Creating high-quality audiobooks that meet ACX standards is not just a technical requirement—it directly affects your ability to reach a wider audience, earn positive reviews, and build a professional reputation as a narrator or producer. ACX (Audiobook Creation Exchange) is the primary marketplace connecting authors with narrators and rightsholders, and its strict quality guidelines exist to ensure that every audiobook on Audible, iTunes, and Amazon delivers a consistently excellent listening experience. Proper editing and proofing are the pillars of that quality; they transform raw voice recordings into polished, immersive audio that listeners will enjoy from start to finish. This article provides an in-depth guide to best practices for editing and proofreading your audiobook files to meet ACX approval criteria, with actionable steps that go beyond the basics to help you produce work that stands out in a competitive marketplace.

Understanding ACX Standards

Before you even begin editing, you must thoroughly understand the technical specifications that ACX requires. Meeting these standards is the baseline for acceptance; failing any single one can result in a rejected file and wasted effort. The key parameters are:

  • Sample Rate: 44,100 Hz (44.1 kHz). This is the standard CD-quality sample rate and is non-negotiable.
  • Bit Depth: 16-bit. This provides sufficient dynamic range for spoken word while keeping file sizes manageable.
  • Bitrate: Constant bitrate (CBR) at 192 kbps or higher for mono files. ACX recommends 192 kbps for mono, and 256 kbps for stereo, but mono is standard for audiobooks.
  • Channels: Mono is required (or stereo if the original recording is stereo, but mono is preferred).
  • Noise Floor: Must be at or below -60 dBFS (decibels relative to full scale). This means background noise, hum, hiss, and room echo must be virtually silent during pauses.
  • Peak Levels: The loudest peaks in your audio should fall between -3 dB and -20 dB, with no overs above 0 dB. Typically, aim for peaks around -3 dB to -6 dB for headroom.
  • Loudness (RMS): Average loudness should be between -18 dB and -23 dB, with a target of -20 dB RMS. This ensures consistent listening volume across audiobooks.
  • File Format: MP3 encoded with constant bitrate (CBR) at 192 kbps (mono) or 256 kbps (stereo). File naming conventions: Title_AudioFile.mp3 (e.g., YourBook_01.mp3).
  • Chapter Markers: Each file must contain embedded chapter markers (timestamps) that align with the book’s chapter breaks. ACX provides a chapter marker template.
  • Audio Length: No single file should exceed 120 minutes. For longer books, split into parts.

ACX’s official audio submission requirements page provides the definitive list. Bookmark it and refer to it often. Many rejections occur from simple oversights: a wrong sample rate, peak levels too low, or a noise floor slightly above -60 dB. Measurement tools within your DAW (Digital Audio Workstation) can help you monitor these specs in real time.

Editing Best Practices

Editing is where the raw take becomes a clean, polished recording. The process should remove distractions—clicks, pops, breaths that are too loud, long pauses, and mistakes—while preserving the natural flow of your narration. Use professional audio editing software such as Pro Tools, Logic Pro, Reaper, Audacity, or DaVinci Resolve’s Fairlight to perform the following tasks.

Noise Reduction

Background noise is the most common cause of ACX rejection. Even a quiet room has ambient noise: computer fans, HVAC hum, traffic, or subtle electrical buzz. Your goal is a noise floor of -60 dB or below. However, aggressive noise reduction can degrade audio quality, introducing artifacts (metallic sounds, warbling, or “swimming” noise). Best practices:

  • Record a clean room tone (10–30 seconds of silence in your recording space) and use it as a noise profile for spectral subtraction tools like iZotope RX’s Noise Reduction or Audacity’s Noise Reduction effect. Apply a moderate amount (up to -20 dB reduction) and audition the result.
  • Use a noise gate to silence gaps between words or during long pauses, but set the threshold carefully so that quiet consonants (like “f” or “s”) don’t trigger the gate. A gate alone is not enough for an ACX-silent floor; combine it with noise reduction.
  • Consider spectral repair for persistent noises (e.g., clicks, pops, electrical crackles) that aren’t broadband. Tools like iZotope RX’s Spectral Repair or De-click can remove isolated noise without affecting overall sound quality.
  • Check your noise floor. Use a metering plugin (like Span or the DAW’s built-in analyzer) to ensure that in silent sections, the level is -60 dB or below. If it’s above, re-evaluate your recording environment and re-record if necessary.

