CrystalSound vs DeepFilterNet
CrystalSound
crystalsound.aiDeepFilterNet
github.comSide by side
| Figure | CrystalSound | DeepFilterNet |
|---|---|---|
| Price from | $19/mo | Free |
| Free plan | Yes | Yes |
| Free trial | Yes | — |
| Open source | — | Yes |
| Windows | Yes | Yes |
| macOS | Yes | Yes |
| Linux | Yes | Yes |
| API | — | Yes |
| Self-hosted | — | Yes |
| Plans published | 4 | 2 |
| Spec score · AI Speech Enhancement Software | 8.3 (#11) | 8.7 (#5) |
Plans
CrystalSound
- StarterFree
- Elite (Coming soon)$228 a year
- EnterprisePriced on request
- PremiumPrice not published
In their own words
CrystalSound
CrystalSound is a desktop app for reducing call noise and recording meetings on Windows, macOS and Linux. Its bidirectional noise cancellation is designed to reduce background noise from both the user and the other party. My Voice Only uses a registered voice profile to filter unwanted noise and other voices. The app can record the screen and both sides of audio, then provide meeting minutes, keywords and insights. It also enhances MP3, WAV and FLAC recordings and offers pitch shift, reverb and robotic voice effects. CrystalSound is described as working with conference apps, with OBS, Discord, Loom, Slack, Teams, Skype and Meet shown. The installation guide lists 4 GB RAM for each operating system, plus Windows 10 (20H1), supported recent macOS versions, or Linux with glibc 2.28 or later. Audio and VoiceID are processed locally. The company also collects app version, settings, email, operating system, usage times and which apps use its virtual speaker or microphone. Starter is free, and a 7-day trial is offered; Elite is listed at $228 a year and marked “Coming soon.”
DeepFilterNet
DeepFilterNet is a free, open-source framework for speech enhancement on full-band 48 kHz audio, using deep filtering. Its command-line tool suppresses noise in WAV files, with 48 kHz as its only supported sampling rate. A LADSPA plugin provides real-time noise suppression and can be used with PipeWire filter chains to create a virtual microphone or audio output, including for VoIP apps such as Zoom or Discord. The plugin's minimum latency is 20 ms from STFT processing, plus any delay added by the LADSPA host. The Rust command-line tool defaults to DeepFilterNet2; the Python interface documents DeepFilterNet and DeepFilterNet2. An optional post-filter may slightly over-attenuate very noisy sections. The framework supports Linux, macOS and Windows, though training is tested only under Linux. The Python package installs as deepfilternet, and an optional train extra adds data-loading functionality for Linux training. Code is available under either the MIT License or Apache License 2.0.