This translation has not been editorially reviewed yet. The German version is authoritative. Deutsche Fassung
Accessibility
How loud can background music be under language?
Music or sounds under a speaker’s voice sound atmospheric to well-hearing people. For hard-of-hearing people, they hide the language: The brain can no longer separate the voice from the background. Criterion 1.4.7 of AAA requires for audio-only content recorded with voice that the background sound is missing, switched off or at least 20 decibels quieter than the voice.
What the Criterion Requires
For recorded sound recordings whose main content is spoken language, one of three conditions applies: no background sound, a control to switch off the background sound, or a level spacing of at least 20 decibels between speech and background. 20 decibels means that the language appears about four times as loud as the background. Excluded are recordings whose purpose is the sound itself, such as music with singing or radio plays with conscious soundscape.
For whom this counts
About one in six adults hears restricted, at the age of one in two. Hearing aids amplify everything, including music. For these people, a clean level of speech without support is the difference between understanding and guessing. The criterion also works in loud environments, small speakers and when listening with headphones in the train.
This is how we deal with it
The level distance can be measured in a tone analysis, but not from the source text of a website. The test service shall find and identify audio elements; the assessment is carried out by a person with an audio program. The criterion belongs to the category AAA in knowledge.
Examples
Music under the voice of the speaker
Mix: Sprache −12 dB, Musikbett −18 dB (Abstand 6 dB)Mix: Sprache −12 dB, Musikbett −32 dB (Abstand 20 dB)
<!-- oder: zweite Fassung ohne Musik -->
<a href="folge-12-ohne-musik.mp3">Fassung ohne Musik</a>A distance of 20 decibels makes the language appear four times as loud as the background; a regulator for switching off also meets the criterion.
- Record language and music as separate tracks; then the distance can be set during mixing.
- Lower the background under Language by at least 20 decibels or hide it during the language.
- For podcasts, offer a version without music if music belongs to the brand.
- Check the mix with a level knife, not by ear; Your own hearing is lenient when familiar with the content.
