I have the problem of masking tinnitus. I have the following questions:

  1. Can broadband noise mask tone tinnitus, or any kind of tinnitus (such as narrowband, or chirp)?

  2. Assume an audiogram test has identified tinnitus as tonal with loudness level of $x$ dB and frequency/pitch of $y$ Hz. How can I create broadband noise using this information (note that we don't know the variance of the noise)? How could this be implemented in Matlab?



  • $\begingroup$ actually, for me, it seems that broadband "noise" seems to set off the tinnitus. like if i am in a room that is silent, it often remains silent. but if i am out in the woods and i can hear low-level background sounds (from distant traffic to nearby birds), it seems like these low-level sounds have to compete with the noisy tinnitus. perhaps this is just the character of my hearing. $\endgroup$ Mar 17 '15 at 18:07

Read about tinnitus masking on Wiki. Yes, you can mask it with various kinds of noise and/or natural sounds that are "noisy" (ocean, rainfall, etc)

For the synthesis of masking noise, there is plenty of information via google & StackOverflow. Keep in mind that broadband noise (versus bandlimited noise) is present across the spectrum and thus, you only need to be approximately in the same bark frequency band of the tone you want to mask.

Here is a plot showing one measurement of Tinnitus spectrum (from http://www.thelancet.com/cms/attachment/2021743353/2041580240/gr1.jpg). I wasn't able to access the full thelancet article:

enter image description here

  • $\begingroup$ Hi Ruoho, I cannot be using truly broadband noise, but limit it to critical frequency band, but is it possible to obtain spectrum of tinnitus from audiogram? $\endgroup$
    – user915783
    Mar 17 '15 at 20:52
  • $\begingroup$ I've updated my answer with a diagram, to show the location of Tinnitus pitch on an audiogram $\endgroup$ Mar 17 '15 at 21:53
  • $\begingroup$ @ Ruoho I am new to audiometery..can you point me to source where you found above picture? Also, as per my first question, if I know my tinnitus tone is at "x" db which may be "y" in linear scale, can i assume that y^2 would be the power that I need and so i create white noise of Power spectral density = y1^2, and pass it through Filter with bandwidth DelF so that y^2 = y1^2 * DelF? $\endgroup$
    – user915783
    Mar 18 '15 at 21:14
  • $\begingroup$ Okay, I added the link: thelancet.com/cms/attachment/2021743353/2041580240/gr1.jpg I wasn't able to access the full article for free. And yes, if you convert from dB to linear, the power is the square (^2), and you'll want to have a filtered (band-passed) noise with wide-enough bandwidth to mask. $\endgroup$ Mar 18 '15 at 21:58
  • $\begingroup$ @ Ruoho Thanks for reply. Can you give the link of the article. I can get it downloaded $\endgroup$
    – user915783
    Mar 19 '15 at 0:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.