Plateau masking, or the Hood plateau method, is the time honored, gold-standard of masking approaches. Masking is put into the non-test ear and gradually increased while determining if there is a shift in the test ear threshold. The procedures are fairly straight-forward. This chapter will discuss the procedure and overview use of the audstudent.com audiometer simulator, which can be used to practice plateau masking.

Before the plateau approach can be understood, there are a few fundamentals to cover.

While thresholds are measured in dB HL, the masking level intensity calibration is dB EM, which stands for “effective masking,” an unfortunate name choice in my opinion. If you say “I used X dB effective masking” you might think it means you were effectively preventing cross hearing, but dB EM is just an intensity standard, and it does not imply that you are correctly/effectively masking.

Even if the name is not ideal, the way the audiometer is calibrated is convenient. A 20 dB EM masking noise will mask a 20 dB HL tone, if the two sounds are put into the same ear. For example, route a 20 dB HL pure tone to your right ear, and set the left channel of the audiometer to 20 dB EM and then route that to the right ear instead of the left ear. If you are correctly calibrated, you hear the noise, but are not be able to detect the tone. (If you increase the tone to 25 dB HL, it becomes audible.) That is a way to check that your audiometer’s calibration, but of course it is not the way you use masking: the noise is routed to the contralateral ear

When masking is needed, we are concerned about the test ear signal having crossed to the non-test ear cochlea. If you have crossover to the non-test ear cochlea that is 20 dB HL, that sound will be masked with 20 dB EM that reaches the non-test ear cochlea. And of course, this holds at all intensities: if you have 40 dB HL of crossover to the non-test ear (NTE) cochlea, then it will be rendered inaudible by 40 dB EM noise at that same cochlea. In the formula masking section of this e-book, we’ll use this concept to our benefit. If we can calculate how much crossover is at the non-test ear cochlea, then we can figure out how much noise to put into the non-test ear to ensure the crossover is masked. However, before formula masking can be mastered, it helps to fully understand plateau masking.

They key point at this moment is that dB EM is just an intensity level, and “effective” doesn’t necessarily mean you have effectively masked.

The brain receives input from both ears, having masking noise in the opposite ear makes it harder to detect near-threshold level sounds in the test ear.

When you put masking into the non-test ear, you may see an initial 5 dB threshold increase that is just due to the increased difficulty of “signal detection” task. That minor increase in threshold is not proof that the signal was crossing over without masking.

Central masking can increase the non-test ear threshold or the test ear threshold. The signal will need to be louder to be heard – regardless of whether it is detected in the test ear or in the non-test ear.

If the test ear signal is crossing over and is heard in the non-test ear, thresholds are likely artificially low: You aren’t yet measuring the test ear’s threshold. With plateau masking, you will put in noise into the non-test ear, initially at a low intensity level – a level just 10 dB above the non-test ear threshold. If there was cross hearing, this noise level should elevate your measured threshold, but it may not be enough to eliminate all cross hearing. Figure 5-1 begins to illustrate.

Figure 5-1. A person with unilateral hearing loss, who has 50 dB interaural attenuation (IA), is tested. The unmasked test ear threshold is 50 dB HL, the crossover (Co) is 0 dB HL, which is heard in the non-test ear by bone-conduction. On the right side of the figure, 10 dB EM of masking noise is put into the non-test ear. This masks the crossover. The patient would no longer hear the 50 dB HL tone presented to the test ear.

Let’s examine the case where the true test ear threshold is 70 dB HL in a patient with unilateral sensorineural hearing loss. (In this first case, we will not show any central masking effect.) As was illustrated in Figure 5-1, without masking, the patient who has 50 dB interaural attenuation (IA) will detect the crossed-over tone when it is 50 dB HL: The 50 dB HL tone lost 50 dB as it crossed to the non-test ear cochlea. Because 0 dB HL is the bone-conduction threshold of the non-test ear, the crossed-over tone was heard. That crossover could be masked with as little as 0 dB EM, but in the first step of plateau masking, you start with the masking noise at 10 dB above the air-conduction threshold. As shown in 5-1, that prevents cross hearing of the 50 dB HL tone. But, with only 10 dB of noise in the non-test ear, we will not yet be able to accurately measure the threshold. When the signal level is raised to 65 dB HL, the crossover is 15 dB HL. That would be audible in the presence of 10 dB EM – the signal would “pop out” over the noise and be heard (Figure 5-2).

