Audacity is the most popular free audio editor in the world, and it's capable of producing audiobook files that pass ACX's quality control. The catch: Audacity wasn't designed specifically for audiobook mastering, so you need to know which effects to use, in what order, and with what settings. Get the sequence wrong and you'll chase your tail fixing one spec while breaking another.
This tutorial walks through the complete Audacity audiobook mastering process, from opening a raw recording to exporting an ACX-ready MP3. Every step includes specific settings. For the condensed version of just the ACX-specific Audacity settings, see the Audacity ACX settings guide. For context on what each ACX specification means, the complete ACX audio requirements guide covers the technical background. And for an overview of the full production pipeline, from script to distribution, see the audiobook production workflow guide.
What Are the Prerequisites?
Before starting, you need:
- A clean, edited recording. All mistakes removed, long pauses trimmed, clicks and pops cleaned up. Mastering processes a finished recording and cannot fix editing problems.
- Audacity 3.x or later installed. Older versions may have different menu locations.
- Your raw recording as a WAV or FLAC file. Do not start from an MP3, since each MP3 encode/decode cycle degrades quality.
- A quiet recording environment. If your raw noise floor is above -55 dBFS, you'll struggle to meet ACX's -60 dBFS requirement after loudness adjustment.[1] The noise floor guide explains why.
The mastering order matters. Each step assumes the previous step is complete. Do not skip or rearrange.
Step 1: How Do You Import and Set Project Settings?
Open Audacity and import your WAV file (File → Import → Audio, or drag the file into the Audacity window).
Check the project settings in the lower-left corner:
- Project Rate: 44100 Hz
- Channel: should show Mono (if stereo, convert in Step 2 first)
If your file is at a different sample rate (48000 Hz is common from USB microphones), you'll convert in the next step.
Step 2: How Do You Convert to ACX Format Requirements?
ACX requires 44,100 Hz mono.[1] Convert now before any processing.
If stereo: Select all (Ctrl+A / Cmd+A) → Tracks → Mix → Mix Stereo Down to Mono
If wrong sample rate: Select all → Set Project Rate (bottom-left) to 44100 → Tracks → Resample → confirm 44100 Hz
Convert format first because all subsequent processing should happen at the delivery sample rate and channel configuration.
Step 3: How Do You Apply a High-Pass Filter?
A high-pass filter removes low-frequency rumble that doesn't contribute to speech clarity: building vibration, HVAC systems, plosive energy, traffic.
- Select all audio (Ctrl+A / Cmd+A)
- Effect → High-Pass Filter
- Settings:
- Frequency: 80 Hz
- Roll-off: 12 dB/octave (second-order)
This removes energy below 80 Hz, which is below the fundamental frequency of even the deepest male speaking voices (typically 85–180 Hz).[2] The 12 dB/octave slope avoids audible artefacts at the crossover point.
Why this matters: Low-frequency rumble inflates RMS measurement without contributing to intelligible speech. Removing it gives you more headroom for actual speech content.
Step 4: How Do You Measure Your Current Noise Floor?
Before adjusting loudness, check noise floor. Select a section of pure silence (at least 1 second, room tone only) and run:
- Select a silent section
- Analyze → ACX Check
Look at the noise floor reading. If already below -65 dBFS, you have good headroom and can skip noise treatment. Between -60 and -65 dBFS, apply conservative treatment. Above -60 dBFS, noise reduction is necessary before proceeding.
Critical: Your noise floor rises when you boost loudness. A raw noise floor of -60 dBFS becomes approximately -50 dBFS after a 10 dB gain boost. You need headroom now to stay under -60 dBFS after mastering.[1]
Step 5: How Do You Handle Noise Treatment?
If noise floor needs treatment, use Audacity's tools conservatively. Aggressive noise reduction creates artefacts worse than the noise itself.
Noise Reduction
- Select a section of pure silence (at least 0.5 seconds room tone only)
- Effect → Noise Reduction → Get Noise Profile
- Select all audio (Ctrl+A / Cmd+A)
- Effect → Noise Reduction:
- Noise reduction (dB): 6 to 8 (no more)
- Sensitivity: 6
- Frequency smoothing (bands): 3
Do not exceed 8 dB in a single pass. Two passes of 6 dB produces fewer artefacts than one pass of 12 dB.
Noise Gate (Alternative)
A noise gate silences audio below a threshold. It works well for pauses between sentences but doesn't reduce noise during speech.
- Select all audio
- Effect → Noise Gate
- Settings:
- Threshold: -40 dBFS
- Level reduction: -100 dBFS
- Attack: 10 ms
- Hold: 50 ms
- Decay: 100 ms
Listen to several transitions after applying to check for audible switching artefacts.
Step 6: How Do You Set Compression?
Compression reduces dynamic range, bringing whispered passages closer to projected passages. This makes it easier to hit a consistent RMS target.
