---
id: resemble-perth
version: "1.0.1"
license: unclear
license_treatment: permissive
maintenance: aging
---

# resemble-perth — Audio Watermarking and Detection Library

License: permissive · Maintenance: aging · Downloads: 239.9K/mo

## What it is and what it does

Perth is a Python library for embedding and extracting imperceptible watermarks in audio files. It implements neural network-based watermarking techniques—primarily the Perth-Net Implicit approach—designed to survive audio transformations like compression and resampling while maintaining perceptual audio quality. The library provides both a command-line interface for quick tasks and a Python API for integration into applications.

You load audio, initialize a watermarker object, apply or extract a watermark, and save the result. The package includes utilities to measure audio quality metrics and compare original and watermarked audio. It supports Python 3.8, 3.9, 3.10, and 3.11 and has no runtime dependencies, making installation straightforward.

Use it for:

- Embed ownership or authenticity markers in audio content for copyright protection or provenance tracking.
- Detect whether audio has been watermarked as part of a content verification or anti-piracy workflow.
- Test audio watermarking robustness by applying transformations and checking if watermarks survive.
- Evaluate perceptual quality of watermarked audio using built-in metrics before deploying to production.
- Integrate watermarking into an audio processing pipeline via the Python API.

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Embeds imperceptible watermarks into audio files and detects them afterward, using neural network-based techniques that survive common audio transformations.

Yes, if you need audio watermarking. The library has low install friction, permissive licensing, no known vulnerabilities, and an active repository. However, the aging maintenance status and lack of recent updates suggest you should verify that the neural network models and robustness claims meet your specific use case before committing to production deployment.

## Install

pip install resemble-perth
uv add resemble-perth
poetry add resemble-perth

## Installing resemble-perth

Before you install: Low friction installation with no runtime dependencies. Maintenance status is aging—last release was in May 2025 and the repository has not been updated since then, though it remains active with 533 stars.

License in practice: Licensed under MIT (permissive), so you can use, modify, and distribute the package freely with minimal restrictions.

Quickstart:

pip install resemble-perth

import perth
watermarker = perth.PerthImplicitWatermarker()
watermarked_audio = watermarker.apply_watermark(wav, watermark=None, sample_rate=sr)
watermark = watermarker.get_watermark(watermarked_audio, sample_rate=sr)

Audio input must be loaded separately; the package does not include audio I/O dependencies.

Verify before relying:

- Whether the neural network models are bundled or require separate download/initialization.
- GPU/CUDA requirements and performance characteristics for the PerthImplicitWatermarker.
- Specific audio formats and sample rates supported beyond the examples shown.
- Robustness claims against particular audio transformations (compression codecs, bit rates, etc.).

## Package facts

- License: not declared (permissive)
- Python support: supports_current
- Install friction: low
- Maintenance: aging
- Downloads: 239.9K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

audio watermarking library, embed watermarks in audio, detect audio watermarks, neural network audio watermarking, robust audio watermark detection, audio fingerprinting, imperceptible audio marking, audio-processing, watermarking, neural-networks

[View on SkillFed](https://skillfed.io/packages/resemble-perth) · [View on PyPI](https://pypi.org/project/resemble-perth/)