How to edit a podcast episode for consistent volume levels

A podcast can contain excellent interviews, thoughtful storytelling and polished music, yet still feel uncomfortable if listeners are constantly reaching for the volume control. One speaker may sound distant, another may dominate the conversation, and an advert or music sting may suddenly jump several decibels. Consistent volume levels make an episode easier to follow through headphones, car speakers, smart speakers and mobile devices.

Editing for even loudness is more than turning every waveform up to the same height. Speech has natural changes in energy, microphones respond differently, and background noise can become distracting when quiet sections are heavily amplified. A reliable process combines clip gain, compression, limiting, noise control and loudness measurement so the final file sounds balanced without becoming flat or fatiguing.

Start with a clean and organised session

Before processing audio, organise the episode in your digital audio workstation. Put each person on a separate track where possible, and place music, advertisements, sound effects and room tone on their own tracks. Label tracks clearly and use colour coding so you can quickly identify the host, guests and non-speech elements. This is particularly helpful when an Australian podcast includes a remote guest calling from Perth while the host is recording in Melbourne or Sydney.

Listen through the complete recording before making detailed adjustments. Mark interruptions, long silences, clipped words, microphone bumps, overlapping speech and sections where a guest moves away from the microphone. The first pass should reveal the episode’s real problems rather than encouraging you to process every track identically. Keep an unprocessed duplicate of each original recording so edits can be reversed without losing the source.

Remove obvious distractions, but preserve natural pauses and breathing where they support the conversation. Tight editing can make a podcast sound rushed, especially when a thoughtful guest needs a moment to answer. Add short fades to every cut, use room tone to cover abrupt gaps, and avoid cutting between noticeably different background ambiences. A consistent noise bed is usually less distracting than a sequence of complete digital silences.

Balance voices before using loudness tools

Clip gain is the best first step for managing uneven speech. Raise a quiet sentence or lower an unexpectedly loud phrase before it reaches a compressor. This allows the compressor to work more gently and prevents it from reacting dramatically to one isolated word. Aim to bring the performance into a reasonably narrow range by hand, rather than trying to fix a large volume difference with a limiter at the end.

A practical starting point is to make ordinary speech sit around the same level on the DAW meters, then adjust by ear. Depending on the voice and recording environment, peaks may commonly fall somewhere around -12 to -6 dBFS during the editing stage, leaving sufficient headroom for later processing. These figures are working targets rather than rules. A soft-spoken narrator and an energetic presenter may need different treatment even when their meters appear similar.

Use automation for longer changes, such as a guest becoming quieter during a serious answer or a presenter moving closer to the microphone. Volume rides can sound more transparent than heavy compression. When two people speak at once, decide whose words carry the meaning and lower the other track slightly. For a panel discussion, gentle ducking of secondary voices can preserve clarity without making the conversation sound artificially edited.

Shape speech with compression and limiting

Compression reduces the difference between loud and quiet parts of a voice. For spoken-word material, a moderate ratio such as 2:1 or 3:1 is often a useful starting point, with attack and release times adjusted to retain consonants and natural rhythm. Aim for a few decibels of gain reduction on ordinary speech rather than forcing every syllable into the same level. If the compressor is constantly working hard, return to clip gain and correct the source balance.

Listen for signs of over-processing. Excessive compression can bring up lip noises, computer fans, air conditioning and room reflections. A release time that is too fast may create audible pumping after every phrase, while an attack that is too quick can dull the clarity of consonants. Compare the processed track with its bypassed version at the same perceived volume, because louder nearly always sounds better during a careless comparison.

After compression, use a true-peak limiter as a safety stage rather than a volume-recovery shortcut. Set a ceiling around -1 dBTP to reduce the likelihood of inter-sample peaks becoming distortion after encoding or playback conversion. The limiter should catch occasional peaks, not remove the performance’s dynamics. If it is reducing level repeatedly by several decibels, lower the track or bus feeding it and investigate the loudest moments.

