Podcast Audio Cleanup Without Hiring a Sound Engineer

You do not need a sound engineer to fix most podcast audio

Most rough podcast audio has the same problems: room hum, computer fan hiss, harsh Zoom compression, plosives, keyboard clicks, or a phone recording that sounds distant. A sound engineer helps when you have a difficult room, multiple mics, or a client deadline. For a solo podcast, YouTube voiceover, or gaming clip, you can usually get 80–90% of the way there with a repeatable cleanup chain and a little restraint. The goal is not perfect silence. The goal is intelligible, comfortable speech that does not sound processed.

Step 1: Fix what you can before cleanup

Cleanup plugins work better when they are not rescuing a bad capture. Before you touch noise removal, check the recording itself.

• Record 10–15 seconds of room tone at the start. You will use it to profile the noise.
• Keep your mouth 6–8 inches from the mic and slightly off-axis. Too close means plosives; too far means room echo.
• Set gain so loud peaks hit around −12 to −6 dBFS. If the waveform is flat-topped, re-record. Clipping cannot be undone.
• For Zoom or Google Meet, use local recording if the app supports it, not cloud recording. Cloud versions often add compression and dropouts.
• Turn off fans, air purifiers, and noisy keyboards during the take. A $10 blanket behind you is more useful than a $200 plugin.

Step 2: Run cleanup in the right order

Order matters. If you compress first, you make the noise louder and harder to remove. Here is a practical chain that works in Audacity, Reaper, Adobe Audition, or most editors.

1. Edit and trim first

Remove false starts, long pauses, and obvious clicks before processing. Do not over-edit breaths; natural breathing sounds better than a choppy gate. If you need a fast first pass on a noisy voice memo or gaming clip, ClearVoice can remove steady hiss without installing a DAW. Just do not stack it with three other noise reducers.

2. High-pass filter

Set a high-pass filter at 80 Hz for deeper voices or 100 Hz for higher voices, with a 12 dB/octave slope. This removes rumble, desk bumps, and HVAC low end without thinning the voice. If the result sounds thin, lower the cutoff by 10–20 Hz.

3. Noise removal with restraint

In Audacity, select a quiet room-tone section, open Noise Reduction, and click Get Noise Profile. Then select the whole track and use Reduction 6–10 dB, Sensitivity 5–6, and Frequency Smoothing 3. Apply it once. If you still hear hiss, fix the source or lower the noise floor instead of pushing reduction to 20 dB. Over-reduction creates watery, metallic artifacts and can make consonants lisp.

If you want a free browser-based option for that first pass, try ClearVoiceHQ (clearvoicehq.com), then finish with EQ and compression in your editor. Treat it as a noise-removal step, not a magic mastering button.

4. EQ for clarity

Use a narrow cut of 2–3 dB around 200–400 Hz if the voice sounds muddy. Add 1–2 dB around 2–5 kHz for presence. If the voice sounds harsh, cut 2–4 dB around 6–8 kHz. Make small moves and bypass the EQ often. If you cannot hear the difference, you probably do not need the move.

5. Compression

Start with a ratio of 2:1 to 3:1, threshold around −18 to −12 dB, attack 10–20 ms, and release 100–150 ms. Add makeup gain so the average level comes back up. You want consistent volume, not a crushed wall of sound. If you hear pumping or breathing, raise the threshold or lower the ratio.

6. De-ess

Target the 5–8 kHz range. A narrow 3–6 dB reduction on harsh S sounds is usually enough. Too much de-essing makes people sound like they have a lisp. Listen on headphones and phone speakers before committing.

7. Loudness normalize

For a mono podcast, aim for about −16 LUFS integrated. For stereo, −14 LUFS is common. Set true peak to −1 dBTP. Export a 24-bit WAV at 48 kHz for archive, then make a 192–320 kbps MP3 for distribution. Do not normalize to 0 dBFS; streaming platforms will turn it down and you will lose headroom.

Quick fixes for Zoom calls, phone memos, and gaming clips

• Zoom calls: Use the local audio recording if available. If not, ask each guest to record their own track. Cloud audio often has compression artifacts that noise removal cannot fix. Apply a gentle gate at −40 to −35 dB only if background noise is constant; aggressive gating makes speech choppy.
• Phone voice memos: High-pass at 100 Hz, noise reduction at 6 dB, and a light de-esser. Phone mics exaggerate proximity boom and sibilance.
• Gaming clips: Separate your mic track from game audio if your capture software allows it. Use a gate or expander to reduce keyboard clicks, then compress lightly. If you cannot separate tracks, accept more room noise; removing it will damage the game audio.

Listen for the tradeoffs, not just the noise

Every cleanup move has a cost. Noise removal removes some voice detail. Compression raises room tone. Gates chop breaths and tails. EQ can make a warm voice sound thin. The test is simple: bypass the whole chain and A/B it. Listen on headphones, a phone speaker, and a laptop speaker. If the processed version sounds more fatiguing after two minutes, back off. It is better to leave a little room tone than to make the host sound like a robot.

A repeatable 15-minute cleanup workflow

1. Trim obvious mistakes and silence, but keep natural pauses.
2. Apply high-pass filter at 80–100 Hz.
3. Profile room tone and apply noise reduction at 6–10 dB.
4. Make 2–3 small EQ moves for mud, presence, and harshness.
5. Compress at 2:1–3:1 with 10–20 ms attack and 100–150 ms release.
6. De-ess 3–6 dB around 5–8 kHz.
7. Normalize to −16 LUFS mono or −14 LUFS stereo, true peak −1 dBTP.
8. Export WAV and MP3, then listen once on phone speakers.

That is the whole job. You do not need a sound engineer for every episode. You need a quiet recording, a short processing chain, and the discipline to stop before the cleanup becomes the problem.