Generating mp3 waveform with php

Question

I need to build a web script that converts a mp3 file to a waveform image, and I came across this website.

I’ve uploaded the PHP file and .dll and .exe file from LAMB to my website (hosted by host monster).

But when I try uploading the MP3 file on the website, it just generates a black image instead of the waveform image.

I have little knowledge of PHP, so I don’t exactly know what’s wrong, but I’m guessing it has to do with the uploading part of the program.

Could anybody tell me what is wrong and how I can get over it?

The following is the php code as is (also available in this repository on GitHub).

  <?php

  ini_set("max_execution_time", "30000");

  // how much detail we want. Larger number means less detail
  // (basically, how many bytes/frames to skip processing)
  // the lower the number means longer processing time
  define("DETAIL", 5);

  define("DEFAULT_WIDTH", 500);
  define("DEFAULT_HEIGHT", 100);
  define("DEFAULT_FOREGROUND", "#FF0000");
  define("DEFAULT_BACKGROUND", "#FFFFFF");

  /**
   * GENERAL FUNCTIONS
   */
  function findValues($byte1, $byte2){
    $byte1 = hexdec(bin2hex($byte1));                        
    $byte2 = hexdec(bin2hex($byte2));                        
    return ($byte1 + ($byte2*256));
  }

  /**
   * Great function slightly modified as posted by Minux at
   * http://forums.clantemplates.com/showthread.php?t=133805
   */
  function html2rgb($input) {
    $input=($input[0]=="#")?substr($input, 1,6):substr($input, 0,6);
    return array(
     hexdec(substr($input, 0, 2)),
     hexdec(substr($input, 2, 2)),
     hexdec(substr($input, 4, 2))
    );
  }   

  if (isset($_FILES["mp3"])) {

/**
 * PROCESS THE FILE
 */

// temporary file name
$tmpname = substr(md5(time()), 0, 10);

// copy from temp upload directory to current
copy($_FILES["mp3"]["tmp_name"], "{$tmpname}_o.mp3");

    // support for stereo waveform?
$stereo = isset($_POST["stereo"]) && $_POST["stereo"] == "on" ? true : false;

    // array of wavs that need to be processed
$wavs_to_process = array();

/**
 * convert mp3 to wav using lame decoder
 * First, resample the original mp3 using as mono (-m m), 16 bit (-b 16), and 8 KHz (--resample 8)
 * Secondly, convert that resampled mp3 into a wav
 * We don't necessarily need high quality audio to produce a waveform, doing this process reduces the WAV
 * to it's simplest form and makes processing significantly faster
 */
if ($stereo) {
        // scale right channel down (a scale of 0 does not work)
  exec("lame {$tmpname}_o.mp3 --scale-r 0.1 -m m -S -f -b 16 --resample 8 {$tmpname}.mp3 && lame -S --decode {$tmpname}.mp3 {$tmpname}_l.wav");
        // same as above, left channel
  exec("lame {$tmpname}_o.mp3 --scale-l 0.1 -m m -S -f -b 16 --resample 8 {$tmpname}.mp3 && lame -S --decode {$tmpname}.mp3 {$tmpname}_r.wav");
  $wavs_to_process[] = "{$tmpname}_l.wav";
  $wavs_to_process[] = "{$tmpname}_r.wav";
} else {
  exec("lame {$tmpname}_o.mp3 -m m -S -f -b 16 --resample 8 {$tmpname}.mp3 && lame -S --decode {$tmpname}.mp3 {$tmpname}.wav");
  $wavs_to_process[] = "{$tmpname}.wav";
}

// delete temporary files
unlink("{$tmpname}_o.mp3");
unlink("{$tmpname}.mp3");

// get user vars from form
$width = isset($_POST["width"]) ? (int) $_POST["width"] : DEFAULT_WIDTH;
$height = isset($_POST["height"]) ? (int) $_POST["height"] : DEFAULT_HEIGHT;
$foreground = isset($_POST["foreground"]) ? $_POST["foreground"] : DEFAULT_FOREGROUND;
$background = isset($_POST["background"]) ? $_POST["background"] : DEFAULT_BACKGROUND;
$draw_flat = isset($_POST["flat"]) && $_POST["flat"] == "on" ? true : false;

$img = false;

// generate foreground color
list($r, $g, $b) = html2rgb($foreground);

// process each wav individually
for($wav = 1; $wav <= sizeof($wavs_to_process); $wav++) {

  $filename = $wavs_to_process[$wav - 1];

  /**
   * Below as posted by "zvoneM" on
   * http://forums.devshed.com/php-development-5/reading-16-bit-wav-file-318740.html
   * as findValues() defined above
   * Translated from Croation to English - July 11, 2011
   */
  $handle = fopen($filename, "r");
  // wav file header retrieval
  $heading[] = fread($handle, 4);
  $heading[] = bin2hex(fread($handle, 4));
  $heading[] = fread($handle, 4);
  $heading[] = fread($handle, 4);
  $heading[] = bin2hex(fread($handle, 4));
  $heading[] = bin2hex(fread($handle, 2));
  $heading[] = bin2hex(fread($handle, 2));
  $heading[] = bin2hex(fread($handle, 4));
  $heading[] = bin2hex(fread($handle, 4));
  $heading[] = bin2hex(fread($handle, 2));
  $heading[] = bin2hex(fread($handle, 2));
  $heading[] = fread($handle, 4);
  $heading[] = bin2hex(fread($handle, 4));

  // wav bitrate 
  $peek = hexdec(substr($heading[10], 0, 2));
  $byte = $peek / 8;

  // checking whether a mono or stereo wav
  $channel = hexdec(substr($heading[6], 0, 2));

  $ratio = ($channel == 2 ? 40 : 80);

