The recordings were divided by acoustic features that used speech processing techniques to track pitch and voice warble and intensity.

使用语音处理技术追踪语调的起伏、颤动和强度，然后根据声学特征对录音进行区分。