A visualiser for Sonic Pi 3

Every since Sonic PI had a transparency mode added (not available on the Raspberry Pi versions) I have been interested in adding a Visualiser graphics display as a backdrop. The built in Scope can give attractive displays, but I wanted something with a bit more colour, and which covered the whole screen. I did some experiments with the iTunes visualiser which added quite a nice backdrop, but only with a significant delay between the audio and the fluctuations of the display. The recent arrival of Sonic Pi 3 allowed for further possibilities, because it enabled the use of OSC messaging. I did a trawl of the internet and came across a promising looking processing sketch which produced a pattern which reacted to incoming audio. It was written several years ago and was only monochrome based. I played around with this, upgrading it to work on the latest version of Processing, and added some colour. I experimented with adding further  basic shapes to the display (it originally used an ellipse primitive, set up to give concentric circles, which could also be driven to produce a star burst effect). I added rectangles and star shapes and experimented with off-setting these as the program ran, and also with using more than one basic shape at the same time. I then added some code so that the sketch could receive incoming OSC messages sent from Sonic Pi 3, which could be used to control various parameters for the shapes, such as stroke width, colour, and offsets on the screen. I added further flexibility such as the ability to rotate the shapes, and to shift the whole display vertically and horizontally across the screen. The final setup works well with Sonic Pi 3, which controls the patterns both with the Audio signal it produces, and also with OSC messages which can be sent to the sketch display code. These can be timed appropriately with the musical content.

The code for the sketch is shown below

//Visualiser for use with Sonic Pi 3 written by Robin Newman, September 2017
// based on an original sketch https://github.com/andrele/Starburst-Music-Viz
//Changes: changed to full screen, updated for Processing 3, added colour, added rectangle and star shapes
//added OSC input (from Sonic Pi) to alter parameters as it runs, removed slider inputs.
//input OSC:   /viz/float       updates STROKE_MAX, STROKE_MIN and audioThresh
//             /viz/pos         updates XY random offset values (can be zero)
//             /viz/col         updates RGB values
//             /viz/shapes      sets shapes to be used from S, E and R (or combinations thereof)
//             /viz/stardata    sets data for star shapes
//             /viz/rotval      turns rotation of shapes on/off
//             /viz/shift       turns XY shift across screen on/off
//             /viz/stop        initiates sending stop all signal back to Sonic Pi port 4557

import ddf.minim.analysis.FFT;
import ddf.minim.*;
import oscP5.*; //to support OSC server
import netP5.*;

Minim minim;
AudioInput input;
FFT fftLog;

int recvPort = 5000; //can change to whatever is convenient. Match with use_osc comand in Sonic Pi
OscP5 oscP5;
NetAddress myRemoteLocation; //used to send stop command b ack to Sonic PI

// Setup params
color bgColor = color(0, 0, 0);

// Modifiable parameters
float STROKE_MAX = 10;
float STROKE_MIN = 2;
float audioThresh = .9;
float[] circles = new float[29];
float DECAY_RATE = 2;
//variables for OSC input
float [] fvalues = new float[5]; //STROKE_MAX, STROKE_MIN,audioThresh values
int [] cols = new int[3]; //r,g,b colours
int [] positions = new int[2];// random offset scales for X,Y
int [] stardata = new int[4];// data for star shape, number of points, random variation
int shiftflag = 0; //flag to control xy drift across the screen set by OSC message
int two = 0; //variable to force concentric shapes when more than one is displayed
String shapes = "E"; //shapes to be displayed, including multiples from S,E,R
int rotval =0;
int xoffset = 0,yoffset = 0;
int xdirflag = 1,ydirflag = 1;

void settings() {
  fullScreen(P3D);
}

void setup() {  
  frameRate(60);
   myRemoteLocation = new NetAddress("127.0.0.1",4557); //address to send commands to Sonic Pi
  minim = new Minim(this);
  input = minim.getLineIn(Minim.MONO, 2048); //nb static field MONO referenced from class not instance hence Minim not minim

  fftLog = new FFT( input.bufferSize(), input.sampleRate()); //setup logarithmic fast fourier transform
  fftLog.logAverages( 22, 3); // see http://code.compartmental.net/minim/fft_method_logaverages.html

  noFill();
  ellipseMode(RADIUS); //first two coords centre,3&4 width/2 and height/2
  fvalues[0]=1.0;
  fvalues[1]=0.0;
  fvalues[2]=0.32;
  cols[0] = 255;
  cols[1]=0;
  cols[2]=150;
  positions[0] = 50;
  positions[1]=40;
  stardata[0]=2;
  stardata[1]=4;
  stardata[2]=3;
  stardata[3]=5;
  /* start oscP5, listening for incoming messages at recvPort */
  oscP5 = new OscP5(this, recvPort);
  background(0);
}

