Velvet is an iZotope de-essing and vocal refinement plugin that splits sibilance and tonal bands, features Sibilance Learn, a six-band dynamic EQ, a De-Click module, and tone shaping controls for smoother, cleaner vocals.
Purpose and Core Concept
Velvet is a de-esser designed to address harsh sibilance and unwanted mouth sounds while also offering dynamic tone shaping. Its core innovation is the intelligent split into Sibilance and Tonal channels, allowing independent processing of problematic high-frequency hiss and the musical body of the vocal.
The plugin targets both corrective and creative workflows. It provides a single tool to clean up raw recordings, but also includes features to sculpt the vocal’s tonal character. The Learn function automates setup, making it accessible for quick fixes, while manual controls offer flexibility for detailed adjustments.
Key Capabilities and Modules
The intelligent de-essing section uses a six-band dynamic EQ. Pressing Learn automatically identifies sibilant frequencies and configures the EQ bands. Users can then add or adjust nodes, scale the overall effect, and choose filter types. This transparent approach removes hiss without dulling the vocal.
The De-Click module targets clicks, pops, and other transient mouth noises. A single fader controls intensity, and a repair counter provides visual feedback on detected and processed events. This is particularly useful for fast removal of distracting artifacts.
Tone Shaping and Workflow
Dynamic tone shaping is achieved with independent Lift and Tame knobs, which boost or cut frequencies only when needed. Eight target profiles and frequency sliders let you mirror or avoid specific parts of the target curve, enabling vocal signature enhancement without static EQ.
Workflow aids include fifty pro presets that can be personalized with Learn, soloing of either channel to hear the split, threshold adjustment for the channel separation, a global Delta for auditioning changes, and a mix fader for parallel processing. These streamline both setup and fine-tuning.
Practical Applications and Limitations
In practice, Velvet is suited for vocal cleanup in mixing and mastering or podcast post-production. It excels when sibilance is inconsistent or when mouth clicks are frequent. The tone shaping can add presence or smooth harshness without excessive processing, making it a versatile final stage.
As a de-esser, its primary function is vocal-focused; using it on instruments may require manual frequency selection. There is no mention of batch processing or sidechain capabilities. The plugin’s learning and dynamic features reduce guesswork but still rely on individual program material for optimal hearing.
Key features
- Intelligent split into Sibilance and Tonal channels
- Sibilance Learn to auto-detect frequencies with a six-band dynamic EQ
- De-Click module for removing pops and clicks with a repair counter
- Dynamic tone shaping with Lift and Tame knobs and eight target profiles
- Fifty pro presets, solo, threshold, global Delta, and mix fader controls
Best for
Best for vocal producers, mixers, and podcast editors who want an efficient, transparent de-esser that also adds subtle tonal control. It shines in cleaning up sibilant or clicky recordings and equipping them with a polished sheen, suitable for both correction and creative vocal reshaping.