audio_signal_handling.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Madmom tutorial\n",
    "\n",
    "This tutorial explains some of the basic design patterns of `madmom`.\n",
    "\n",
    "Before we start, we have to import some packages:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import madmom"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Functions / classes\n",
    "\n",
    "One of the basic design principles of `madmom` is to provide the same functionality both on functions as well as on class level. The classes provide the same functionality but at the next higher abstraction layer. The main differences are:\n",
    "\n",
    "* functions can return multiple values or a tuple\n",
    "* classes encapsulate these additional values as attributes\n",
    "* functions usually operate on and return numpy arrays\n",
    "* classes are usually subclasses of numpy arrays, i.e. they can be used like numpy arrays\n",
    "* functions require the input data to be of a specific type\n",
    "* classes can handle various input types\n",
    "\n",
    "All these features will be discussed in this tutorial.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading an audio file\n",
    "\n",
    "Since `madmom` is a library with a strong focus on music information retrival (MIR), let's first load an audio file.\n",
    "The `load_wave_file` function returns a memory mapped numpy array of the audio file and the sample_rate as an integer value."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(memmap([-2494, -2510, -2484, ...,   520,   655,   639], dtype=int16), 44100)"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "signal, sample_rate = madmom.audio.signal.load_wave_file('data/sample.wav')\n",
    "signal, sample_rate"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Classes provide the same functionality as functions but add some abstraction\n",
    "A `Signal` object is basically the same as the returned `signal` array but it is enhanced with the the `sample_rate` as an attribute:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(Signal([-2494, -2510, -2484, ...,   520,   655,   639], dtype=int16), 44100)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sig = madmom.audio.signal.Signal('data/sample.wav')\n",
    "sig, sig.sample_rate"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The class uses exactly the same function to load the aufio file, but it encapsulates the sample rate as an attribute of the object. Since the sample rate is important for a lot of consecutive calculations (e.g. to determine the bin frequencies of a STFT) it is necessary to keep it around, but when using the `Signal` class, there is no need to pass it to all the functions which depend on the sample rate of the signal as a separate variable. The `Signal` class takes care of this automatically (or: the other classes are smart enough to access it correctly).\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Classes are subclasses of numpy arrays\n",
    "The `Signal` class is a subclass of a numpy array, i.e. it can be used exactly the same way as any numpy array. \n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Framing an audio signal\n",
    "\n",
    "We can then split the signal into overlapping frames. Again, there is both a function and a class performing the same task. Since we consider the classes the more convenient way, we use only them in the ongoing. A look into the source code helps you to discover the equivalent function :)\n",
    "\n",
    "To split the signal into frames of 2048 samples length which overlap by 441 samples, we use the existing `Signal` object as the first argument for instantiation of the `FramedSignal` object."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<madmom.audio.signal.FramedSignal at 0x113362128>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fs = madmom.audio.signal.FramedSignal(sig, frame_size=2048, hop_size=441)\n",
    "fs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### If classes are not subclasses of numpy arrays, they behave similar\n",
    "\n",
    "The `FramedSignal` is implemented as an indexable/iterable class, so we can access the individual frames exactly like a numpy array, e.g. slice the signal. The only difference is that the index -1 does not refer to the last frame, but to the frame left of the first one.\n",
    "\n",
    "Note: the reason why the `FramedSignal` is implemented this way and not as a strided view on the signal numpy array is that we also want to be able to use float hop sizes, e.g. to get a constant frame rate. E.g. to get a frame rate of 200, we could either set the `hop_size` to 220.5 samples or simply pass the `fps` argument instead.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<madmom.audio.signal.FramedSignal at 0x113362048>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fs_200 = madmom.audio.signal.FramedSignal(sig, frame_size=2048, fps=200)\n",
    "fs_200"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(100.0,\n",
       " 441.0,\n",
       " Signal([    0,     0,     0, ..., -4553, -4666, -4589], dtype=int16))"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fs.frame_rate, fs.hop_size, fs[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(200.0,\n",