Figure 5-2. With 10 dB EM in the non-test ear, the measured threshold is 65 dB HL. This crossover will not be audible when the masking noise is raised to 15 dB HL.

If the noise is raised further, to 15 dB EM, the 65 dB HL test ear signal crossover will not be audible, and since that is below the test ear threshold, it is not heard in the non-test ear either. When the test ear signal is 70 dB HL it is heard in the test ear (because in our example, that was the test ear threshold). But note the 70 dB HL test ear signal is also crossing over and can heard in the non-test ear. Increasing the masking noise intensity another 5 dB will prevent the cross hearing, but the tone will remain audible in the test ear (Figure 5-3).

Figure 5-3. With 15 dB EM in the non-test ear, the crossed-over signal is heard. Increasing the masking noise to 20 dB EM eliminates the cross hearing. The patient is now hearing the signal only in the test ear and the threshold is accurately measured.

At this point, the noise can be increased again another 5 dB (or more, e.g. to 80 dB EM) and the same result will occur. The patient will hear the signal in the test ear and respond. The crossover is masked, so the NTE will not be stimulated. (Figure 5-4)

Figure 5-4. Once the masking noise is adequate, increasing the noise levels modestly (e.g. to 25 dB EM, or even to 80 dB EM) has no further effect. The test ear threshold is the same. The masking noise is more than enough to prevent the crossover from being heard.

The masking noise cannot be increased indefinitely, however. When the noise in the non-test ear is sufficiently intense, the noise will cause a vibration of the skull, and that noise will “cross back (CB)” to the test ear, and prevent hearing of the test tone. This is called over masking (OM). In Figure 5-5, 120 dB EM in the left ear loses 50 dB due to interaural attenuation. It is creating a right ear bone-conducted cross-back signal of 70 dB HL. The masking noise having crossed back will prevent hearing of the 70 dB HL threshold-level signal, and the test ear threshold increases. Each 5 dB increase in masking noise will increase the test ear measured threshold by 5 dB.

Figure 5-5. Overmasking occurs when the non-test ear noise level is intense enough to cross back (Cb) to the test ear and mask the test ear tone. The interaural attenuation value is the same as for crossover, in this example it is 50 dB. The crossback of the noise to the test ear elevates the measured threshold.

If you plot the test ear threshold as a function of the contralateral noise level (as in Figure 5-6), you observe a period where each 5 dB increase in noise level causes a corresponding increase in threshold. This is the “undermasking” portion of the plateau masking procedure. As you increase the noise levels further you hopefully find the “plateau”: a period during which increasing the masking noise level has no effect on threshold. The plateau ends when overmasking begins. You may not always see the undermasking “side” of the plateau function. In some cases, your initial masking level is enough to “put you on the plateau”.

Figure 5-6. A diagram of the masking plateau showing the increase in threshold when the masking noise level is not sufficient, the plateau portion where increase in masking noise does not change threshold, and the overmasking portion, where each 5 dB increase in masking level causes a 5 dB increase in threshold. The plateau width in this illustration is from 15 to 115 dB EM – a 100 dB wide plateau.

• 1  Recognize the need for masking. This implies that the non-test ear (better ear) has been tested first and the test ear’s (poorer ear’s) unmasked threshold has been obtained.
• 2  Set the initial masking level at 10 dB above the non-test ear threshold. Re-obtain threshold. In many cases, the threshold will shift due to either the initial threshold having been due to the signal crossing to the non-test ear, or if it was the test ear responding, the central masking effect may cause a slight elevation in threshold.
• 3  Increase the masking in 5 dB steps, obtaining threshold each time. Since increasing the masking noise will not improve the hearing threshold, you do not need to decrease the test signal intensity below the threshold obtained previously. If the tone is reliably heard at the previous threshold level, that is sufficient. (You don’t need to use the full Hughson-Westlake technique.)

Example: The right ear air-conduction threshold is 5 dB HL. The left ear unmasked air-conduction threshold is 70 dB HL. Masking is needed. The initial masking level is 15 dB EM. The threshold (measured in the left ear) is now 75 dB HL. The next step is to increase the masking 5 dB, to 20 dB EM. Rather than decreasing the stimulus level, determine if 75 is still heard. If it is, then the masking is increased another 5 dB, and threshold is again checked; but, if the 75 dB stimulus was not heard, then threshold would be re-established.

