Compression in hearing devices alters the intended loudness hierarchy of digital alert sounds

What it is

Designers routinely use loudness differences to convey priority — a quiet chime for low-priority information, a loud alarm for an emergency. This design assumption relies on one premise: the volume ratio arriving at the user's ear matches the ratio set at design time. For users wearing hearing aids or cochlear implant processors, that premise breaks down: these devices commonly apply wide dynamic range compression (WDRC) to sound, and the compression process systematically alters the loudness differences originally built into a set of sounds — the loudness hierarchy carefully designed at production time can arrive at the user's ear noticeably flattened.

Why it happens

Sensorineural hearing loss commonly narrows the usable loudness range (see the sibling entry); to fit the wide range of everyday sound levels — from very quiet to very loud — into this narrowed usable window, hearing devices generally apply gain that varies by input level: more amplification for quieter sounds, less or none for louder sounds, often processed separately across frequency bands. The direct consequence is that the original loudness difference between input sounds gets compressed smaller after passing through this nonlinear gain curve — a sound effect deliberately made 10 dB softer to signal "minor notification" might end up only two or three decibels softer than the "primary alert" after compression, sharply eroding what was once a clear priority hierarchy.

Most hearing devices also layer on noise-suppression and transient-handling algorithms, which often treat short, brief notification sounds (a single chime) differently from sustained sounds (a ringtone), further making loudness-based hierarchy relationships unpredictable.

Studying it

Assessing this effect involves feeding a set of interface alert sounds, designed with intentional loudness differences, through an actual hearing-device processing chain (a real device or the manufacturer's fitting/simulation software), measuring the output loudness of each sound before and after processing, and comparing how well the loudness differences between sounds survive compression — rather than assuming linear amplification.

Where it stops holding

• Compression parameters vary by hearing device brand, model, and the specific fitting prescription used (e.g., NAL-NL2, DSL formulas) — there is no single universal compression curve to design against, and any "compensation approach" needs to be verified separately for the specific device class.
• This effect mainly undermines priority designs coded through loudness difference alone; if urgency is instead coded through pitch change or rhythmic pattern, the impact of compression is typically much smaller — which is exactly why loudness shouldn't be the only usable coding dimension.
• This entry covers compression's disruption of an existing loudness hierarchy; it doesn't address compression's overall effect on speech intelligibility, which is a separate evaluation dimension in hearing device fitting.

Applying it

• If a product's notification priority design depends only on loudness difference, recognize that this hierarchy may not hold for hearing-device users, and encode priority information redundantly through pitch or rhythmic pattern as a compression-resistant backup channel.
• For products with a substantial share of hearing-device users (older-adult user bases, hearing-health apps), before finalizing alert sound design, run it through common hearing-device fitting/simulation software to check whether the intended loudness hierarchy survives.
• Verification: measure the output loudness of a designed set of alert sounds both unprocessed and after typical hearing-device compression, comparing whether the relative loudness difference between sounds noticeably narrows under compression, and use that measured data to decide whether pitch or rhythm should be added as a supplementary coding dimension.

Related

• Same group: A3.17.1 Conductive hearing loss affects loudness perception; sensorineural loss also degrades frequency resolution · A3.17.2 Unilateral hearing loss completely eliminates localization ability that depends on interaural time and level differences · A3.17.4 A normal audiogram with difficulty understanding speech in noise is an easily overlooked hearing impairment
• Nearby: A3.11 Pitch discrimination · A3.13 Temporal resolution and rhythm perception · A3.02 Loudness perception and equal-loudness contours
• Search terms: wide dynamic range compression · hearing aid fitting · WDRC · cochlear implant processing