       " 220.5,\n",
       " Signal([    0,     0,     0, ..., -4553, -4666, -4589], dtype=int16))"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fs_200.frame_rate, fs_200.hop_size, fs_200[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The `fps` argument is converted into the appropriate `hop_size`. This conversion could be performed automatically since the `Signal` object has also the needed the sample rate. Thus constant frame rates could be achieved independently of the sample rate of the signal. No need to resample all audio files to the same sample rate beforehand."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Automatic object instantiation\n",
    "Another design feature of the classes is that if they are instantiated with an argument which they don't expect, they try to create an instance of what they expect. E.g. the `FramedSignal` expects a `Signal` instance as the first argument, but we can also pass a file name instead. The `FramedSignal` then instantiates a `Signal` object:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "fs_ = madmom.audio.signal.FramedSignal('data/sample.wav')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A reference to the automatically instantiated `Signal` object is kept as the `signal` attribute."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Signal([-2494, -2510, -2484, ...,   520,   655,   639], dtype=int16)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fs_.signal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(100.0,\n",
       " 441.0,\n",
       " Signal([    0,     0,     0, ..., -4553, -4666, -4589], dtype=int16))"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fs_.frame_rate, fs_.hop_size, fs_[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "All three ways of obtaining a `FramedSignal` instance are equivalent."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Complex STFT\n",
    "\n",
    "Based on this `FramedSignal` we can now continue to calculate the `STFT` of it. The `STFT` class is (of course) a subclass of a numpy array, i.e. it has a `dtype`, can be sliced, transposed etc.."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ShortTimeFourierTransform([[-3.15249181 +0.00000000e+00j,  2.62215710 -3.02424717e+00j,\n",
       "         0.96471846 +5.57876539e+00j, ...,  0.03642104 -3.61841870e-04j,\n",
       "        -0.03633529 -5.19269597e-05j,  0.03670426 +2.85875227e-04j],\n",
       "       [-4.28428888 +0.00000000e+00j,  2.02009392 +2.01263618e+00j,\n",
       "        -1.54068637 -8.30089092e+00j, ...,  0.01295261 -1.75991822e-02j,\n",
       "        -0.01980834 -9.33106802e-03j, -0.00536164 +2.16161963e-02j]], dtype=complex64)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stft = madmom.audio.stft.STFT(fs)\n",
    "stft[0:2]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Magnitude Spectrogram\n",
    "More often than the complex `STFT`, the magnitudes of it are used for MIR tasks. Magnitude spectrograms are represented by the `Spectrogram` class."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "spec = madmom.audio.spectrogram.Spectrogram(stft)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x113ed2be0>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAD7CAYAAAB68m/qAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n6tJVQv8fb7K/ugfhb5QAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113ad1f98>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(spec[:, :200].T, aspect='auto', origin='lower')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As before, although the `Spectrogram` class expects a `STFT` instance as its only argument, we could also have passed any other data to it as long as a `STFT` object can be instantiated from it. This can be a `FramedSignal` or `Signal` instance or just a file name (or file handle).\n",
    "\n",
    "Additional keywords are passed accordingly, i.e. we can get the magnitude spectrogram of a file with a frame size of 2048 samples, a hop size of 200 samples, a FFT size of 4096 samples simple as:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "spec = madmom.audio.spectrogram.Spectrogram('data/sample.wav', frame_size=2048, hop_size=200, fft_size=4096)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((618, 2048), array([  0.00000000e+00,   1.07666016e+01,   2.15332031e+01, ...,\n",
       "          2.20177002e+04,   2.20284668e+04,   2.20392334e+04]))"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spec.shape, spec.bin_frequencies"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Additionally, the `Spectrogram` class also keeps their underlying objects as attributes, e.g. the factor to which the individual frames overlap, can be accessed as if we go up the instantiation chain."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.90234375"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spec.stft.frames.overlap_factor"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For a complete list of attributes and methods of the classes, please refer to the API documentation. These are just the basic things needed to continue with other (higher level) tasks."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}