How to implement effects in java sound

Question

I have read the programmer's guide in java sound and it said there that it can generate effects such as distortion, delay and others. I would like to implement some of those, I can already capture and play sound from a microphone but I don't know how to manipulate the data to get my desired effect. Can someone explain the content of the array of bytes? I also hope you can give me some hints or anything that could help me implement my own effect. Thanks in advance!

Answer 1

going_infinity wrote:
The contents of the double array do not exceed the range. They are mostly of -0.07 o 0.07.
I forgot to post this code, this is how byte was transformed to a double array. I am just extending an existing codeAh, yes...that makes a lot more sense. I assumed you were reading from the line as double, and that wouldn't work. But you're already converting it, which is why everything stays in the range you'd expect.
"Weirdness" solved.
If I use the double array or the byte array, how can I implement some effects like echo or reverb? I understand that since the "raw data" provides amplitude, I can get it's frequency and there is some FFT involved. All I can see here is determining the pitch of the sound but still, no effects. I'm still confused on where to start. Well, it depends on the effect. I'm not an effects expert, so you'll probably want to go do some Google-ing for what all of the affects are...but I think I can explain an echo.
Echo is a function of a variable, decay, which is how quickly an echo will fade away. From a physics perspective, think of decay as friction, it'll eventually stop a moving object.
So, you might implement an echo kinda like the following
for (int i = 0; i < buffer.length; i++) {
     double threshhold = 0.01;
     double decay = 0.5;
     double echo = sample;
     int j = i+1;
     while((Math.abs(echo) > threshhold) && (j < buffer.length)) {
          buffer[j++] += echo;
          echo *= decay;
Where some decreasing-fraction of a sample is added to all samples in the future until it drops below some threshhold.
That may or may not be an "echo" effect, necessarily, as I think that a "true" echo also has a peroid (the delay between the sound and its echo) that I'm not taking into effect, so that effect is probably more like overdrive...but it illistrates the point. You'll have to figure out what each effect "adds" or "subtracts" from the sound, and then loop-through the sound data sequentially adding the effects of the effect.