Volume Leveling

Consistent loudness ensures listeners don’t constantly adjust volume. You should even out variations in your voice due to emphasis, proximity to the mic, or emotional intensity. Steps:

  • Use compression (with a ratio of 2:1 to 4:1, moderate threshold, slow attack, and fast release) to reduce the dynamic range. Spoken word compression preserves natural dynamics while smoothing out peaks.
  • Apply normalization to bring the overall level up to a target RMS of -20 dB. Most DAWs offer a “normalize to RMS” function. Some producers prefer to normalize peak to -3 dB first, then adjust RMS gain.
  • Check RMS consistency. Use a loudness meter (like Youlean Loudness Meter or the DAW’s Loudness module) to verify that your entire file averages between -18 and -23 dB. If sections are too quiet, manually boost them; if too loud, reduce the gain.
  • Use clip gain (not automated fades) to fine-tune level changes between sections or scenes. This is especially helpful for books with multiple characters or narrative shifts.
  • Aim for an integrated loudness of -20 LUFS (Loudness Units relative to Full Scale) as per ACX guidelines. This is effectively your RMS target.

Removing Mistakes

False starts, stumbles, repeated words, and long breaths are common and must be removed. Use the following techniques:

  • Punch-and-roll recording (re-recording over the mistake in real time) eliminates the need for extensive cleanup. If you can’t re-record, use the “cut” tool to delete the error and crossfade the surrounding audio to avoid a click.
  • Remove or reduce mouth clicks and tongue pops using De-click tools (iZotope RX, Waves De-click, or the free Audacity Click Removal). Apply sparingly to avoid degrading sibilants.
  • Edit out heavy breaths that are distracting—but keep light, natural breaths for realism. A good rule is to remove breaths longer than 200 ms or those that sound like “gasping.”
  • Check for microphone handling noise (rumbles, bumps) and edit them out. Use a high-pass filter (80 Hz roll-off) to remove low-frequency rumble from handling or room resonance.

Silence Trimming

Silence must be added intentionally for pacing; any silence that isn’t part of the narrative should be removed. ACX requires:

  • No silence at the beginning of each file. Start the speech immediately or within 0.5 second of the file start. A short, natural silence of up to 0.2 seconds is acceptable for a pause before the first word.
  • No silence at the end of each file. The audio should fade seamlessly into the next file (or into silence if it’s the last file). ACX recommends ending with a short tail of ambient noise (no more than 0.5–1 second of room tone) to end naturally.
  • Internal pauses between sentences or paragraphs should be consistent and not too long. Aim for about 0.3–0.5 seconds for a comma, and 0.5–1 second for a period. For major section breaks, up to 2 seconds is acceptable but must be justified by the text.
  • Trim all leading and trailing silence at the beginning and end of each file. Use the “Truncate Silence” or “Remove Silence” feature in your DAW, but manually verify that you haven’t cut into speech.

iZotope’s guide to ACX audio editing offers deeper insights into using spectral tools for noise reduction and silence management.

Proofing Your Audiobook

Proofing (also called quality control, or QC) is the final validation before submission. It is a separate, rigorous process from editing and should be done after all edits are complete. The goal is to catch flaws that affect the listening experience: audio glitches, inconsistent volume, mispronunciation, pacing issues, and alignment with the written text.

Listening on Different Devices

Your studio monitors or headphones may sound great, but your audience will listen on car stereos, earbuds, laptop speakers, and Bluetooth headsets. Listen to your finished files on at least three different playback systems:

  • High-quality headphones (studio monitors) to detect subtle noise and editing artifacts.
  • Smartphone speakers or laptop speakers to check how the audio translates to small, limited-frequency speakers.
  • Car stereo or Bluetooth speaker (if available) to evaluate how the narration sounds in a noisier environment.
  • Use a playback app that displays metadata (like Audacity or VLC) to confirm that chapter markers are correctly placed and that file properties (bitrate, sample rate) match ACX specs.

Checking for Consistency

Listeners will notice if your voice changes tone, speed, or proximity to the mic between chapters. Watch for:

  • Voice consistency: Your tone and energy should stay uniform for the same character or narrative passages. Re-recordings made on different days must match the original session’s mic placement, EQ, and ambience.
  • Pronunciation: Check any uncommon words, foreign names, or places. Use a pronunciation guide or dictionary. If you correct a mistake, ensure the replacement sounds natural and doesn’t stick out.
  • Pacing: Maintain a steady reading speed. Too fast becomes hard to follow; too slow becomes boring. The ideal pace for audiobooks is roughly 150–160 words per minute. Use a script timer to check.
  • Emotional tone: Match the mood of the scene—do not sound overly cheerful during a sad passage, or monotone during an action sequence.

