Sound radiation of trained vocalizers

Current research at NYU has focused on computational estimation of vocal loudness of George Whitefield, an Anglican preacher in the 18th century who reportedly spoke to crowds of 30,000 or more. After having established an overall level for his voice, we have begun investigating how his voice would have radiated spatially. Existing literature on the radiation of the spoken voice has focused extensively on the context of conversation in workspaces. These studies typically examine one soft, one normal, and one raised voice condition. Trained actors and orators, however, employ more methods of projection than are used in conversational speech and can achieve higher loudness as well. The radiation patterns from these types of communication have not been quantifiably studied yet. This paper investigates the radiation patterns of different methods of projection in trained vocalizers using relative intensity levels at 1 m from the front of the speaker. The results are compared to the existing data for conversational speech, and the implications to open-air oratory are discussed.