shithub: aubio

Download patch

ref: 7b7a58e087d43e0586b52412034bc6fd87e60f71
parent: b99e2a51435c08e7caf437fe05f14b9da1c39b53
parent: cfb7fb7bf50e9b25e486d0d9a5c1cc75f1d4055d
author: Paul Brossier <piem@piem.org>
date: Sat Sep 16 19:55:51 EDT 2017 

Merge branch 'master' into dct


--- a/aubio.pc.in
+++ b/aubio.pc.in
@@ -7,4 +7,4 @@
 Description: a library for audio labelling
 Version: @VERSION@
 Libs: -L${libdir} -laubio
-Cflags: -I${includedir} 
+Cflags: -I${includedir}
--- a/python/demos/demo_bpm_extract.py
+++ b/python/demos/demo_bpm_extract.py
@@ -3,7 +3,7 @@
 from aubio import source, tempo
 from numpy import median, diff
 
-def get_file_bpm(path, params = None):
+def get_file_bpm(path, params=None):
     """ Calculate the beats per minute (bpm) of a given file.
         path: path to the file
         param: dictionary of parameters
@@ -10,19 +10,26 @@
     """
     if params is None:
         params = {}
-    try:
-        win_s = params['win_s']
-        samplerate = params['samplerate']
-        hop_s = params['hop_s']
-    except KeyError:
-        """
-        # super fast
-        samplerate, win_s, hop_s = 4000, 128, 64 
-        # fast
-        samplerate, win_s, hop_s = 8000, 512, 128
-        """
-        # default:
-        samplerate, win_s, hop_s = 44100, 1024, 512
+    # default:
+    samplerate, win_s, hop_s = 44100, 1024, 512
+    if 'mode' in params:
+        if params.mode in ['super-fast']:
+            # super fast
+            samplerate, win_s, hop_s = 4000, 128, 64
+        elif params.mode in ['fast']:
+            # fast
+            samplerate, win_s, hop_s = 8000, 512, 128
+        elif params.mode in ['default']:
+            pass
+        else:
+            print("unknown mode {:s}".format(params.mode))
+    # manual settings
+    if 'samplerate' in params:
+        samplerate = params.samplerate
+    if 'win_s' in params:
+        win_s = params.win_s
+    if 'hop_s' in params:
+        hop_s = params.hop_s
 
     s = source(path, samplerate, hop_s)
     samplerate = s.samplerate
@@ -44,19 +51,29 @@
         if read < hop_s:
             break
 
-    # Convert to periods and to bpm 
-    if len(beats) > 1:
-        if len(beats) < 4:
-            print("few beats found in {:s}".format(path))
-        bpms = 60./diff(beats)
-        b = median(bpms)
-    else:
-        b = 0
-        print("not enough beats found in {:s}".format(path))
-    return b
+    def beats_to_bpm(beats, path):
+        # if enough beats are found, convert to periods then to bpm
+        if len(beats) > 1:
+            if len(beats) < 4:
+                print("few beats found in {:s}".format(path))
+            bpms = 60./diff(beats)
+            return median(bpms)
+        else:
+            print("not enough beats found in {:s}".format(path))
+            return 0
 
+    return beats_to_bpm(beats, path)
+
 if __name__ == '__main__':
-    import sys
-    for f in sys.argv[1:]:
-        bpm = get_file_bpm(f)
+    import argparse
+    parser = argparse.ArgumentParser()
+    parser.add_argument('-m', '--mode',
+            help="mode [default|fast|super-fast]",
+            dest="mode")
+    parser.add_argument('sources',
+            nargs='*',
+            help="input_files")
+    args = parser.parse_args()
+    for f in args.sources:
+        bpm = get_file_bpm(f, params = args)
         print("{:6s} {:s}".format("{:2f}".format(bpm), f))
--- a/python/demos/demo_tapthebeat.py
+++ b/python/demos/demo_tapthebeat.py
@@ -44,7 +44,7 @@
 click = 0.7 * np.sin(2. * np.pi * np.arange(hop_s) / hop_s * samplerate / 3000.)
 
