Overview

Give Matchering a target and a reference, and it produces a mastered target aimed at the reference’s RMS, frequency response, peak amplitude and stereo width. That makes it useful when a particular recording provides a meaningful benchmark, but less suitable when you want to build a sound without a reference or need a broad mastering environment.

Matchering 2.0 is available as a containerized web application, a Python library and a ComfyUI node, and it is integrated into the UVR5 Desktop App. Songmastr, MVSEP and Moises host it for users who want to try it without installing it. The project says the algorithm uses digital signal processing rather than a neural network, and users may use their mastered tracks wherever they want.

Key features

Reference matching is the central capability: the algorithm targets four measurable aspects of the reference rather than offering an open-ended set of controls. This can help producers pursue a consistent level and tonal profile across tracks, but results depend on choosing a reference that suits the material. The algorithm is described as working well with almost all genres, especially EDM; experimental music with a very specific musical form is the stated exception.

WAV and MP3 are supported input formats, and WAV export is available. MP3 loading has an extra setup requirement: FFmpeg must be installed separately. That is a modest hurdle for a local workflow, but worth accounting for in automated or shared deployments.

Developers can connect the Python library to the broader Python ecosystem or use its command-line application. The library requires Python 3.8.0 or higher and a machine with 4 GB of RAM, so it is a reasonable fit for compatible systems but not a zero-dependency option.

Pricing

Matchering’s Open-source software plan costs 0.00 USD per free under the GNU General Public License v3 (GPLv3). It includes the open-source software, and no paid plans are offered. That makes it a strong fit for users comfortable running or integrating the software themselves; there is no paid tier to trade setup work for.

The project also says users can try Matchering through Songmastr, MVSEP and Moises without installing it. The web application itself is designed for home and in-house use, not public Internet hosting as-is. Anyone considering a public service should account for the security and scalability work the maker says is necessary.

Platforms

Matchering supports Linux, macOS and Windows, as well as web and self-hosted use. The Python library, containerized web application and ComfyUI node provide distinct routes to use it; the choice depends on whether the priority is integration, a web interface or a node-based workflow. MP3 input requires separately installed FFmpeg.

Who it's for

Matchering suits musicians and producers with a reference track in mind, particularly those working in genres the algorithm is said to handle well, such as EDM. It also suits developers who want reference-matching processing in a Python workflow, provided their machine meets the library’s runtime requirements.

It is a poor fit for operators who plan to expose the supplied web application publicly without making changes. The maker warns that its current design is for home and in-house use: Django, SQLite, Redis and the Matchering worker run in one container, SQLite is non-scalable, DEBUG is enabled, and there is no production web server. Those conditions call for security and scalability changes before public exposure.

Pros and cons

Pro: Matching RMS, frequency response, peak amplitude and stereo width gives users a clear reference-based mastering target.
Pro: The Python library, command-line application, containerized web application, ComfyUI node and UVR5 integration accommodate several technical workflows.
Pro: The free GPLv3 offering and permission to use mastered tracks wherever users want make it an accessible option for personal and commercial output.
Con: The result depends on the choice of reference, and the project cautions against experimental music with a very specific musical form.
Con: MP3 loading requires a separate FFmpeg installation, adding a dependency to workflows that need MP3 input.
Con: The web application is not suitable for public Internet hosting without security and scalability changes.

Alternatives

For a broader directory of options, see Audio Mastering Software and AI mastering tools.

IK Multimedia ReSing is worth considering for macOS or Windows users who want its free tier or a one-time paid plan; its free tier is limited to 2 voices, 2 instruments and 1 RVC import, with no model generation.
Sonoris DDP Creator is a paid alternative for users seeking DDP Creator Standard or Pro licenses, with a free trial and one-time prices of 249.00 EUR and 349.00 EUR respectively.
DSP-Quattro may fit macOS users comparing a free trial and paid 6.x licenses, including a 99.00 USD per month new-license plan.
FUSER is an option for users who want a paid mastering tool for macOS or Windows; it has a free trial and costs 59.00 GBP per once.
Tunr is another free macOS and Windows option, with a free key sent by email, counted per machine, and output choices including 16-bit and 24-bit.
Brainworx bx_masterdesk may suit users comparing a paid subscription for macOS or Windows, with CORE at 14.99 USD per month and PRO at 29.99 USD per month.
Voxengo SPAN is a free real-time FFT spectrum analyzer plugin for macOS and Windows rather than a reference-matching mastering workflow.
Waves Curves AQ is an alternative for users considering a broader paid plugin subscription: Waves Ultimate costs 24.99 USD per month or 249 per year and includes Curves AQ.

Verdict

Choose Matchering if you want free, reference-led mastering and can use its local, Python or hosted routes within their limits. Its focused matching targets and flexible integrations are compelling; the public-hosting warning, MP3 dependency and reliance on a suitable reference are the main reasons to choose another tool.

	Matchering	ShevtoneAudio Mastering
Free plan	Yes	No
Free trial	No	No
Paid from	Free	—
Open source	No	No
Platforms	Linux, macOS, self-hosted, Web, Windows	macOS

Matchering vs ShevtoneAudio Mastering

Matchering

ShevtoneAudio Mastering