The “-6 dB Rule” for music

When fitting hearing aids with music in mind, one has to have music in mind. Of the many differences between speech as an input to a hearing aid, and music as an input to a hearing aid, is the crest factor.  Assuming that you have selected the hearing aid appropriately in order to ensure that overly intense music does not overdrive your analog to digital converter (i.e., the “front end”) then what else can be done?

The crest factor has little to do with cavities, although we should brush and floss regularly.  It has to do with how peaky a response is.  Specifically the crest factor is the difference (in decibels) between the peak of a signal and the average (or RMS of the signal).  We deal with this every day in the realm of hearing aids for speech, although we may not explicitly be aware of it.  When the reference test gain is specified (or assessed in a hearing aid test box) it is set to 77 dB below the OSPL90.  Or, if you wanted to calculate it, the reference test gain is OSPL90  –  77 dB.  Why 77 dB?…. well, glad you asked!  Its 65 dB + 12 dB.

And, why is it 65 dB + 12 dB?  Its average conversational speech at 1 meter (about 65 dB SPL) plus the crest factor for speech…. Which is about 12 dB.  The difference between the average level of speech and its peaks is about 12 dB.  And the reason it’s so low is because our vocal tracts have a heck of a lot of damping.  We have soft lips and cheeks, soft and gooey tongues, saliva, and rather narrow openings on either side of our nasal cavities.  In short, when something emanates from the vocal tract the peaks are highly damped and brought down to about 12 dB higher than the average of the speech signal.

There is no gooey tongue or saliva in a violin or a drum.  Clarinets and guitars don’t have a nasal cavity (although bag pipes do function with a parallel tube as our nose).  Instrumental music is less damped than our speech.  And its crest factor is about 18-20 dB… much peakier than speech.

So…. What does this mean for hearing aids and music? Since instrumental music is 6 dB peakier than for speech (18 dB crest factor – 12 dB crest factor = 6 dB difference) this means that that the OSPL90 (and the gain) need to be 6 dB lower for music than for speech of a similar input.  We set the OSPL90 to prevent tolerance problems which really means that we set the OSPL90 to prevent overly intense signals (which really means the peaks) from being transmitted.  A reduction of 6 dB for the OSPL90 for the “music program” relative to the “speech in quiet” program makes sense, at least as an initial start point.

And assuming that both music and speech should be set similarly with WDRC, this implies that the gain should also be set 6 dB lower for the “music program” than for the “speech in quiet” program.   There is some clinical evidence (see for example, Davies-Venn, E., Sousa, P., and Fabry, D. JAAA, 2007, 18(8), 688-699) for this as well evidence on theoretical grounds.   WDRC is primarily an “outer hair cell processing repair strategy” rather than dependent on the nature of the input to the hearing aid so this does follow nicely.

The -6 dB rule for music as an input to a hearing aid is based on some assumptions, but tends to be a good first approximation relative to one’s “speech in quiet” program setting.

About Marshall Chasin

Marshall Chasin, AuD, is a clinical and research audiologist who has a special interest in the prevention of hearing loss for musicians, as well as the treatment of those who have hearing loss. I have other special interests such as clarinet and karate, but those may come out in the blog over time.