 # pyaudio callback
-def pyaudio_callback(in_data, frame_count, time_info, status):
+def pyaudio_callback(_in_data, _frame_count, _time_info, _status):
     samples, read = a_source()
     is_beat = a_tempo(samples)
     if is_beat:
--- a/python/ext/py-phasevoc.c
+++ b/python/ext/py-phasevoc.c
@@ -155,9 +155,28 @@
   return self->routput;
 }
 
+static PyObject *
+Pyaubio_pvoc_set_window (Py_pvoc *self, PyObject *args)
+{
+  uint_t err = 0;
+  char_t *window = NULL;
+
+  if (!PyArg_ParseTuple (args, "s", &window)) {
+    return NULL;
+  }
+  err = aubio_pvoc_set_window (self->o, window);
+
+  if (err > 0) {
+    PyErr_SetString (PyExc_ValueError, "error running aubio_pvoc_set_window");
+    return NULL;
+  }
+  Py_RETURN_NONE;
+}
+
 static PyMethodDef Py_pvoc_methods[] = {
   {"rdo", (PyCFunction) Py_pvoc_rdo, METH_VARARGS,
     "synthesis of spectral grain"},
+  {"set_window", (PyCFunction) Pyaubio_pvoc_set_window, METH_VARARGS, ""},
   {NULL}
 };
 
--- a/python/lib/aubio/cmd.py
+++ b/python/lib/aubio/cmd.py
@@ -20,154 +20,234 @@
             action="store_true", dest="show_version")
 
     subparsers = parser.add_subparsers(title='commands', dest='command',
+            parser_class= AubioArgumentParser,
             metavar="")
 
+    parser_add_subcommand_help(subparsers)
+
+    parser_add_subcommand_onset(subparsers)
+    parser_add_subcommand_pitch(subparsers)
+    parser_add_subcommand_beat(subparsers)
+    parser_add_subcommand_tempo(subparsers)
+    parser_add_subcommand_notes(subparsers)
+    parser_add_subcommand_mfcc(subparsers)
+    parser_add_subcommand_melbands(subparsers)
+    parser_add_subcommand_quiet(subparsers)
+    parser_add_subcommand_cut(subparsers)
+
+    return parser
+
+def parser_add_subcommand_help(subparsers):
+    # global help subcommand
+    subparsers.add_parser('help',
+            help='show help message',
+            formatter_class = argparse.ArgumentDefaultsHelpFormatter)
+
+def parser_add_subcommand_onset(subparsers):
     # onset subcommand
     subparser = subparsers.add_parser('onset',
             help='estimate time of onsets (beginning of sound event)',
             formatter_class = argparse.ArgumentDefaultsHelpFormatter)
-    parser_add_input(subparser)
-    parser_add_buf_hop_size(subparser)
+    subparser.add_input()
+    subparser.add_buf_hop_size()
     helpstr = "onset novelty function"
     helpstr += " <default|energy|hfc|complex|phase|specdiff|kl|mkl|specflux>"
-    parser_add_method(subparser, helpstr=helpstr)
-    parser_add_threshold(subparser)
-    parser_add_silence(subparser)
-    parser_add_minioi(subparser)
-    parser_add_time_format(subparser)
-    parser_add_verbose_help(subparser)
+    subparser.add_method(helpstr=helpstr)
+    subparser.add_threshold()
+    subparser.add_silence()
+    subparser.add_minioi()
+    subparser.add_time_format()
+    subparser.add_verbose_help()
     subparser.set_defaults(process=process_onset)
 