void draw() {
  background(0);
  pushMatrix();
  //calculate changing xy offsets: shiftflag set to 0 to siwtch this off
  xoffset += 10*xdirflag*shiftflag;
  yoffset += 10*ydirflag*shiftflag;
  if(shiftflag==0){xoffset=0;yoffset=0;} //reset offset values to zero if shifting is off
  //reverse directions of shifting when limits reached
  if (xoffset >displayWidth/3){xdirflag=-1;}
  if (xoffset < -displayWidth/3){xdirflag=1;} if (yoffset > displayHeight/3){ydirflag=-1;}
  if (yoffset < -displayHeight/3){ydirflag=1;}
  //transform to new shift settings
  translate(displayWidth/2+xoffset, displayHeight/2+yoffset); //half of screen width and height (ie centre) plus shift values

  //optional rotate set by OSC call
  rotate(float(rotval)*(2*PI)/360);
  
  //get limits for stroke values and audiThreshold from OSC data received
  STROKE_MIN=fvalues[0];
  STROKE_MAX=fvalues[1];
  audioThresh=fvalues[2]; 
  //println("fvalues: ",STROKE_MIN,STROKE_MAX,audioThresh); //for debugging

  // Push new audio samples to the FFT
  fftLog.forward(input.left);

  // Loop through frequencies and compute width for current shape stroke widths, and amplitude for size
  for (int i = 0; i < 29; i++) {

    // What is the average height in relation to the screen height?
    float amplitude = fftLog.getAvg(i);

    // If we hit a threshold, then set the "circle" radius to new value (originally circles, but applies to other shapes used)
    if (amplitude < audioThresh) { circles[i] = amplitude*(displayHeight/2); } else { // Otherwise, decay slowly circles[i] = max(0, min(displayHeight, circles[i]-DECAY_RATE)); } pushStyle(); // Set colour and opacity for this shape circle. (opacity depneds on amplitude) if (1>random(2)) {
      stroke(cols[0], cols[1], cols[2], amplitude*255);
    } else {
      stroke(cols[1], cols[2], cols[0], amplitude*255);
    }
    strokeWeight(map(amplitude, 0, 1, STROKE_MIN, STROKE_MAX)); //weight stroke according to amplitude value

    if (shapes.length()>1) { //if more than one shape being drawn, set two to 0 to draw them concentrically
      two = 0;
    } else {
      two = 1;
    }
    // draw current shapes
    if (shapes.contains("e")) {
      // Draw an ellipse for this frequency
      ellipse(random(-1, 1)*amplitude*positions[0]*two, random(-1, 1)*amplitude*positions[1]*two, 1.4*circles[i], circles[i]);
    }
    if (shapes.contains("r")) {
      rectMode(RADIUS); 
      rect( random(-1, 1)*amplitude*positions[0]*two, random(-1, 1)*amplitude*positions[1]*two, 1.4*circles[i], circles[i]);
    }
    if (shapes.contains("s")) {
      strokeWeight(3); //use fixed stroke weight when drawing stars
      //star data Xcentre,Ycentre,radius1,radius2,number of points
      star(random(-1, 1)*amplitude*positions[0]*two, random(-1, 1)*amplitude*positions[1]*two, circles[i]*stardata[0], circles[i]/stardata[1], int(stardata[2]+random(stardata[3])));
    }
    popStyle();

    //System.out.println( i+" "+circles[i]); //for debugging
  } //end of for loop
  popMatrix();
}

void oscEvent(OscMessage msg) { //function to receive and parse OSC messages
  System.out.println("### got a message " + msg);
  System.out.println( msg);
  System.out.println( msg.typetag().length());

  if (msg.checkAddrPattern("/viz/float")==true) {
    for (int i =0; i<msg.typetag().length(); i++) {
      fvalues[i] = msg.get(i).floatValue();
      System.out.print("float number " + i + ": " + msg.get(i).floatValue() + "\n");
    }
  }

  if (msg.checkAddrPattern("/viz/pos")==true) {
    for (int i =0; i<msg.typetag().length(); i++) {
      positions[i] = msg.get(i).intValue();
      System.out.print("pos number " + i + ": " + msg.get(i).intValue() + "\n");
    }
  }

