Most new podcasters do not quit because their ideas run out. They quit because editing a single episode eats a whole evening, and they are never quite sure the result sounds right. The truth is that podcast editing is not one big mysterious task; it is a fixed sequence of small ones, and once you have a repeatable order, a forty-minute episode becomes a ninety-minute job instead of an all-nighter. This is that order, from raw recording to a published file, with the settings that make each stage work.
Set up for a clean recording first
Editing cannot fix a bad recording, so five minutes of setup saves an hour later.
- Pick a quiet, soft room. A bedroom with a rug, curtains, and a bed beats a bare kitchen every time. Soft surfaces soak up reflections that make a voice sound boxy.
- Mic position. Place the mic 10–15 cm from your mouth, slightly off-axis so plosives miss the capsule. Use a pop filter if you have one.
- Gain. Aim for peaks around −12 to −6 dBFS. If you record too quietly, you will raise the noise floor when you boost later.
- Record a few seconds of room tone. Sit still and silent for five seconds at the start. You will use this as a noise sample.
- Record a separate track per person if you can. One file per host makes every later step easier.
If you are recording over the internet, have each person record locally rather than relying on the call software, which compresses audio heavily. A local file at 48 kHz and 24-bit is plenty.
Stage 1: the rough cut
Do not polish anything yet. Your first job is to make the episode the right length and the right shape.
- Listen once at normal speed and drop markers wherever something goes wrong — a flubbed sentence, a long tangent, a phone buzzing.
- Cut the big mistakes, but be ruthless about tangents. If a story adds nothing, remove it now while the structure is still flexible.
- Trim the starts and ends so the file opens on speech and closes shortly after the last word.
- Do not fix small ums and breaths yet. That is a different pass, and mixing the two makes both slower.
The goal of the rough cut is content, not sound. Resist the urge to adjust levels here.
Stage 2: remove silence and tighten the pacing
Now you make the episode feel tighter. Dead air is the biggest reason a podcast drags.
Use a silence-removal pass rather than deleting every gap by hand. Set the threshold a few dB above your noise floor — if room tone sits at −55 dBFS, start around −45 dBFS. Set the minimum duration to about 400 ms so you only catch genuine gaps, not the small beats between words. Keep 80–150 ms of padding on each side so the tails of words survive, and add short 10–20 ms crossfades so cuts do not click.
Then listen back and loosen anything that sounds rushed. A half-second pause before a punchline is not wasted time; it is the setup. You want to remove unintentional silence, not the rhythm of speech.
Stage 3: clean the noise
Apply noise reduction after you have trimmed, because the tool now sees less material to work with.
- High-pass filter at 80–100 Hz for a male voice, 100–120 Hz for a female voice. This removes rumble and desk thumps with no effect on clarity.
- Notch any hum at 50 Hz (mains frequency in India) and its harmonics at 100 and 150 Hz if a buzz survives.
- Learn a noise profile from your room-tone clip, then reduce by a gentle 6–10 dB. This is the number beginners get wrong; pushing to 20 dB is what creates that underwater, robotic voice.
- Gate lightly in the pauses — threshold just above the noise floor, with a hold of 100–200 ms and a smooth release — so the gaps go truly quiet without chopping words.
If the recording was made in a noisy place, an AI denoiser handles the hard cases with almost no setup. Offex Audio's free online cleaner removes background noise from voice recordings without installing anything, which is handy for a quick rescue before the full edit.
Stage 4: shape the voice with EQ and compression
Now you make it sound professional. Two tools do most of the work.
EQ. A simple three-band approach is enough for speech:
- High-pass at 80–100 Hz (if you have not already).
- Cut 2–4 dB around 200–400 Hz if the voice sounds muddy or "cardboardy."
- Boost 2–4 dB around 3–5 kHz for presence and intelligibility.
- Gentle cut around 6–8 kHz if sibilance — hard "s" sounds — is harsh.
Compression. This evens out the loud and quiet parts so the listener is not constantly reaching for the volume. Start with a ratio of 3:1, a threshold that gives 4–6 dB of gain reduction, a fast attack around 10–20 ms, and a release around 100–200 ms. Follow it with a makeup gain so the level returns to where it started. If you can hear the compressor working, you have overdone it.
For a two-person show, apply the same chain to each voice track so they match. Do not compress the whole mix together — that ties the two voices to each other.
Stage 5: set consistent levels and loudness
Consistency across episodes matters more than absolute loudness. Podcasts are commonly delivered around −16 LUFS integrated (Apple and Spotify target roughly this range; −14 LUFS is a fine target for YouTube). True peak should stay below −1 dBTP to avoid distortion after lossy encoding.
Use a loudness meter to check, not your ears alone. Two rules:
- Match episode to episode. Listeners should never adjust volume when your next episode starts.
- Keep dialogue roughly even. If one host is 4 dB louder than the other, fix it on the individual tracks before the final mix.
Stage 6: add music and export
If you use an intro and outro, keep the music 12–18 dB below the voice so it never fights the speech. Fade it out under the first words. If you have a voice-over-music file you want split, stem separation can pull the two apart cleanly.
Export as a WAV at 48 kHz, 24-bit as your archive master, then make an MP3 at 128 kbps or higher (192 kbps is safer) for publishing. Name the file clearly with the episode number and date so you can find it later.
Conclusion
A repeatable workflow is what turns podcast editing from a chore into a routine. Record cleanly, cut the content, remove the dead air, clean the noise, shape the voice, set the levels, and export. Each stage has one job, and doing them in this order means you never fight a problem twice. Do it a few times and the sequence becomes muscle memory — and you get your evenings back. Start with the recording setup and the silence pass, since those two alone fix the majority of what makes a beginner podcast sound like a beginner podcast.
Frequently asked questions
How long should editing one podcast episode take?
With a repeatable workflow, a one-hour conversation usually edits down in under an hour once you stop re-deciding every step. Trim silence, clean noise, level, then export.
Do I need paid software to edit a podcast?
No. You can remove silence, clean background noise and prepare files with free browser tools, then use any editor for final polish if you want more control.
What order should I edit in?
Trim dead air first, then reduce steady noise, then balance loudness between speakers, and export last. Doing noise removal after levelling undoes your level work.