Addressable RGB LED Music and Sound Visualizer

Contributors: Michael Bartlett
Favorited Favorite 9


Let’s face it: nowadays, most musical performances are complimented by some fancy light shows. Go to any concert, festival, club – they all have a corresponding visual performance or effects. Why not add your own home to that list? Here’s a simple yet effective project to make your very own son et lumière!

All palettes work with every visualization, but, for timeliness, not every combination is shown.

Required Materials

To follow along with this tutorial, you’ll need the following:

Any microcontroller with 3.3V and 5V pins will suffice, any analog potentiometer should work, and any resistor between 300–500 Ω can be used. The resistor and capacitor are not required, but they will help prevent possible damage to the LEDs.

If you’re compiling from the Arduino IDE or similar, you’ll want to snag the the NeoPixel Library since the code used is heavily based on it.

Depending on your intent, the trimpot and buttons may not be necessary. The trimpot is only used to adjust the brightness threshold, so, if you want maximum brightness, you don’t have to worry about incorporating it. The three buttons cycle visualizations, color schemes, and shuffle mode respectively, so, if you want to do without those features (and just use shuffle mode all the time), that’s also a possibility.

It is also suggested that you use an Arduino and Breadboard Holder to simplify wiring and to mount the LED strip:

Full Visualizer

A small notch was cut in the BReadboard Holder to hold a piece of MDF, on which the LEDs are attached.

Recommended Reading

Before embarking upon this tutorial, you may find the following links useful:

Since we’re using the NeoPixel library, it may also be a good idea to get familiar with the NeoPixel Documentation.


This project requires virtually no soldering! The few exceptions will probably be soldering some pins to the sound detector, and, if you’ve cut a roll of addressable LEDs in the middle, you’ll have to solder some wires to the starting LED’s pins. If you’ve never soldered before, I highly suggest taking a look at this guide.

Below is also a general chart for how the pin(s) on each component should be routed and an accompanying diagram. Before you begin, here are some things to keep in mind:

  • Be conscious of the orientation you think would allow the sound detector to take optimal readings for your intentions. Bending the pins to hold the sound detector perpendicular to the breadboard is a recommended option.
  • Electrolytic capacitors are polarized, so how they are oriented is important. Make sure to place the side with a white stripe and a negative symbol into a negative current (ground) and the other into positive current.
  • Resistors and pushbuttons are not polarized.
  • Trimpots are not polarized either, however their middle pin is the analog output, so don’t power that directly.

The pins used in the diagram and the code are in parentheses. If you use a different pin, don’t forget to change it in the code as well:

Sound DetectorAddressable LED stripTrimpotPushbutton1 mF (1000 µF) Capacitor300–500 Ω Resistor
Envelope → Analog (A0)Digital/Analog (A5) → Resistor → DIN5V → left or right pinGND → Either sideBetween ground and 5VBetween Digital/Analog (A5) and DIN on LED strip
3.3V → VCC5V →5V>Middle pin → Analog (A1)Other side → Digital (4, 5, 6)
Remaining left or right pin → GND

alt text

Sound Detector Notes: The microphone used is not a sophisticated, logarithmic sound receiver like your ear; it is only measuring compressional waves in the air. Consequently, the microphone is more likely to detect and/or prioritize lower-frequency sounds since they require more energy to propagat, and therefore oscillate the air more intensely. Also, a resistor can be placed in the “GAIN” slots to modify the gain. Standard gain should be sufficient for our purposes, but, for more info, visit this tutorial.

The entire circuit should look something like the diagram below.

alt text

Fritzing diagram for the circuit as described above. Click the image for a closer look.


Below is a small sample program to test if everything is connected properly. It only contains one visualizer and one color palette to keep things concise. It doesn’t need buttons since there’s nothing to toggle, but you can still use it to test your potentiometer.

The complete program featured in the video can be found at this GitHub repository.

Things to remember before you compile:

  • If you didn’t use a potentiometer, don’t forget to remove all references to the variable knob in the code (ctrl+F will come in handy for that). Otherwise, the program will think you still have a potentiometer that is set to a very low value (i.e. everything will be very dim).
  • If you didn’t use buttons, change the initialization bool shuffle = false; to bool shuffle = true;. The code should compile and run properly, but for good practice you should remove all blocks the code says to delete since they reference the BUTTON constants.

Note: This example assumes you are using the latest version of the Arduino IDE on your desktop. If this is your first time using Arduino, please review our tutorial on installing the Arduino IDE.

