Masking
“Of all the clinical procedures used in auditory assessment, masking is probably the most often misused and the least understood. For many clinicians the approach to masking is a haphazard hit-or-miss bit of guesswork with no basis in any set of principles.”
Purpose
When a stimulus is presented to one ear of a listener, there are occasions when the sound may be heard in the other ear due to bone conduction of the signal. The "fix" for this problem is to use noise in the opposite ear to mask the signal, so that the tester knows that the patient responses are due to hearing the signal in the ear being tested. Masking allows us to accurately evaluate the degree and type of hearing loss in each ear independently.
The test ear (TE) is the ear on which an audiogram is being determined. The non-test ear (NTE) is the ear not being tested. The goal of masking is to eliminate the ability of the non-test ear to "hear" the test signal by stimulating it with noise intense enough to disguise or "mask" the test signal in the non-test ear. Masking ensures that, when a test signal is presented to the patient, the response of the patient is due to the test signal being perceived in the test ear instead of in the non-test ear.
When testing hearing in the test ear, some sounds may be intense enough to stimulate the non-test ear. This is called crossover. Crossover happens by bone conduction. When the sound in the test ear is of sufficient intensity, it vibrates the bones of the skull, thereby stimulating the cochlea of the non-test ear.
When the intensity of the sound that reaches the non-test ear exceeds the bone-conduction threshold in that ear, the sound will be perceived. This is called cross hearing
The test ear (TE) is the ear on which an audiogram is being determined. The non-test ear (NTE) is the ear not being tested. The goal of masking is to eliminate the ability of the non-test ear to "hear" the test signal by stimulating it with noise intense enough to disguise or "mask" the test signal in the non-test ear. Masking ensures that, when a test signal is presented to the patient, the response of the patient is due to the test signal being perceived in the test ear instead of in the non-test ear.
When testing hearing in the test ear, some sounds may be intense enough to stimulate the non-test ear. This is called crossover. Crossover happens by bone conduction. When the sound in the test ear is of sufficient intensity, it vibrates the bones of the skull, thereby stimulating the cochlea of the non-test ear.
When the intensity of the sound that reaches the non-test ear exceeds the bone-conduction threshold in that ear, the sound will be perceived. This is called cross hearing
Examples of Why Masking is Necessary
The following examples demonstrate why masking is necessary in audiometric evaluation. Imagine that a patient has normal hearing in the left ear and no hearing in the right ear (a "dead" ear). The audiogram may show no responses to masked air or bone conduction stimuli for the right ear at equipment limits.
Using insert earphones, the intensity of the signal presented by air conduction eventually will be great enough that the bones of the skull will vibrate, and the signal will be heard in the non-test (left) ear. Using the bone vibrator, the signal is heard in the non-test (left) ear at the same intensity level as if the stimulus were presented from the left mastoid because interaural attenuation can be as low as 0 dB for a bone-vibrator transducer. The resulting audiogram suggests that the patient has a severe conductive hearing loss in the right ear. This is not true. The presence of unmasked thresholds that occur as a result of a sound being heard by crossover is known as a shadow curve.
The clinician must determine whether the hearing loss in the right ear is sensorineural, conductive, or mixed. To determine this, it is necessary to have accurate bone-conduction thresholds for the right ear. Without masking, the bone-conduction responses with the bone vibrator on the right mastoid are in the normal range. It may be that these thresolds are true and that the patient has a conductive hearing loss. Alternatively, it may be that the right bone-conduction thresholds are true and that the patient has a conductive hearing loss. Alternatively, it may be that the right bone-conduction thresholds are higher than these levels, reflecting a mixed or sensorineural hearing loss, but the bone-conducted signal is being perceived by the patient in the left ear, due to cross hearing. It is impossible to be certain of the validity of the right ear bone-conduction thresholds without masking the left ear.
Masking the left ear allows the clinician to determine that the hearing loss is conductive or sensorineural.
Using insert earphones, the intensity of the signal presented by air conduction eventually will be great enough that the bones of the skull will vibrate, and the signal will be heard in the non-test (left) ear. Using the bone vibrator, the signal is heard in the non-test (left) ear at the same intensity level as if the stimulus were presented from the left mastoid because interaural attenuation can be as low as 0 dB for a bone-vibrator transducer. The resulting audiogram suggests that the patient has a severe conductive hearing loss in the right ear. This is not true. The presence of unmasked thresholds that occur as a result of a sound being heard by crossover is known as a shadow curve.
