Cathar vs Selekt Audio Restoration
Cathar
github.comSelekt Audio Restoration
selektaudio.comSide by side
| Figure | Cathar | Selekt Audio Restoration |
|---|---|---|
| Price from | — | $4.17/mo |
| Free plan | — | Yes |
| Free trial | — | Yes |
| Open source | — | — |
| Web | — | Yes |
| macOS | Yes | — |
| Linux | Yes | — |
| Plans published | 0 | 6 |
| Spec score · AI Audio Restoration Software | 7.6 (#11) | 8.3 (#5) |
| purpose | Cathar is an audio restoration toolkit for noisy recordings and video audio tracks. | The browser-based restoration tool cleans isolated vocals, voiceovers and podcasts in modern voice mode, or complete vintage recordings and transfers in vintage mode. |
Plans
Selekt Audio Restoration
- FreeFree
- Free TrialFree
- Audio Restoration processing$0.25, paid once
- Early Access$5 a month
- Early Access monthly$5 a month
- Early Access annual$50 a year
In their own words
Cathar
Cathar is a free audio restoration toolkit for Linux and macOS, with tools to clean, enhance, and level recordings. It accepts WAV, MP3, FLAC, OGG, and M4A files, along with audio tracks inside MP4 and MKV video, and writes WAV, FLAC, or AIFF according to the output filename extension. Restoration options include denoising, de-humming, de-winding, de-clicking, de-clipping, de-reverberation, spectral repair, de-plosive processing, and rustle removal. Enhancement tools handle de-essing, breath removal, voice isolation, and bandwidth extension; loudness and peak normalization prepare delivery levels. The Rust toolkit uses inspectable and tunable DSP stages. Effects can run through the Rust crate or as command-line steps in a media pipeline. An optional machine-learning feature adds a deterministic GRU denoiser with safetensors weights. Optional terminal tools provide an audio player and spectrogram viewer. Cathar is licensed under Apache 2.0 or MIT, and its latest listed release is v0.8.0; the project describes itself as pre-1.0.
Selekt Audio Restoration
Selekt Audio Restoration is a browser-based tool for cleaning vocal recordings and older recordings or transfers. Modern voice mode targets steady room noise in isolated vocals, voiceovers and podcasts; vintage mode addresses hiss, hum, crackle, clicks and surface noise. Users can optionally master the cleaned track at no charge and compare or download the original, restored and mastered versions. The tool accepts MP3, WAV, FLAC, M4A and OGG files up to 250 MB. Modern voice mode treats non-voice audio as noise, so a full mix may require separating the vocal stem first. Users can also choose stems to split restored audio. Uploads are processed privately and are not added to the catalog. Results appear in My Library → Studio Results and remain available to stream or download for 48 hours. The restoration workflow is part of Selekt’s connected audio workspace, where results can move to tools such as Stem Lab and Drum Kit. A free plan is available; paid access starts at $5 per month, with an annual option at $50 per year.