If you have not previously installed an Arduino library, please check out our installation guide.

//Program by Michael Bartlett

#include <Adafruit_NeoPixel.h>  //Library to simplify interacting with the LED strand
#ifdef __AVR__
#include <avr/power.h>   //Includes the library for power reduction registers if your chip supports them. 
#endif                   //More info:

//Constants (change these as necessary)
#define LED_PIN   A5  //Pin for the pixel strand. Does not have to be analog.
#define LED_TOTAL 36  //Change this to the number of LEDs in your strand.
#define AUDIO_PIN A0  //Pin for the envelope of the sound detector
#define KNOB_PIN  A1  //Pin for the trimpot 10K

//  These values either need to be remembered from the last pass of loop() or 
//  need to be accessed by several functions in one pass, so they need to be global.

Adafruit_NeoPixel strand = Adafruit_NeoPixel(LED_TOTAL, LED_PIN, NEO_GRB + NEO_KHZ800);  //LED strand objetc

uint16_t gradient = 0; //Used to iterate and loop through each color palette gradually

uint8_t volume = 0;    //Holds the volume level read from the sound detector.
uint8_t last = 0;      //Holds the value of volume from the previous loop() pass.

float maxVol = 15;     //Holds the largest volume recorded thus far to proportionally adjust the visual's responsiveness.
float knob = 1023.0;   //Holds the percentage of how twisted the trimpot is. Used for adjusting the max brightness.
float avgVol = 0;      //Holds the "average" volume-level to proportionally adjust the visual experience.
float avgBump = 0;     //Holds the "average" volume-change to trigger a "bump."

bool bump = false;     //Used to pass if there was a "bump" in volume


//////////<Standard Functions>

void setup() {    //Like it's named, this gets ran before any other function.

  Serial.begin(9600); //Sets data rate for serial data transmission.

  strand.begin(); //Initialize the LED strand object.;  //Show a blank strand, just to get the LED's ready for use.  

void loop() {  //This is where the magic happens. This loop produces each frame of the visual.
  volume = analogRead(AUDIO_PIN);       //Record the volume level from the sound detector
  knob = analogRead(KNOB_PIN) / 1023.0; //Record how far the trimpot is twisted
  avgVol = (avgVol + volume) / 2.0;     //Take our "average" of volumes.

  //Sets a threshold for volume.
  //  In practice I've found noise can get up to 15, so if it's lower, the visual thinks it's silent.
  //  Also if the volume is less than average volume / 2 (essentially an average with 0), it's considered silent.
  if (volume < avgVol / 2.0 || volume < 15) volume = 0;

  //If the current volume is larger than the loudest value recorded, overwrite
  if (volume > maxVol) maxVol = volume;

  //This is where "gradient" is reset to prevent overflow.
  if (gradient > 1529) {

    gradient %= 1530;

    //Everytime a palette gets completed is a good time to readjust "maxVol," just in case
    //  the song gets quieter; we also don't want to lose brightness intensity permanently 
    //  because of one stray loud sound.
    maxVol = (maxVol + volume) / 2.0;

  //If there is a decent change in volume since the last pass, average it into "avgBump"
  if (volume - last > avgVol - last && avgVol - last > 0) avgBump = (avgBump + (volume - last)) / 2.0;

  //if there is a notable change in volume, trigger a "bump"
  bump = (volume - last) > avgBump;

  Pulse();   //Calls the visual to be displayed with the globals as they are.

  gradient++;    //Increments gradient

  last = volume; //Records current volume for next pass

  delay(30);   //Paces visuals so they aren't too fast to be enjoyable

//////////</Standard Functions>

//////////<Helper Functions>

//Pulse from center of the strand
void Pulse() {

  fade(0.75);   //Listed below, this function simply dims the colors a little bit each pass of loop()

  //Advances the gradient to the next noticeable color if there is a "bump"
  if (bump) gradient += 64;

  //If it's silent, we want the fade effect to take over, hence this if-statement
  if (volume > 0) {
    uint32_t col = Rainbow(gradient); //Our retrieved 32-bit color

    //These variables determine where to start and end the pulse since it starts from the middle of the strand.
    //  The quantities are stored in variables so they only have to be computed once.
    int start = LED_HALF - (LED_HALF * (volume / maxVol));
    int finish = LED_HALF + (LED_HALF * (volume / maxVol)) + strand.numPixels() % 2;
    //Listed above, LED_HALF is simply half the number of LEDs on your strand. ↑ this part adjusts for an odd quantity.

