Podcast Audio Chain: Silence → Normalize (-16 LUFS) → EQ → MP3

The chain (in this order — order matters)

1. Top and tail.Cut mic checks, countdowns, and trailing room tone. Edit content first — never process audio you'll delete.
2. Silence removal.Collapse long pauses to breaths (~0.5s), strip dead air. Tighten, don't sterilize — zero-silence speech sounds robotic.
3. EQ for voice. High-pass 80Hz, small wide cut at 300–500Hz if muddy, gentle presence lift at 3–5kHz. If you hear the EQ working, back it off.
4. Normalize to -16 LUFS. The directory standard (true peak ≤ -1 dBTP). Do this after EQ — anything level-changing later invalidates it.
5. Fades + MP3 128k export. Short fades on intros/outros, then joint-stereo MP3 at 44.1kHz (~1MB/min). Archive a WAV master first; derive MP3s from it, never from MP3s.

ffmpeg -i raw.wav -af "silenceremove=start_periods=1:start_duration=0.5:start_threshold=-50dB,highpass=f=80,loudnorm=I=-16:TP=-1.5:LRA=11" -c:a libmp3lame -b:a 128k episode.mp3

That line is genuinely useful for one file — and genuinely miserable for forty. Which is the entire point of what follows.

Quick Solution

Save the chain once — finish every episode identically.

Five steps × forty episodes is where podcasts go to die in Audacity menus. Formatif Pro ($8.95 one-time) saves this exact chain — silence removal, voice EQ, -16 LUFS normalization, fades, MP3 export — as a reusable recipe and runs whole seasons through it offline, with per-file status. Consistency across episodes is half of “sounding professional,” and consistency is what batching guarantees.