  if (msg.checkAddrPattern("/viz/col")==true) {
    for (int i =0; i<msg.typetag().length(); i++) {
      cols[i] = msg.get(i).intValue();
      System.out.print("col number " + i + ": " + msg.get(i).intValue() + "\n");
    }
  }
  if (msg.checkAddrPattern("/viz/shapes")==true) {
    shapes=msg.get(0).stringValue();
    //for(int i =0; i<msg.typetag().length(); i++) {
    // shapes += msg.get(i).stringValue().toLowercase();      
    //}
    System.out.print("shapes code "+ shapes + "\n");
  }
  if (msg.checkAddrPattern("/viz/stardata")==true) {
    for (int i =0; i<msg.typetag().length(); i++) {
      stardata[i] = msg.get(i).intValue();
      System.out.print("stardata number " + i + ": " + msg.get(i).intValue() + "\n");
    }
  }
  if (msg.checkAddrPattern("/viz/rotval")==true) {
    rotval =msg.get(0).intValue();
    System.out.print("rotval code "+ rotval + "\n");
  }
  if (msg.checkAddrPattern("/viz/shift")==true) {
    shiftflag =msg.get(0).intValue();
    System.out.print("shiftflag code "+ shiftflag + "\n");
  }
  if (msg.checkAddrPattern("/viz/stop")==true) {
    kill(); //stop Sonic Pi from running
  }
}

//function to draw a star (and polygons)
void star(float x, float y, float radius1, float radius2, int npoints) {
  float angle = TWO_PI / npoints;
  float halfAngle = angle/2.0;
  beginShape();
  for (float a = 0; a < TWO_PI; a += angle) {
    float sx = x + cos(a) * radius2;
    float sy = y + sin(a) * radius2;
    vertex(sx, sy);
    sx = x + cos(a+halfAngle) * radius1;
    sy = y + sin(a+halfAngle) * radius1;
    vertex(sx, sy);
  }
  endShape(CLOSE);
}

void kill(){ //function to send stop message to Sonic Pi on local machine
  OscMessage myMessage = new OscMessage("/stop-all-jobs");
   myMessage.add("RBN_GUID"); //any value here. Need guid to make Sonic PI accept command
  oscP5.send(myMessage, myRemoteLocation); 
}

The basis of a visualiser depends on doing a fast fourier transform analysis of the incoming audio, and calculating amplitudes related to the different frequency components of the audio, which is continuously monitored input buffer by input buffer. I don’t intend togo anywhere near the complex mathematics involved, but there are a lot of useful articles at different levels which you can read on the subject. I quite liked this one on the Fourier Transform. https://betterexplained.com/articles/an-interactive-guide-to-the-fourier-transform/. Also it is useful to look at the documentation of the minim library analysis section http://code.compartmental.net/minim/index_analysis.html if you want more detailed information on the calls employed.
You may find it easier to look at the original sketch from which I started https://github.com/andrele/Starburst-Music-Viz before taking on board the additions I have added to make the sketch more flexible and wide ranging.
I have added various transformations. Starting at the beginning of the main draw() function there are x and y offsets which increase each time the loop iterates, until a maximum offset is reached, when they reverse in direction. This causes the shapes to move smoothly left and right and up and down the screen. a shiftflag eanbles this to be siwtched on and off by one of the control OSC messages.
There follows an optional rotate command which can rotate the axis of the shapes being drawn, again controlled by an incoming OSC message.
Next values for setting the limits on the stroke size being used to render the shapes are read from data received by an OSC message, together with an audioThreshold setting.
A buffer from the input audio is now processed, and amplitude values for different frequency ranges are stored in an array of “circle” settings. NB the name circles is used for the variable as this was the only shape used in the original.Perhaps it might be better names shapes() now as there are three different classes of shapes used. A new value for the “radius” is stored if it is less than the audioThreshold setting, otherwise, a “decayed” value of the currently stored value from the previous iteration is stored. (suitable minimum values are set).
rgb colour values are set using values received from an OSC message, and these are then swapped at random, before setting the stroke colour attributes  to be used.
Next a flag two is set according to whether  one or more than one shapes have been selected. In the latter case the shapes are forced to be drawn concentrically, by nullifying the offset values by setting two = 0. The selected shapes are then drawn, before the loop starts a further iteration.

The oscEvent(OscMessage msg) function is triggered when an OSC message is received. It contains sections to parse the various OSC messages which can be sent from Sonic PI to the processing sketch. The parameters for each command are used to update various lists containing information used by the draw function.eg cols[ ] holds rgb values, startdata[ ] holds the parameters for the star shapes, fvalues[ ] holds the floating point values for the STROKE_MAX, STROKE_MIN  and audioThresh settings. These and other similar settings are updated when the relevant OSC messages are received, so Sonic PI can control the operation of the sketch as it runs.
The star functions draws the star shape. It is lifted straight from the star example here https://processing.org/examples/star.html.
The final function kill( ) is used to send a /stop-all-jobs OSC message back to Sonic PI to stop all programs running on that machine. It can be triggered by a /viz/stop OSC message being sent from Sonic PI to the Sketch.