    for (int i = start; i < finish; i++) {

      //"damp" creates the fade effect of being dimmer the farther the pixel is from the center of the strand.
      //  It returns a value between 0 and 1 that peaks at 1 at the center of the strand and 0 at the ends.
      float damp = float(
                     ((finish - start) / 2.0) -
                     abs((i - start) - ((finish - start) / 2.0))
                   / float((finish - start) / 2.0);

      //Sets the each pixel on the strand to the appropriate color and intensity
      //  strand.Color() takes 3 values between 0 & 255, and returns a 32-bit integer.
      //  Notice "knob" affecting the brightness, as in the rest of the visuals.
      //  Also notice split() being used to get the red, green, and blue values.
      strand.setPixelColor(i, strand.Color(
                             split(col, 0) * pow(damp, 2.0) * knob,
                             split(col, 1) * pow(damp, 2.0) * knob,
                             split(col, 2) * pow(damp, 2.0) * knob
    //Sets the max brightness of all LEDs. If it's loud, it's brighter.
    //  "knob" was not used here because it occasionally caused minor errors in color display.
    strand.setBrightness(255.0 * pow(volume / maxVol, 2));

  //This command actually shows the lights. If you make a new visualization, don't forget this!;

//Fades lights by multiplying them by a value between 0 and 1 each pass of loop().
void fade(float damper) {

  //"damper" must be between 0 and 1, or else you'll end up brightening the lights or doing nothing.
  if (damper >= 1) damper = 0.99;

  for (int i = 0; i < strand.numPixels(); i++) {

    //Retrieve the color at the current position.
    uint32_t col = (strand.getPixelColor(i)) ? strand.getPixelColor(i) : strand.Color(0, 0, 0);

    //If it's black, you can't fade that any further.
    if (col == 0) continue;

    float colors[3]; //Array of the three RGB values

    //Multiply each value by "damper"
    for (int j = 0; j < 3; j++) colors[j] = split(col, j) * damper;

    //Set the dampened colors back to their spot.
    strand.setPixelColor(i, strand.Color(colors[0] , colors[1], colors[2]));

uint8_t split(uint32_t color, uint8_t i ) {

  //0 = Red, 1 = Green, 2 = Blue

  if (i == 0) return color >> 16;
  if (i == 1) return color >> 8;
  if (i == 2) return color >> 0;
  return -1;

//This function simply take a value and returns a gradient color
//  in the form of an unsigned 32-bit integer

//The gradient returns a different, changing color for each multiple of 255
//  This is because the max value of any of the 3 LEDs is 255, so it's
//  an intuitive cutoff for the next color to start appearing.
//  Gradients should also loop back to their starting color so there's no jumps in color.

uint32_t Rainbow(unsigned int i) {
  if (i > 1529) return Rainbow(i % 1530);
  if (i > 1274) return strand.Color(255, 0, 255 - (i % 255));   //violet -> red
  if (i > 1019) return strand.Color((i % 255), 0, 255);         //blue -> violet
  if (i > 764) return strand.Color(0, 255 - (i % 255), 255);    //aqua -> blue
  if (i > 509) return strand.Color(0, 255, (i % 255));          //green -> aqua
  if (i > 255) return strand.Color(255 - (i % 255), 255, 0);    //yellow -> green
  return strand.Color(255, i, 0);                               //red -> yellow

//////////</Helper Functions>

Final Touches

With the electronics and the code working, you can now add your visualizer to a variety of enclosures or art pieces. For the final touches on this project, an elk was laser etched on a piece of acrylic. The LED strip was then wrapped around the outer perimeter of the piece.


Add some music, and you have yourself a beautiful piece of interactive art.

Resources and Going Further

To see all the code used int his project, visit the GitHub repository.

Need more inspiration? Check out these other SparkFun tutorials:

WAV Trigger Hookup Guide V11

An introduction to being able to trigger music and sound effects based on buttons, sensors, or switches using the WAV Trigger board.

Addressable RGB LED Music and Sound Visualizer

Use an Arduino and the SparkFun Sound Detector to create visualizations on Addressable RGB LED strips.

Proto Pedal Chassis Hookup Guide

Prepare the Proto Pedal Chassis by drilling holes for controls and painting it.

THAT InGenuis and OutSmarts Breakout Hookup Guide

Learn about the benefits of balanced signal transmission, and how to apply the THAT InGenius and OutSmarts breakouts.