Understanding audio pathways – why latency matters
Sending clean audio directly to someone’s hearing devices is great. If they only hear that audio then it will sound very good. However, the same sound also arrives at the listener’s ears via other pathways.
If the delay is long enough, hearing the same sound twice in quick succession sounds like an echo. If the delay is a bit shorter, we only hear one sound, but it is distorted. If the delay is shorter still, the distortion disappears and the two sounds are perceived as one.
To ensure that an Assistive Listening system sound good, it is important to consider the different paths that audio can take and how the different components combine to form the overall auditory input.
- Acoustic propagation from a performer and/or the PA system directly to the listener’s ear.
- Acoustic propagation from a performer and/or the PA system via the microphones and amplification of the listener’s hearing device.
- Assistive listening signal, streamed directly to, and amplified by, their hearing device.
How much sound someone hears via routes A and B will depend on the nature and extent of hearing loss. With less hearing loss, more of the unaided sound is heard naturally, via route A. With more hearing loss, more aided sound is heard, via route B.
The processing latency of hearing devices is exceptionally short. It is typically less than 10 ms, so sound via paths A and B combine without any noticeable delay. This is particularly important for good ‘own voice perception’ when the listener hears themselves talking.
Latency considerations for venues
The latency of an Assistive Listening system will be perceived if the sound delivered by path C introduces too much delay compared to paths A and B. How much latency is acceptable depends on how loud A and B are compared to C and the type of sounds. Transient sounds like percussion and impacts are the most noticeable. Keeping relative delay below 40 ms is considered adequate.
The detrimental impact of excess latency can be mitigated by reducing the sound level from paths A and B.
Path B can often be reduced using a hearing aid app. The app may offer separate volume controls for the surroundings (path B) and the streamed signals (path C), or to adjust the balance between the two.
Path A can only be reduced by preventing the ambient sound from reaching the ear drums. This can be achieved by wearing noise cancelling headphones over, or instead of, hearing aids.
For very large venues, the acoustic propagation time for A and B can be significant to the point where C could precede them. If this is the case, it may be necessary for a separate Assistive Listening stream with extra delay added to be used for more distant areas.