For independent authors and voice actors, ACX represents the primary gateway to distributing an audiobook across Audible, Amazon, and iTunes. The platform's technical requirements are designed to ensure a consistent, high-quality listening experience, but failure to comply is the leading cause of submission rejection. A compelling narration cannot compensate for an audio file that fails automated checks for loudness, noise floor, or peak level.

Building a repeatable, engineering-focused post-production workflow targeted specifically at ACX standards transforms guesswork into a precise, predictable process. This guide provides a comprehensive breakdown of those standards, a step-by-step mastering chain, and the quality control procedures required to achieve first-pass approval.

Decoding the ACX Technical Mandate

ACX relies on automated quality control (QC) systems to scan every submission. Understanding these parameters is the first step to building a compliant workflow. The system measures file format, integrated loudness, true peak level, and noise floor against specific thresholds.

File Format and Core Specifications

  • Codec: MP3 (MPEG-1 Audio Layer 3)
  • Bitrate: 192 kbps Constant Bitrate (CBR). Variable Bitrate (VBR) is not accepted.
  • Sample Rate: 44.1 kHz. Files recorded at 48 kHz must be converted to 44.1 kHz during export.
  • Channels: Mono or Stereo. Mono is standard for single-narrator audiobooks, ensuring consistent phase alignment across playback systems. Stereo is acceptable but provides no fidelity benefit for spoken word.
  • Bit Depth: 16-bit is the standard for WAV masters before MP3 encoding.

Loudness: The LUFS Standard

ACX measures loudness using the LUFS (Loudness Units Relative to Full Scale) standard, specifically ITU-R BS.1770. Unlike peak normalization, which only looks at the highest amplitude sample, LUFS measures the perceived average energy over time.

  • Integrated Loudness: Must fall between -23 dB and -18 dB LUFS.
  • Short-Term Loudness: Should average around -18 dB LUFS.
  • True Peak: Must not exceed -3 dB. True peak accounts for inter-sample peaks that occur during digital-to-analog conversion, which standard sample peak meters can miss.

This loudness window is intentionally restrained compared to music. It is optimized for extended listening sessions, reducing listener fatigue and maintaining dynamic expression. A file mastered too loud will likely be rejected.

Noise Floor Requirements

The noise floor of your audio, measured during silent passages, must be -60 dB or lower. This includes background room tone, electronic hum, and breaths. The noise floor must also remain consistent throughout the book. Fluctuating background noise, such as a fan cycling on and off, is a frequent cause of rejection.

The ACX Post-Production Workflow: A Six-Step Process

The sequence in which you apply processing tools directly impacts the final audio quality. The following chain represents a professional mastering workflow optimized for audiobooks.

1. Raw Audio Editing and Cleanup

Post-production begins with editing. All mistakes, long pauses, and excessive mouth clicks must be removed. Use a spectral frequency display to visually locate plosives and clicks, repairing them with a spectral repair tool rather than simply cutting silence. At this stage, you are building a clean foundation for the processors that follow. Managing harsh breaths is also critical. Reduce the gain on breaths that are too loud rather than removing them entirely, as some ambient context preserves the natural rhythm of the narration.

2. Spectral Noise Reduction

Capture a noise print from a silent section of audio. Apply noise reduction with a gentle hand. Over-processing creates an unnatural, hollow quality and can remove the high-frequency air from your voice. The target is a noise floor below -60 dB. If you can hear the noise reduction pumping or pumping artifacts, your settings are too aggressive. Tools like iZotope RX provide advanced spectral editing capabilities for targeted noise removal without affecting the dialogue.

3. Equalization (EQ) for Tone Shaping

EQ shapes the tonal balance of your voice to improve clarity and warmth. Apply these adjustments before compression to ensure the compressor reacts to the balanced signal.

  • High-Pass Filter: Roll off everything below 80-100 Hz to eliminate subsonic rumble and handling noise.
  • Low-Mid Cut: Reduce frequencies around 200-400 Hz to remove "muddy" or "boxy" resonances that can accumulate in untreated rooms.
  • Presence Boost: A gentle 2-4 dB shelf or broad bell around 3-5 kHz adds intelligibility and helps the narrator cut through background noise.
  • Air Band: A subtle shelf above 10 kHz adds a sense of openness and polish.
  • De-Essing: Tame sibilant frequencies (5-10 kHz) with a dedicated de-esser before compression. This prevents high-frequency sibilance from over-activating the compressor.