- Select all audio (Ctrl+A / Cmd+A)
- Effect → Compressor
- Settings:
- Threshold: -20 dBFS
- Noise Floor: -60 dBFS
- Ratio: 2:1
- Attack Time: 0.20 seconds (200 ms)
- Release Time: 1.0 seconds
- Check: "Make-up gain" (checked)
- Check: "Compress based on Peaks" (unchecked)
For spoken word, 2:1 to 3:1 ratio is standard. Above 4:1 destroys speech dynamics and makes narration sound flat and fatiguing.[2] The mastering guide explains why gentle compression is critical for spoken word.
Step 7: How Do You Adjust Loudness?
Audacity 3.x Loudness Normalization supports both LUFS and RMS modes. Use RMS mode for ACX, since ACX measures RMS not LUFS.[3]
- Select all audio (Ctrl+A / Cmd+A)
- Effect → Loudness Normalization
- Settings:
- Normalize to: -20 dBFS RMS
- Type: RMS
- Treat mono as dual-mono: unchecked
Targeting -20 dBFS puts you in the centre of ACX's -23 to -18 dBFS range, giving 2 dB of margin on each side.[1]
Step 8: How Do You Apply Peak Limiting?
After loudness adjustment, some peaks may exceed -3 dBFS. You need a limiter to catch them.
- Select all audio (Ctrl+A / Cmd+A)
- Effect → Limiter
- Settings:
- Type: Hard Limit
- dB limit: -3.5
- Hold (ms): 10
- Apply Make-up Gain: No
Set the limiter to -3.5 dBFS, not -3.0 dBFS. Audacity's limiter operates on sample peaks, not true peaks.[3] The 0.5 dB margin accounts for inter-sample peaks that the sample-peak limiter cannot detect. Without this margin, your file may pass Audacity's ACX Check but fail ACX's true peak check.
For a more accurate true peak assessment, use a dedicated meter like Youlean Loudness Meter.[4] The free audiobook quality checker guide covers your options.
Step 9: How Do You Add Silence Padding?
ACX requires specific silence at the beginning and end of each chapter:[1]
- Head silence: 0.5 to 1 second before narration begins
- Tail silence: 1 to 5 seconds after narration ends
The silence should be room tone, not digital silence. Digital silence creates audible clicks at transitions.
Adding Head Silence
- Place cursor at the very beginning of the track
- Generate → Silence → 0.75 seconds
- If you have clean room tone, paste that instead of generated silence
Adding Tail Silence
- Place cursor at the very end of the track
- Generate → Silence → 3 seconds
The silence padding guide covers room tone, crossfade techniques, and why digital silence causes problems.
Step 10: How Do You Verify Before Export?
Run ACX Check one more time on the processed file.
- Select all audio (Ctrl+A / Cmd+A)
- Analyze → ACX Check
- Verify:
- RMS between -23 and -18 dBFS (should be near -20)
- Peak below -3 dBFS (should be at -3.5 or lower)
- Noise floor below -60 dBFS
If any metric fails, the loudness troubleshooting guide covers specific fixes, and the ACX rejection for loudness guide explains common RMS failure patterns.
Step 11: How Do You Export as MP3?
ACX requires MP3 with specific encoding settings:[1]
- File → Export → Export as MP3
- Settings:
- Bit Rate Mode: Constant
- Quality: 192 kbps
- Channel Mode: Mono
After export, verify the MP3 too. Import the exported MP3 back into Audacity and run ACX Check again. MP3 encoding can shift peak levels by 0.5–1 dB.[2] If the exported MP3 fails, adjust your limiter ceiling lower (try -4.0 dBFS) and re-export.
Step 12: How Do You Process Remaining Chapters?
Every chapter must be processed through the same chain with the same settings for inter-chapter consistency, which ACX's human reviewers check.
For efficiency, use Audacity's Macro feature (Tools → Macros → New) to record your processing chain. The recommended macro sequence: High-Pass → Compressor → Loudness Norm → Limiter.[3] The Audacity ACX settings guide includes a ready-to-use macro configuration.
What Are Common Audacity Mastering Mistakes?
Processing in the Wrong Order
Applying loudness normalisation before noise treatment amplifies the noise. Always handle noise first, then loudness, then peak limiting.
Using Music Compression Settings
Music mastering uses heavier compression (4:1 to 10:1 ratios) for competitive loudness. Audiobooks need gentle compression (2:1) to preserve natural speech dynamics.
Forgetting to Verify the Final MP3
The WAV project file and the exported MP3 are different files with potentially different measurements. Always verify the exported MP3. That's what ACX measures.
Setting the Limiter at Exactly -3.0 dBFS
Audacity's limiter uses sample peak detection. True peaks can exceed sample peaks by 0.5–1 dB.[2] Set the limiter at -3.5 dBFS minimum to account for this.
What Are Audacity's Limitations?
Audacity is capable but has real limitations:
- No true peak limiting. The built-in limiter operates on sample peaks only.[3]
- No batch processing without macros. Processing 30 chapters manually requires repeating every step 30 times.
- Limited noise reduction. Less sophisticated than dedicated tools like iZotope RX.
- No inter-chapter consistency checking. No built-in way to compare RMS or noise floor across chapters.
These limitations don't prevent ACX-compliant output, but they make the process slower for large projects. The mastering tools comparison evaluates alternatives for different workflows.