Intel Spatial Audio for Java* Package Tutorial

Lesson 7: Streaming

Goal: This lesson covers the use of streaming media in the Intel Spatial Audio for Java* package. The package provides support for streaming sound sources and streaming listeners.

The Streaming Sound Source

The applet generates its tune by filling up a buffer with data for a random note, then submitting that buffer to the streaming sound source using the submitBuffer() method. When you submit a buffer, you can specify an argument to an event which will be signaled when the audio renderer is finished with the buffer. In this way, you can re-use the buffers rather than allocate more and more memory.

This applet uses four steps to accomplish the streaming:

1. Create the buffers and the data. The data for seven different notes and a rest are stored in byte arrays -- byte dataRest, byte dataA[], etc. We only use three instances of the BufferHolder class, though. The idea is to submit all three to the streaming sound source, then catch the event indicating that the audio renderer is done with each buffer, set the data to the next value, and resubmit it. When we create each BufferHolder, we specify the argument to the completion event. The Event.id will be SoundSource.BUFFER_EVENT. We must also specify the AWT Component which will be the target of the event; in an applet, you can often pass this.

buffer1 = new BufferHolder(dataRest, (Object)"event1", null, this);

2. Remove any buffers from the streaming sound source's playback queue:

streamingSoundSource.flush();

3. Start up the sound source. We create a method called playNextNote() in which we set the data variable of a BufferHolder to the byte array representing the next note and submit the buffer to the streaming sound source. Then we call this method once on each of the three BufferHolders.

b.data = getCurrentNote();
streamingSoundSource.submitBuffer(b);

4. Keep playing until the tune is done (15 notes). We use the applet's handleEvent() method to catch the BUFFER_EVENTS. When we receive one, we know that the audio renderer is finished with the data, so we set the BufferHolder's data variable to the byte buffer representing the next note and submit the buffer again.

public boolean handleEvent(Event evt) {

switch(evt.id) {

case SoundSource.BUFFER_EVENT:

if (evt.arg     == "event1") {
	playNextNote(buffer1);
	return true;
}
if (evt.arg     == "event2") {
	playNextNote(buffer2);
	return true;
}
if (evt.arg     == "event3") {
	playNextNote(buffer3);
	return true;
}

} // switch

...

}

The Streaming Listener

While a direct listener sends the data data generated by the audio renderer directly to the audio output device (usually headphones or speakers), the streaming listener receives buffers of data via its requestBuffer() method. Please refer to the documentation in the User's Guide for details and a simple usage model.

This page was last updated on Feb 11th, 1997.

Legal Stuff