noise condition was highest for the Earbuds1HPDs configuration and lowest for the Earbuds and CanalPhones configurations. In theN-ONE noise condition, the SNRs followed the same pattern as the N-ALL noise condition with the Earbuds1 HPDs resulting in the most favorable SNR and the Canal- Phones and Earbuds resulting in the least favorable SNR.

Attenuation

During the study, recordings were made of the noise arriving in the listener's ear canal for each noise andearphonecombination. One-third-octave band FFT analyses were conducted for each individual in eachearphoneand noise combination through 01dB-Metravib sound analysis software. Figure 4 shows the average attenuation (calculated as the difference between the noise in theearphonecondition and the open ear condition) along with the estimates of variability.

Following the calculations of attenuation and analyses of the PLLs, the relationship between attenuation and PLL was explored. The correlation between individual attenuation provided by theearphoneconfiguration and the individual increases in PLL was conducted to determine if therewas a significant change in PLL as a function of attenuation. The correlation was found to be significant (r 5 2.40, p , .01), indicating that as attenuation increased, there were smaller increases in PLL.

DISCUSSION

There were two objectives to this experiment. The first was to determine if an alternative to in-theearearphonesfor noise attenuation (the addition of earmuffs to earbuds) would allow for lower PLLs through a PMP than earbuds. The second was to determine if a surrounding background noise would yield different PLLs than a directional noise source. There is little argument that users of PMPs increase their PLL when in the presence of background noise as compared to quiet, but we were interested in some of the reasons for the increase and if attenuating the background noise through an alternative method to in-the-earearphoneswould affect the degree of increase.

Effects ofEarphoneConfiguration and Attenuation

Listeners in our study set their PLL to 74 dBA (freefield equivalent 5 65.6 dBA) in quiet, which agreed well with those found by other authors: 75-78 dBA (Hodgetts et al, 2007), 72 dBA (Hodgetts et al, 2009), and 69 dBA (free-field equivalent) (Airo et al, 1996). Listeners in our study selected an average PLL of 84 dBA (freefield equivalent 5 75.1 dBA) in noise, which agreed well with those found by other authors: 81-89 dBA (Hodgetts et al, 2007), 89 dBA (Hodgetts et al, 2009); and 82-85 dBA (free-field equivalent values) (Airo et al, 1996).

In our study, males selected PLLs that were higher on average than those selected by female listeners. This is in agreement with prior studies that have found that male listeners tend to select higher PLLs than females (Torre, 2008).

Our finding confirmed those of previous investigators that found that listeners significantly increase their PLL in the presence of background noise as opposed to quiet listening environments. On average, our participants increased their PLL by 10 dB in the presence of noise. This is in line with increases in PLL foundfromprevious studies, rangingfrom4 to 20 dB (Rice et al, 1987; Airo et al, 1996; Williams, 2005; Fligor and Ives, 2006; Ahmed et al, 2007; Hodgetts et al, 2007; Torre, 2008; Hodgetts et al, 2009).

Our study demonstrated a significant effect ofearphoneconfiguration that follows the levels of attenuation provided by the differentearphones. The amount of attenuation provided by theearphonesis shown in Figure 4. The attenuation was measured objectively through probe tube recordings in the different noise andearphonecombinations. As seen in the figure, the attenuation averaged 5 dB for the Earbuds, 13 dB for the CanalPhones, and 20 dB for the Earbuds1HPDs. The increase in PLL was highest for the Earbuds condition (12 dB), smaller for the CanalPhones condition (11 dB), and lowest for the Earbuds1HPDs condition (4 dB). Comparing the attenuation provided by the differentearphonesto the differences in levels selected in quiet and noise revealed that the increase in PLL was higher when the attenuation was lower, and the correlation for this relationship was significant.

