My Theories
This section summarises four key sound-based tinnitus approaches I use:
- Amplitude modulation
- Acoustic neuromodulation
- Randomised sequences
- Notched audio
For each of them you’ll see the theory, hypothesis on a mechanism of action, how they can be customised and how I make them into music for a more universal effect.
1. Amplitude Modulation
Theory
Amplitude modulation (AM) applies a periodic change in loudness to a carrier sound (tone, narrowband noise, broadband noise, filtered noise, music etc). Or more simply, it turns sound on and off.
The theory is that temporally patterned sound can interact with tinnitus-related neural activity more effectively than unmodulated sound, potentially promoting short-term suppression and (in some protocols) longer-term reduction in perceived tinnitus.
Which means turning sound on and off in a reasonably fast way (10 times per second or more usually) can potentially turn down the noise of tinnitus. Short term effect has been tested and it may also work longer term but we need trials to confirm this.
The research by Neff et al found short-term tinnitus loudness suppression following AM sounds, with stronger suppression for AM near the tinnitus frequency compared with general noise.
Mechanism of Action
The mechanism isn’t fully established but there are several possibilities:
Residual Inhibition
Many sounds can temporarily reduce tinnitus after sound offset (seconds to minutes). These are usually sounds that have a match to the tinnitus frequency of the person. AM may strengthen this effect for some people, possibly by increasing salience and neural engagement in the relevant auditory channels.
Disrupting abnormal neural synchrony
Rhythmic modulation can influence oscillatory activity and temporal coding in auditory pathways. This is plausible, but direct evidence linking AM to longer lasting tinnitus improvement remains limited. Most of the robust effects measured to date are short-term suppression.
Attention and perceptual re-weighting
Modulated sounds can reduce tinnitus prominence by capturing attention or reducing contrast between tinnitus and silence (a masking or reclassification effect), often improving coping even if the actual tinnitus is unchanged. This sits with broader sound-therapy frameworks (like the classic white noise therapy).
Neural Entrainment
AM sounds, particularly those modulated at 10 Hz (in the alpha frequency band), are believed to entrain neural oscillations in the auditory cortex, helping to normalise abnormal cortical activity. This links to findings that people with tinnitus are often deficient in Alpha brain wave activity (wakeful relaxation).
Customising Audio Therapy
Carrier Selection
For tonal tinnitus we should ideally be aiming to have the carrier (the sound we use AM on) to be at the tinnitus frequency.
For noise-like tinnitus we would be using it on something with plenty of frequency content, like white noise or other variants, or a narrow band noise.
AM Rate
Common experimental / clinical choices range from low rates (eg 4–20 Hz) up to faster modulation, but comfort and perceived benefit vary. Modulating at higher rates also introduces audio artefacts – which can be interesting to experiment with.
The practical approach is to have a small set of modulation is up to around 20Hz (max 40Hz), tied into the music being used.
Modulation Depth
Deeper modulation increases the perception of pulsing but can become irritating, depending on the nature of the carrier. Generally using noise as the carrier increases irritation and by association discomfort. You can experiment from a lower depth, raising until comfort levels have been breached.
Sessions
Many protocols use repeated daily sessions, usually time limited (which may be related to tolerance of the quite basic carrier sounds).
Some outcomes are short-lived, so it should not promise persistent effects without evidence.
Safety
Any listening will need to enforce level limits for listening so as not to risk damaging hearing, there’s a temptation for people to increase the volume to completely cover tinnitus.
In more advanced cases it should include calibration guidance for headphone usage and caveats / guidance for people with hyperacusis.
My Implementation
I generally aim for 10Hz as the main AM frequency when I put it into music. It doesn’t create too many additional frequencies and it’s quite pleasing to listen to. I also use it for the brain entrainment principle, to try and encourage the frequencies that are missing in those of us with tinnitus.
I’ve experimented in various pieces with different rates. In some sequences where the extra frequencies it creates aren’t an issue I’ve modulated at 40Hz. These higher modulations seem to work quite well, though they naturally create quite buzzy, electronic sounds that not everyone may like (I personally do).
2. Acoustic Neuromodulation
Theory
The Acoustic Neuromodulation Coordinated Reset device delivers sequences of sine wave tones at specific frequencies around the tinnitus pitch, arranged in temporal patterns. 2 tones are played either side of the tinnitus tone in a random pattern.
The rationale is based on computational and neurophysiological models suggesting that appropriately timed stimulation can reduce pathological synchrony in tinnitus-related networks.
The theory and findings are in the original Tass et al paper.
Mechanism of Action
Desynchronisation of abnormal neural activity: The goal is to “reset” phase relationships across neuronal populations implicated in tinnitus. The initial controlled study reported improvements versus placebo.
Network-level connectivity changes: Neuroimaging and modelling work describes changes consistent with reduced pathological coupling after CR-like stimulation.
More simply, it aims to re-tune the way certain groups of brain cells fire when tinnitus is present. To prevent the abnormal amplification of tinnitus by the brain and reduce the unhelpful connections that reinforce tinnitus.
Customising Audio Therapy
According to the theory it requires tonal tinnitus with a reasonably stable pitch and reliable pitch-matching. In practice (and from experience of being on the trial for the device) if you have multiple tones then you can choose the dominant one.
