JOAL学习笔记 第二课 循环与淡出

JOAL学习笔记

 

先是例行的连续代码页

import java.nio.ByteBuffer;
import com.jogamp.openal.AL;
import com.jogamp.openal.ALC;
import com.jogamp.openal.ALFactory;
import com.jogamp.openal.util.ALut;
public class LoopingAndFadeaway {
  	static int[] buffer = new int[1];
  	static int[] source = new int[1];
  	static float[] sourcePos = { 0.0f, 0.0f, 0.0f };
  	static float[] sourceVel = { 0.0f, 0.0f, 0.1f };
  	static float[] listenerPos = { 0.0f, 0.0f, 0.0f };
  	static float[] listenerVel = { 0.0f, 0.0f, 0.0f };
  	static float[] listenerOri = { 0.0f, 0.0f, -1.0f, 0.0f, 1.0f, 0.0f };
  	static AL al;
  	static ALC alc;
        static int loadALData() {
            if (al.alGetError() != AL.AL_NO_ERROR) {
                return AL.AL_FALSE;
            }
            int[] format = new int[1];
            int[] size = new int[1];
            ByteBuffer[] data = new ByteBuffer[1];
            int[] freq = new int[1];
            int[] loop = new int[1];
   	
            al.alGenBuffers(1, buffer, 0);
            if (al.alGetError() != AL.AL_NO_ERROR)
                return AL.AL_FALSE;
            ALut.alutLoadWAVFile("wavdata/Footsteps.wav",format,data,size,freq,loop);
            al.alBufferData(buffer[0], format[0], data[0], size[0], freq[0]);


            al.alGenSources(1, source, 0);
            al.alSourcei(source[0], AL.AL_BUFFER, buffer[0]);
            al.alSourcef(source[0], AL.AL_PITCH, 1.0f);
            al.alSourcef(source[0], AL.AL_GAIN, 1.0f);
            al.alSourcefv(source[0], AL.AL_POSITION, sourcePos, 0);
            al.alSourcefv(source[0], AL.AL_VELOCITY, sourceVel, 0);
            al.alSourcei(source[0], AL.AL_LOOPING, AL.AL_TRUE);
            if (al.alGetError() != AL.AL_NO_ERROR) {
                return AL.AL_FALSE;
            }
            return AL.AL_TRUE;
        }
        static void setListenerValues() {
            al.alListenerfv(AL.AL_POSITION,	listenerPos, 0);
            al.alListenerfv(AL.AL_VELOCITY,    listenerVel, 0);
            al.alListenerfv(AL.AL_ORIENTATION, listenerOri, 0);
        }
	
        static void killALData() {
            al.alDeleteBuffers(1, buffer, 0);
            al.alDeleteSources(1, source, 0);
            ALut.alutExit();
        }
    public static void main(String[] args) {
        ALut.alutInit();
        al = ALFactory.getAL();

        if(loadALData() == AL.AL_FALSE) {
            System.exit(1);
        }; 
         setListenerValues();
        al.alSourcePlay(source[0]);
        long startTime = System.currentTimeMillis();
        long elapsed = 0;
        long ticker = 0;
        long lastTime = 0;
        while (elapsed < 10000) {
            elapsed = System.currentTimeMillis() - startTime;            
            if (ticker > 100) {
                ticker = 0;
                sourcePos[0] += sourceVel[0];
                sourcePos[1] += sourceVel[1];
                sourcePos[2] += sourceVel[2];
                al.alSourcefv(
                    source[0],
                    AL.AL_POSITION,
                    sourcePos, 0);
            }
            ticker += System.currentTimeMillis() - lastTime;
            lastTime = System.currentTimeMillis(); 
        }
        ALut.alutExit();
    }
}

之后是一些值得注意的问题

 

首先，如果你想通过修改声源或听众的位置来达到不同的声音效果，请确定你用于测试的音频是单声道音频，如果使用了双声道音频，OpenAL依然会播放，但会忽略其位置量造成的声音改变，笔者在这里走了很多弯路，最终在一个国外论坛的一角发现了这个原因。

 

之后一点，这个程序是由C语言翻译过来的，因此带了一点C语言的特征，例如最后循环使用了忙等待，对于一个高性能CPU忙等待不是一个好的选择，可以使用sleep修改为闲等待，虽然这样精确性有所下降，但声音的淡出过程的确对精确性要求不高，以此节约大量CPU资源是值得的，比如说这样写：

while (true) {
	sourcePos[0] += sourceVel[0];
	sourcePos[1] += sourceVel[1];
	sourcePos[2] += sourceVel[2];
		
	al.alSourcefv(source[0], AL.AL_POSITION, sourcePos, 0);
	try {
		Thread.sleep(100);
	} catch (InterruptedException e) {System.out.println("Intp!");}
}