Meeting ACX (Audiobook Creation Exchange) sound quality standards is non-negotiable for audiobook narrators and producers who want their work accepted by Amazon and Audible. The ACX audio specifications are strict, and failing to comply means rejection or poor listener experience. Fortunately, modern software plugins offer precise, repeatable tools to shape your raw recording into a polished, ACX-compliant master. This guide explains how to use these plugins intelligently, covering every essential step from noise reduction to final loudness metering. It expands on foundational concepts, introduces additional plugin categories, and provides deeper troubleshooting strategies to help you pass ACX quality control on the first attempt.

Understanding ACX’s Sound Quality Requirements in Detail

Before you reach for any plugin, you must internalize the ACX audio QC criteria. These specifications exist so every audiobook on Audible sounds consistent across playback systems — from high‑end headphones to car speakers. ACX expects a mono WAV file (or MP3 192 kbps or higher) with the following technical specs:

  • Sample rate: 44.1 kHz or 48 kHz. Minor differences in sample rate conversion can introduce aliasing, so settle on one rate and stick with it for the entire project.
  • Bit depth: 16-bit or higher (24-bit recommended for headroom). Recording at 24-bit gives you 144 dB of dynamic range, which is essential when your voice has sudden loud passages.
  • Format: Mono (not stereo). Stereo files double the data and add phase cancellation risks; always record and deliver in mono.
  • Maximum peak level: −3 dBFS (no sample may exceed −1 dBFS on a true‑peak meter). This headroom prevents inter‑sample peaks from causing distortion on playback.
  • Average loudness: −18 LUFS to −23 LUFS (integrated loudness), with an ideal target of −18 LUFS. This range matches the loudness of most commercial audiobooks.
  • Noise floor: ≤ −60 dBFS (unweighted). The RMS level of your background room tone during silent sections must be at least 60 dB below full scale.
  • Plosives, sibilance, clicks: must be minimal or absent. Hard p and t pops, excessive s and sh sounds, and mouth clicks are common rejection triggers.
  • No clipping, distortion, or digital artifacts: Even one clipped sample can fail automated checks. Always use a limiter as a safety net.

These standards are not arbitrary. They ensure that audiobooks sound natural at any volume, that background noise doesn’t distract listeners, and that dynamic range remains wide enough for expressive reading without fatigue. Meeting them requires more than a quiet recording space; it demands thoughtful processing with plugins that preserve natural voice quality while eliminating problems.

Why Plugins Are Essential for ACX Compliance

Raw microphone recordings almost never fall within ACX limits. Room tone, breath noise, plosives, and uneven vocal dynamics are unavoidable, even in professionally treated spaces. Software plugins let you surgically correct these issues without degrading the recording. The right plugin chain can:

  • Erase low‑frequency rumble and HVAC noise
  • Compress dynamics to keep the voice consistently present
  • Equalize the tone for warmth and clarity
  • Limit peaks to prevent overshooting −3 dBFS
  • Display real‑time loudness measurements
  • De‑click and de‑mouth everyday mouth sounds that would require hundreds of manual edits

Because ACX automated checking is unforgiving — it scans every sample — a reliable plugin workflow saves you from repeated re‑uploads and ensures your audio passes every quality gate. Moreover, plugins bring repeatability: once you dial in settings for your voice and room, you can save presets and apply them to every chapter with minimal tweaking. This consistency is the hallmark of a professional audiobook producer.

Building Your Plugin Toolbox for Audiobook Production

Not every plugin is suitable for spoken word. Avoid heavy‑handed multiband compressors or creative effects designed for music. Instead, choose transparent, precision‑oriented tools. Below are the core categories you need, along with recommended commercial and free options. Invest in one or two tools per category; mastering a few plugins yields better results than owning a hundred you barely understand.

1. Noise Reduction and De‑Noising

Background noise is the most common reason for ACX rejection. A noise floor above −60 dBFS fails immediately. Use a noise reduction plugin that learns a noise profile from a silent section of your recording and subtracts it. The industry standard is iZotope RX (Spectrum De‑Noise module), which offers intelligent adaptive reduction with real‑time preview and adjustable smoothing. For a simpler, real‑time solution, Waves NS1 suppresses noise automatically without messing with voice tonality — it’s especially good for consistent room hum. Free options include Reaper’s ReaFir (FFT subtraction mode) and Audacity’s built‑in noise reduction. Audacity’s tool works well if you carefully set the sensitivity to avoid artifacts.