As far as Sonic Pi is concerned, it produces two sorts of input. First the audio that it produces is fed back to the sketch which looks at the default audio input. On my Mac I used the program Loopback ( https://rogueamoeba.com/loopback/ ) to perform this. This is a paid for program, but you can use it for free for 10 minutes or so at a time.It is based on the previous free SoundFlower utility, but this has not been fully updated for recent MacOS and you may find it difficult to get it to work instead. The setup I used is shown below:Note that Sonic Pi is added as a Source and that its output is monitored through the speakers, so that the audio is fed both there and to the processing sketch. This loopbackdevice will appear in the list of devices in the MIDI audio setup program, and should be selected as the default input device,

Secondly, Sonic PI is used to send OSC messages to the processing sketch to control its various parameters. this can be done by sending “one off” messages or by putting the message sender into a live loop, sending messages at regular intervals. An example is shown below, where the OSC messages are combined with the program producing the music, but they can  be run as a separate program in a different buffer, which is an advantage if you are visualising a linear piece and do not want to restart it every time you press run to update the OSC messages sent.

#Program to drive Sonic Pi 3 visualiser written in "processing"
#by Robin Newman, September 2017
#see article at https://rbnrpi.wordpress.com
#set up OSC address of processing sketch
use_osc '127.0.0.1',5000
#select shapes to show
osc "/viz/shapes","e"  #"s" "e" "r" Star,Ellipse, Rectangle or combination
sleep 0.1

live_loop :c do
  #choose starting colour for shapes
  osc "/viz/col",rrand_i(0,64),rrand_i(128,255),rrand_i(0,255)
  sleep 0.1
end

live_loop :f do
  #set Stroke max min widths and audioThreshold
  osc "/viz/float",([8.0,5.0,3.0].choose),[1.0,2.0].choose,(0.4+rand(0.3))
  sleep 2
end

#set range of random positional offset (can be 0,0)
#automatically disabled when showng more than one shape
osc "/viz/pos",20,0

#control "bouncing" shapes around the screen 1 for on 0 for off
osc "/viz/shift",0

live_loop :s do
  #setup star data inner/outer circle radius, number of points
  #and random variation of number of points
  osc "/viz/stardata",[1,2].choose,[1,2,3].choose,5,2
  sleep 4
end

rv=0 #variable for current rotation
live_loop :r do
  rv+=5*[-8,1].choose # choose rotation increment
  rv=rv%360
  osc "/viz/rotval",rv #change rv to 0 to disable rotation
  sleep 0.1
end

#Now setup the sounds to play which will trigger the visualiser
use_bpm 60
set_volume! 5
use_random_seed 999

with_fx :level do |v|
  control v,amp: 0 #control the volume using fx :level
  sleep 0.1
  
  in_thread do #this loop does the volume control
    control v,amp: 1,amp_slide: 10 #fade in
    sleep 140
    control v,amp: 0,amp_slide: 10 #fade out
    sleep 10
    osc "/viz/stop" #send /viz/stop OSC message to sketch
    #sketch sends back a /stop_all_jobs command to port 4557
  end
  
  #This drum loop is written by Eli see https://groups.google.com/forum/#!topic/sonic-pi/u71MnHnmkVY
  #used with his permission. I liked it, and it has good percussive output
  #to drive a visualiser
  live_loop :drums do
    this_sample = [:loop_compus, :loop_tabla, :loop_safari].ring
    start = [ 0.0 , 0.125 , 0.25 , 0.375 , 0.5 , 0.625 , 0.75 , 0.875 ].ring
    sample this_sample.look , beat_stretch: 4, start: start.look, rate: 0.5
    sleep 1
    tick
  end
end

As this program runs, you can alter the parameters eg shape(s) chosen, shift parameter etc. and rerun. Effectively live coding the visualiser. The program uses a great percussion loop written by Eli, which he said I might use. https://groups.google.com/forum/#!topic/sonic-pi/u71MnHnmkVY

There is a video of the program in operation on youtube here

and you can download the code for the sketch and Sonic Pi programs here

You can download and install Processing from https://processing.org/download/

You will also have to install the libraries Minim and oscP5 from the Sketch=>Import Library… menu.
To use, set up the loopback audio and select it as the default input device. Load and play the sketch in processing, then run the Sonic Pi program on the same computer. You can adjust the transparency of the Sonic Pi screen on the Preferences Visuals tab to make it semi transparent, or minimise Sonic Pi once it is running, if you are not going to do live coding of the control OSC messages.

