Audacity

Audacity is the industry-standard free audio editor for self-narrating indie authors. Downloaded more than 200 million times across all platforms (114M+ via FossHub since 2015 alone), it is the default starting point for any author recording their own audiobook — and remains a capable production tool even for experienced narrators. It runs on Windows, macOS, and Linux at zero cost, with no subscription, no expiry, and no feature gating.

What Audacity Does for Audiobook Authors#

The core workflow for ACX audiobook production maps directly onto Audacity's built-in tools. Authors record chapter by chapter using the punch-and-roll feature (built in since version 3.0) to fix mistakes mid-session without stopping: Audacity backs up a configurable pre-roll (typically five seconds), plays it back so you can match your tone, then records from that point — seamlessly replacing the flubbed passage. This alone eliminates hours of post-production editing compared to fixing mistakes after a full session.

After recording, Audacity's built-in Audiobook Mastering Macro automates the full ACX compliance chain in a single click: it applies a low-roll-off speech EQ filter, normalizes RMS loudness to -20dB, then applies a soft limiter. From there, the free ACX Check plugin (a Nyquist analyzer, download separately from plugins.audacityteam.org) reads your selection and reports all three ACX technical requirements at once:

• Peak level: must be no louder than -3dBFS
• RMS (average loudness): must fall between -18dBFS and -23dBFS
• Noise floor: must be no louder than -60dBFS

If your noise floor fails, Audacity's Noise Reduction effect handles most cases — the community-recommended starting point is a gentle "6,6,6" setting to avoid voice artifacts. Persistent USB microphone whine responds better to the free Mosquito-Killer4 plugin. Unpredictable environmental noise (traffic, pets) requires re-recording; no plugin can fix it cleanly enough to pass ACX human review.

For multi-chapter projects, Audacity's Macro system allows you to batch-process multiple files through the same mastering chain, export them at the correct spec (WAV 16-bit, 44.1kHz, mono, CBR), and name them to match ACX file requirements. The LabelTrack feature marks chapter edit points for navigation during editing sessions — note that labels are for editing reference only and are not automatically embedded as chapter metadata in exported audio files (a separate ffmpeg step is needed for that, though ACX does not require embedded chapters — it requires individual files per chapter instead).

Current Version and Status#

The current stable release is 3.7.7 (December 2025). Audacity's 3.7.x series has been actively maintained throughout 2024–2025, with releases in October 2024 (3.7.0), December 2024 (3.7.1), March 2025 (3.7.2/3.7.3), June 2025 (3.7.4), and August 2025 (3.7.5). An Audacity 4 Alpha 2 pre-release is available for Windows and macOS — it brings a rebuilt interface (migrated from wxWidgets to Qt, matching MuseScore 4's framework) and new clip-editing capabilities including ripple editing and grouped clips. However, Audacity 4 projects are not backwards-compatible with version 3.x files, and the alpha is not recommended for production audiobook work.

AI Features — Optional, Windows and macOS#

Audacity itself has no built-in AI features. However, Intel's OpenVINO AI plugin suite (free, separate download from audacityteam.org/download/openvino) adds optional AI capabilities that run 100% locally:

• AI Noise Suppression: Neural network-based noise reduction, works best on spoken word audio. Comparable to but different from the built-in Noise Reduction effect.
• Transcription: Powered by OpenAI's Whisper model (Whisper.cpp). Converts spoken audio to a LabelTrack with timestamped text — useful for proofing narration against the manuscript.

Platform support has expanded: the OpenVINO AI plugins arrived on macOS as of June 11, 2025 (originally Windows-only), and now cover music separation, noise suppression, music generation, transcription, and super resolution. Linux support is still in development. Because the AI features are optional, external, and require a separate download, Audacity's core offering remains non-AI.

Privacy Policy — Resolved Controversy#

Muse Group acquired Audacity in May 2021 and published an updated privacy policy that triggered significant community backlash (some calling it "spyware"). The policy has been revised multiple times. The current privacy notice (effective August 6, 2025, applying to version 3.7.2+) makes all data collection optional: update checking is on by default but disableable in preferences; error reporting and usage analytics are both opt-in. IP addresses are immediately anonymized (only the first three octets are retained). This is not a material concern for typical indie authors.

System Requirements#

• Windows: Windows 10/11 (tested). Windows 8.1, 7, and Vista may work but are not officially supported. No specific CPU or GPU requirement. Fast local HDD or SSD required — network drives and USB thumb drives are not reliable enough for recording.
• macOS: Universal Binary (Intel + Apple Silicon). Minimum macOS version not explicitly listed on the current download page; macOS 10.13+ typically cited in community documentation.
• Linux: 64-bit AppImage (runs on most modern distributions without installation).
• Storage: ~30MB for installer; project files vary by recording length (uncompressed WAV for a full audiobook runs several gigabytes).

How It Compares to Alternatives#

Hindenburg Narrator ($15/mo, $150/yr, or $499 perpetual) is the only purpose-built audiobook recording application, and its text-sync feature — which displays your manuscript and follows along as you record — is a genuine workflow advantage Audacity cannot match. Hindenburg also auto-checks ACX compliance on export. For authors who plan to narrate multiple books per year, the subscription may pay for itself in time saved.

Reaper ($60 discounted / $225 commercial one-time license) is a professional-grade DAW that offers non-destructive editing and deep customization. Experienced audio engineers prefer it, but the learning curve is steep for authors new to audio production.

Adobe Audition ($22.99/month) is the professional broadcast standard with the most polished editing interface, but the subscription cost and feature complexity make it overkill for most self-narrating authors.

GarageBand (free, macOS/iOS only) is accessible for Mac beginners but is primarily a music tool with no ACX-specific workflow, no equivalent of the ACX Check plugin, and no Windows or Linux support.

For budget-conscious authors — and especially those starting out — Audacity's combination of zero cost, cross-platform availability, and a massive library of audiobook-specific tutorials makes it the default recommendation. Authors who narrate frequently and want a more streamlined workflow should evaluate Hindenburg Narrator.

Why It Earns an Editor's Pick#

No other free, cross-platform audio editor offers Audacity's combination of punch-and-roll recording, built-in mastering macros tuned to ACX specifications, ACX Check plugin integration, and the depth of community knowledge (tutorials, courses, forum support) accumulated over two decades. At $0, it sets the baseline that every self-narrating author should learn before deciding whether to invest in a paid alternative.