ACX Technical Specifications: The Foundation for Audiobook Approval

Producing an audiobook for ACX (Audiobook Creation Exchange) demands more than a great voice and compelling narration. Every submission must pass a strict technical review before it reaches audible.com, Amazon, or iTunes. Understanding and applying ACX’s recording specifications is the difference between a fast approval and a frustrating rejection. This article dissects each requirement, explains how to configure your recording software, and offers production techniques that ensure your files meet the standard.

ACX’s technical requirements are designed to guarantee consistent playback quality across all devices, from high-end headphones to car stereos. The specifications cover audio format, sample rate, bit depth, noise floor, and loudness levels. Ignoring any one of these can cause your audiobook to be rejected, wasting hours of editing time. By mastering these settings you produce a professional, distribution-ready product.

ACX Technical Specifications Deep Dive

ACX publishes a clear set of requirements. Before adjusting any dial, internalize these numbers. Each parameter affects the final sound quality and compatibility with Audible’s playback systems.

File Format and Bit Rate

ACX accepts only MP3 files encoded with a constant bit rate (CBR) of 192 kbps or higher. Variable bit rate (VBR) is not allowed because it can cause inconsistent playback on some devices. When exporting, always select MP3, CBR, and a minimum of 192 kbps. Higher bit rates (e.g., 256 kbps or 320 kbps) are acceptable and may preserve more detail, but 192 kbps is the proven sweet spot for file size and quality.

Sample Rate

The required sample rate is 44,100 Hz (44.1 kHz). This is the standard for CD-quality audio and matches Audible’s encoding pipeline. Some producers work at 48 kHz for video projects, but for ACX you must resample to 44.1 kHz during export. Failing to do so triggers immediate rejection.

Bit Depth

Record and edit at 16-bit. While 24-bit recording gives you more headroom during production, ACX mandates a final 16-bit output. If you record in 24-bit, you must dither down to 16-bit during export. Avoid simply truncating bits, as that introduces noise and distortion. Proper dithering adds a tiny amount of shaped noise that preserves the signal’s integrity.

Noise Floor

ACX requires the noise floor to be below -60 dB (decibels). This means the background noise (room tone, electronic hum, breath sounds) must be at least 60 dB quieter than the peak level of your voice. In practice, aim for a noise floor of -65 dB to -70 dB to build in margin. A higher noise floor produces audible hiss, especially during quiet passages, and will fail validation.

RMS Level and Peak Level

ACX’s loudness guidelines are often misunderstood. The average RMS (root mean square) level should be between -23 dB and -18 dB. Your peak level must not exceed -3 dB (some sources say -3.5 dB to be safe). This ensures the audio is loud enough for comfortable listening but leaves headroom to prevent distortion. No clipping is allowed anywhere in the file.

For comparison, a typical ACX-approved narration RMS is around -20 dB to -19 dB with peaks hitting -3 dBFS. Use a loudness meter or an ACX validation plugin to check these levels.

Setting Up Your Recording Environment for ACX

Great recording settings start before you hit record. Your physical space and equipment directly influence whether you can hit the noise floor and dynamic range targets.

Acoustic Treatment

Record in a quiet room with minimal reverberation. Hard surfaces (bare walls, floors, windows) create slap echo and comb filtering, which add artifacts that are hard to remove later. Use acoustic panels, corner bass traps, or even heavy moving blankets to absorb reflections. A portable vocal isolation shield placed behind your microphone can further reduce room ambience. The goal is a dry, dead sound that leaves you full control during post-production.

Microphone Selection and Placement

A large-diaphragm condenser microphone with a cardioid pickup pattern is the standard for audiobook narration. It captures the voice with detail while rejecting sound from the sides and rear. Avoid dynamic microphones (they often require a preamp with high gain, raising the noise floor). Place the microphone 6–12 inches from your mouth, slightly off-axis to reduce plosives. Use a pop filter to block bursts of air that cause low-frequency thumps.

Monitoring and Cables

Use closed-back headphones to prevent audio from leaking back into the microphone. Keep cables away from power cords to avoid electromagnetic interference. A balanced XLR cable is preferred over unbalanced 1/4-inch cables for noise rejection. Ensure your audio interface has clean preamps with an equivalent input noise (EIN) rating below -129 dBu.

Configuring Your Digital Audio Workstation (DAW)

Each DAW has its own menus, but the core settings are universal. Below are specific steps for the three most common applications used by ACX narrators.

