Masking is the technique that keeps an audiogram honest. It ensures that the ear you think you are testing is the ear that actually responded. Skipping it when it is needed is one of the most common — and most consequential — errors in audiometry.

The problem: cross-hearing

A tone presented loudly to one ear does not stay there. Some of its energy travels through the skull and reaches the opposite cochlea. If that opposite ear hears well, it can respond to the tone meant for the test ear. The patient presses the button, and you record a threshold — but it belongs to the wrong ear. This is cross-hearing.

Figure 1. Cross-hearing across the skull. A tone presented to the test ear loses energy crossing the head — the interaural attenuation — but the remainder can still reach the non-test cochlea.

Interaural attenuation

Sound does lose energy crossing the head. The amount of that loss is the interaural attenuation (IA). For air-conduction tones through supra-aural earphones, the IA is conventionally taken as at least 40 dB; insert earphones, sitting in the canal, give a larger and more variable IA, so a higher single value is used.⁷ For bone conduction the IA is essentially 0 dB — a vibrator on the skull drives both cochleae almost equally.

TraineeThe 0 dB bone-conduction IA has a sharp consequence: an unmasked bone-conduction threshold tells you the better cochlea’s sensitivity, but not which ear it came from. This is why bone-conduction masking is governed by a different rule from air conduction — the presence of an air–bone gap in the test ear, rather than a between-ear difference.

When is masking needed?

The principle is a single comparison. Masking the non-test ear is required when the tone, after losing the interaural attenuation crossing the skull, could still be heard by the non-test cochlea — that is, when:

(air-conduction level in the test ear) − (interaural attenuation) ≥ (bone-conduction threshold of the non-test ear)

In words: if a poor test ear needs a loud tone, and the non-test ear hears well, the loud tone will cross over and masking is needed. The interactive masking calculator in the Tools module lets you try this comparison with any pair of values.

TraineeA useful shortcut for air conduction: masking is needed whenever the test-ear air-conduction threshold exceeds the non-test ear’s bone-conduction threshold by the interaural attenuation or more. Note that it is the non-test ear’s bone conduction that matters, not its air conduction — the cross-heard tone reaches that cochlea by bone.

How masking works: the plateau method

Masking is done by playing a narrow band of noise into the non-test ear while the test ear is re-tested. The noise keeps the non-test ear busy so it cannot respond to the cross-heard tone. The challenge is choosing the right amount of noise: too little and cross-hearing persists; too much and the noise itself crosses back and shifts the test ear’s apparent threshold.

The plateau method solves this. The masker is raised in steps while the test-ear threshold is re-measured at each step. At first, adding noise pushes the apparent threshold up — the non-test ear is being silenced. Then comes a plateau: a range of masker levels over which the test-ear threshold does not change. That stable plateau value is the true masked threshold. Raising the masker further eventually pushes the threshold up again — that is overmasking.

ClinicianThe minimum and maximum useful masker levels can be stated explicitly. The minimum effective masking is the non-test ear’s air-conduction threshold plus a safety margin. The maximum, before the masker overmasks, is the test ear’s bone-conduction threshold plus the interaural attenuation. When the minimum exceeds the maximum there is no usable plateau at all — the masking dilemma, discussed below.⁴

The masking dilemma

Occasionally masking is impossible. This happens classically with large bilateral conductive losses: each ear needs so much noise to be masked that the noise inevitably crosses back and contam- inates the test ear. There is no masker level that silences one ear without disturbing the other — the plateau has vanished. Insert earphones, with their larger interaural attenuation, widen the usable range and are the first remedy for a threatened masking dilemma.

The shadow curve

The visible sign of unmasked cross-hearing is the shadow curve. When a dead or very poor ear is tested without masking, the audiogram shows an apparent threshold curve that simply mirrors the good ear, offset by roughly the interaural attenuation. It looks like a real, if poor, audiogram, but it is the good ear responding. Recognising a shadow curve — and re-testing with masking — is a core interpretive skill, and it is the reason masking cannot be treated as an optional extra.