• 4  Continue increasing the masking in 5 dB steps until three consecutive increases in the masking level occur with no change in hearing threshold. As you are increasing the noise level, you might want to make a mental count “that was up once, that was up twice, that was the third time: done.”

Masking level  	TE Threshold
(dB EM) to NTE    (dB HL)
 None               75
 15                 75
 20                 75  “up once” (first time increasing noise 
                         does not change threshold)
 25                 75  “up twice”
 30                 75  “up third time – done

Masking level  	TE Threshold
(dB EM) to NTE    (dB HL)
 None               70
 30                 80
 35                 85
 40                 90
 45                 95
 50                 95  “up once”
 55                 95  “up twice”
 60                 95  “up third time – done

A special circumstance occurs when the patient has no measurable hearing in the test ear. You don’t need to continue to increase the masking noise intensity once you have reached the audiometer’s maximum output level, and the patient no longer responds. Let’s use the example of a patient with 80 dB interaural attenuation and a 10 dB threshold in the non-test ear. Masking would start with 20 dB EM in the nontest ear.

Masking level  	TE Threshold
(dB EM) to NTE    (dB HL)
 None               90
 20                 105
Note: the 10 dB SL of masking raised the threshold 15 dB. Why?  
There was a 5 dB central masking effect plus 10 dB elevation 
of threshold due to the masking noise.     
 25                 110
 30                 115
 35                 120
 40                 No Response

Note that clinically there is no need to determine the end of the plateau – you do not have to find the point where overmasking occurs. However, as we will discuss next, sometimes you will have a narrow plateau and sometimes you will find the point where overmasking begins.

When there is conductive loss in the test ear, by definition bone-conduction is normal. This next section describes why this narrows the masking plateau; how the end of the plateau comes at a lower intensity level than if the loss is sensorineural

Figure 5-7. Typically, the bone-conduction threshold would not be known at this point in testing. You may be able to make the assumption that the loss is conductive based on the patient’s history/symptoms, and if you have found abnormal immittance test results. In this case (Example D), masking is needed when testing air-conduction in the left ear. The initial masking level would be 20 dB EM, put into the right ear.

As shown in Figure 5-7, the left ear air-conduction threshold needs to be masked. This patient’s right ear bone-conduction threshold is 10 dB, so this patient has an interaural attenuation of 60 dB. In this case, the true masked air-conduction threshold will be 75 dB HL. Let’s next examine what would happen as we increase the masking noise levels.

Masking level  	TE Threshold
(dB EM) to NTE    (dB HL)
 None               70
 20                 75
 25                 75
 30                 75
 35                 75

Now, let’s examine the complication of having conductive loss in the non-test ear (Figure 5-8). Let’s again assume that the masked air-conduction threshold will be 75 dB HL, but this time, the loss is sensorineural in the left ear. The interaural attenuation is 60 dB in this example. Note that the non-test ear (right ear) has conductive loss with a 40 dB air-bone gap, and 50 dB air-conduction threshold. Again, the rule for starting the masking process is to use a level 10 dB above the non-test ear air-conduction threshold, so the starting masking level will be 60 dB EM in the right ear.

Figure 5-8. In this example, we examine the case of a patient who has right ear conductive loss and left ear sensorineural loss, with an eventual 75 dB HL air-conduction threshold.

The masking approach would be as follows.

Masking level  	TE Threshold
(dB EM) to NTE    (dB HL)
 None               70
 60                 75
 65                 75
 70                 75
 80                 75

Examine Figures 5-9 and 5-10. Note that in both the case of the test ear conductive loss (5-9) and non-test ear conductive loss (5-10), the plateau width is narrower than what was seen in Figure 5-6 where the test ear had sensorineural loss and the non-test ear had normal hearing. In that case, the plateau width was 100 dB HL. Figure 5-11 shows the three cases superimposed.

Figure 5-9. Masking plateau illustration in the case of test ear conductive loss. The end of the plateau is at a lower intensity. The plateau extends from 20 to 65 dB EM: The test ear conductive loss reduces the plateau width to 45 dB.

Figure 5-10. Masking plateau illustration in the case of non-test ear conductive loss. The start of the plateau is at a higher intensity. The plateau extends from 60 to 135 dB EM: The non-test ear conductive loss reduces the plateau width to 75 dB.

