Podcast Editing Tutorial for Journalists

1. Planning the edit before you start

Editing goes far faster when you know what you are building before you open the software. Listen back to your material, or read the transcript, and mark the strong moments: the clips that carry the story, the best answers, the points where the argument turns. Sketch a running order — the sequence of segments, where narration or links go, where interview clips land, where any music or actuality sits. For a news podcast this might follow a fixed format; for a narrative piece it is the shape of the story. Either way, a plan turns editing from aimless scrubbing into executing a structure.

Good editing also depends on good raw material, so a few recording habits pay off enormously in the edit. Record in the quietest space you can, get the microphone close to the speaker, and capture a few seconds of room tone — the sound of the silent room — at every location, which you will use later for noise reduction and to fill gaps naturally. Keep your original recordings safe and untouched; you will edit copies, and you may need the unedited source later to demonstrate that your edit was fair.

2. The tools: Audacity, Audition, Descript, Reaper

You do not need expensive software to edit a professional-sounding podcast. Audacity is free, open-source, cross-platform, and more than capable of the entire workflow in this guide — cutting, level changes, noise reduction, EQ and export. It is the recommended starting point. Adobe Audition is a full professional audio workstation, worth using if your newsroom already has an Adobe subscription, with strong restoration and multitrack tools.

Descript works differently: it transcribes your audio and lets you edit the sound by editing the text, deleting a sentence in the transcript to delete it in the audio, which many journalists find fast and natural for interview-heavy work. Reaper is a low-cost, highly capable digital audio workstation popular with more advanced producers who want fine control. The advice is simple: learn the fundamentals in one tool — Audacity is ideal — before worrying about the others, because the concepts of cutting, levelling and cleaning transfer directly between all of them.

3. The rough cut

The rough cut is your first assembly: the segments in order, roughly trimmed, giving the episode its shape and rough length. Import your recordings, lay them out on the timeline in the running order you planned, and cut away the obvious waste — the false starts, the long setup chatter, the section where the phone rang. Do not polish at this stage. The aim is to hear the whole thing end to end, confirm the structure works, and see whether it runs long or short, so you know how ruthless the next pass needs to be.

Work non-destructively wherever your tool allows: edit on copies of your clips, and use the undo history freely so experimenting is cheap. If your software supports multiple tracks, keep voice, music and actuality on separate tracks so you can adjust each independently later. Save often, and save versions, so you can step back to an earlier assembly if a bold cut turns out to be a mistake. A rough cut you can listen to all the way through is the milestone that tells you the episode will work.

4. Tightening: filler, pauses and pace

Tightening is where a serviceable recording becomes a pleasure to listen to. Speech is full of filler — ums, ers, false starts, repeated words — and long pauses that feel natural in the room but drag on tape. Remove them carefully, closing the gaps so the result still sounds natural rather than clipped. Trim overlong tangents. The goal is momentum: every second that does not earn its place is a second a listener might drift away. Ruthless tightening is often the single biggest improvement an amateur edit can make.

Restraint matters, though. Cut too tight and speech sounds unnatural and breathless; leave a little breathing space so it still sounds human. When you remove a pause, replacing dead silence with your recorded room tone keeps the background consistent and hides the join — an abrupt cut to true digital silence is jarringly obvious. And every tightening edit inside an interview answer must respect the meaning of what was said, a point the ethics section returns to: removing an um is fine, removing a crucial qualifier is not.

5. Balancing levels

Different sources are recorded at different volumes: your narration, a studio guest, a phone interview and a clip of actuality will rarely match. Balancing levels means adjusting each so the listener never has to lunge for the volume control. Aim for a consistent perceived loudness across voices, with music and background sitting clearly below speech so it never fights the words. Use your editor's level tools and watch the meters, rather than trusting your ears alone at one fixed monitoring volume, which can deceive you.

Two tools help here. Normalisation raises or lowers a whole clip to a target level. Gentle compression reduces the gap between the loudest and quietest parts of a voice, evening out a speaker who veers between a whisper and a boom, which makes speech easier to follow on phone speakers and in noisy places. Apply compression modestly; over-compressed audio sounds flat and fatiguing. Leave a little headroom below the maximum so nothing distorts, and save the final overall loudness matching for the export stage covered next.

6. Noise reduction and EQ basics

The best noise reduction happens before you record — a quiet room and a close microphone — but the edit can rescue a lot. Most tools have a noise reduction feature that samples a short passage of pure background (this is where your room tone earns its keep), builds a profile of that noise, and subtracts it from the whole track. Use it gently: push it too far and voices take on a watery, artificial quality that is worse than the hum you were removing. A modest reduction that leaves a trace of natural background usually sounds best.

EQ (equalisation) shapes the tone of a voice by boosting or cutting particular frequency ranges. For speech, a gentle high-pass filter that rolls off the very low frequencies removes rumble, handling noise and air-conditioning hum without touching the voice. Small, careful boosts can add clarity or warmth, but heavy-handed EQ quickly sounds unnatural. As with everything in cleaning, less is more: the aim is a clear, natural voice that does not call attention to the processing. Make these fixes on copies so you can always compare against the untouched original.

7. Loudness standards for podcasts

Podcasts are measured for loudness in LUFS — loudness units relative to full scale — which reflect how loud audio actually sounds to a listener, rather than just its peak level. The measure that matters is integrated loudness: the average perceived loudness across the whole episode. Matching a consistent integrated loudness is what stops your podcast sounding dramatically quieter or louder than the show a listener played before yours, which otherwise sends them reaching for the volume every time.

