Table of Contents

Understanding ACX Audio Specifications: The Technical Foundation

ACX mandates uncompressed or lossless audio formats to guarantee consistent playback across all Audible, Amazon, and iTunes platforms. Understanding the exact digital parameters is non-negotiable. Let's break down the required specifications.

Accepted File Formats: WAV and FLAC

The only accepted formats are WAV (Waveform Audio File Format) and FLAC (Free Lossless Audio Codec). While FLAC offers smaller file sizes due to compression, the industry standard is WAV. WAV is uncompressed, preserving the full dynamic range of your recording without any potential artifacts. Never submit MP3, AAC, or other lossy formats, as they will be rejected immediately.

Why WAV is Preferred

WAV files store audio data as raw PCM (Pulse-Code Modulation) without any lossy compression. This ensures that every nuance of your narration—from subtle emotional shifts to plosive consonants—is captured exactly as recorded. FLAC uses lossless compression, which can reduce file size by 50% or more, but it requires decoding during playback. Some production environments introduce slight delays with FLAC, so most audio engineers stick to WAV for consistency. ACX accepts both, but choose WAV for maximum compatibility with their quality control systems.

Sample Rate and Bit Depth: The Digital Blueprint

Your audio must adhere to these exact digital specifications:

  • Sample Rate: 44.1 kHz. This is a non-negotiable standard for all audiobooks. Recording at 48 kHz is common for video, but you must down-sample to 44.1 kHz for ACX.
  • Bit Depth: 16-bit or 24-bit. 16-bit is the minimum acceptable standard. 24-bit provides significantly more headroom for dynamic range, which is strongly recommended during recording and editing. You can dither to 16-bit during the final export if required.
  • Channels: Mono or Stereo. While ACX accepts stereo, the vast majority of audiobooks are listened to in mono. Recording and delivering in mono simplifies your workflow, cuts file sizes in half, and eliminates potential phase correlation issues.

Understanding Sample Rate Conversion

If you record at 48 kHz, you must convert to 44.1 kHz using a high-quality sample rate converter (SRC). Most DAWs like Reaper and Pro Tools use r8brain or similar algorithms. Avoid low-quality SRC that introduces aliasing or high-frequency roll-off. When converting, export at 32-bit float intermediate format to prevent dithering errors, then finalize to 16-bit or 24-bit at 44.1 kHz.

Loudness and Dynamics: Meeting the ACX Meter Standards

Loudness is measured using RMS (Root Mean Square) and Peak levels. ACX requires a specific balance to ensure consistent volume across all listening environments.

  • RMS Level: Your audio must average between -18 dB and -23 dB. This is often called the "average loudness." Aim for the center of this range (-20 dB) for the safest pass. A voice that is too quiet or too loud will be rejected.
  • Peak Level: The loudest instantaneous sound in your audio must not exceed -3 dB. This provides essential headroom and prevents digital clipping, which causes harsh distortion.

Use a dedicated loudness meter plugin, such as the free YouLean Loudness Meter, to monitor these levels during your mastering stage. Note that ACX does not use LUFS like some streaming platforms; RMS is the specified metric. However, you can use a LUFS meter as a reference, but always double-check with an RMS meter.

The Importance of Headroom

Peak levels at -3 dB leave 3 dB of headroom below digital zero (0 dBFS). This margin prevents inter-sample peaks that could cause distortion during playback on devices with lower-quality DACs. If your audio hits 0 dBFS at any point, the file will fail ACX's quality check. Use a brickwall limiter with the ceiling set to -3 dB to enforce this limit.

Preparing Your Workstation and Recording Environment

Your recording environment directly impacts your ability to meet ACX's noise floor requirements. A quiet space and clean signal chain are essential.

Acoustic Treatment and Noise Floor

ACX requires a noise floor of -60 dB or lower. This includes room tone, computer fan hum, electrical interference, and external environmental noise. To achieve this, you need a properly treated space. Use absorption panels to reduce reverb and a reflection filter around your microphone. Record a few seconds of silence in your treated space to measure the noise floor. If it is above -60 dB, you must identify and eliminate the source (e.g., turn off HVAC systems, move away from computer fans, use a noise gate during editing).