  // start putting together the initial canvas
  // $data_size = (size_of_file - header_bytes_read) / skipped_bytes + 1
  $data_size = floor((filesize($filename) - 44) / ($ratio + $byte) + 1);
  $data_point = 0;

  // now that we have the data_size for a single channel (they both will be the same)
  // we can initialize our image canvas
  if (!$img) {
    // create original image width based on amount of detail
            // each waveform to be processed with be $height high, but will be condensed
            // and resized later (if specified)
    $img = imagecreatetruecolor($data_size / DETAIL, $height * sizeof($wavs_to_process));

    // fill background of image
    if ($background == "") {
      // transparent background specified
      imagesavealpha($img, true);
      $transparentColor = imagecolorallocatealpha($img, 0, 0, 0, 127);
      imagefill($img, 0, 0, $transparentColor);
    } else {
      list($br, $bg, $bb) = html2rgb($background);
      imagefilledrectangle($img, 0, 0, (int) ($data_size / DETAIL), $height * sizeof($wavs_to_process), imagecolorallocate($img, $br, $bg, $bb));
    }
  }

  while(!feof($handle) && $data_point < $data_size){
    if ($data_point++ % DETAIL == 0) {
      $bytes = array();

      // get number of bytes depending on bitrate
      for ($i = 0; $i < $byte; $i++)
        $bytes[$i] = fgetc($handle);

      switch($byte){
        // get value for 8-bit wav
        case 1:
          $data = findValues($bytes[0], $bytes[1]);
          break;
        // get value for 16-bit wav
        case 2:
          if(ord($bytes[1]) & 128)
            $temp = 0;
          else
            $temp = 128;
          $temp = chr((ord($bytes[1]) & 127) + $temp);
          $data = floor(findValues($bytes[0], $temp) / 256);
          break;
      }

      // skip bytes for memory optimization
      fseek($handle, $ratio, SEEK_CUR);

      // draw this data point
      // relative value based on height of image being generated
      // data values can range between 0 and 255
      $v = (int) ($data / 255 * $height);

      // don't print flat values on the canvas if not necessary
      if (!($v / $height == 0.5 && !$draw_flat))
        // draw the line on the image using the $v value and centering it vertically on the canvas
        imageline(
          $img,
          // x1
          (int) ($data_point / DETAIL),
          // y1: height of the image minus $v as a percentage of the height for the wave amplitude
          $height * $wav - $v,
          // x2
          (int) ($data_point / DETAIL),
          // y2: same as y1, but from the bottom of the image
          $height * $wav - ($height - $v),
          imagecolorallocate($img, $r, $g, $b)
        );         

    } else {
      // skip this one due to lack of detail
      fseek($handle, $ratio + $byte, SEEK_CUR);
    }
  }

  // close and cleanup
  fclose($handle);

  // delete the processed wav file
  unlink($filename);

}

header("Content-Type: image/png");

// want it resized?
if ($width) {
  // resample the image to the proportions defined in the form
  $rimg = imagecreatetruecolor($width, $height);
  // save alpha from original image
  imagesavealpha($rimg, true);
  imagealphablending($rimg, false);
  // copy to resized
  imagecopyresampled($rimg, $img, 0, 0, 0, 0, $width, $height, imagesx($img), imagesy($img));
  imagepng($rimg);
  imagedestroy($rimg);
} else {
  imagepng($img);
}

imagedestroy($img);

  } else {

?>

  <form method="post" action="<?php print $_SERVER["REQUEST_URI"]; ?>"     enctype="multipart/form-data">

  <p>MP3 File:<br />
    <input type="file" name="mp3" /></p>

  <p>Image Width:<br />
    <input type="text" name="width" value="<?php print DEFAULT_WIDTH; ?>" /></p>

  <p>Image Height:<br />
    <input type="text" name="height" value="<?php print DEFAULT_HEIGHT; ?>" /></p>

  <p>Foreground Color: <small>(HEX/HTML color code)</small><br />
    <input type="text" name="foreground" value="<?php print DEFAULT_FOREGROUND; ?>" /></p>

  <p>Background Color: (Leave blank for transparent background) <small>(HEX/HTML color code)</small><br />
    <input type="text" name="background" value="<?php print DEFAULT_BACKGROUND; ?>" /></p>

  <p>Draw flat-line? <input type="checkbox" name="flat" /></p>

  <p>Stereo waveform? <input type="checkbox" name="stereo" /></p>

  <p><input type="submit" value="Generate Waveform" /></p>

  </form>

<?php

  }    

Answer 1

Have a look at this here, it is a PHP only implementation of what you are trying to do: http://www.phpclasses.org/package/482-PHP-Extracts-attributes-of-audio-files-Visualization.html

Answer 2

If for whatever reason this script is unable to open any one of the audio files, it will fail and produce a black box. So assuming the host isn't blocking exec(), and depending on your server configuration, it could be an issues with the paths. You might want to try one or both of the following:

• Use the full path to the 'lame' executable in the exec() calls
• Use the full path for the audio files, in the exec() calls and elsewhere in your script

exec("/usr/local/bin/lame /path/to/{$tmpname}_o.mp3 -m m -S -f -b 16 --resample 8 /path/to/{$tmpname}.mp3 && /usr/local/bin/lame -S --decode /path/to/{$tmpname}.mp3 {$tmpname}.wav");

Answer 3

I've found good library for doing just that: JustWave Project All mentioned above is true do...

1. You need an executable library to generate the waveimage on serverside. This script uses FFMEG
2. For cache this library uses SQLite
3. You clould generate your images and filength data locally and then upload if you don't have FFMEG on actual sever.
4. Nice simple output, player included, color tweakable.

Hope this helps someone with the same needs.