Review for Errors

Even the best editor misses something. Systematic proofing techniques include:

  • Listen while following the script – either on paper or with a synchronized text-to-speech tool (like the ACX Master Audio Player). Mark any discrepancy in a QC log.
  • Isolate potential problem areas: Listen specifically for plosives (breathy “p” and “b” sounds), sibilance (harsh “s” and “sh”), mouth clicks, and lip smacks. These can be fixed with EQ (gentle high-pass for plosives) or de-essing.
  • Check for unintended room noise – a door slam, car horn, or cough may creep in. Use a spectrogram view in your DAW to spot noise bursts visually.
  • Verify chapter markers – each marker should correspond exactly to the chapter title text in the audio. ACX’s QC tool checks this automatically, but you should also listen to ensure the transition is smooth.

Compare with Script

Accuracy to the text is critical. Missing words, added words, or changed phrasing can alter meaning or introduce legal issues (if you misquote a copyrighted work). Use the following approach:

  • Read the script while listening to the audio at a moderate volume. Highlight any place where the audio deviates—then decide if it’s a harmless paraphrase or an error that requires a retake.
  • Use a text-to-speech sync tool (like Descript or applications that align text with audio) to automate much of this process, but always manually spot-check.
  • Re-record corrections if the error is more than a minor stumble. Do not try to patch over a blatant misread with a crossfade—it will sound unnatural.

For a thorough proofing checklist, refer to Voices.com’s audiobook proofing checklist, which includes many of these points.

Additional Tips for Success

Beyond the editing and proofing workflow, several overarching practices will help you consistently produce ACX-compliant audiobooks that delight listeners and pass inspection on the first try.

Use Quality Equipment

Your tools directly influence the time you spend cleaning up. A good microphone and proper acoustic treatment reduce noise and reverb, making editing easier. Essentials:

  • Microphone: A large-diaphragm condenser mic (like the Audio-Technica AT2020 or Rode NT1-A) is popular for spoken word, but some prefer dynamic mics (Shure SM7B) if the room is untreated. Use a pop filter and shock mount.
  • Acoustic treatment: Sound-absorbing panels, blankets, or even a homemade vocal booth (PVC frame with blankets) minimize room reflections. Avoid recording in a completely bare room; the echo will be difficult to remove.
  • Audio interface: A USB interface (Focusrite Scarlett, Universal Audio Volt) with a clean preamp and at least 48V phantom power for condensers.
  • Headphones: Use closed-back headphones to prevent microphone bleed during recording. Open-back headphones are acceptable for mixing but not for recording.

Maintain a Consistent Pace

Listeners expect a steady, engaging delivery. Practice your pacing by rehearsing passages and timing yourself. If you tend to speed up when excited, use a metronome or a click track at 150 BPM as a guide during recording (just don’t let the click bleed into the mic). Record at a comfortable speed—slower than normal conversation—to allow for natural pauses.

Follow ACX Formatting Guidelines

Before submitting, double-check these details:

  • File naming: Use the format Title_PartXX.mp3 (e.g., TheGreatGatsby_01.mp3). No spaces, special characters, or underscores in place of spaces.
  • Chapter markers: Export MP3 files with embedded chapter markers. In Pro Tools, create a marker track; in Audacity, use the label track then export with “Import metadata from label.”
  • ID3 tags: Fill in the title (Part X), author, narrator, album, genre, and track number via your software’s metadata editor.
  • Submission process: Upload your files to the ACX portal; if you are a narrator, the rights holder will also need to approve. ACX runs an automated check; any spec violation will show as a warning or rejection.

ACX’s guide to creating a great audiobook covers the submission process in detail.

Keep Backup Copies

Never work on a single file. Maintain these versions:

  • Raw unedited takes in case you need to redo a section from scratch.
  • Edited files (before normalization/compression) to avoid losing the original dynamics.
  • Final mastered versions ready for submission. Use a naming convention like Title_PartXX_v2.mp3 to track revisions.
  • Cloud backup (Google Drive, Dropbox, or external hard drive) to prevent data loss. A lost hard drive can set your project back weeks.

Conclusion

Producing a professional audiobook that meets ACX standards is a multifaceted process, but with careful attention to editing and proofing, you can deliver a listening experience that keeps audiences engaged and earns your work high ratings. Start by thoroughly understanding the technical requirements, then move methodically through noise reduction, volume leveling, mistake removal, and silence trimming. Proofing—listening on multiple devices, checking consistency, reviewing against the script—is the safety net that catches the remaining flaws. Invest in good equipment, establish a consistent workflow, and always keep backups. Dedication to these best practices will not only help you pass ACX approval on the first attempt but also build your reputation as a reliable, high-quality audiobook producer in a competitive marketplace.