Why speech intelligibility should guide the mix
A mix can sound polished on studio monitors and still be difficult to understand for someone with hearing loss. The problem is not always volume. Speech may be masked by music, reverberation, competing voices, sudden level changes or excessive low-frequency energy. Increasing the master level can make the programme louder without making the words clearer—and may create listening fatigue or discomfort.
According to the World Health Organization, more than 1.5 billion people worldwide live with some degree of hearing loss. Their listening needs vary considerably: some use hearing aids or cochlear implants, while others rely on captions, assistive listening systems or careful control of volume and background sound. Good mixing practice should therefore improve intelligibility without assuming that every listener hears in the same way.
Start with the relationship between voice and background
The most effective clarity decisions are often made before detailed processing. Identify the main speech element and protect it throughout the arrangement and mix. Music, atmospheres and sound effects should support the message rather than compete with the frequencies and rhythms that carry consonants.
Useful approaches include:
- Reducing accompaniment during important dialogue, either manually or with gentle sidechain ducking.
- Leaving space around speech transients instead of placing dense effects directly underneath every word.
- Checking whether multiple speakers overlap, particularly at the ends of phrases.
- Using a shorter or more controlled room sound when reverberation blurs syllables.
Automatic ducking can be helpful, but it should be slow and musical enough to avoid audible pumping. Manual automation remains preferable for interviews, documentaries and scripted content where the editorial importance of each line changes.
Use EQ to reveal consonants, not simply to brighten the voice
Speech intelligibility depends heavily on consonants, which provide information about word boundaries and meaning. Vowels carry much of the energy, but consonants such as f, s, t and k are often quieter and easier to mask.
As a starting point, remove unnecessary rumble below the useful speech range with a carefully chosen high-pass filter. The exact frequency depends on the speaker, microphone and recording, so avoid applying a fixed preset to every voice. A small reduction in muddy low-mid energy can create space for speech, while a modest presence lift may improve definition. Excessive high-frequency boost, however, can make sibilance, hiss and hearing-aid feedback more noticeable.
Listen in context rather than in solo. A voice that sounds slightly thin on its own may be much easier to follow against music. Conversely, a rich, warm solo vocal can become indistinct when combined with a dense soundtrack.
Control dynamics without removing natural expression
Large differences between quiet and loud words are especially challenging for listeners who depend on hearing aids or who are listening in noisy environments. Moderate compression can narrow the gap and keep speech present, but over-compression may bring up breaths, room noise and background effects.
A practical workflow is to combine clip gain or volume automation with gentle compression. Set the level of individual phrases first, then use compression for consistency rather than asking it to solve every problem. Always check:
- quiet speech at the lowest realistic playback level;
- loud words and laughter for distortion or discomfort;
- transitions between speakers with different vocal levels;
- the mix on headphones, small speakers and typical television or laptop playback.
For broadcast and streaming work, loudness compliance is important, but an integrated loudness target alone cannot guarantee intelligibility. Two programmes can measure similarly while one has much clearer dialogue because its speech-to-background relationship is better controlled.
Make accessibility part of delivery, not an afterthought
Mixing is only one part of accessible communication. Accurate captions provide essential support, especially when audio quality, accents, overlapping dialogue or environmental noise make listening difficult. The Royal National Institute for Deaf People explains that hearing loss can affect communication in different ways, so accessible content should offer more than a single listening route.
When preparing a programme, verify that captions match the spoken words and identify meaningful non-speech information, such as a warning tone, telephone ring or music cue. If the production includes a live event, consider a live captioning or speech-to-text option. For recorded content, review captions against the final mix rather than an earlier edit: a late change to timing or dialogue can make them inaccurate.
A repeatable quality-control checklist
Before approval, ask another listener to follow the content without visual context. A fresh listener may identify problems that the mixer has stopped noticing after repeated playback. Then test the mix systematically:
- Play the programme at a quiet, realistic listening level.
- Check speech over the busiest music and effects sections.
- Listen for words lost at edits, fades and speaker changes.
- Compare several playback systems and, where possible, a noisy domestic environment.
- Confirm that captions, transcript and final audio remain synchronised.
The goal is not to make every voice bright, loud or heavily processed. It is to create a stable hierarchy in which the listener can identify who is speaking, follow the words and understand the programme without unnecessary effort.
FAQ: speech mixing and hearing loss
Does making dialogue louder always make it clearer?
No. If music, reverberation or competing voices are masking the dialogue, increasing volume raises the competing material too. Improving separation and controlling dynamics is usually more effective.
Should every speech track receive the same EQ settings?
No. Age, voice, microphone, room and delivery all affect the appropriate treatment. Use presets only as starting points and make decisions while listening to the complete mix.
Is compression always beneficial for listeners with hearing loss?
Moderate compression can make quiet phrases easier to hear, but too much may increase background noise, sibilance and fatigue. Combine gentle compression with automation and careful editing.
Are captions still necessary when the mix is clear?
Yes. Clear audio cannot meet every listener’s needs, particularly in noisy environments or when hearing loss is severe. Accurate captions provide an additional, reliable way to access the content.