VARIATIONS ON A THEME

At the heart of many of the modules lie machine learning aided algorithms. Some need cues in the form of sections of the noise they need to learn patterns from, such as the De-noise modules: Spectral, Voice and Guitar. Many also offer an adaptive mode where they are performing their own analysis. The Deconstruct module allows the user to split noise from tonal sounds and transients, and this is, to a large extent, what most other Repair modules do, whether that’s the transient removal of De-click and Mouth De-click or the noise reduction of De-rustle, De-wind and De-plosive. With this in mind, a user can experiment with modules not necessarily aimed at their source, such as using De-wind or De-plosive on a boomy drum mix.