The finding of smaller increases in PLL withearphonesthat provide attenuation of the background noise is not a new finding. Fligor and Ives (2006) indicated that use of in-the-earearphonesallowed for lower PLLs in background noise. However, theFligor and Ives study did not control for potential differences in themethod of attenuation provided by theearphonesor the frequency response of theearphones. Either of these factors could have contributed to their findings.Furthermore, no directmeasuresweremade of the noise levels arriving to the listener; only an estimate of attenuationwas provided. The present study extends the Fligor and Ives finding of lower PLLs with attenuatingearphonesand documents that when frequency response is controlled and the level of the music provided to the two ears is also controlled, listeners select a smaller PLL. The PLL that is selected is correlated with the amount of attenuation provided by theearphones.

The finding of lower PLLs in noise with attenuatingearphonesagrees with Hodgetts et al (2007) who found that over-the-earearphonesresulted in a smaller increase in PLL than earbuds and that the incorporation of noise reduction in the over-the-earearphonesresulted in even smaller increases in PLL. The passive process of covering the ears resulted in lower PLLs, and the addition of noise reduction resulted in further benefit to the listener as seen in lower PLL selection. Although earbuds do not provide sufficient attenuationfrombackground noise, over-the-earearphones, in-the-earearphones, noise reduction alogorithms, or HPDs do.

Differences exist betweenearphonesused in the literature in cost and method of attenuation. The overthe- earearphonesused in the Hodgetts et al (2009) study are no longer available, but when they were, they cost around $25. It appears that over-the-earearphoneshave dropped in popularity among consumers; users preferearphonesthat sit in their ear canals. The placement ofearphonesin the ears allows for less interference with sunglasses, hats, and so on. The ER-6i noise isolatingearphonesused in the Fligor and Ives studyfromEtymotic are rather expensive (z$100). By comparison, theearphonesused in the present study were rather inexpensive, costing about $10 and purchasedfroma mass merchandiser.

The smaller increase in PLL between quiet and noise for the CanalPhones over the Earbuds indicates that use of CanalPhones can significantly reduce a listener's overall noise exposure. The fact that the CanalPhones were purchasedfroma mass merchandiser suggests that they are widely available, and the cost of the Canal- Phones is reasonable compared to the cost of the PMP itself. Therefore, use of in-the-earearphonesinstead of the Earbuds that are provided with thePMPis a simple, inexpensive step in reducing noise exposure.

In selecting appropriateearphones, the greater the degree of attenuation provided, the smaller the anticipated increase in PLL in background noise. Depending on the listener's typical listening situation, selection of in-the-earearphoneswith moderate degrees of attenuation may be advantageous. Although the use of earmuffs is not practical and would be rejected by most users in the same way that over-the-earearphoneshave dropped in popularity, use of HPDs is viable, and this study demonstrates that attenuation of background noise through any means would allow listeners to select lower PLLs.

Directional versus Surrounding Noise

Our second question was related to differences in directionality of background noise. It was hypothesized that listenerswould select different PLLs in surrounding noise environments than in environments where noise originatesfromonly one location. In surrounding noise, both the music (signal) and the noise would be the same level at both ears, resulting in equal SNRs between the ears. In a directional noise, although the music level may be the same level in both ears, the noise level arriving to the two ears could differ, such that the noise level would be higher to the near ear and lower to the far ear. This would result in different SNRs between the ears, which could influence the listener's selection of a PLL.

If differences in background noise directionality would affect a listener's selection of a PLL,wewould have found significant differences between the PLLs selected in the two noise environments. However, this was not the case; the PLLs selected between the two noise conditions were not significantly different. The small effect size indicates that any differences that would be found would not be meaningful in everyday situations. Some real-world environments may involve a directional noise that may be off-axis to the listener. In the configuration of a single noise source located at 0, two loudspeakers located at 45 and 315, or a configuration of a surrounding noise source, the SNR for the music and noise would likely be the same in both ears. In the case of a single noise source located off-axis, the SNRs between the two ears would not be equal. One would wonder, in this case, how a listener would go about selecting their PLL. It was possible that they would either increase the SNR in one ear for their desired SNR or accept some sort of compromise between the SNRs in the two ears.

