FFT music analyser

Processing Electronics (Arduino, etc.)
1

Hello everyone,
I am very new to processing, and to programming at all really. I’ve done some basic things on my arduino with LED’s, and I am now trying to extend my bluetooth LED project on my arduino.
I want to add the functionality to make the LED’s move on the music I am listening to. I am having a bit of trouble though…

I copied this code from the internet

import ddf.minim.analysis.*;
 import ddf.minim.*;
 import processing.serial.*;
 
 Serial port;
 Minim minim;
 AudioInput in;
 FFT fft;
 PFont p;

 int buffer_size = 1024; //sets FFT size (frequency resolution)
 float sample_rate = 44100;
 
 int spectrum_height = 380; // determines range of dB shown

 float[] freq_array = {0,0,0,0,0,0,0};
 float[] freq_height = {0,0,0,0,0,0,0};
 float max = 0;
 
 float dB_scale = 5.0; // px per dB
 float gain = 50; // in dB for all bands
 float[] fine_gain = {0,0,0,0,0,0,0};
 boolean CALIB = true; //wether or not to use calibrated gain
 float[] CALIB_GAIN = {50.0, 50.0, 56.0, 61.0, 64.0, 69.0, 75.0};
 
 void setup()
 {
   size(400,400);
   p = createFont("Arial",16,true); // Arial, 16 point, anti-aliasing on
   
   minim = new Minim(this);
   //port = new Serial(this,Serial.list()[0],9600); //initial serial comm. not used for now.
   
   in = minim.getLineIn(Minim.MONO,buffer_size,sample_rate);
   fft = new FFT(in.bufferSize(), in.sampleRate()); // create a FFT obj in the same time domain sas the line in sample buffer.
   fft.window(FFT.HAMMING);
   
   if (CALIB == true){
   fine_gain = CALIB_GAIN ;
   gain = 0 ;
   }
 }
 
 void draw()
 {

 // clear background
 background(0);
    
 // forward FFT on the samples in input buffer
 fft.forward(in.mix);
 
 //frequency band ranges
 freq_height[0] = fft.calcAvg((float) 20, (float) 53);
 freq_height[1] = fft.calcAvg((float) 54, (float) 144);
 freq_height[2] = fft.calcAvg((float) 145, (float) 386);
 freq_height[3] = fft.calcAvg((float) 387, (float) 1036);
 freq_height[4] = fft.calcAvg((float) 1037, (float) 2779);
 freq_height[5] = fft.calcAvg((float) 2780, (float) 7455);
 freq_height[6] = fft.calcAvg((float) 7456, (float) 20000);
 
 if (max(freq_height) > max){max = max(freq_height);}
 
 stroke(64,192,255);
 noFill();

for(int i = 0; i < 7; i++)  {
    float val = abs(dB_scale*(20*((float)Math.log10(freq_height[i]/max)) + (fine_gain[i] + gain)));     // draw the line for frequency band i using dB scale
   if (freq_height[i] < 0.01) {   val = 0;   }  // avoid log(0)
    rect(15 + (i * (spectrum_height)/7),spectrum_height,(400-20)/7,-val);
    freq_array[i] = val;
    text(val/dB_scale,15 + (i * (spectrum_height)/7),80);
  }

textFont(p,15);
fill(255);
text(finegain(fine_gain, gain),20,40);
text(str(max),20,60);


 //send to serial
/*
port.write(0xff);
for(i=0;i<7;i++){
 port.write((byte)freq_array[i]);
}
*/

 }

String finegain(float[] g, float glob){
return str(g[0] + glob) + " - " + str(g[1] + glob) + " - " + str(g[2] + glob) + " - " + str(g[3] + glob) +" - "+ str(g[4] + glob) + " - " + str(g[5] + glob) +" - "+ str(g[6] + glob);
}


void keyReleased()
{
  // +/- used to adjust gain on the fly
  if (key == '+' || key == '=') {
    gain = gain + 5.0;
  } else if (key == '-' || key == '_') {
    gain = gain - 5.0;
  } else if (key =='1') {
    fine_gain[0] += 1.0;
  } else if (key =='q') {
    fine_gain[0] -= 1.0;
  } else if (key =='2') {
    fine_gain[1] += 1.0;
  } else if (key =='w') {
    fine_gain[1] -= 1.0;
  } else if (key =='3') {
    fine_gain[2] += 1.0;
  } else if (key =='e') {
    fine_gain[2] -= 1.0;
  } else if (key =='4') {
    fine_gain[3] += 1.0;
  } else if (key =='r') {
    fine_gain[3] -= 1.0;
  } else if (key =='5') {
    fine_gain[4] += 1.0;
  } else if (key =='t') {
    fine_gain[4] -= 1.0;
  } else if (key =='6') {
    fine_gain[5] += 1.0;
  } else if (key =='y') {
    fine_gain[5] -= 1.0;
  } else if (key =='7') {
    fine_gain[6] += 1.0;
  } else if (key =='u') {
    fine_gain[6] -= 1.0;
  }
}

void stop()
{
in.close();
minim.stop();
super.stop();
}

The website I copied it from said it analyses the music on the PC, and then sends it through serial to the arduino, which then drives the LEDS.
But I havn’t been able to test it yet, because I’ve ran in a problem already;
Processing takes the input from my microphone on my headset, which isn’t what I want.
I want to analyse all the sounds of the pc that goes to my headset, but I can’t select my headset, because it doesn’t see it as an input device. How do make processing analyse the music on my headset? I am running windows 10.

Any help would be appreciated,
Staijn

1 Like
2

It’s not directly possible with processing, to be more specific with minim. You need a driver which creates a loopback interface like loopback. Then you can choose it as your input and analyse it.

But you did not tell us on what operating system you work, neither formatted your code. So please check out the guide on how to ask a question and edit your post.

1 Like
3

Edited, OS is Windows 10

4

Check this: http://code.compartmental.net/minim/audioinput_class_audioinput.html

It seems that on a desktop, you need to choose what is your input. I am assuming you do this using the sound manager of your OS. Then Minim should be able to use the current selected line input as your input source in your code. Let us know if it works for you.

Kf

1 Like
5

Here are three ways described to re-record your computer output on windows. And as @kfrajer correctly mentioned: minim will choose the default recording device of your os, as it is written in your example code.

1 Like