Audiometry Simulator

You start with the better ear and at 1000 Hz. From an audible level you proceed in down 10 dB, up 5 dB steps (the modified Hughson-Westlake procedure). If a response is obtained on at least half of the presentations at the same level (typically two out of three), that level is the threshold.

The usual order is 1000 → 2000 → 4000 → 8000 → (1000 again) → 500 → 250 Hz. Repeating 1000 Hz is a reliability check: if it differs from the first measurement by more than 5 dB, the test is repeated. Once air conduction is done, bone conduction is measured.

In air conduction: if the difference between the air-conduction threshold of the test ear and the bone-conduction threshold of the non-test ear is equal to or greater than the interaural attenuation (40 dB with supra-aural phones, 60 dB with inserts), the sound can cross the head and be heard in the better ear — masking is required.

In bone conduction: the sound reaches the opposite cochlea with almost no attenuation (IA ≈ 0). So masking is used whenever the test ear shows a meaningful air-bone gap (> 10 dB) or the bone-conduction thresholds of the two ears differ by ≥ 10 dB.

When you present a loud sound to the poorer ear, it vibrates the skull and can be heard in the better ear. The patient responds, but not with the ear you are testing. The misleading curve you plot without masking is called the shadow curve ; it typically runs parallel to the better ear's thresholds, one IA (40–60 dB) below them.

The unilateral profound loss case in the simulator sets exactly this trap: without masking you will find a "60 dB threshold in the left ear" — while the left ear hears nothing at all.

It is the standard up-down staircase used to search for a threshold. After a response the level is dropped by 10 dB; after no response it is raised by 5 dB. This "down 10 / up 5" oscillation produces repeated ascending approaches to the threshold. The threshold is defined as the lowest level at which a response is obtained on at least 50% of the ascending trials .

The simulator runs this check in the background: if you record a threshold without enough repetitions, it warns you.

If the thresholds of two neighbouring octaves differ by 20 dB or more, the half-octave frequency in between is measured as well. Otherwise the real shape of the curve is missed — in a noise notch in particular, skipping 3000 and 6000 Hz hides the notch completely.

To mask during bone-conduction testing you place a phone on the non-test ear. When the ear canal is closed, the bone-conduction threshold at low frequencies improves — that is, the patient hears better than expected. This gain has to be added when the masking level is calculated.

The effect is marked with supra-aural phones (~30 dB at 250 Hz) and very small with deeply seated inserts. It appears in ears with a normal middle ear; in a conductive loss it is not expected, because a "natural occlusion" is already present.

The noise in the non-test ear becomes so loud that it too crosses the head and starts to mask the test ear. The threshold of the test ear then worsens artificially. The safe window between under-masking and over-masking is called the plateau ; the true threshold lies there.

No. The simulator runs entirely in your browser; no measurement is sent to a server. Your progress is kept only in your own browser's local storage. The classroom code is generated in your browser too — whether you share it is up to you.

The audiometer console, the view of the booth and the live audiogram all have to be on screen at once. At phone width that layout stops being usable; pressing the wrong key and recording the wrong threshold becomes far too easy. Small screens are therefore blocked on purpose. A tablet works in landscape.

No. What is tested here is a virtual patient; you are the tester. The sounds are not calibrated and depend on your computer's volume, so draw no conclusions about your own hearing. If you have concerns about your hearing, see an audiology clinic.