In various embodiments, an audio streaming system may provide bi-directional communication, in which content data is streamed from a server associated with a streaming service to a plurality of client devices and user generated audio data can also be transmitted from a client device to the streaming service over a network such as the Internet. In some embodiments, the client device includes both audio playback and voice communications capabilities such that a communication channel can be established between listeners and the streaming service via the client device and over the network. Automated speech recognition may also be used to translate the user-generated audio data into text such that the content can be further analyzed, such as to sort, categorize, rank, or filter the user-generated audio data. The user-generated audio data may also be analyzed for audio quality, such as to filter out poor quality data.