The Interview Sounded Clear. Why Were Viewers Still Turning Up the Volume?

The interview has gone live, and the first message reaches the Nairobi newsroom before the social editor has finished posting the link: “Good conversation, but please increase the sound.” At the editing desk, that feels unfair. The guest's words were clear, the recording never clipped, and the producer heard every answer comfortably through headphones. Yet on a phone, between two other videos, the interview makes the viewer reach for the volume button. The team starts discussing a better microphone when the more useful question is what happened between a clean recording and a consistently listenable programme.

Picture the editor's working day. A quiet guest leans back during an answer, the presenter interrupts with a laugh, and the opening music arrives with considerably more energy than either voice. To hear the guest clearly, the editor turns up the headphone output. The whole session now feels comfortable, so the mix is approved. What has changed, however, is the listening level at the desk, not necessarily the level or balance in the exported programme. A headphone volume knob can make a quiet file sound satisfying without adding anything to the file a viewer will receive.

This is where three different ideas need separating: clarity, peak level and perceived loudness. Clarity concerns how intelligible and free from unwanted interference the speech is. A peak meter shows brief signal maxima, helping the operator identify moments approaching the available ceiling. Perceived loudness involves how the sound is experienced over time, influenced by its energy, frequency content and dynamics. A recording can be clear and contain one tall peak while the conversation around that peak remains much quieter than expected. Keeping a meter out of the red is important, but it does not by itself produce a balanced programme.

Think about that single loud laugh. If it is the highest moment in the interview, raising the entire file brings the laugh and the quiet guest up together. The laugh may approach the output ceiling before the guest becomes comfortable to hear. Peak normalisation works on this relationship: it applies a gain change based on the chosen peak target, rather than understanding which parts of the conversation need attention. It does not automatically reduce the distance between the loud interruption and the soft answer. That is why “normalise the file” is incomplete advice when nobody has established which kind of normalisation is being used or what problem it is meant to solve.

Loudness measurement offers another view. A loudness meter can estimate programme loudness using a defined method, commonly expressed in LUFS, rather than relying only on the highest sample peak. An integrated measurement describes the measured programme over time, while shorter measurement windows help reveal changes within it. These readings are useful because two programmes with similar peaks can still have noticeably different average listening levels. They are not a replacement for listening, and one number cannot tell you whether a quiet answer is being masked by music or whether processing has made a presenter sound strained.

Delivery requirements also matter. A broadcaster, commissioning client and online destination may specify different loudness and peak conditions, and their requirements should be checked before preparing the final version. There is no single number that should be copied into every project simply because it appears in a tutorial. Likewise, do not assume that every service will raise a quiet upload until it sounds right. Playback normalisation can behave differently across services, formats and listening settings. Even where it is applied, it cannot reliably repair badly balanced voices, excessive background music or a guest whose words keep disappearing within the programme.

The first useful correction often happens before a loudness processor is added. Listen to the conversation itself. If the guest becomes much quieter whenever they turn away, microphone placement and speaking consistency need attention at the recording stage. Where separate tracks are available, adjust the voices individually rather than forcing the entire mix to follow the louder person. Clip-gain changes or careful level automation can bring quiet answers into a more usable relationship with the rest of the discussion. Raising a noisy recording also raises its noise, so an extremely weak source cannot be treated as though it contains unlimited clean speech waiting to be turned up.

Compression can then help control variation, but it needs a defined purpose. Rather than lifting every quiet passage indiscriminately, a compressor reduces level according to its settings when the signal meets its operating conditions. Appropriate output gain may then bring the controlled signal to a more useful level. Push that process too hard and breaths, room sound and other background noise may become more conspicuous, while natural speech can lose its ease. The aim is not to make every syllable equally loud. It is to keep the listener following the conversation without repeatedly adjusting the volume.

A limiter serves a different role again. It can help control peaks at the output, but it is not a substitute for balancing the interview. Driving an uneven mix harder into a limiter may squeeze the laugh while still leaving the underlying editorial problem unresolved. For delivery work, a suitable true-peak meter or limiter can help assess or manage reconstructed peaks that sample-peak readings alone may not reveal. Follow the destination's current requirements and leave the specified headroom. A loud file with audible distortion is not an improvement over a quiet one.

The opening music needs its own decision rather than being allowed to set the audience's expectations accidentally. If the ident sounds powerful and the first spoken sentence suddenly drops back, viewers experience that change before they have heard the subject of the interview. If music continues underneath speech, a guest who sounds clear in isolation can become difficult to follow in the mix. Adjusting the music beneath the voice may solve more than another increase on the master output. Judge the transition and the complete programme, not only a clean soloed microphone track.

Try one practical comparison with a completed interview. Choose a representative passage containing the presenter, a quieter guest answer and the opening transition. Listen to it beside a professionally delivered spoken-word reference intended for the same destination, using the same device and leaving the playback volume unchanged. Avoid advertisements or music clips as the reference because their content can make the comparison misleading. Notice where you want to reach for the volume control, then use loudness and peak measurements to investigate that impression. If only the guest requires adjustment, that points towards internal balance. If the whole interview is consistently quieter, review the overall mix and delivery preparation rather than replacing the microphone on that evidence alone.

Repeat the check on the exported file, because the session and the deliverable are not always identical. Confirm the intended mix, inspect its measured loudness and peaks, and listen at a sensible level through both headphones and a typical phone speaker. The phone test is not permission to discard tonal balance or chase maximum volume. It is a way to discover whether the words remain usable on a common listening device. Where possible, review the actual uploaded playback as well, keeping the test device and volume setting consistent so that the comparison remains meaningful.

A small media team benefits from turning this into a repeatable handover rather than a last-minute rescue. Agree which destination requirements apply, establish a consistent monitoring routine, and record which version has passed the final checks. Keep voice balance, music transitions, measured loudness and output peaks as separate decisions instead of hoping one preset will handle them all. A clean microphone capture gives the editor room to work, but a dependable delivery process is what helps that recording reach the audience without making them fight the volume control.

When planning podcast and interview equipment in Kenya with Zamar Audio Solutions, start with the complete job: capturing usable voices, keeping the necessary tracks separate, monitoring accurately and preparing the final programme for its destination. Confirm the capabilities the workflow needs before choosing the equipment, because hardware alone does not establish a delivery standard. Getting it right means knowing why the audience can follow the interview comfortably, rather than buying another microphone and hoping the next upload will sound louder.