The levels of noise measured in the open ear conditions demonstrate that in the N-ALL conditions, the noise arriving at the two ears was nearly equivalent. In contrast, in the N-ONE conditions, the head served to protect the far earfromnoise. The average difference in noise levels measured in the two ears in the open canal conditions was 6 dB, and this difference was maintained in allearphoneconditions. We expected that the listener would experience different SNRs in their ears in the two noise conditions. Specifically, we assumed that the noise arriving to the two ears in the N-ALL condition would be equal and that we would measure a substantial difference in the noise levels between the two ears in the N-ONE condition. What was surprising was that the average PLL selected in the two noise conditions did not differ significantly. We had assumed that listeners would increase the PLL in the N-ONE condition to a larger degree than the N-ALL condition in order to optimize their SNR. Presumably, the listener would use the ear closer to the noise to determine the PLL setting and this would result in higher PLLs in the N-ONE condition than the N-ALL condition. However, no difference was found in the PLLs selected in the two conditions In Table 3 (SNRs), the average of the SNRs for the two ears in the N-ONE conditions is approximately equal to the SNRs obtained in the N-ALL conditions. It is possible that listeners aimed for an average SNR between their two ears for the N-ONE condition to be equal to that experienced in the NALL condition.

Listeners in the present study set the SNR at an average of 19 dB in the surrounding noise conditions (NALL) for the Earbuds and CanalPhones and around 113 dB for the Earbuds1HPDs, which agrees well with several previous studies (Airo et al, 1996; Williams, 2005). Airo et al measured a range of SNRsfrom112 to 117 dB, and Williams calculated an SNR of 113 dB based on the difference in background noise and music levels as measuredfroma KEMAR. Our SNR measure is a bit high compared to Rice et al (1987), who found an SNR of 14 dB, but this is likely due to the differences in background noise level between the two studies. Rice et al used a background noise of 70 dBA, and ours was 80 dBA.

A previous study (Hodgetts et al, 2009) stated that use of a single loudspeaker in a sound-treated booth resulted in a relatively diffuse noise environment for the listener as measured at the center of the room with the listener absent. Our objective measures indicate that the presence of the head disrupts the diffuseness of the sound field when a single loudspeaker is located to the side (315) sufficiently to create differences in the noise levels arriving at the two ears. Table 2 shows the overall levels of the noisemeasured in the ear canals for each noise andearphonecombination. As seen in the table, there is a substantial difference between the two ears for the open ear condition in the noise levels measured in the N-ONE noise condition but not in the N-ALL condition. In the open ear scenario, the noise in the close ear was nearly 6 dB higher than in the far ear. These differences fade away once theearphonesare placed on the ears. The levels of the music remain the same regardless of noise condition, as was expected based on the preparation of the music stimulus to be a split monaural signal. The lack of a difference in PLL setting between the N-ONE and N-ALL conditions suggests that the differences in noise arriving at the two ears was not sufficient to cause the listener to change his or her PLL.

In summary, individual listeners significantly increased their PLL in noise compared to where they set it in quiet. Furthermore, individuals increased their PLL to a significantly lesser extent in the two conditions where attenuation of background noise was provided through either the HPDs or the canalearphonesthan the condition where essentially no attenuation was provided (earbuds). The greater the attenuation of the background noise, the lower the PLL in noisy listening environments.

CONCLUSION

The findingsfromthis study demonstrate that listeners increase their PLL in background noise, some to levels considered potentially hazardous. Our study supports the notion found by others that if a listener usesearphonesthat attenuate background noise, they will increase their PLL to a lesser extent than if they are using earbuds. An alternative method of obtaining attenuation without the use of in-the-earearphonesis the addition of HPD earmuffs over earbuds. Within the ranges examined in the present study, the greater the attenuation provided, the smaller the increase in PLL.

