Hello list,
Can anyone point me towards a freely downloadable corpus of anechoically-recorded spoken digits? (American-accent English spoken by a native speaker) I'm designing a cocktail party streaming test and my recording conditions are less than ideal. If the corpus is RMS-normalized with synchronized onsets, that is even better (but can be done in Matlab just as well :) ) Thanks for any help!
Best,
Anthony Parks
PhD Student in Architectural Acoustics
Rensselaer Polytechnic Institute
Program in Architectural Acoustics