Measuring Noise Floor Accurately

Use your DAW's noise print function or a spectrum analyzer. Record 10 seconds of silence with your microphone at your normal recording position. The RMS level of that silent section should be below -60 dB. If not, check for ground loops (use a balanced cable), electrical hum (move away from power supplies), or environmental noise (close windows, turn off appliances). A noise gate can remove residual low-level hiss, but it must be set carefully to avoid cutting out the end of words.

Essential Gear Recommendations

While you don't need a professional studio, quality gear is required to meet ACX standards.

  • Microphone: A large-diaphragm condenser (e.g., Audio-Technica AT2020, Rode NT1-A) or a dynamic microphone (e.g., Shure SM7B). Avoid USB microphones for professional submissions due to lack of control.
  • Audio Interface: Provides clean preamps and stable analog-to-digital conversion (e.g., Focusrite Scarlett, Universal Audio Apollo).
  • Monitoring: Closed-back headphones are essential to prevent audio bleed into your microphone. Audio-Technica ATH-M50x or Sony MDR-7506 are industry standards.

Budget-Friendly Options

If you are starting out, consider the Audio-Technica AT2020 (condenser) paired with a Focusrite Scarlett Solo. This combination delivers professional-grade sound for under $300. For dynamic microphones, the Shure SM58 is a durable alternative to the SM7B, though it may require more gain. Always use a shock mount and pop filter to reduce handling noise and plosives.

Software Setup: Configuring Your DAW

Set your DAW (Digital Audio Workstation) project to the correct specifications from the start. For example, in Reaper or Audacity, create a project template with a sample rate of 44.1 kHz and a bit depth of 24-bit. Name your tracks clearly and set your input levels to peak no higher than -6 dB to -3 dB during recording. This gives you plenty of headroom for editing and mastering.

Template Best Practices

Save a project template that includes:

  • A mono audio track with your microphone input.
  • A limiter on the master bus with ceiling at -3 dB (for monitoring only, not recording).
  • A loudness meter (RMS) on the master.
  • A small amount of reverb on a send if you want to create a sense of space for your own monitoring (but do not print it to the final file).

This template ensures you never accidentally record at the wrong sample rate or forget to check levels.

File Naming, Structure, and Metadata

Proper file management is critical for ACX processing and for the final retail product. Disorganized files cause delays.

Chapter-by-Chapter Naming Conventions

Each chapter must be delivered as a separate file. ACX recommends a clear, consistent naming structure. Avoid spaces, special characters, or inconsistent capitalization.

Recommended Format:

  • BookTitle_ChapterXX_AuthorName_NarratorName.wav
  • Example: TheGreatAdventure_Chapter01_JaneDoe_JohnSmith.wav

Ensure chapters are numbered sequentially (Chapter01, Chapter02, etc.) to maintain order. Do not use "Part 1" or "Section A" unless specified by the rights holder.

Folder Structure for Large Projects

For audiobooks with 50+ chapters, organize your files into folders: Raw_Recordings, Edited_Chapters, Final_Exports, and Sample. This prevents confusion and makes it easy to locate specific files when re-editing. Use a batch renaming tool like Advanced Renamer (free) to apply naming conventions uniformly.

The Retail Sample

Your submission must include a retail sample that matches the audio specifications of the final book. This sample is the first 10–15 minutes of the audiobook. It must be a single, continuous file that showcases your narration, the audio quality, and the overall production value. The sample is often used by Audible for marketing, so make it perfect. It must pass the same ACX noise and loudness checks as the full book.

Sample Content Tips

Choose a section that demonstrates your range: a dialogue with multiple characters, a descriptive passage, and a moment of high emotion. Avoid starting with a slow chapter or one with excessive technical terminology. If the book has a prologue, consider including it if it’s compelling. The sample is your audition to the listener, so edit it even more rigorously than the rest.

Embedded Metadata

While ACX handles most metadata in their portal, embedding basic information in your WAV files using the BWF (Broadcast Wave Format) chunk is a professional best practice. Include the track title (e.g., "Chapter 01"), the artist (narrator name), and the album title (book title). This helps in organization if files are misplaced or handled by a third party.

