Multiple Voices

Unmix › Multiple Voices…​

video New Speech Processing Options | New Features in SpectraLayers 10

Unmix multiple voices, including overlapping voices (up to 3 simultaneous overlap).

In a dialogue, the AI tries to assign voices to different layers based on identity.

  • Voices Discrimination: The higher this value, the more the AI will try to discriminate similar voices.
  • Maximum Voices: Maximum number of speakers in the recording. Ideally always set this value to the known number of speakers.
[Note] Note

When used before Transcription this can be used for dialogue transcription with identified speakers.

[Note] Note

If the recording contains background noise, it’s best to first use Noisy Speech or Voice Denoiser.