How to use it: Select a few seconds of pure room tone (no speaking), capture the noise profile, and apply reduction of 12–18 dB. Listen critically — over‑reduction causes warbling or “underwater” artifacts. If you hear artifacts, reduce the reduction amount or use a gentler algorithm (e.g., RX’s Adaptive mode). For noisy recordings, apply noise reduction in two passes: first reduce broadband noise, then use spectral editing to remove tonal buzz (like a refrigerator).

2. Equalization (EQ)

Voice equalisation in audiobooks is subtle. The goal is to remove resonant peaks and add presence, not to colour the sound. A parametric EQ like FabFilter Pro‑Q 3 offers surgical precision with spectrum analyser feedback. Start by cutting frequencies below 80 Hz with a high‑pass filter to eliminate rumble and mic proximity effect. If the voice sounds muddy, cut around 250–400 Hz by 1–2 dB. For clarity (without harshness), a gentle shelf boost at 3–5 kHz can help, but stay under 2 dB. Always use a linear‑phase mode when possible to avoid phase smearing, which can be noticeable on spoken word. If you’re using a cheaper EQ, a minimum‑phase high‑pass filter is fine.

Key tip: Never boost frequencies that already have sibilance or nasal qualities. Instead, cut their problem neighbours. For example, if the voice sounds honky at 800 Hz, cut that region by 1–3 dB with a wide Q. Use the spectrum analyser to match your voice to a reference commercial audiobook.

3. Compression and Dynamic Control

Compression tightens the dynamic range so that soft syllables are audible and loud outbursts never clip. For ACX, a moderate compression ratio of 2:1 to 3:1 with a slow attack (~10 ms) and medium release (~50–100 ms) works well. Use a leveling amplifier‑style compressor (e.g., Waves R‑Comp, Softube FET, or the free TDR Kotelnikov) for transparent control. Place compression after EQ but before limiting. If your voice has very wide dynamic range (e.g., you whisper and shout in the same chapter), consider using two compressors in series: a low‑ratio compressor for gentle evening, followed by a higher‑ratio limiter for peak control.

Setting the threshold: Aim for 3–6 dB of gain reduction on peaks. Over‑compression leads to a fatiguing, “squashed” sound that ACX’s human reviewers will flag. Use the compressor’s make‑up gain to bring the level back to original — the integrated loudness meter will tell you if you’re too quiet or too loud.

4. De‑essing

Sibilance — harsh s, sh, z, and ch sounds — is a frequent issue in audiobook narration. A de‑esser is a frequency‑dependent compressor that attenuates sibilant frequencies (typically 5–8 kHz). Use a gentle de‑esser like Waves Renaissance DeEsser or the built‑in tools in your DAW. Set the range to reduce 2–4 dB and the threshold so that only the harshest sibilants trigger reduction. Over‑de‑essing creates a lisp that sounds unnatural and can fail ACX review. For problem recordings, use a dynamic EQ (like FabFilter Pro‑Q 3’s dynamic mode) to target only the sibilant bursts.

5. Limiting and Peak Control

A limiter ensures no sample exceeds ACX’s maximum peak of −3 dBFS (or −1 dBFS using true‑peak meters). A transparent brickwall limiter like iZotope Ozone (Limiter) or the free LoudMax works equally well. Set the output ceiling to −3 dBFS and adjust the input gain so the limiter catches only transient overshoots (1–2 dB of gain reduction). Do not push the limiter hard — loudness is achieved through proper gain staging, not limiting. If you see more than 3 dB of gain reduction on the limiter, go back and adjust your compressor settings rather than trying to fix everything with the limiter.

6. Loudness Metering

You cannot rely on your ears alone. ACX specifies integrated loudness in LUFS. Use a dedicated loudness meter such as Youlean Loudness Meter 2 (free, with a paid upgrade for true‑peak and histogram) or iZotope Insight. These tools display the integrated loudness over the entire file, momentary loudness, and true‑peak levels. They also show loudness range (LRA) and momentary maximums. Adjust your compression and gain accordingly to hit −18 LUFS ±2 dB. Export a short segment first to check your metering; don’t rely on real‑time meters alone because they could be affected by playback latency.

Step‑by‑Step Plugin Workflow for ACX Compliance

Follow this order to build your processing chain from start to finish. Deviating can introduce artifacts that are hard to undo. Each step builds on the previous one, so stick to the sequence.

