Speech modulated noise master audio - HARVARD speech corpus
mediaposted on 16.10.2019 by Philippa Demonte
Media is any form of research output that is recorded and played. This is most commonly video, but can be audio or 3D representations.
Master audio .wav file of speech modulated noise (SMN), created from envelope of speech applied in time domain to speech-shaped noise (SSN) generated by 52nd order linear predictive coding (see Matlab script).
Sampling rate: 48 kHz
For use in testing of speech intelligibility with Demonte (2019): HARVARD speech corpus - audio recording 2019. figshare. collection.