Measure loudness across the whole episode

Peak meters show the highest instantaneous level, but they do not describe how loud a podcast feels over time. Loudness meters based on LUFS or LKFS are more useful for delivery. Integrated loudness measures the episode as a whole, short-term loudness shows a shorter window, and momentary loudness reacts to very brief changes. Using all three helps distinguish an isolated laugh from an episode that is broadly too quiet.

Many podcast producers work towards an integrated loudness near -16 LUFS for stereo programmes and around -19 LUFS for mono programmes. These are common reference points, not a universal law, and hosting platforms may apply their own normalisation. A stereo interview exported as a dual-mono file can create confusion, so check the channel format and listen in mono before delivery. A guest’s voice should remain intelligible even through a single smart speaker or a phone held at arm’s length.

Measure the complete episode, including the opening music, sponsor message, transitions and closing credits. A twenty-second sting can distort the overall reading if it is much louder than the spoken content. Check loudness range and short-term readings around every transition, then adjust clip gain or automation where the change feels abrupt. Treat the meter as evidence, not as a replacement for listening; two episodes with the same integrated LUFS can feel very different because of speech density, spectral balance and background noise.

Finish, check and deliver a listener-friendly file

Music and advertising need their own level strategy. Lower music beneath speech with automation or sidechain ducking, and create a short fade before the presenter begins. The music should support the mood without forcing the listener to turn down the programme. Pay particular attention to sponsor segments inserted into an interview, since a separately produced advert may have been mastered louder or with much denser compression than the rest of the episode.

Review the final mix on studio headphones, ordinary earbuds, a laptop or phone speaker, and a car system if available. Australian listeners often consume podcasts during the morning commute, while walking, cooking or travelling between suburbs, so speech must survive changing background noise and imperfect listening positions. A check in a car can reveal low-frequency rumble, harsh sibilance and an intro that disappears as soon as road noise enters the environment.

Verify that the beginning and end are clean, metadata is correct, and the export uses the required sample rate and file format. AAC can be suitable for some publishing workflows, while WAV is generally preferred when another party will perform the final encoding. Listen to the exported file rather than assuming the DAW session represents it perfectly. Encoding can expose clipped peaks, stereo problems or music that feels unexpectedly loud.

For creators who need recording, editing, mixing or mastering support, a professional studio can provide a controlled monitoring environment and a repeatable delivery process. LnL Recording’s studio rates can be useful when budgeting for an episode that includes remote interviews, narration, advertising inserts or a larger series workflow. The value is often in the final consistency: every episode should sound as though it belongs to the same programme.

Keep speech natural while meeting local responsibilities

Consistent loudness should never erase personality. A presenter’s emphasis, a guest’s laugh and a deliberate pause all carry meaning. The goal is to control distracting extremes while keeping the performance alive. If a voice sounds squeezed, brittle or unnaturally close, reduce the processing and return to the individual clips rather than adding another plug-in to compensate.

Australian production teams also need to consider consent and privacy when recording conversations. The rules for listening to or recording a private conversation differ between states and territories, so a remote guest should be told clearly when recording begins and how the material will be used. The Privacy Act 1988 may apply when personal information is collected, particularly in commercial or organisational settings, and permissions should cover publication, editing and promotional extracts where relevant.

Copyright clearance matters when adding music, archival audio or excerpts from another programme. A track that is acceptable for private editing may not be licensed for a public podcast, paid advertisement or branded series. Keep written records for commissioned music, guest releases and third-party clips. Good documentation protects the production and prevents a technically polished episode from needing to be withdrawn later.

Create a repeatable mastering template, but do not let it replace episode-by-episode judgement. Save a version with the final loudness reading, true-peak result and export settings. If the programme is distributed through several platforms, compare the uploaded result after processing and document any platform-specific requirements. A measured, carefully monitored workflow gives Australian audiences a comfortable listening experience whether they are using headphones in Brisbane, a kitchen speaker in Adelaide or a car on a regional highway.