The current cost of noise-attenuating in-the-earearphonesis quite reasonable. PMP users who aim to decrease their noise exposure have an inexpensive option for noise attenuatingearphonesthat would allow them to listen to their music at a lower level when in the presence of background noise. Audiologists should query their patients about PMP use and recommend the use of in-the-earearphonesfor decreasing noise exposure.

ReferencesREFERENCES

Ahmed S, Fallah S, Garrido B, et al. (2007) Use of portable audio devices by university students. Can Acoust 35:35-52.

Airo E, Pekkarinen J, Olkinuora P. (1996) Listening to music withearphones: an assessment of noise exposure. Acta Acustica united with Acustica 82:885-894.

American National Standards Institute (ANSI). (1996) Specifications for Audiometers. ANSI S3.6-1996. New York: ANSI.

Apple. (2009) Apple Reports First Quarter Results. Retrieved July 7, 2009 http://www.apple.com/pr/library/2009/01/21results.html (accessed July 7, 2009).

Berger E, Megerson S, StergarM. (2009) Personalmusic players: are we measuring the sound levels correctly? August 11 The ASHA Leader.

Catalano PJ, Levin SM. (1985) Noise-inducedhearing lossand portable radioswith headphones. Int J PediatrOtorhinolaryngol 9:59-67.

Clark W. (1991) Noise exposurefromleisure activities: a review. J Acoust Soc Am 90:175-181.

Danhauer JL, Johnson CE, Byrd A, et al. (2009) Survey of college students on iPod use andhearinghealth. J Am Acad Audiol 20:5-27.

Farina A. (2007) A study ofhearingdamage caused by personal MP3 players. Proceedingsfromthe 123rd Convention of the Audio Engineering Society, New York.

Fligor BJ, Cox CL. (2004) Output levels of commercially available portable compact disc players and the potential risk tohearing. Ear Hear 25:513-527.

Fligor BJ, Ives T. (2006) Doesearphonetype affect risk for recreational noise-inducedhearing loss? Paper presented at the Noise InducedHearing Lossin Children at Work and Play Conference, Covington, KY. http://www.hearingconservation.org/displaycommon. cfm?an51&subarticlenbr5112.

Gaba C. (2008) iPod Sales: Quarterly and Total. http://www. systemshootouts.org/ipod_sales.html (accessed July 7, 2009).

Hodgetts WE, Rieger JM, Szarko RA. (2007) The effects of listening environment andearphonestyle on preferred listening levels of normalhearingadults using an MP3 player. Ear Hear 28:290-297.

Hodgetts WE, Szarko RA, Rieger JM. (2009) What is the influence of background noise and exercise on the listening levels of iPod users? Int J Audiol 48:825-832.

Hoover A, Krishnamurti S. (2010) Survey of college students' MP3 listening: habits, safety issues, attitudes, and education. Am J Audiol 19:73-83.

International Organization for Standardization (ISO). (2002) Technique Using a Microphone in a Real Ear (MIRE Technique). ISO 11904-1. Part 1 of Acoustics: Determination of Sound EmissionfromSources Placed Close to the Ear. Geneva: ISO.

International Organization for Standardization (ISO). (2004) Technique Using a Manikin. ISO 11904-2. Part 2 of Acoustics: Determination of Sound EmissionfromSound Sources Placed Close to the Ear. Geneva: ISO.

Kumar A, Mathew K, Alexander S, Kiran C. (2009) Output sound pressure levels of personal music systems and their effect onhearing. Noise Health 11:132-140.

Levey S, Levey T, Fligor B. (2011) Noise exposure estimates of urban MP3 players. J Speech Lang Hear Res 54(1):263-277.

Occupational Safety and Health Administration (OSHA). (1983) Occupational noise exposure:hearingconservation amendment; final rule. Fed Regist 48:9738-9785.