How to Embed Metadata

In Reaper, use the Region Render Matrix to add metadata. In Audacity, use the "Edit Metadata" dialog before export. For batch processing, tools like MetaFlac (for FLAC) or BWF MetaEdit (free from the Audio Engineering Society) allow you to tag multiple files at once. Always double-check that the title matches the ACX project exactly—capitalization and punctuation included.

Mastering Your Audio for ACX Acceptance

Mastering for ACX is about consistency and compliance. Your goal is to make every file sound equal in volume, tone, and clarity while strictly adhering to the loudness specs.

The Three-Pass Editing Workflow

Professional narrators use a structured workflow to ensure no errors slip through:

  1. Pass 1 (Cleanup): Remove retakes, long pauses, and obvious mouth clicks. Apply noise reduction if necessary. Do not over-process; a natural room tone is expected.
  2. Pass 2 (Leveling): Use compression or volume automation to smooth out dynamic inconsistencies in your voice. A light 2:1 or 3:1 ratio is usually sufficient. The goal is to bring the quietest and loudest parts of the narration closer together so the average RMS lands within -18 dB to -23 dB.
  3. Pass 3 (Limiting & Export): Apply a limiter with the ceiling set firmly to -3 dB. This ensures no peaks exceed the maximum allowed level. Export as WAV (44.1 kHz, 16-bit or 24-bit, Mono).

Compression Settings for Voice

Start with a threshold around -20 dB, ratio 2:1, attack 10 ms, release 50 ms. Adjust based on your voice's dynamic range. Avoid heavy compression; the human ear expects some variation. Use volume automation (rubber-band editing) for particularly quiet or loud sentences instead of pushing compression too far.

Using the ACX Check Plugin

Before uploading, analyze your files using the free ACX Check Plugin for Audacity or Reaper. This tool instantly analyzes your file and reports whether it passes the Peak, RMS, and Noise Floor requirements. If the plugin gives a green light, you can submit with confidence. Do not rely on your ears alone; this level of verification is essential.

Interpreting the Plugin Results

The plugin shows three metrics: Peak (must be ≤ -3 dB), RMS (must be between -18 and -23 dB), and Noise Floor (must be ≤ -60 dB). If any fails, go back to the mastering stage. For noise floor failure, check for room tone or electrical hum. For RMS failure, adjust compression or volume levels. The plugin also flags potential issues like DC offset (which should be close to 0).

The Submission Process and Common Pitfalls

Once your audio files are prepared, uploading them correctly is the final hurdle. Many submissions are rejected due to minor but easily avoidable errors.

Cover Art Specifications

Your audiobook cover art is a separate but mandatory deliverable.

  • Dimensions: Minimum 2400 x 2400 pixels.
  • Format: JPEG or PNG.
  • Color Space: RGB (not CMYK).
  • Resolution: 300 DPI.
  • The cover must be the final version, with no text errors, and must match the book's title and author name exactly. A cover rejection is a common delay.

Cover Design Tips

Audible displays covers at various sizes, so ensure text is legible when scaled down. Avoid thin fonts, small details, or dark images with black text. Use a service like Canva or hire a professional designer. Remember that the cover is your book's first impression; a high-quality cover reduces listener bounce rate.

Top 10 Reasons for ACX Rejection

Avoid these frequent mistakes:

  1. Clipping or Distortion: Peaks exceeding -3 dB or hitting 0 dB.
  2. High Noise Floor: Background noise (hiss, hum, room tone) above -60 dB.
  3. Incorrect Loudness: Average RMS outside the -18 dB to -23 dB range.
  4. Mouth Noises: Unedited clicks, smacks, and lip pops.
  5. Plosives: "P" and "B" sounds causing low-frequency bursts on the microphone.
  6. Sibilance: Harsh "S", "Sh", and "T" sounds.
  7. Inconsistent Pacing or Performance: Obvious edits that disrupt the natural flow of the narrative.
  8. Incorrect File Format: Submitting MP3, AAC, or files with wrong sample rate (e.g., 48 kHz).
  9. Poor File Naming: Using spaces, special characters (&, #, %), or inconsistent naming schemes.
  10. Metadata Mismatch: The title, author, or narrator in the metadata does not match the project page exactly.

