Provided are an audio data processing method and a device. The method comprises the steps of: acquiring the voice of the photographed person according to at least two different sampling parameters so as to at least acquire first audio data and second audio data (S101); and sending audio data collected according to the minimum sampling parameters to a terminal device (S102). The sampling parameter is small so that the audio data of the photographed person acquired according to the minimum sampling parameter are ensured to be less, the transmission bandwidth occupied by the audio data is small when the audio data are transmitted, and thus voice recording equipment is ensured to transmit the acquired audio data to the terminal equipment in real time.