Audacity (Free and Open-Source)

Audacity is popular for its price and simplicity. To set up for ACX:

  1. Set sample rate: Go to Edit > Preferences > Quality. Under Default Sample Rate, enter 44100 Hz. Under Default Sample Format, select 24-bit if you want extra headroom during recording, then dither to 16-bit when exporting.
  2. Record at 24-bit: In the same Preferences window, set Recording > Sample Format to 24-bit. This gives you lower quantization noise.
  3. Export settings: When you’re done editing, go to File > Export > Export as MP3. Under Format Options, choose Constant Bitrate (CBR) and set the bit rate to 192 kbps or higher. Before exporting, ensure the project sample rate is 44100 Hz (visible in the bottom-left corner of the Audacity window).
  4. Dithering (optional but recommended): If you recorded at 24-bit, use File > Export > Export Audio and enable Dither > Shape Noise. This reduces truncation distortion.

Audacity does not natively include a real-time loudness meter. Use the Analyze > Contrast tool to check RMS levels, or install a free plugin like Youlean Loudness Meter.

Adobe Audition (Paid)

Audition offers robust metering and a dedicated ACX profile in the Amplitude Statistics window. Follow these steps:

  1. Sample rate: Go to Edit > Preferences > Audio Hardware and set the default sample rate to 44100 Hz. Also check that your audio interface is set to the same rate in its control panel.
  2. Recording: Create a new audio file (Ctrl+N) and choose 44100 Hz, 16-bit or 24-bit stereo. For narration, always choose mono to avoid phase issues unless your script requires stereo effects (rare for ACX).
  3. Export: Go to File > Export > File. Choose MP3 Audio (mp3). Under Format Settings, set Bitrate to 192 kbps or higher and Bitrate Type to Constant. Make sure the sample rate remains 44100 Hz and the bit depth converts to 16-bit (Audition will dither automatically if you check Dither).
  4. Validate levels: Select the entire file, then go to Window > Amplitude Statistics. Click Scan Selection. Audition displays RMS, peak, and noise floor. Compare against ACX standards.

Reaper (Affordable and Powerful)

Reaper is a flexible DAW with a tiny footprint. Configuration:

  1. Project settings: File > Project Settings > Audio Settings. Set Sample Rate to 44100. Use 24-bit recording depth.
  2. Monitoring FX: Add a limiter or compressor to your monitoring chain (not recorded) to hear how your voice will sound after processing.
  3. Render settings: File > Render. Choose Output Format > MP3 (libmp3lame). Set Bitrate to 192 kbps and check Constant bitrate. Under Resample mode, pick R8Brain Pro (512pt) for highest quality.
  4. Dither: In the render dialog, enable Dither and select Noise-shaped (6% order). Set target bit depth to 16-bit.

Recording Techniques for Consistent ACX-Quality Audio

Settings alone don’t guarantee approval. Your technique during recording defines how much work is needed in post-production.

Mic Technique and Plosives

Position the pop filter at least 2 inches from the microphone grille. Speak across the capsule, not directly into it. For plosives like “P” and “B,” drop your head slightly or turn your mouth a few degrees away from the mic axis. This reduces the air puff without changing your vocal tone.

Monitoring Levels in Real Time

Keep your recording level so that the peak hits no higher than -6 dBFS to -3 dBFS. Most narrators aim for an average RMS around -20 dBFS to -18 dBFS during recording, which leaves room for compression or normalization later. If your DAW has a peak meter, watch for red indicators – any clipping is permanent and will fail ACX’s check.

Consistency Across Sessions

Record chapters in one session if possible, or at least in the same acoustic environment with identical mic placement. A slight change in chair height or mouth-to-mic distance alters the tone and RMS level. Take notes on your setup (distance, gain knob position, filter settings) so you can replicate it for retakes or second books.

Post-Processing: Polishing Your Audiobook to ACX Specs

Even with perfect raw audio, post-processing verifies and corrects small errors. The goal is to satisfy ACX’s numerical criteria without introducing audible artifacts.

Noise Reduction

First, capture a sample of your room noise (5–10 seconds of silence). Use a noise gate to kill silence between phrases, but be careful not to chop off the tails of words. For continuous background hiss, use a spectral noise reduction tool (like Audacity’s Noise Reduction or Audition’s Adaptive Noise Reduction). Aim to lower the noise floor to -65 dB or lower. Over-reduction creates a robotic “underwater” sound, so keep the reduction to 12–18 dB maximum.

Normalization and Compression

Do not normalize to 0 dB. Instead, normalize the file so that the peak reaches -3 dBFS (or -3.5 dB). This prevents true peaks from exceeding the ACX limit. Use a compressor with a gentle ratio (2:1 or 3:1) to even out loud and soft passages. Set the threshold so that the loudest parts are only slightly attenuated, then raise the overall gain. The final RMS should fall between -23 dB and -18 dB. A limiter set to -3 dB FS can catch occasional overshoots.