The clinician must determine whether the hearing loss in the right ear is sensorineural, conductive, or mixed. To determine this, it is necessary to have accurate bone-conduction thresholds for the right ear. Without masking, the bone-conduction responses with the bone vibrator on the right mastoid are in the normal range. It may be that these thresolds are true and that the patient has a conductive hearing loss. Alternatively, it may be that the right bone-conduction thresholds are true and that the patient has a conductive hearing loss. Alternatively, it may be that the right bone-conduction thresholds are higher than these levels, reflecting a mixed or sensorineural hearing loss, but the bone-conducted signal is being perceived by the patient in the left ear, due to cross hearing. It is impossible to be certain of the validity of the right ear bone-conduction thresholds without masking the left ear.
Masking the left ear allows the clinician to determine that the hearing loss is conductive or sensorineural.
How does masking work?
A pure-tone signal is a signal of a particular frequency, such as a 1000 Hz tone of hypothetical amplitude. A masking noise is a band of noise with a particular center frequency. In our example, the test signal is 1000Hz, and the masking noise is composed of a narrow band of frequencies around 1000Hz.
The masking noise works to "cover up" the test signal. The test signal is still reaching the non-test ear due to crossover and can still be heard by cross hearing but the masking noise does not allow the patient to perceive the test signal because it is embedded in the noise.
When using masking, crossover and cross hearing still are occurring, but the masking noise is covering up the test signal so that the patient does not respond to the signal in the non-test ear. Because crossover and cross hearing are still ocurring, the masking noise needs to be sufficient to mask the test signal.
The masking noise works to "cover up" the test signal. The test signal is still reaching the non-test ear due to crossover and can still be heard by cross hearing but the masking noise does not allow the patient to perceive the test signal because it is embedded in the noise.
When using masking, crossover and cross hearing still are occurring, but the masking noise is covering up the test signal so that the patient does not respond to the signal in the non-test ear. Because crossover and cross hearing are still ocurring, the masking noise needs to be sufficient to mask the test signal.
Interaural Attenuation
Crossover occurs by bone conduction. The test signal vibrates the bones of the skull, stimulating the cochlea of the non-test ear. When the sound energy vibrates the bones of the skull, some sound energy is absorbed as the vibratory energy passes through the mass of the skull. The amount of sound energy lost as a result of this absorption is called interaural attenuation (IA). The amount of interaural attenuation for a given patient can be determined by subtracting a given patient can be determined by subtracting the intensity of the signal reaching the non-test ear from the intensity of the test signal. The equation for this is
Intensity of test signal - intensity of signal reaching non-test ear = Interaural Attenuation
In an example, an 80 db HL signal is presented to the test ear. 20 dB HL crosses over to the non-test ear. This means that 60 dB of sound energy was lost during crossover. The interaural attenuation is 60 dB.
Knowledge of the amount of interaural attenuation is necessary for determining whether the test signal is capable of stimulating the non-test ear. If the intensity level of test signal minus the interaural attenuation value is greater than or equal to the bone-conduction threshold of the non-test ear, then the test signal is capable of being heard by crossover and masking must be used.
An 80 dB HL signal is presented to the test ear. 20 dB crosses over to the non-test ear. This means that 60 dB of sound energy was lost during crossover and interaural attenuation is 60 dB. Now, consider what is happening in the non-test ear. It is known that 20 dB has crossed over to the non-test ear. But is the patient actually hearing the sound in the non-test ear? In other words, is cross hearing occurring? That depends on the patient's bone-conduction threshold for the non-test ear. Why the bone-conduction threshold? Because crossover happens by bone conduction. So, assume that the patient has a bone-conduction threshold of 30 dB HL in the non-test ear. Is the test signal being heard in the non-test ear? The answer is "No." Only 20 dB HL of sound intensity reached the non-test ear and the patient requires at least 30 dB HL to hear the sound.
Assume now that the patient has a bone-conduction threshold of 10 dB HL in the non-test ear. Is the test signal being heard in the non-test ear? The answer is "Yes." There is 20 dB HL of sound intensity reaching the the non-test ear. This is more intense than the softest sound that the patient can hear. The patient perceives the sound and will respond.
To review, the interaural attenuation value is used to determine whether the sound will reach the other cochlea via crossover. If the sound reaches the non-test cochlea, then we must understand whether the sound may be heard by the non-test ear, which depends on the bone-conduction threshold of the non-test ear and how much the interaural attenuation has attenuated the intensity of the test signal. If the sound may be heard in the non-test ear, masking of the non-test ear is necessary.
Interaural attenuation values vary depending on several factors. The amount of interaural attenuation depends on the transducer that is used. Bone vibrators hae the lowest interaural attenuation. Supra-aural earphones have intermediate values of interaural attenuation. Insert earphones have the highest interaural attenuation. So, a sound is most likely to be heard by crossover when using a bone vibrator. A sound is least likely to be heard by crossover when using insert earphones.
The amount of interaural attenuation also varies depending on the frequency of the test signal. In general, lower frequency sounds have lower interaural attenuation and higher frequency sounds have higher interaural attenuation, although this relationship is somewhat different for insert earphones.
