Normalize audio levels

“Airstation” (a podcast network like Gimlet) publishes shows from dozens of creators. Each submits audio recorded on different equipment at different levels. Listeners shouldn’t have to adjust volume between episodes.

LUFS (Loudness Units Full Scale) normalization measures perceived loudness rather than peak amplitude, giving consistent results across content.

API

ittybit audio \
  -i https://airstation-app.com/submissions/show-14-ep-8.wav \
  --normalize lufs \
  --target_lufs -16 \
  --format mp3 \
  --quality high

const task = {
  input: 'https://airstation-app.com/submissions/show-14-ep-8.wav',
  kind: 'audio',
  options: {
    normalize: 'lufs',
    target_lufs: -16,
    format: 'mp3',
    quality: 'high',
  },
};

const res = await fetch('https://api.ittybit.com/jobs', {
  method: 'POST',
  headers: {
    Authorization: `Bearer ${process.env.ITTYBIT_API_KEY}`,
    'Content-Type': 'application/json',
  },
  body: JSON.stringify(task),
});
const data = await res.json();

import requests

task = {
    "input": "https://airstation-app.com/submissions/show-14-ep-8.wav",
    "kind": "audio",
    "options": {
        "normalize": "lufs",
        "target_lufs": -16,
        "format": "mp3",
        "quality": "high",
    },
}

res = requests.post(
    "https://api.ittybit.com/jobs",
    headers={"Authorization": f"Bearer {api_key}"},
    json=task,
)
data = res.json()

TASK='{
  "input": "https://airstation-app.com/submissions/show-14-ep-8.wav",
  "kind": "audio",
  "options": {
    "normalize": "lufs",
    "target_lufs": -16,
    "format": "mp3",
    "quality": "high"
  }
}'

curl -X POST https://api.ittybit.com/jobs \
  -H "Authorization: Bearer $ITTYBIT_API_KEY" \
  -H "Content-Type: application/json" \
  -d "$TASK"

CLI

ittybit audio \
  -i show-14-ep-8.wav \
  -o show-14-ep-8.mp3 \
  --normalize lufs \
  --target_lufs -16 \
  --quality high

Recommended LUFS targets

If unsure, use -16 for speech and -14 for music. These match what major platforms normalize to internally — sending audio at the target level avoids double-normalization artifacts.

True peak limiting

Normalization can push peaks above 0 dBTP, causing clipping on some devices. Set a true peak ceiling:

ittybit audio \
  -i show-14-ep-8.wav \
  -o show-14-ep-8.mp3 \
  --normalize lufs \
  --target_lufs -16 \
  --true_peak -1

A ceiling of -1 dBTP is safe for all codecs and playback systems. Apple Podcasts recommends -1 dBTP.

Batch normalize a catalog

Normalize all episodes in a series to the same level:

for file in episodes/*.wav; do
  ittybit audio \
    -i "$file" \
    -o "normalized/$(basename "${file%.wav}.mp3")" \
    --normalize lufs \
    --target_lufs -16 \
    --true_peak -1 \
    --quality high
done

Peak vs LUFS normalization

LUFS is the right choice for consistent loudness. Peak normalization is only useful when you need to maximize headroom before further processing.