Editing Breath, Mouth Noise, and Pops

Remove loud breaths that are not part of the performance. Use spectral editing to clip out clicks, lip smacks, and tongue noises. For plosives, find the low-frequency burst (around 50–150 Hz) and attenuate it with a high-pass filter set to 80–100 Hz, or manually reduce the volume envelope. Be sparing – too much editing creates an unnatural, sterile narration.

Multiband Compression (Advanced)

If sibilance (harsh “S” sounds) is a problem, use a de-esser or multiband compressor to reduce frequencies around 5–8 kHz. ACX does not forbid de-essing, but over-processing can dull the voice. Apply it in short bursts only where needed.

Validating Your Files: ACX Audio File Checker and Third-Party Tools

Before submitting, run your final MP3 through validation. ACX provides a free ACX Audio File Checker (online or downloadable tool) that scans for all specifications: sample rate, bit rate, bit depth, noise floor, RMS, peaks, and clipping. It produces a pass/fail report with specific error messages.

Using the ACX Checker

Upload your MP3 or open it in the tool. Review each category. If anything is flagged, adjust your export settings or re-process the audio. Common failures include:

  • Noise floor too high: Apply more aggressive noise reduction or record in a quieter space.
  • Peak too high: Lower the output gain and apply a limiter at -3 dBFS.
  • RMS outside range: Use compression or normalize to bring RMS into -23 to -18 dB.
  • Wrong sample rate (e.g., 48 kHz): Resample to 44.1 kHz in your DAW before export.

Plugin-Based Validation

Some DAWs support ACX-specific plugins. For example, Izotope RX includes an ACX-compliant loudness meter and batch processor. Metering plugins like TBProAudio dpMeter can monitor RMS and peak in real time. Using these during production saves time by catching issues early.

Common Pitfalls and How to Avoid Them

Even experienced narrators run into problems. Here are the most frequent ACX rejection reasons and their fixes.

Plosives and Sibilance

Plosives cause low-frequency bursts that push the peak level artificially high, making the file appear clipped. A pop filter and off-axis technique are essential. For sibilance, position the microphone slightly above your mouth (not directly in front) and use a de-esser plugin.

Room Tone Inconsistencies

If you record different chapters on different days, the background noise may vary. This creates audible jumps during playback. Unify room tone by recording a few seconds of silence in your space, then use it to fill gaps or as a noise print for removal. Alternatively, use a noise gate with a fast release to kill silence entirely, but ensure the room itself is quiet enough.

Clipping or Distortion

Clipping occurs when the input level exceeds 0 dBFS in the digital domain. Even one clipped sample is enough to fail ACX validation. Always leave headroom during recording – peaks should never exceed -6 dBFS. Use a limiter only on the master bus of your DAW, not during recording.

Wrong Export Settings

A common mistake is exporting as a stereo MP3 file instead of mono. ACX requires mono files for narration unless the book includes sound effects or music that demands stereo. Check the export dialog: if you see two waveforms, you’re exporting stereo. Convert to mono before rendering.

Final Export Settings for ACX: A Checklist

Before you export your final chapter files, run through this checklist to ensure every setting is correct.

  • File format: MP3
  • Bitrate mode: Constant (CBR)
  • Bitrate: 192 kbps or higher (e.g., 192, 256, or 320 kbps)
  • Sample rate: 44100 Hz (44.1 kHz)
  • Bit depth: 16-bit (dithered from 24-bit if necessary)
  • Channels: Mono
  • Peak level: No higher than -3 dBFS (prefer -3.5 dBFS)
  • RMS level: Between -23 dB and -18 dB
  • Noise floor: Below -60 dB (target -65 dB or lower)
  • No clipping: Zero samples at 0 dBFS

After exporting, run the ACX Audio File Checker on every file. Some narrators batch-process with a custom export preset to maintain consistency across all chapters.

Conclusion: Consistency Is the Key to ACX Success

Meeting ACX’s technical specifications is not a one-time configuration – it requires disciplined workflow at every stage, from microphone placement to final validation. By internalizing the requirements detailed above and using the tools provided by your DAW and ACX, you can submit files that pass on the first attempt.

For additional guidance, consult ACX’s official technical requirements page (ACX Audio Submission Requirements), and explore the Audacity Manual for recording and exporting tutorials. If you use Adobe Audition, the Audio Hardware Preferences guide helps confirm sample rate settings. For acoustic treatment basics, review Sweetwater’s Acoustic Treatment 101.

Each audiobook you produce reinforces good habits. Over time, checking these specs becomes automatic, letting you focus on what matters most: delivering a compelling performance that listeners enjoy.