Spectral Recovery
ADV
Overview
Spectral Recovery automatically adds missing upper frequencies to bandwidth-limited speech content and patches holes in the frequency spectrum due to compression artifacts. Spectral Recovery is particularly useful in handling VoIP recordings, such as from Skype or Zoom, where there is a hard cutoff above which no audio exists.
Controls
Learn
When Learn is pressed, Spectral Recovery analyzes your selection to determine the suggested values for the Cutoff frequency and Smoothing amount.
Spectral Patching
When selected, this automatically fills in holes in the spectrogram below the Cutoff frequency by sampling the area around the missing audio. This control is very subtle audibly, but produces a visually natural spectrogram image.
Amount
Controls how much signal energy is added above the Cutoff frequency.
Cutoff
Determines the frequency where the original audio signal stops or fades. Spectral Recovery synthesizes new signals above this frequency and fills the spectral holes below this frequency.
Vowel/Sibilant Balance
Controls the relative level of synthesized vowel/sibilant sounds above the Cutoff frequency. Use it to increase or decrease sibilant sounds to your liking as a positive or a negative percentage.
De-ess
De-ess attenuates or reduces sibilance, the harsh high-frequency sounds that come from “S”, “F”, “X”, “SH”, and a soft “C”.
Smoothing
Smoothing is similar to a crossfade, but happens across frequencies near the Cutoff. It controls the width of transition band between the original audio and the synthesised audio.
Render
To apply Spectral Recovery processing to the current selection, click the Render button in the bottom right hand corner of the window.