The AudioNotch Tinnitus Treatment Blog

What are the most common tinnitus frequencies?

I’ve been reading the literature on tinnitus for several years, and I have yet to come across a paper that provided data on what the distribution of tinnitus frequencies was among people who have tinnitus. Obviously there’s not a lot of money available to ask a question like this (since there’s no drug or treatment available). So I spoke to our CTO, Adrian Green, and we decided to mine the AudioNotch tinnitus treatment database to see if we could get an estimate of this data.

When people sign up for AudioNotch, their sound therapy is customized for their unique tinnitus frequency. They use our tinnitus tuner to match a computer generated tinnitus frequency with the tinnitus frequency that they hear inside of their head. Although it’s not perfect, and up to 50% of people can have a very difficult time determining their tinnitus frequency, it’s still¬†a useful measure in the aggregate sense.

We took the information from our active¬†accounts (to filter out people who were just playing around with the tuner), and then we took their most recently found tinnitus tone. We also cut out tones below 50 hZ and above 20,000 hZ (since that’s effectively the limit of human hearing). The data we got was as follows:

This is what the distribution of tinnitus tones looks like from our database

This is what the distribution of tinnitus tones looks like from our database


tones-analysisThe total sample size was about 1,300 frequencies.

Some insights into the data:

  • The tones are distributed roughly in a normal curve distribution. This is a bit unexpected since we’d think that the tinnitus tones would be clustered in higher frequency distributions, since most hearing loss is high frequency, and tinnitus tones seem to map onto regions of hearing loss. Then again, a lot of natural phenomenon seem to follow a normal distribution, so it’s not altogether that surprising.
  • The sound energy of most environmental music (a huge driver of hearing loss) drops off rapidly around 8,000 hZ, so it’s interesting that so many tinnitus tones are still in the greater than 8,000 hZ region (for the aforementioned reason). That said, there’s a huge amount of tinnitus around that frequency range, which seems to match our expectations since there’s so much music in that region of sound.
  • Limitations of the data were several: the tuning process is often inaccurate, the sample of people who matched their frequency was self-selected by individuals looking to pay for a treatment service and who were computer literate, and it doesn’t include the initial tinnitus tones from people whose tinnitus frequencies shifted. Still, I suspect the approximate normal distribution roughly tracks along the actual real-world population data.
  • There’s a little uptick in tinnitus frequencies around the 20,000 hZ region, making me wonder if people are hearing frequency tones beyond the range of perceptible hearing (perhaps there’s some sort of pathological neurological cause for this?).

This data can also be compared to our internal success rate data that we got from e-mails surveys.

In conclusion, it seems like tinnitus frequencies appear to track a normal distribution centered around approximately 8,000 hZ. Let us know your thoughts in the comments below!