Listen to speech, music, the Ling sounds or your own recording through the hearing loss you choose. Pick a ready-made audiogram or mark the thresholds on the audiogram yourself; the sound is reprocessed according to the conductive and sensorineural components, loudness recruitment and reduced frequency selectivity. You can also add the listening environment, a hearing aid, tinnitus and age-related change.
For education · not an exact copy of what a person with hearing loss hears
First, in Normal hearing mode, set the volume so the sound is as comfortable as someone talking in front of you. Then switch to With hearing loss; each change reprocesses the sound and playback continues where it was.
Empty frequencies are estimated from their neighbours. Without bone conduction values, the loss is treated as sensorineural.
✓ audible · ~ partly: some of it is audible but not the part that tells it apart · ✕ inaudible
Start the sound with the player below, then change the settings: the sound is reprocessed with each change and carries on from where it was with the new setting. Switch between Normal and With loss to compare. The player stays at the top of the screen while you move through the settings; go back to section 1 to change the sound.
In sensorineural loss, damage to the outer hair cells disturbs the cochlea's natural compression. The threshold rises, but loudness catches up quickly at high levels: soft sounds disappear, loud sounds are still uncomfortable. Conductive loss has no such effect; the sound is only attenuated.
As auditory filters widen, nearby frequencies can no longer be separated; vowel formants blur and noise fills the gaps in speech more easily. That is why raising the volume does not fully restore clarity.
Gain follows the NAL-R rule from the shape of the loss; in compression mode soft sounds get more gain and loud sounds less. This is a simplified version of a real hearing aid fitting: there is no real-ear measurement, feedback or noise reduction.
Tinnitus is sound heard with no outside source, and it is often perceived near the frequency region where the hearing loss is greatest. In matching studies its loudness is usually 5-15 dB above threshold; even so, how bothersome it is does not scale with loudness.
Tinnitus is not added to the “Normal hearing” version; it is only heard in the hearing loss and hearing aid versions. With the aid on, more background sound can make the tinnitus less noticeable.
The curves use the age-related median thresholds of ISO 7029 and, when noise is added, the median noise-induced shift of ISO 1999. A median is the value that half of people that age hear better than and half worse; it does not predict any one person's audiogram.
In auditory neuropathy spectrum disorder the cochlea responds to sound, but neural transmission cannot preserve timing. Here the timing of the sound is disturbed with small random shifts; even with a mild audiogram, speech is much harder to follow than expected, especially in noise.
A concept proposed for a loss of some auditory nerve fibres while thresholds stay normal; its diagnosis in humans is still debated. Here the gaps in the envelope are filled and internal noise is added; the effect shows most in noise.
These two settings are conceptual simulations. They are inspired by published methods but do not reproduce any patient's experience in a measured way. They exist to show why the audiogram falls short in these two conditions.
The “Auditory neuropathy” and “Hidden hearing loss” presets turn these settings on automatically.
A sentence is played through a random hearing loss. Type what you heard; then see the correct sentence, the words you missed and which loss was used. It is an awareness activity for families, teachers and students.
Record a sentence with the microphone and type how it is written. The sentence is added to the pool only while this page is open; it is never uploaded. Useful for sentences a teacher prepares in their own voice.
Share your settings as a link, bring an audiogram over from the Audiogram Tool, or download the processed sound to use in teaching.
The audiogram, the sound source (except your own recording) and all settings are written into the link.
Paste the link you copied with “Copy link” in the tool; air and bone conduction thresholds are brought over. You can also open this page's audiogram in the tool.
Downloaded as a two-channel (right and left ear) WAV. All versions share the same scaling, so the difference between them is kept.
In sensorineural loss the problem is not only audibility. Outer hair cells amplify soft sounds and separate frequencies sharply; when they are damaged the threshold rises, loudness catches up quickly at high levels, and frequencies blur together. That is why the simulator splits sensorineural loss into two parts: the outer hair cell part is processed as level expansion (loudness recruitment), the inner hair cell part as plain attenuation.
When bone conduction is entered, the air-bone gap is treated as the conductive component and only attenuates; the bone conduction threshold itself is the sensorineural component. Without bone conduction, the whole loss is treated as sensorineural. Type and degree use the pure-tone average of 500, 1000, 2000 and 4000 Hz, as in the site's other tools.
The simulator assumes the chosen sound reaches the ear at a given level (for example 65 dB SPL for normal speech) and applies the thresholds accordingly. Because the level of your headphones is unknown, setting the normal hearing version to a natural loudness brings the result closer to reality.
SII gives how much of speech is audible, weighting each frequency band by its contribution to intelligibility, on a 0 to 1 scale (ANSI S3.5-1997). The standard speech spectrum and the one-third octave method are used here; spread of masking is not included. SII is not a percentage of understanding: an SII of 0.5 can correspond to very different scores depending on context and talker.
During playback the audiogram shows the momentary level of the sound reaching the ear in one-third octave bands. On an audiogram, lower means louder: bands that drop below the threshold line are audible, bands that stay above it (in the shaded area) are not. Levels are given relative to each band's normal hearing threshold so they can be compared on the chart, and are approximate.
This is a teaching tool and is not for diagnosis or choosing a device. The frequency response of your headphones, room reverberation and your own hearing change the result. Central auditory processing, attention, cognitive load and lip-reading are not modelled. Two people with the same audiogram do not hear in the same way either.
No. It simulates audibility, loudness recruitment and reduced frequency selectivity with published methods, but the brain adapts to a loss over years and two people with the same audiogram do not hear alike. The aim is to give an intuitive sense of what a loss does to sound.
The right and left ears are processed separately; one-sided or asymmetric loss, the talker's direction and ear-specific tinnitus can only be heard correctly with headphones. On speakers the two channels mix in the room, and small speakers cannot reproduce low frequencies.
Play the female talker in the “Normal hearing” version and set your device volume so it sounds like someone talking normally in front of you. Then switch to the other versions without touching it. Turning it up too far makes the loss sound milder than it is.
No. Recording and all processing happen inside your browser; the sound is never uploaded and is gone when you close the page. Recordings are not written into the share link either.
The aid restores audibility, but the suprathreshold effects of sensorineural loss, recruitment and blurred frequencies, continue in the cochlea. The aid's upper frequency limit and compression also limit some detail. The difference is clearer in noise.
Speech samples: Mozilla Common Voice (CC0). Music, Ling sounds and everyday sounds were synthesised for this page; “Für Elise” (Beethoven) is in the public domain.
Try it in the simulator
Auditory Scene
