Quick Reference
| Parameter | Requirement | Notes |
|---|---|---|
| RMS level | -24 to -14 dB | Wider than ACX on both ends; an ACX master passes automatically |
| True peak | Not specified | -3 dBTP recommended to stay ACX-compatible |
| Noise floor | Under -60 dB | No background noise, hiss or excessive mouth noise |
| Head silence | 0.5 – 1 second | Before the audio starts |
| Tail silence | 1 – 5 seconds | After the audio ends |
| File format | MP3 preferred, WAV, FLAC | MP3 192+ kbps CBR preferred; FLAC compression level 5+ |
| Sample rate / depth | 44.1 kHz / 16-bit | Stated for MP3 and WAV deliveries |
| Channel format | Mono or stereo | Either works, but not a combination across files |
| File length | Max 120 minutes per file | Longer chapters split as "Chapter X, continued" |
| Opening credits | Required, separate file | Title and author/narrator names only |
| Closing credits | Single separate file | Must reference the completion of the book |
| Retail sample | Max 5 minutes | No music, credits or explicit content; named ISBN_sample.mp3 |
How Spotify Audiobooks Measures These Values
RMS, wider than ACX
Loudness is specified as -24 to -14 dB RMS, a raw signal power measurement like ACX uses, not BS.1770 LUFS. The window is wider than ACX's -23 to -18 dB on both ends, which makes Spotify the most forgiving of the audiobook platforms on level: a master leveled for ACX is inside this range before you check anything.
No true peak ceiling, but keep one anyway
The guide publishes no true peak or sample peak limit. That does not make peaks irrelevant: MP3 encoding raises inter-sample peaks, and a master that slams 0 dBFS will clip audibly after transcoding. Keeping -3 dBTP costs nothing on spoken word and keeps the same file valid for ACX and Findaway, which do enforce a ceiling.
Consistency is a hard rule
Two consistency requirements are explicit: every file is mono or stereo, never a combination, and chapter headings are all-or-nothing - every file starts with a verbal chapter indication or none does. Like ACX, the title is evaluated as a set, so one deviating file holds up the whole book.
Structure: credits, chapters, sample
The delivery is more than chapters. Opening credits are a required standalone file (title and names only), closing credits close the book in a single file, and the retail sample is capped at 5 minutes with no music, credits or explicit content. Files longer than 120 minutes must be split. Naming the files {ISBN}_{track number}.mp3 keeps the upload ordering itself.
What Spotify Checks Beyond the Numbers
The guide also asks for subjective quality: high clarity, no distracting edits, no background hiss, no excessive mouth noise, not harsh or tinny. Those are reviewer judgments, not meter readings. The measurable part - levels, noise floor, silences, format, consistency - is what you can verify completely before upload, and missing it is what delays availability in stores.
FAQ
Does Spotify Audiobooks have a LUFS or true peak requirement?
No. The official Spotify for Authors guide specifies loudness as RMS between -24 and -14 dB and sets no true peak ceiling at all. If you also deliver to ACX or Findaway, keep true peak at -3 dBTP anyway: it costs nothing on a voice master and keeps one render valid for every platform.
Can I reuse my ACX master for Spotify Audiobooks?
Yes. The ACX working range of -23 to -18 dB RMS sits entirely inside Spotify's -24 to -14 dB window, the noise floor requirement is the same -60 dB, and the head/tail silence windows are identical. Spotify even prefers the same MP3 format ACX requires. A passing ACX file passes Spotify Audiobooks.
Mono or stereo - which should I deliver?
The guide accepts both, with one hard rule: not a combination. Every file in the audiobook must share the same channel format. For single-narrator titles mono is the sensible choice: identical listening result, half the data, no channel-balance risk.
What happens to chapters longer than two hours?
No single file may exceed 120 minutes. The guide's instruction is to split long chapters and introduce the continuation file as "Chapter X, continued". Books without chapters should be split into segments of 30 to 120 minutes each.
What are the credits and retail sample rules?
Opening credits are required as a separate file containing only the title and author/narrator names, in the form "[Title], Written by [Author], Narrated by [Narrator]". Closing credits are a single separate file that must reference the completion of the book. The retail sample is capped at 5 minutes, may not contain music, credits or explicit content, and should be named ISBN_sample.mp3. If you skip it, Spotify auto-generates one from a chapter.