If you've tuned the control channel correctly your "tsbks" counter should be continuously incrementing and your stderr.2 log should not contain any "tuning error" or "timeout" messages. With that done, audio should decode cleanly (provided you have specified the -T and -U options) unless your virtual machine is cpu starved due to too many background processes or insufficient memory. If that happens, try reducing the -S sample size (minimum of approx 274000) so there is less data to transfer and decode.
Logs are a good way to troubleshoot most issues. Are you seeing lots of "PCM underrun" messages?
Graham