How to Fix Common Audio Issues

For sibilance, use a de-esser plugin (set to around 5-7 kHz). For plosives, employ a pop filter and adjust microphone placement (off-axis). For mouth noises, use a spectral editing tool like iZotope RX De-click or manually zoom in and delete clicks. For inconsistent pacing, practice reading with a metronome or use a time-compression/expansion tool sparingly.

Advanced Tips for Professional Narrators

Once you have mastered the basics, these workflow enhancements will save you hours of editing time.

Punch and Roll Recording

Instead of recording an entire chapter and then editing out mistakes, use "Punch and Roll." This technique involves recording in short segments. When you make a mistake, you stop, back up a sentence or two, and the DAW plays the last few seconds of audio before seamlessly dropping into record mode. This eliminates the need for most post-recording editing and maintains the natural energy of the performance. Most DAWs like Reaper and Pro Tools have dedicated Punch and Roll features.

Setting Up Punch and Roll

In Reaper, enable "Record mode: time selection auto-punch" and set a pre-roll of 2-3 seconds. In Pro Tools, use QuickPunch. Practice the workflow until it becomes second nature. The key is to keep the recording momentum—don't stop to edit mid-session. Punch and roll reduces editing time by up to 70% for experienced users.

Macros and Batch Processing

Automate repetitive tasks to ensure consistency. Create macros in your DAW that apply your specific noise reduction, compression, and limiting settings with a single keystroke. Use batch processing to export all your chapter files at once, ensuring they all have the correct sample rate, bit depth, and file format. This prevents human error when exporting 50+ files manually.

Example Macro in Reaper

Create a custom action that: (1) applies noise reduction (using a saved preset), (2) runs a compressor with your standard settings, (3) applies a limiter at -3 dB, and (4) renders to WAV with your preferred naming format. Bind this to a keyboard shortcut (e.g., Ctrl+Shift+E). For batch renaming, use Reaper's Render Matrix to export all regions as individual WAVs with automatic naming.

Maintaining Consistency Across Sessions

Audiobooks often take weeks to record. To maintain a consistent sound across multiple sessions, save your microphone placement, gain staging, and DAW project template. Record a short calibration file at the start of every session and compare its wave form to previous sessions. This ensures the audio from Chapter 1 matches Chapter 30 perfectly, avoiding embarrassing jumps in tone or volume for the listener.

Session Documentation

Keep a log of: microphone distance from mouth, preamp gain setting, room temperature (humidity affects microphone performance), and any gear changes. Use the same headphones and listening volume for editing decisions. If you use different microphones for different characters (e.g., a darker mic for villains), note that in your documentation and apply EQ matching if necessary.

Final Verification and Submission

Before hitting submit, perform a final checklist:

  • Have all files been run through the ACX Check Plugin? Check each file individually, not just one representative.
  • Is the retail sample meeting the same standards as the full book? The sample is often the first thing Audible QA checks.
  • Are your files named correctly with no spaces or special characters? Double-check that chapter numbers are sequential.
  • Is the cover art uploaded at the correct size and format? Verify dimensions with an image editor.
  • Does the metadata on the ACX portal match the book details exactly? Title, author, narrator, and ISBN (if applicable) must match the rights holder's listing.

Submitting Through the ACX Portal

Log in to your ACX account, navigate to your project, and upload each file to its corresponding chapter slot. The system will automatically run a quick validation. If you see a red "Fail" icon, review the error message and fix the issue. Common portal errors include "File too large" (WAV files over 2 GB—split the chapter if necessary), or "Sample rate mismatch" (if the file header contains incorrect metadata).

Mastering ACX's technical requirements is a mark of a true professional. By adhering to these file format, loudness, and delivery standards, you demonstrate respect for the platform and the rights holders. A smooth approval process allows your work to reach listeners faster and builds a reputation for reliability in the competitive audiobook industry. For additional reading, consult the official ACX Submission Requirements and the Audible Quality Standards for further details.