Portnuff C, Fligor B. (2006) Sound output levels of the iPod and other MP3 players: is there potential risk tohearing? Paper presented at the Noise InducedHearing Lossin Children at Work and Play Conference, Covington, KY. http://www.hearingconservation. org/displaycommon.cfm?an51&subarticlenbr5112.

Rice C, Breslin M, Roper R. (1987) Sound levelsfrompersonal stereo players. Br J Audiol 21:273-278.

Scharine A, Tran P, Binseel M. (2004) ARL Acoustic Measurements in Buildings 518 and 520 at APG. U.S. Army Research Laboratory Memorandum Report ARL-MR-0580. Aberdeen Proving Ground, MD: U.S. Army Research Laboratory.

Shah S, Gopal B, Reis J, Novak M. (2009) Hear today gone tomorrow: an assessment of portable entertainment player use andhearingacuity in a community sample. J Am Board Fam Med 22:17-23.

Torre P, III. (2008) Young adults' use and output level settings of personal music systems. Ear Hear 29:791-799.

Turunen-Rise I, Flottorp G, Tvete O. (1991) Personal cassette players ('Walkman'). Do they cause noise-inducedhearing loss? Scand Audiol 20:239-244.

Williams W. (2005) Noise exposure levelsfrompersonal stereo use. Int J Audiol 44:231-236.

Williams W. (2009) Trends in listening to personal stereos. Int J Audiol 48:784-788.

Wood W, Lipscomb D. (1972) Maximum available sound-pressure levelsfromstereo components. J Acoust Soc Am 52:484-487.

Zemplenyi J, Gilman S, Dirks D. (1985) Optical method for measurement of ear canal length. J Acoust Soc Am 786:2146-2148.

Zogby J. (2006) Survey of teens and adults about the use of personal electronic devices and head phones. http://www.asha.org/ (accessed on July 2, 2009).

AuthorAffiliationDOI: 10.3766/jaaa.23.3.5

Paula Henry*

Ashley Foots*

* U.S. Army Research Laboratory, Aberdeen Proving Ground, MD

Paula Henry, RDRL-HRS-D, Building 520, Room 22, Aberdeen Proving Ground, MD 21005-5425; Phone:Portions of this manuscript were presented as a poster at AudiologyNOW! 2010, San Diego, CA.

Word count:7321Copyright American Academy of Audiology Mar 2012

Indexing (details)

CiteSubject

Hearingimpairment;Occupational safety;Ears &hearing;NoiseMeSH

Adolescent,Adult,Ear Protective Devices,Female,Humans,Male,Signal-To-Noise Ratio,Young Adult,Hearing-- physiology(major),Hearing Loss, Noise-Induced -- prevention & control(major),Loudness Perception(major),MP3-Player -- standards(major),Music(major),Noise -- adverse effects(major)Company / organization

Name:

Occupational Safety & Health Administration--OSHANAICS:

926150Title

Comparison of User Volume Control Settings for Portable Music Players with ThreeEarphoneConfigurations in Quiet and Noisy Environments

Author

Henry, Paula;Foots, AshleyPublication title

Journal of the American Academy of AudiologyVolume

23Issue

3Pages

182-91Number of pages

10Publication year

2012Publication date

Mar 2012Year

2012Publisher

American Academy of AudiologyPlace of publication

McLeanCountry of publication

United StatesPublication subject

Medical Sciences--OtorhinolaryngologyISSN

10500545Source type

Scholarly JournalsLanguage of publication

EnglishDocument type

Feature,Comparative StudyDocument feature

Photographs;Graphs;References;TablesAccession number

22436116ProQuest document ID

1017871726Document URL

http://search.proquest.com/docview/1017871726?accountid=49910Copyright

Copyright American Academy of Audiology Mar 2012Last updated

2013-02-24Database

ProQuest Medical Library