A widely used target for stereo podcasts is an integrated loudness of around -16 LUFS, with mono episodes commonly aimed a little lower. Treat this as a sensible, commonly cited target rather than a single universal rule — different platforms and producers vary, and you should check any specific figure your hosting or distribution service publishes. Most editors include loudness normalisation: you set the integrated target and the tool adjusts the whole file to hit it. Also keep an eye on the true peak so the loudest moments do not clip and distort.

Common podcast loudness targets (widely cited, not a single law):
  Integrated loudness (stereo):  around -16 LUFS
  Integrated loudness (mono):    often a little lower
  True peak:                     keep below 0 dBFS to avoid clipping

Set the integrated LUFS target in your editor's loudness
normalisation tool, then check the true peak before export.

8. Music, clips and licensing

Music and sound design lift a podcast, but almost all music is protected by copyright, and using a commercial track without permission can expose you and your publisher to legal action regardless of the platform you release on. Use only music you are properly licensed for: royalty-free or production-music libraries where you buy or subscribe to a licence; tracks released under a licence such as Creative Commons, whose terms — including any required attribution — you follow exactly; or music you commission or own outright. Read each licence for how it may be used, and never assume a track is free because it is easy to find.

The same care applies to sound effects, and especially to clips of other broadcasts or recordings — a snatch of a news bulletin, a song, a film line. These carry their own rights, and any reliance on fair-dealing exceptions for reporting or criticism is a legal judgement to check rather than assume. The practical discipline is a music and audio log: for every piece of third-party audio in the episode, record what it is, where it came from, and the licence that permits your use. If a rights question ever arises, that log is your evidence that you had permission to use what you used.

9. Exporting and file formats

When the edit is finished, cleaned and loudness-matched, export the final mix. Podcasts are almost universally distributed as MP3, which every player supports and which keeps file sizes reasonable for download and streaming; a moderate bitrate is fine for speech-led content and keeps episodes light. Keep your editable project file and, ideally, a high-quality master export as well, so you can re-export or make changes later without going back to raw clips.

Fill in the episode's metadata — title, episode number, description — and add ID3 tags if your workflow uses them, so the episode displays correctly in listening apps. Give a final, full listen to the exported file on a couple of devices, ideally including phone speakers and earphones, since that is how most people will hear it. This last check catches the problems that slip past on studio monitors: a level that is too low on a phone, a join that clicks, a music bed that buries a crucial word.

10. Accuracy and consent when editing interviews

Editing an interview is not just a technical task; it is an editorial and ethical one. Editing for length and clarity is entirely legitimate and expected — raw interviews are long and full of hesitation. The firm line is that editing must never change the meaning of what a contributor said. You may remove filler, tighten pauses, cut tangents and reorder for flow, but you must not splice words to manufacture a sentence the person never uttered, cut away essential context or qualification, or place answers next to each other to imply a connection that was not there. The test: would the speaker recognise the edit as a fair account of what they meant?

This is reinforced by the standards journalism works to. Broadcasters regulated by Ofcom must not materially mislead audiences, and the same principle underpins the press codes and simple professional honesty: audio is powerful precisely because a voice feels like direct testimony, so distorting it is a serious breach of trust. Protect yourself and your contributor by keeping the original unedited recording, so you can always show the source if an edit is challenged.

Consent runs alongside accuracy. Contributors should understand they are being recorded for publication and in what context, and vulnerable interviewees deserve particular care. Where your interview involves anything sensitive, agreeing the terms of use in advance protects everyone; our broadcast interview request pack offers templates for setting those expectations clearly. Handling consent and accuracy well is what separates journalistic audio from mere content.

A useful discipline is to keep a short edit log for any contentious interview, noting what you cut and why. It is rarely needed, but if a contributor later claims they were misrepresented, being able to show the original recording alongside a record of your editing decisions is a powerful defence — and the act of keeping the log tends to make your editing more careful in the first place.

11. Pre-publish checklist

Run through this before you release an episode:

• The original, unedited recordings are saved and untouched
• Every interview edit preserves the speaker's meaning and context
• Filler, dead air and tangents removed without sounding clipped
• Levels balanced so voices are consistent and music sits below speech
• Noise reduction and EQ applied gently, with no watery artefacts
• Exported to a sensible integrated loudness target (around -16 LUFS for stereo) with true peak below clipping
• All music, effects and clips are properly licensed and logged
• Episode metadata and description completed
• Final file checked on phone speakers and earphones, not just monitors
• Contributor consent and any agreed conditions of use are on record

12. Common mistakes to avoid

• Editing that changes meaning — the one unbreakable rule; never splice words or cut context so a contributor appears to say something they did not.
• Over-tightening speech — cutting every breath leaves audio sounding unnatural and breathless; leave a little human space.
• Pushing noise reduction too hard — heavy processing makes voices sound watery, which is worse than a faint natural background.
• Cutting to true digital silence — fill gaps with recorded room tone so joins do not jump out.
• Using unlicensed music or clips — almost all music is under copyright; use only material you are licensed for and log it.
• Ignoring loudness — an episode far off a sensible target such as around -16 LUFS forces listeners to reach for the volume.
• Checking only on studio monitors — most people listen on phone speakers and earphones, so audition there before release.
• Not keeping the original recording — you may need the unedited source to show an edit was fair if it is ever challenged.

Frequently asked questions