CPU usage ~2–4% on modern i7. Minimal latency (under 10ms). Artifacts are controlled well in “Modern” mode.
Trained on thousands of voice and noise samples. Uses deep learning to isolate speech even under heavy background (e.g., traffic, café noise). noise removal plugin free
Records a noise print from a selection of audio (silence section), then subtracts that exact spectral shape. Works like iZotope RX’s Voice De-noise but free. CPU usage ~2–4% on modern i7