4. Dynamic Range Compression

Compression narrows the dynamic range between the quietest and loudest passages, ensuring consistent volume. Set your compressor with a ratio of 3:1 to 4:1. Adjust the threshold so the compressor engages on louder phrases, providing 3-6 dB of gain reduction. Use a medium attack (10-20 ms) to preserve transient clarity for hard consonants, and a medium release (50-100 ms) to avoid audible pumping. Follow the compressor with makeup gain to bring the average level up to a usable range.

5. Loudness Normalization and Limiting

This is the most common point of ACX failure. Do not simply normalize to -3 dB peak. You must target the LUFS window.

  1. Measure: Play a representative loud section of your audio and read the Short-Term LUFS value on your meter.
  2. Adjust: If your audio is too quiet (below -23 LUFS), apply gain. If it is too loud (above -18 LUFS), reduce gain. Tools like Auphonic can automate this process to target specific loudness levels.
  3. Limit: Place a brickwall limiter at the very end of your chain with the ceiling set to -3.5 dB. This provides a safety buffer against inter-sample peaks and guarantees you stay below the -3 dB true peak limit.

Pro Tip: Target an integrated loudness of -19 to -20 LUFS. This places you safely within the -23 to -18 window, accounting for variances in different reading systems.

6. Exporting and MP3 Encoding

Export your master to a 16-bit 44.1 kHz WAV file. To create the final MP3, use a dedicated encoder (such as LAME). Set the format to Constant Bitrate (CBR) at 192 kbps. Verify the sample rate remains 44.1 kHz. Always check the final MP3 file, not the WAV master, as the encoding process can slightly alter levels.

Common ACX Rejection Codes and Fixes

Understanding why ACX QC fails helps you target your revisions effectively.

  • Clipping (Peak exceeds -3 dB): Lower your brickwall limiter ceiling to -3.5 dB or reduce makeup gain on your compressor.
  • Noise Floor too high: Revisit your noise reduction settings. Ensure your noise print is captured from a silent section and that the reduction is sufficient to bring the floor below -60 dB.
  • Inconsistent Loudness: This usually occurs when chapters are processed individually without reference. Use batch processing to apply identical gain and compression settings across all files.
  • Format Error: Verify your export settings. Common errors include VBR encoding, 48 kHz sample rate, or stereo files being submitted when mono is expected.
  • Excessive Mouth Clicks: Manual spectral repair is often required for loud clicks. Automated click removers can miss transient events. Listen at low volume to spot checks.

Quality Control and Verification

Before uploading, run every chapter through an ACX checker tool. For Audacity users, the ACX Check plugin provides a comprehensive pass/fail report for loudness, peak, and noise floor. For other DAWs, export a short segment and use the official ACX Check page.

Listen to the entire book at a low volume to ensure consistent intonation and level. Check playback on multiple systems: studio headphones, car speakers, and earbuds. If the audio sounds good on reference monitors but distorted on consumer earbuds, you may need to revisit your limiting or compression settings.

Advanced Workflow Efficiency

For series production or long books, manual processing is not scalable. Build a DAW template (Reaper, Audition, Pro Tools) with your processing chain pre-configured. Use batch processing to apply identical settings to all chapters. Automatic loudness levelers can process an entire book in a single pass, guaranteeing consistency across every file.

Accurate metadata tagging is the final step. Use tools like Chapter and Verse to embed Audible-compatible chapter markers, book title, author, and runtime. A technically perfect file with incorrect metadata will still cause errors in distribution.

Conclusion

Mastering ACX submission standards transforms post-production from a source of anxiety into a repeatable, predictable process. When your audio passes quality control on the first attempt, you maintain creative momentum and build a professional reputation for reliability. Use the workflow outlined here as your technical foundation. Adapt it to your specific voice and recording environment, and consistently deliver audiobooks that meet the highest industry standards.