The stimulation frequencies are played around the tinnitus pitch (2 either side) in repeating sequences with controlled inter-tone timing.
Build in comfort controls (level, timbre, duty cycle) and manage fatigue (sessions split across the day can improve tolerability).
My Implementation
Probably one of my most used things. I developed a sequence after a lot of trial and error that for a lot of people knocks down their tinnitus. It’s based on the papers but uses frequency spaced across the hearing range as a universal therapy.
Technically it shouldn’t work because it doesn’t have any pitch matching.But in practice it does work.
I tried all sorts of timings and tunings, creating elaborate sampled instruments in various tone spacings. The thing that worked the best ended up being tones spaced at either 5 or 7 semitones apart. That’s kind of my start point, a simple musical sequence. I went all around the world and tried so many variations to get back to where I began. However, that’s the scientific approach for you..
3. Randomised Tones
Theory
The theory of ‘fractal tones’ is of algorithmically generated sequences that sound musical but are non-repeating enough to remain gently engaging.
Widex Zen implemented this into the hearing aid. In their approach, fractal tones are used as a calming sound therapy option, often integrated with counselling and relaxation strategies rather than positioned as a direct neural “reset”. The principle is distraction, with the randomness keeping a level of attentive listening.
Mechanism of Action
The peer-reviewed evidence base is smaller than for some other approaches, it includes small studies and combined interventions (fractal tones plus counselling/amplification), which makes the specific effect of the tone sequences harder to isolate.
Attention redirection without fatigue
Because patterns do not repeat predictably, the sound can remain mildly engaging, drawing attention away from tinnitus without becoming irritating.
Arousal reduction
Many users report benefit through relaxation and reduced stress response, which can reduce tinnitus intrusiveness.
Sound enrichment and habituation
Provides low-level sound that reduces silence-driven tinnitus attention, consistent with broader sound therapy principles.
Customising Audio Therapy
Sequences can offer multiple variations via scale, register, tempo density, and timbre.
The sequences can be blended with low-level noise, other active sounds, using audio effects,whilst keeping control over brightness to avoid aggravation in hyperacusis.
For hearing loss they can be individualised by integrating with EQ compensation for equal volume perception and comfort.
My Implementation
Although the evidence base is low my personal experimentation and reports from listeners show that these tone sequences are very effective when used within music, they keep attention which can lead the brain away from the tinnitus sound.
This is one of the most universal sound options because it does not require tinnitus pitch identification.
They can be combined with other audio technique, I usually place them around the active technique to add interest. I began playing with these inspired by the example sequences Widex put out into a web player.
I asked one of their reps about the tones at a conference, because I thought they should cover the audible frequency range. Their answer got my interest when they told me it wasn’t about masking but about distracting, which then got me looking and experimenting deeply with them.
4. Notched Audio
Theory
Notched audio removes (or strongly attenuates) a narrow frequency band centred on the individual’s tinnitus frequency from a broadband sound (often music). The goal is to reduce stimulation at the tinnitus frequency while maintaining stimulation in adjacent regions, encouraging adaptive reorganisation and reduced tinnitus percept over time.
A study found reduced tinnitus loudness
Another found it superior to TRT
A meta-analysis found it to not be effective however they only included 3 studies and didn’t consider loudness, only THI scores.
An influential study reported reduced tinnitus loudness and reduced tinnitus-related auditory cortex activity after listening to tailor-made notched music.
Later randomised controlled evidence has shown mixed results depending on outcomes and study design.
Mechanism of Action
Notched audio is primarily for stable, tonal tinnitus where pitch can be identified with reasonable accuracy. there are two mechanisms.
Reversal of maladaptive cortical reorganisation
The hypothesis is that enhanced input to frequencies surrounding the tinnitus pitch, paired with reduced input at the tinnitus pitch region, can shift auditory cortical representation and reduce tinnitus-related activity.
Reduced gain at the tinnitus frequency channel
By deprioritising stimulation in that band over long listening periods, the system may reduce salience of that frequency region.
Customising Audio Therapy
Pitch matching is often error-prone, listening to your tinnitus tone can often suppress it and make matching hard. Mismatch of the tinnitus tone can plausibly reduce efficacy.
Notch width is important, generally used at a 1 octave range. Too narrow may be ineffective, whereas too wide can degrade music quality and reduce listening comfort.
The source material needs to have enough frequency content for the notch to be effective. Custom music is often best so you can control how much content there is. Broadband content is typically used so adjacent frequencies remain well stimulated.
Studies often used daily listening over weeks to months. During this time the user will need to keep checking their tinnitus frequency to make sure the therapy is still at the right target.
My Implementation
I use notching in broadband sounds. With it being a universal way to implement I use moving notches that go up and down the frequencies, usually in a random pattern.
Many chronic tinnitus patients struggle with pure white noise (me included) so I’ve used boosts in the past to create a sweep effect. I now mainly use notches, creating a similar perceived effect
Next Theories?
I’m working on the tinnitus ripple therapy and how to implement something into audio. It’s looking very much like a comb filter at the moment, it’s quite musically pleasing so should work well alongside everything else that I do.