8Minimum Masking Level

FDefinition and how it is measured

The minimum masking level (MML) is the lowest intensity of a masking sound at which the tinnitus is just rendered inaudible (or, by some definitions, just covered). It is the bedside answer to the question “how much sound does it take to bury this tinnitus?”

To measure it, a broadband or narrowband noise is presented — commonly to the ipsilateral ear, sometimes contralateral — and its level is raised in small (1–2 dB) steps until the patient reports that the tinnitus has disappeared into the noise. The level at that point, expressed in dB SL above the masker-frequency threshold, is the MML [2000].

FMML and the loudness match: not the same number

It is tempting to expect the MML to equal the loudness match, since both seem to index “how loud” the tinnitus is. In practice they differ. Many patients have a low MML — their tinnitus is covered by only a modest noise, often 5–20 dB SL — and MML tends to track, but not duplicate, the matched loudness [2003].

The gap between them is informative. A patient whose tinnitus matches to a few dB SL but needs a high MML to mask it is behaving differently from one in whom the two values coincide, hinting at how broadly the tinnitus interacts with external sound across frequency.

TFeldmann’s masking curves

Heinrich Feldmann showed that, unlike masking of one external tone by another, tinnitus does not obey the orderly “masking spreads upward in frequency” rules of normal psychoacoustics. By plotting the masking level needed across a range of masker frequencies he produced tinnitus masking curves with characteristic shapes [1981].

He described several patterns: a convergent type, in which masker and tinnitus thresholds come together; a divergent type; a congruent type that parallels the audiogram; a resistant type that cannot be masked at any reasonable level; and a persistent type. The crucial conceptual point is that tinnitus is frequently masked by sounds remote from its pitch, and that some tinnitus resists masking altogether — behaviour that ordinary external sounds never show, and that argues for a central generator [2014].

TWhat a high versus low MML means

A low MML (broadly ≤20 dB SL) is the common, favourable pattern: the tinnitus yields easily to external sound. A high MML (>30–40 dB SL), or a tinnitus that cannot be masked at all (Feldmann’s resistant type), is associated with greater central gain or neural synchrony, with hyperacusis and recruitment, and often with greater distress [2010].

MML is state-dependent: attention, ambient noise and tinnitus fluctuation move it several decibels between sessions, so it is reported as a representative value rather than a single reading [2000].

CWhat MML predicts for sound therapy

The clinical pay-off of MML is prognostic. A readily maskable tinnitus — low MML, a non-resistant masking curve — tends to do better with sound-based approaches such as maskers, sound enrichment and tinnitus retraining therapy, because the very phenomenon the therapy exploits (external sound interacting with the percept) is demonstrably present [2003].

Conversely, a high or un-maskable MML, especially with hyperacusis, flags a patient for whom pure masking may be uncomfortable or ineffective and who may need a different emphasis — lower-level sound enrichment, habituation-focused counselling, or attention to the limbic and somatosensory contributors rather than louder masking [2013]. MML is therefore best read not as a severity score but as a clue to which therapeutic lever is likely to work.