tonhwa.blogg.se

I have an acoustic piano, an old Sony Vaio PC and a newly-purchased Behringer XENYX Q802USB. I have been using Jamkazam predominantly to rehearse with an acoustic bass player who lives not far from me. My observations in this posting are based on Jamkazam sessions with the bass player, though I aspire to play with others at greater distances as well. The bass player has been exclusively using Realtek WMD so far, though he too has recently acquired a Behringer audio interface and may be confronted with similar issues. I have been considering the following options, which I will discuss below. I first tried the Realtek WMD after initially installing Jamkazam, before purchasing the Behringer, using my PC mic to pick up audio. My latency was low, around 5-6 ms., and my jitter was generally acceptable. Jamkazam’s only “complaint” was that I was using WMD, not ASIO. Nonetheless, I bought the Behringer hoping to be able to use a better microphone while retaining low latency, so far without success. While Behringer recommends using ASIO4ALL for its Q802USB, many Jamkazam users have noted problems with it, so I tried using the Windows WMD USB Audio driver first. Latency increased to 20ms, making it difficult to play together. There was a tendency for our tempo to slow down during a tune, so I decided to go back to using Realtek for rehearsals.ĭeciding to check out ASIO4ALL for myself, I discovered that I could not successfully use the smaller frame sizes (around 2.5 ms.), needing to set my frame size to 10 ms, to get intelligible audio. Inexplicably it seemed easier for us to play together when I used ASIO4ALL than WMD USB Audio.

If I retry WMW USB again, would we still have problems playing together?Īfter reading an excellent Jamkazam forum posting by marhdeth ( ), I followed the recommendation, trying the Reaper DAW.

Using a 2.5 ms frame size, I brought latency down in the jamkazam test to 5-6 ms. However I was not able to successfully add this gear configuration to my audio profile, failing the I/O rate and variance test. In a response to the marhdeth posting, Nick Blair reported having the same problem, and the thread does not posit a solution.