A secondary sound buffer that contains an entire sound. Typically it is filled with data only once.