Measuring Loudness Levels for Consistent Podcast Episodes

Why loudness consistency matters more than raw volume

A listener settling into a podcast episode should never reach for the volume dial halfway through. When one voice booms and the next sits somewhere near a whisper, the brain has to work harder than it should. Attention drifts, and the episode starts to feel unpolished even if the content itself is excellent. Loudness is the invisible craft skill behind that effortless listening experience.

Human hearing is sensitive to relative differences rather than absolute numbers. A 3 dB jump between two segments reads as noticeably louder, and a 6 dB jump can feel jarring. If you record interviews, host links and adverts across different days, rooms and microphones, those differences accumulate quietly. Measuring before you publish lets you catch them while they are still easy to fix.

Peaks and averages: two numbers worth knowing

Audio level measurement splits into two broad ideas, and confusing them is the most common cause of inconsistent episodes. Peaks describe the loudest instant in a recording. Averages describe the perceived loudness across a longer stretch of time, which is much closer to how a listener actually experiences your show.

• Sample peak level is measured in dBFS and tells you how close any single moment came to the digital ceiling of 0 dBFS. Clipping lives here.
• True peak accounts for peaks that occur between samples once your file is converted back to analogue. It is always equal to or slightly higher than the sample peak.
• Average level, often shown as RMS, gives a rough sense of how loud a passage feels over a few seconds.
• Loudness units, or LUFS, weight frequencies and timing to mimic human perception. This is the number that matters most for cross-platform consistency.
• Loudness range describes how much the level moves within your episode. A dramatic piece of storytelling can handle more range than a two-person chat.

If you remember only one pair of ideas, make it these: keep your true peaks under control, and land your average loudness in a sensible place.

Setting a target and picking your meters

For a stereo podcast, a widely used target is around -16 LUFS integrated, with a true peak ceiling of -1 dBTP. Mono episodes are usually aimed nearer -19 LUFS because mono playback tends to be perceived as louder at the same reading. You do not need to chase these figures to the decimal place, but choosing a target and sticking to it across every episode is what creates consistency.

You do not need expensive tools for this. Any modern editor with a loudness meter will do the job. Look for options that display:

• Integrated loudness across the whole programme
• Short-term loudness over a rolling few-second window
• True peak in dBTP
• Loudness range, ideally

Free plug-in suites exist that cover all of this, and many editors now build a loudness panel straight into the export dialogue. Set the meter to run across the full timeline, play the episode from start to finish, and read the integrated figure at the very end.

A simple pre-publish checking routine

You can fold loudness checks into your existing workflow without much extra effort. Fifteen minutes of measurement per episode is usually enough.

• Normalise the whole mix to your chosen target rather than nudging individual clips by ear.
• Scan for isolated peaks — a single laugh, a slammed door or a plosive can spike well above everything else.
• Apply gentle limiting only where needed, and re-check the true peak afterwards. A little limiting is fine; heavy limiting flattens the life out of a conversation.
• Check the intro and outro separately. Music beds and sponsor reads often sit at a different level from your spoken content.
• Compare against your previous episode using the same meter. This is the single best habit for continuity.

If you publish in mono as well as stereo, check both versions. A mix that reads correctly in stereo can feel a touch hot when summed to mono, particularly if you have used wide stereo processing on music.

Common mistakes that undo good work

Even careful engineers trip over the same few issues. Mixing into a pair of headphones at an unusually loud setting, for instance, skews every judgement you make about balance. Choose a comfortable monitoring level and keep it the same from session to session. Treat your monitoring volume as a fixed reference point, not a mood.

Another frequent trap is measuring only a section. Integrated loudness is calculated across the entire programme, so a quiet five-minute interview will drag the average down and make your loud trailer look fine in isolation. Always run the full episode. Similarly, resist the temptation to normalise every individual clip to your target — that flattens natural dynamics and makes conversation sound mechanical.

Finally, watch out for adverts and trailers supplied by other people. They are frequently mastered louder than your own content, and inserting them untouched will produce exactly the volume jump you have been working to avoid.

Trusting your ears after the meter

Meters are a guide, not a verdict. Once the numbers sit where you want them, listen back on the least flattering setup you can find: a phone speaker, a cheap pair of earbuds, a laptop. If the episode holds together there, it will hold together almost anywhere. Consistent loudness is really just a form of respect for your listener — they came for the conversation, not for a fight with their volume buttons.