Advertisements

Sonic-Pi controlled conversation between two McRoboFaces

IMG_4843

Following on from my previous project with a single McRoboFace, 4Tronix have kindly supplied me with a second face to enable me to develop the idea to control two McRoboFaces with Sonic Pi. I have amended the previous project to feed the outputs of the left and right audio channels to two separate adc inputs on the piconzero board, and daisy chained the two McFaces (you merely connect the Dout pin of the first to the Din pin of the second) and then address the leds on the second McRoboFace with an offset of 17. I have developed routines in the python driver program to control each face separately. Each mouth can be set to a fixed position: closed, open, smile or sad, or can be fed from the audio input via the adc, so that it is triggered to open when the signal exceeds a preset threshold.

In order to provide greater control, and to synchronise it to the audio feed from Sonic Pi, I have added Ruby routines to the Sonic Pi program which can send text strings to the python program via a text file. These strings can set the mouth state for each face, and also alter the colours of the leds. because there is only a common brightness setting for both faces (using pwm) If only one face is receiving audio I use that output to control the brightness of both faces. If both faces are set to receive audio then I set the brightness at a fixed value.

The conversation is entirely controlled from Sonic Pi. It plays the audio for each face via a series of pre-recorded samples, and plays each face with a separate audio channel by setting the pan: value either to -1 or to 1. Before each sample is played, control signals are sent via the text file to set up the required state for each face. At the end of the presentation both faces receive audio input together as they “sing” along to a round of Frere Jaques. Finally a control signal is sent to reduce the brightness to zero, effectively switching off all the leds.

Writing and reading the data via a text file is perhaps not the most elegant way to do things, but does seem to work OK. I used a technique I developed previously when reading in large numbers of sample files to “hold up” the Sonic Pi program utilising a cue and sync while the writing completes. Otherwise you can run into “too far behind errors”. On the reception side, at the start of the main program loop the python program polls for the existence of the text file, and if it finds one, reads the data, then deletes the file. It then alters parameters according to the received data. It took quite a lot of experimentation to get the timings and consistent operation of the two programs correct, but having done so, the final system is quite stable. I boost the audio levels to amp: 4 in Sonic Pi, which gives a good signal for the adc inputs to latch on to.

Setup is fairly straight forward. The calibrate button used in the single face project us utilised again, and sets separate offsets for each channel, and the code used to modulate the mouths is very similar to that used in the previous project. Once set, the Sonic Pi program can be run several times, leaving the python program running continuously..

I have enjoyed this project, which had brought together Sonic Pi, Ruby and Python in an interesting way, not to mention recording and processing the samples with Audacity., and I hope you enjoy the video of the final system. I hope it may be possible to write up teh system more fully in the future, but it will be quite a big job to do so.

You can see the video here

 

Sonic Pi driven Sound Bar Graph built on RasPiO ProHat

SPBarGraphMainLayout

Ive recently added an article describing how to use the RasPiO ProHat to build a 4 led bar graph which can be driven by Sonic Pi. It also uses components from RasPiO’s Experimenter’s Kit. The article contains full constructional details with links to the python program, and videos showing the construction and use of the project.

Read the article here

McRoboFace project to singalong with Sonic Pi

IMG_4773

Recently I was sent a pre-production version of 4Tronix McRoboFace, which is a small face whose features light up with 17 Neopixel leds. It is ideally suited as a companion to their Picon Zero controller board, which already has software in its library to accommodate it. I experimented with the item, and the result was a project to produce a talking/singing face, which could respond to an audio input either fed from a microphone, or internally from a raspberry pi running Sonic Pi.

An article here gives full constructional details and links to the software used.

4Tronix hope to launch a kickstarter project for the McRoboFace soon. If you support it, this will give you a nice project with which to use it.

video link to the project in action, explaining itself!

PS3 controlled Edukit robot

ps3robot

Edukit 3 robot kit, revamped with 4Tronix picon zero board and python program to control it with a ps3 wireless controller. The pizero has the dongle for the ps3 controller in its single usb port, and is arranged to boot automatically into the python program driving the robot. A red led lights when the system is ready for action. The left hand joystick is sensed and used to control the motor speed and direction. A button on the controller is sensed to initiate shutdown when finished, so the whole process is automatic and doesn’t require any keyboard, mouse or screen. software is availabe for download at:
https://gist.github.com/rbnpi/ee3b60f200a4ef9b927d2faa0241f7b0

video of the robot in operation is here