The amount of interaural attenuation also varies somewhat for each individual so there is a range of values that occur as a result of individual differences.
When performing a hearing test, the actual interaural attenuation value for a given patient is unknown. To ensure that the non-test ear is not accidentally tested, we use minimum interaural attenuation values. A minimum interaural attenuation value is the smallest interaural attenuation value for a given frequency and transducer type. These values have been determined empirically (observation).
Note how low the minimum interaural attenuation values are for the bone vibrator: 0 dB across the frequency range. Why is the minimum interaural attenuation value for the bone vibrator so low? It is because crossover and cross hearing happen via bone conduction. When the bone vibrator is directly causing vibration of the skull, it takes very little extra power to reach either cochlea.
As stated before, if the intensity level of the test signal minus the interaural attenuation value is greater than or equal to the bone conduction threshold of the non-test ear, then the test signal is capable of being heard by crossover, and masking must be used. However, during clinical testing, we do not know what the actual interaural attenuation value is for a patient. So, we use the minimum interaural attenuation values to determine when masking is needed. The rule then becomes as follows: When the difference between the test signal and the bone conduction threshold for the non-test ear is greater than or equal to the minimum interaural attenuation value for the test signal and the transducer used, then we must assume that there is a possibility that the test signal could be heard by crossover in the non-test ear and masking must be used.
Intensity of test signal - intensity of signal reaching non-test ear = Interaural Attenuation
In an example, an 80 db HL signal is presented to the test ear. 20 dB HL crosses over to the non-test ear. This means that 60 dB of sound energy was lost during crossover. The interaural attenuation is 60 dB.
Knowledge of the amount of interaural attenuation is necessary for determining whether the test signal is capable of stimulating the non-test ear. If the intensity level of test signal minus the interaural attenuation value is greater than or equal to the bone-conduction threshold of the non-test ear, then the test signal is capable of being heard by crossover and masking must be used.
An 80 dB HL signal is presented to the test ear. 20 dB crosses over to the non-test ear. This means that 60 dB of sound energy was lost during crossover and interaural attenuation is 60 dB. Now, consider what is happening in the non-test ear. It is known that 20 dB has crossed over to the non-test ear. But is the patient actually hearing the sound in the non-test ear? In other words, is cross hearing occurring? That depends on the patient's bone-conduction threshold for the non-test ear. Why the bone-conduction threshold? Because crossover happens by bone conduction. So, assume that the patient has a bone-conduction threshold of 30 dB HL in the non-test ear. Is the test signal being heard in the non-test ear? The answer is "No." Only 20 dB HL of sound intensity reached the non-test ear and the patient requires at least 30 dB HL to hear the sound.
Assume now that the patient has a bone-conduction threshold of 10 dB HL in the non-test ear. Is the test signal being heard in the non-test ear? The answer is "Yes." There is 20 dB HL of sound intensity reaching the the non-test ear. This is more intense than the softest sound that the patient can hear. The patient perceives the sound and will respond.
To review, the interaural attenuation value is used to determine whether the sound will reach the other cochlea via crossover. If the sound reaches the non-test cochlea, then we must understand whether the sound may be heard by the non-test ear, which depends on the bone-conduction threshold of the non-test ear and how much the interaural attenuation has attenuated the intensity of the test signal. If the sound may be heard in the non-test ear, masking of the non-test ear is necessary.
Interaural attenuation values vary depending on several factors. The amount of interaural attenuation depends on the transducer that is used. Bone vibrators hae the lowest interaural attenuation. Supra-aural earphones have intermediate values of interaural attenuation. Insert earphones have the highest interaural attenuation. So, a sound is most likely to be heard by crossover when using a bone vibrator. A sound is least likely to be heard by crossover when using insert earphones.
The amount of interaural attenuation also varies depending on the frequency of the test signal. In general, lower frequency sounds have lower interaural attenuation and higher frequency sounds have higher interaural attenuation, although this relationship is somewhat different for insert earphones.
The amount of interaural attenuation also varies somewhat for each individual so there is a range of values that occur as a result of individual differences.
When performing a hearing test, the actual interaural attenuation value for a given patient is unknown. To ensure that the non-test ear is not accidentally tested, we use minimum interaural attenuation values. A minimum interaural attenuation value is the smallest interaural attenuation value for a given frequency and transducer type. These values have been determined empirically (observation).
Note how low the minimum interaural attenuation values are for the bone vibrator: 0 dB across the frequency range. Why is the minimum interaural attenuation value for the bone vibrator so low? It is because crossover and cross hearing happen via bone conduction. When the bone vibrator is directly causing vibration of the skull, it takes very little extra power to reach either cochlea.