  1. Prepare the raw recording: Trim silence at the beginning and end. Remove mouth clicks, breaths that are too loud, and any extraneous sounds. Use a de‑clicker like iZotope RX De‑Click or the built‑in tools in your DAW. In Reaper, the Spectral Editing or ReaEQ can do light de‑clicking. For heavy clicking, use RX’s mouth de‑click module. Do not overdo it — clicks that are barely audible are better left alone than creating artifacts.
  2. Noise reduction first: Sample the room tone and apply noise reduction (12–18 dB). Listen for artifacts. If the noise is inconsistent (e.g., traffic rumble that changes), use a multiband noise reduction approach (e.g., RX Spectral De‑Noise with frequency‑dependent reduction). Reject the reduction if you hear pumping or phasing.
  3. High‑pass filter and corrective EQ: Cut below 80 Hz with a steep roll‑off (24–48 dB/octave). Snip any boxiness around 200–400 Hz. Optionally add a gentle presence boost around 3 kHz. Use a linear‑phase EQ if available to avoid pre‑ringing on transient words.
  4. De‑essing: Insert a de‑esser after EQ. Set threshold and range to tame only the worst sibilants. Check for unnatural sss sounds — if the de‑esser is triggering on every s, raise the threshold.
  5. Compression: Insert a compressor after the de‑esser. Start with ratio 2.5:1, attack 10 ms, release 80 ms, threshold set to achieve 3–5 dB reduction on peaks. Adjust to taste. Use the make‑up gain to bring the level to around −6 dBFS average. Monitor the gain reduction meter; if it’s pumping unevenly, adjust release time.
  6. Limiting: Place a limiter as the final processor. Set output ceiling to −3 dBFS (or −1 dBFS if you use a true‑peak limiter). Raise the input until you see 1–2 dB of gain reduction on the loudest passages. Never exceed 3 dB gain reduction on the limiter.
  7. Loudness measurement: Run the loudness meter over the entire file. If the integrated loudness is below −20 LUFS, reduce the limiter’s threshold slightly (or compress more). If it’s above −16 LUFS, reduce gain before compression. Re‑render and re‑measure until you land near −18 LUFS. A difference of ±1 dB is acceptable, but aim for exactly −18 LUFS for consistency.
  8. Final QC: Listen on headphones, laptop speakers, and car speakers. Check for any distortion, pumping, or unnatural coloration. Run ACX’s free “Check Your Audio” tool or a third‑party plugin like ACX QC Check (built into some DAWs like Reaper’s JS plugins). Pay attention to the first and last five seconds — common failure points.

Advanced Tips for Consistent, High‑Quality Results

Experienced producers develop habits that prevent common failures. Keep these in mind as you refine your plugin use.

  • Work at 24‑bit / 48 kHz: Record at 24‑bit depth to preserve headroom during processing. Downsample to 44.1 kHz only at the final export if required by your platform’s submission guidelines. Use a high‑quality sample rate conversion algorithm (such as iZotope RX’s resampling or Reaper’s r8brain) to avoid aliasing.
  • Apply gain staging: After every plugin, check that the level hasn’t drifted. Aim for peaks around −6 dBFS before the limiter. Use metering plugins to visualise the chain. In DAWs, route each track to a meter before the limiter to verify cumulative gain.
  • Use buses and automation wisely: If your audiobook has multiple chapters recorded at different levels (different days, different mic distances), apply processing on individual clips rather than the master bus. Use volume automation to smooth out level differences between chapters before compression. Automation can also help reduce breaths or sibilance without affecting the entire track.
  • Don’t de‑ess aggressively: ACX does not allow sibilance that sounds artificial. A gentle de‑esser set to reduce 2–3 dB at 6–8 kHz is usually enough. If you still hear harsh sibilance after de‑essing, try a dynamic EQ with a narrow band at the problem frequency — often 7 kHz.
  • Check your loudness on a live broadcast meter: Some free meters include the ITU‑R BS.1770‑4 standard. Use them to verify both short‑term and integrated loudness before export. The short‑term meter can reveal loudness variation within a sentence.
  • Keep a backup of unprocessed files: If a plugin update changes behaviour or you need to revisit the mix, having the raw recordings lets you start fresh without baked‑in artifacts. Store files with descriptive names including date and gain settings.
  • Batch process chapters: Once you have a preset chain that works for your voice, use your DAW’s batch processing or track templates to apply identical settings to all chapters. Then listen to each chapter individually and adjust trim or de‑esser threshold as needed.

