ACX Audio Requirements Explained: The Complete 2026 Checklist

If you are submitting an audiobook to Audible through ACX (the Audiobook Creation Exchange), your audio files must meet a precise set of technical requirements. Files that fall outside these specifications are rejected, which means re-mastering, re-uploading, and waiting through the review queue again. Understanding each requirement upfront saves time, money, and frustration.

This guide breaks down every ACX audio specification in plain language, explains why each one matters, and covers the most common reasons audiobooks get rejected.

The Complete ACX Specification Checklist

Each Specification Explained

Sample Rate: 44.1 kHz

The sample rate determines how many audio snapshots are captured per second. At 44.1 kHz, the audio captures frequencies up to 22.05 kHz — well beyond the range of human hearing, which tops out around 20 kHz. This is the same sample rate used on CDs and is the standard for spoken word audio.

Why it matters: lower sample rates (like 22 kHz or 16 kHz) produce audio that sounds muffled or thin, especially on high-quality headphones and speakers. Higher rates (48 kHz, 96 kHz) are unnecessary for voice and waste file size. ACX specifically requires 44.1 kHz — not 44 kHz, not 48 kHz.

Bit Rate: 192 kbps Constant

The bit rate controls how much data is used to encode each second of audio. At 192 kbps, you get clear, artifact-free speech without excessive file sizes. ACX requires constant bit rate (CBR), not variable bit rate (VBR). This ensures consistent quality throughout the file.

Why it matters: lower bit rates introduce compression artifacts — subtle buzzing, warbling, or loss of clarity that listeners notice during quiet passages. Variable bit rate can cause playback issues on some devices and is not accepted by ACX.

Channels: Mono

ACX requires mono (single channel) audio, not stereo. This might seem counterintuitive — stereo sounds better for music, right? For spoken word, mono is actually preferable. It ensures the narrator's voice comes through both earbuds or speakers identically, with no spatial effects that could be distracting during long listening sessions.

Why it matters: stereo files are rejected outright. Many recording setups default to stereo, so this is a common mistake. Always verify your files are mono before submitting.

Loudness: -23 LUFS

Integrated loudness is measured in LUFS (Loudness Units Full Scale) and represents the average perceived volume of the entire file. ACX targets -23 LUFS, with an acceptable RMS range of -18 to -23 dB. This ensures that listeners do not have to adjust their volume between chapters or between different audiobooks on the platform.

Why it matters: audio that is too quiet forces listeners to crank up their volume, then get blasted when switching to music or another audiobook. Audio that is too loud sounds compressed and fatiguing. -23 LUFS is the sweet spot for comfortable spoken-word listening across devices.

Peak Level: -3 dB Maximum

The peak level is the loudest instantaneous moment in the audio file. ACX caps this at -3 dB, which provides enough headroom to prevent clipping (distortion from the signal exceeding maximum capacity) across all playback devices and listening environments.

Why it matters: peaks above -3 dB can cause harsh distortion, especially through headphones or car speakers at high volume. Even if the clipping is barely audible on your monitoring setup, it may be very noticeable on a listener's cheap earbuds.

Noise Floor: -60 dB or Lower

The noise floor is the level of background noise present when no one is speaking. ACX requires this to be at or below -60 dB, which means the silence between words and sentences should be genuinely silent — no hum from electronics, no hiss from poor-quality microphones, no ambient room noise.

Why it matters: listeners often use audiobooks in quiet environments (bedtime, early morning) and with headphones that amplify subtle sounds. Background hiss that you can barely hear through studio monitors becomes obvious and annoying through earbuds at night. This is one of the most common rejection reasons for home-recorded audiobooks.

Room Tone: Opening and Closing Silence

Every audio file must begin with 0.5 to 1 second of room tone (silence) and end with 1 to 5 seconds of room tone. This is not dead silence (which would sound unnatural) but rather a brief moment of quiet that matches the ambient characteristics of the recording environment.

Why it matters: room tone provides a natural transition between the player interface and the narration. Without it, the audio starts abruptly, which is jarring. The trailing room tone prevents the audio from cutting off mid-breath or mid-word if there is any playback buffering.

File Structure Requirements

Beyond the audio specifications, ACX has structural requirements:

• Opening credits: The first file must include the book title, author name, and narrator name, spoken aloud.
• Closing credits: The final file must include a brief closing statement (usually a repeat of the title and narrator).
• Chapter files: Each chapter is a separate file, named sequentially.
• Consistent audio quality: All files in a submission must have matching technical characteristics — same loudness, same noise floor, same tonal quality. Inconsistencies between chapters are a rejection reason.

Common ACX Rejection Reasons

Based on publicly shared experiences from narrators and producers, these are the most frequent causes of ACX rejections:

• Loudness outside the -18 to -23 dB RMS range. This is the single most common rejection reason. Many home recordings are either too quiet or too loud.
• High noise floor. Background hum from computers, air conditioning, or electrical interference.
• Missing room tone. Files that start or end abruptly without the required silence padding.
• Stereo files instead of mono. An easy mistake if your recording software defaults to stereo.
• Inconsistent quality between chapters. Recording on different days with different mic positions or room conditions creates audible variation.
• Peaks above -3 dB. Plosive sounds (P, B, T) and sharp breaths can cause peaks that exceed the limit.
• Wrong sample rate. Using 48 kHz (a video standard) instead of 44.1 kHz (the audio standard).

How AudioBookByMe Handles ACX Compliance

Every audio file generated by AudioBookByMe is automatically formatted to meet all ACX specifications. There is no manual audio engineering required. Specifically, the platform handles:

• 44.1 kHz sample rate, 192 kbps constant bit rate, mono output
• Loudness normalization to -23 LUFS with peaks limited to -3 dB
• Noise floor at -60 dB or below (AI-generated audio has no ambient noise by nature)
• Automatic room tone padding at the beginning and end of each chapter file
• Consistent quality across all chapter files (since the same voice model generates every chapter)
• Sequential chapter file naming

This means you can upload your generated files directly to ACX without opening an audio editor, applying compression, normalizing levels, or any of the other post-production steps that traditionally required specialized knowledge and software.

Tools for Checking ACX Compliance

If you are producing audio through other means and want to verify compliance before submitting, these tools can help:

• ACX Check plugin (Audacity): A free plugin for the open-source audio editor Audacity that analyzes your files against ACX specifications.
• Youlean Loudness Meter: A free loudness meter that shows real-time and integrated LUFS measurements.
• FFmpeg: A command-line tool that can verify sample rate, bit rate, channels, and other technical properties of audio files.