As stated before, if the intensity level of the test signal minus the interaural attenuation value is greater than or equal to the bone conduction threshold of the non-test ear, then the test signal is capable of being heard by crossover, and masking must be used. However, during clinical testing, we do not know what the actual interaural attenuation value is for a patient. So, we use the minimum interaural attenuation values to determine when masking is needed. The rule then becomes as follows: When the difference between the test signal and the bone conduction threshold for the non-test ear is greater than or equal to the minimum interaural attenuation value for the test signal and the transducer used, then we must assume that there is a possibility that the test signal could be heard by crossover in the non-test ear and masking must be used.
Materials
Audiometer and transducer to be able to apply masking noise to the non-test ear.
Procedures/Instructions
Audiometer Controls
Stimuli are presented via one channel of the audiometer and masking noise is presented via the other channel. The intensity of the masking noise can be raised or lowered independently of the test ear signal using the attenuator dial. The type of masking noise is dependent on the type of signal being presented to the patient. Narrowband noise signals are used for masking of tonal stimuli. On most audiometers, the frequency band of the narrow band noise changes automatically to match the frequency of the tone selected for presentation. Speech-shaped noise is used to mask speech stimuli.
The intensity level of masking noise stimuli are calibrated to an effective masking level (EML). This level has been predetermined to be a sound pressure level (SPL) that effectively masks a signal of a given intensity in decibel effective masking level (dB EML). For example, a 30 dB EML narrowband noise centered at 1000 Hz effectively masks a 30 dB HL pure tone at 1000 Hz.
When determining threshold using masking noise, the masking noise is introduced using the "continuously on" control. This allows the masking noise to be presented continuously, while the test stimuli are presented continuously, while the test stimuli are presented intermittently. If the masking noise was presented only with the test stimuli, the patient would be alerted to the presentation. Therefore, the masking noise must be presented continously to the non-test ear while threshold is established in the test ear.
The intensity level of masking noise stimuli are calibrated to an effective masking level (EML). This level has been predetermined to be a sound pressure level (SPL) that effectively masks a signal of a given intensity in decibel effective masking level (dB EML). For example, a 30 dB EML narrowband noise centered at 1000 Hz effectively masks a 30 dB HL pure tone at 1000 Hz.
When determining threshold using masking noise, the masking noise is introduced using the "continuously on" control. This allows the masking noise to be presented continuously, while the test stimuli are presented continuously, while the test stimuli are presented intermittently. If the masking noise was presented only with the test stimuli, the patient would be alerted to the presentation. Therefore, the masking noise must be presented continously to the non-test ear while threshold is established in the test ear.
Patient Instructions
The instructions that are provided to the patient for masking are vital to obtaining valid responses from the patient. For many patients, simply providing responses to pure-tone stimuli alone can be a challenging task. Masking adds an auditory signal that the patient must ignore, while responding to test stimuli, which complicates the task further.
Here are some example instructions: "Now you are going to hear those same tones again. This time you will hear some noise in your other ear. Just ignore the noise but continue to say yes every time you hear the tones." It is then helpful to present the masking noise to the patient and to reiterate that this is the type of sound that should be ignored. In some cases, patients will provide a false-positive response when the masking noise is introduced. It is necessary to re-instruct the patient when this occurs.
Here are some example instructions: "Now you are going to hear those same tones again. This time you will hear some noise in your other ear. Just ignore the noise but continue to say yes every time you hear the tones." It is then helpful to present the masking noise to the patient and to reiterate that this is the type of sound that should be ignored. In some cases, patients will provide a false-positive response when the masking noise is introduced. It is necessary to re-instruct the patient when this occurs.
Management
Undermasking
When a masking signal is presented to the non test ear, the intensity of the masking noise must be great enough that it covers up the signal crossing over from the test ear and being heard in the non test ear. If it does not, then "undermasking" occurs.
Overmasking
When a masking signal is presented to the non-test ear that is of a high intensity, it may be so high that the masking noise actually crosses over to the test ear by bone conduction. If it is intense enough, this can cause the test ear to also be masked. When you are testing, this would result in the observed threshold increasing every time the masking noise is increased. This is known as "overmasking."
Effective Masking
In between undermasking and overmasking is a level known as effective masking. This level of masking occurs when the masking noise is of a high enough intensity to effectively cover up the signal crossing over from the test ear but is low enough that the masking noise is not being heard by crossover in the test ear.
References
Gelfand, Stanley A. Essentials of Audiology. Thieme, 2016.
DeRuiter, Mark, and Virginia Ramachandran. Basic Audiometry Learning Manual. Plural Publishing Inc., 2017
Image by Audsim
DeRuiter, Mark, and Virginia Ramachandran. Basic Audiometry Learning Manual. Plural Publishing Inc., 2017
Image by Audsim