Common Plugin Mistakes That Lead to ACX Rejection

Even with the best tools, misapplication will fail QC. Watch out for these pitfalls:

  • Over‑noise‑reduction: Applying more than 20 dB of reduction often thins the voice and introduces metallic artifacts (the “birdies” effect). Use multiple passes with lower amounts if necessary, or combine spectral editing for tonal noises.
  • Over‑compression: When the voice sounds lifeless and the background noise is pumping up, the loudness meter may show compliance, but the listening experience is poor. ACX human reviewers will notice. Keep the compressor’s gain reduction moderate and use makeup gain sparingly.
  • Ignoring true‑peak: Sample‑peak meters may show −3 dBFS, but inter‑sample peaks can exceed the limit and cause clipping on playback. Always use a true‑peak limiter as the last insert. Many free limiters (like LoudMax) include true‑peak mode.
  • Not updating plugins: Plugin developers release updates for bug fixes, new loudness standards (e.g., ITU‑R BS.1770‑4), and compatibility with DAW versions. Regularly check for updates, especially before a critical project.
  • Skipping the final check on different systems: Headphones and studio monitors can mask problems like low‑frequency rumble or sibilance. Always export a test clip and listen on earbuds, a phone speaker, and a car stereo before submitting. What sounds clean on studio monitors might be boomy on consumer headphones.
  • Using stereo plugins on a mono track: Many compressors and limiters default to stereo processing. Ensure your plugin is configured in mono mode. Stereo plugins on a mono track can cause phase or gain discrepancies. Use mono instances when available.
  • Forgetting to dither: When exporting a 24‑bit master to 16‑bit for ACX, you must dither with a noise‑shaping algorithm. While ACX accepts both 16‑ and 24‑bit, most submissions are 16‑bit. Use your DAW’s built‑in dither or a dedicated plugin like iZotope MBIT+ to add dither on the final export. Failing to dither results in quantization distortion at low levels — audible as a grainy quality in quiet passages.

Tailoring Your Plugin Chain to Your Voice and Room

No single plugin chain works for every narrator. Your voice — its resonance, sibilance, and dynamic range — determines which tools you need. Here’s how to customise your chain based on common voice characteristics:

  • Deep, bassy voice: Use higher high‑pass filter (100–120 Hz) to avoid rumble and proximity effect. Cut more at 200–300 Hz to reduce boxiness. Use a slower compressor attack (15–20 ms) to let natural transients through.
  • Thin or nasal voice: Apply a gentle low‑shelf boost at 150–200 Hz (1–2 dB). Cut at 800–1000 Hz for nasality. Use a fast compressor attack (5–10 ms) to even out dynamics. Avoid boosting presence frequencies above 5 kHz unless absolutely necessary.
  • Sibilant voice: Use a multiband de‑esser that targets only the sibilant frequencies (6–8 kHz). Consider using a dynamic EQ rather than a standard de‑esser for more surgical control. Reduce the high‑frequency shelf boost in your EQ.
  • Voice recorded in a reflective room: Add more aggressive high‑pass filtering and use a de‑reverb plugin (like iZotope RX De‑Reverb) before noise reduction. The de‑reverb can clean up slap echo without destroying voice clarity.

Record a one‑minute test sample, run it through your candidate chain, and compare the LUFS levels and spectral balance against a commercial audiobook excerpt. Repeat until the sound matches in both loudness and tone.

Conclusion: Master Your Plugin Chain for Reliable ACX Submissions

Meeting ACX sound quality standards is a repeatable process when you understand the specs and wield plugins with precision. Start by learning the acoustic characteristics of your voice and your recording space. Build a straightforward plugin chain: noise reduction → EQ → de‑essing → compression → limiting → loudness measurement. Use quality metering to hit −18 LUFS and a true‑peak cap of −3 dBFS. Avoid over‑processing; the goal is a natural, polished narration that listeners can enjoy for hours without fatigue.

Invest time in mastering one or two plugins per category rather than jumping between dozens. The producers who consistently pass ACX QC on the first attempt are those who trust their tools and their ears. With the right plugins and a methodical workflow, your audiobooks will stand out for their clarity, consistency, and professional sound. Every new project becomes easier as you refine your chain and build presets that account for your unique voice and environment.