shithub: aubio

ref: 8dbff94cd09295e81aea4b8418cfe47db0739377
dir: /python/demos/demo_tempo_plot.py/

View raw version
#! /usr/bin/env python

import sys
from aubio import tempo, source

win_s = 512                 # fft size
hop_s = win_s / 2           # hop size

if len(sys.argv) < 2:
    print "Usage: %s <filename> [samplerate]" % sys.argv[0]
    sys.exit(1)

filename = sys.argv[1]

samplerate = 0
if len( sys.argv ) > 2: samplerate = int(sys.argv[2])

s = source(filename, samplerate, hop_s)
samplerate = s.samplerate
o = tempo("default", win_s, hop_s, samplerate)

# tempo detection delay, in samples
# default to 4 blocks delay to catch up with
delay = 4. * hop_s

# list of beats, in samples
beats = []

# total number of frames read
total_frames = 0
while True:
    samples, read = s()
    is_beat = o(samples)
    if is_beat:
        this_beat = o.get_last_s()
        beats.append(this_beat)
    total_frames += read
    if read < hop_s: break

if len(beats) > 1:
    # do plotting
    from numpy import array, arange, mean, median, diff
    import matplotlib.pyplot as plt
    bpms = 60./ diff(beats)
    print 'mean period:', "%.2f" % mean(bpms), 'bpm', 'median', "%.2f" % median(bpms), 'bpm'
    print 'plotting', filename
    plt1 = plt.axes([0.1, 0.75, 0.8, 0.19])
    plt2 = plt.axes([0.1, 0.1, 0.8, 0.65], sharex = plt1)
    plt.rc('lines',linewidth='.8')
    for stamp in beats: plt1.plot([stamp, stamp], [-1., 1.], '-r')
    plt1.axis(xmin = 0., xmax = total_frames / float(samplerate) )
    plt1.xaxis.set_visible(False)
    plt1.yaxis.set_visible(False)

    # plot actual periods
    plt2.plot(beats[1:], bpms, '-', label = 'raw')

    # plot moving median of 5 last periods
    median_win_s = 5
    bpms_median = [ median(bpms[i:i + median_win_s:1]) for i in range(len(bpms) - median_win_s ) ]
    plt2.plot(beats[median_win_s+1:], bpms_median, '-', label = 'median of %d' % median_win_s)
    # plot moving median of 10 last periods
    median_win_s = 20
    bpms_median = [ median(bpms[i:i + median_win_s:1]) for i in range(len(bpms) - median_win_s ) ]
    plt2.plot(beats[median_win_s+1:], bpms_median, '-', label = 'median of %d' % median_win_s)

    plt2.axis(ymin = min(bpms), ymax = max(bpms))
    #plt2.axis(ymin = 40, ymax = 240)
    plt.xlabel('time (mm:ss)')
    plt.ylabel('beats per minute (bpm)')
    plt2.set_xticklabels([ "%02d:%02d" % (t/60, t%60) for t in plt2.get_xticks()[:-1]], rotation = 50)

    #plt.savefig('/tmp/t.png', dpi=200)
    plt2.legend()
    plt.show()

else:
    print 'mean period:', "%.2f" % 0, 'bpm', 'median', "%.2f" % 0, 'bpm',
    print 'nothing to plot, file too short?'