Figure 5-11. The three plateau charts superimposed. The point is that conductive loss, either in the test ear or non-test ear, will reduce the plateau width.

If the patient has conductive loss in each ear, the plateau width will be reduced “at both ends.” The non-test ear conductive loss raises the intensity needed to “get onto” the plateau and the test ear conductive loss (normal bone-conduction hearing) means that the end of the plateau will be at a lowered level.

The example is shown in Figure 5-12.

Figure 5-12. A case of asymmetrical conductive loss. The right ear air-conduction threshold requires masking. The true threshold for the right ear is 60 dB HL; however, the presence of contralateral masking evokes the central masking effect (listening with noise in the opposite ear is a more difficult task) and the right ear will have a masked air-conduction threshold of 65 dB HL.

Let’s assume once again that this patient has a 70 dB air-conduction interaural attenuation value. The signal is not crossing over at an audible level; however, you don’t know this, so you plateau mask. Masking would begin at 30 dB EM (10 dB above the left ear’s air-conduction threshold.)

Masking level  	TE Threshold
(dB EM) to NTE    (dB HL)
 None               60
 30                 65 (due to the central masking effect)
 35                 65
 40                 65
 45                 65
 Here’s where we would conclude our testing.  But let’s continue to raise the 
 masking noise level to see the overmasking portion.
 
 50                 65
 55                 65 
 60                 65
 65                 65
 This hypothetical patient had a test ear bone-conduction threshold of 0 dB HL and 70 dB 
 interaural attenuation for air-conduction, so once the masking is 70 dB EM, 
 you will begin to overmask.   To continue  . . . 
 
 70                 70
 75                 75
 80                 80

As illustrated in Figure 5-13, the plateau width (from 30 to 65 dB EM -- 35 dB) is narrow.

Figure 5-13. Effect of minor non-test ear conductive loss and concomitant test-ear conductive loss on the masking plateau. This patient has 70 dB interaural attenuation. The plateau portion is narrowed but present.

There are times when a masking plateau cannot be found. Masking is needed, but as soon as masking is used, overmasking occurs.

Let’s increase the non-test ear conductive loss size in the next example of a patient who has 60 dB of air-conduction interaural attenuation.

Figure 5-14. A masking dilemma example. The right ear air-conduction threshold requires masking. (Yes, so does the left ear.) The true threshold for the right ear is 60 dB HL; however, the presence of contralateral masking evokes the central masking effect (listening with noise in the opposite ear is a more difficult task) and the patient will have a masked threshold of 65 dB HL. The case uses an interaural attenuation value of 60 dB.

In this case, the true threshold remains 60 dB HL, although it would be measured as 65 dB HL due to the central masking effect.

The masking approach would be as follows.

Masking level  	TE Threshold
(dB EM) to NTE    (dB HL)
 None               60 

60                 65 

65                 70

70                 75

75                 80
80                 85
85                 90
90                 95
95                 100
100                105
105                110
110                115
115                120
120                No response at audiometer output limit of 120

Failure to find a masking plateau leaves one uncertain. Is this indication of bilateral conductive loss and a patient who is untestable due to the masking dilemma, or does the patient have a unilateral profound hearing loss?

Note in Figure 5-14, that the left ear threshold is 50 dB HL, and that means that the left ear air-conduction threshold also needs to be masked: there is potential for crossover to the right cochlea. When plateau masking is attempted for the left ear, again there will be a masking dilemma. Since you can’t have one or both ears with moderate/moderately severe conductive loss without masking and then have accurate results that show that both ears have profound loss with masking, you will be alerted to the presence of the masking dilemma.

That point merits repeating. If you start out with an audiogram that shows one or both ears has hearing, and with masking now find two “dead ears,” that’s an indication that you have come across a masking dilemma. Further, the cross-test principle will assist you. Immittance test results will be abnormal with bilateral conductive loss. Masking dilemmas occur with bilateral conductive loss.

Figure 5-15. When a masking dilemma is present, there is no masking plateau.

The AudSim Flex© software program includes example masking patients. This is a separate program from mCalc and mQuest. It is available for purchase at the bargain price of $19.99 at audstudent.com. There are tutorials on this website about how to install and use the software.

Although the software has a “Threshold Assistant” mode for unmasked testing, we have not yet implemented “Masking Assistant”. This feature is planned for an upcoming version, but it has been planned for a long time now. I may have to retire before I have a chance to do that.