+def parser_add_subcommand_pitch(subparsers):
     # pitch subcommand
     subparser = subparsers.add_parser('pitch',
             help='estimate fundamental frequency (monophonic)')
-    parser_add_input(subparser)
-    parser_add_buf_hop_size(subparser, buf_size=2048)
+    subparser.add_input()
+    subparser.add_buf_hop_size(buf_size=2048)
     helpstr = "pitch detection method <default|yinfft|yin|mcomb|fcomb|schmitt>"
-    parser_add_method(subparser, helpstr=helpstr)
-    parser_add_threshold(subparser)
-    parser_add_pitch_unit(subparser)
-    parser_add_silence(subparser)
-    parser_add_time_format(subparser)
-    parser_add_verbose_help(subparser)
+    subparser.add_method(helpstr=helpstr)
+    subparser.add_threshold()
+    subparser.add_pitch_unit()
+    subparser.add_silence()
+    subparser.add_time_format()
+    subparser.add_verbose_help()
     subparser.set_defaults(process=process_pitch)
 
+def parser_add_subcommand_beat(subparsers):
     # beat subcommand
     subparser = subparsers.add_parser('beat',
             help='estimate location of beats')
-    parser_add_input(subparser)
-    parser_add_buf_hop_size(subparser, buf_size=1024, hop_size=512)
-    parser_add_time_format(subparser)
-    parser_add_verbose_help(subparser)
+    subparser.add_input()
+    subparser.add_buf_hop_size(buf_size=1024, hop_size=512)
+    subparser.add_time_format()
+    subparser.add_verbose_help()
     subparser.set_defaults(process=process_beat)
 
+def parser_add_subcommand_tempo(subparsers):
     # tempo subcommand
     subparser = subparsers.add_parser('tempo',
             help='estimate overall tempo in bpm')
-    parser_add_input(subparser)
-    parser_add_buf_hop_size(subparser, buf_size=1024, hop_size=512)
-    parser_add_time_format(subparser)
-    parser_add_verbose_help(subparser)
+    subparser.add_input()
+    subparser.add_buf_hop_size(buf_size=1024, hop_size=512)
+    subparser.add_time_format()
+    subparser.add_verbose_help()
     subparser.set_defaults(process=process_tempo)
 
+def parser_add_subcommand_notes(subparsers):
     # notes subcommand
     subparser = subparsers.add_parser('notes',
             help='estimate midi-like notes (monophonic)')
-    parser_add_input(subparser)
-    parser_add_buf_hop_size(subparser)
-    parser_add_time_format(subparser)
-    parser_add_verbose_help(subparser)
+    subparser.add_input()
+    subparser.add_buf_hop_size()
+    subparser.add_time_format()
+    subparser.add_verbose_help()
     subparser.set_defaults(process=process_notes)
 
+def parser_add_subcommand_mfcc(subparsers):
     # mfcc subcommand
     subparser = subparsers.add_parser('mfcc',
             help='extract Mel-Frequency Cepstrum Coefficients')
-    parser_add_input(subparser)
-    parser_add_buf_hop_size(subparser)
-    parser_add_time_format(subparser)
-    parser_add_verbose_help(subparser)
+    subparser.add_input()
+    subparser.add_buf_hop_size()
+    subparser.add_time_format()
+    subparser.add_verbose_help()
     subparser.set_defaults(process=process_mfcc)
 
+def parser_add_subcommand_melbands(subparsers):
     # melbands subcommand
     subparser = subparsers.add_parser('melbands',
             help='extract energies in Mel-frequency bands')
-    parser_add_input(subparser)
-    parser_add_buf_hop_size(subparser)
-    parser_add_time_format(subparser)
-    parser_add_verbose_help(subparser)
+    subparser.add_input()
+    subparser.add_buf_hop_size()
+    subparser.add_time_format()
+    subparser.add_verbose_help()
     subparser.set_defaults(process=process_melbands)
 
-    return parser
+def parser_add_subcommand_quiet(subparsers):
+    # quiet subcommand
+    subparser = subparsers.add_parser('quiet',
+            help='extract timestamps of quiet and loud regions')
+    subparser.add_input()
+    subparser.add_hop_size()
+    subparser.add_silence()
+    subparser.add_time_format()
+    subparser.add_verbose_help()
+    subparser.set_defaults(process=process_quiet)
 
-def parser_add_input(parser):
-    parser.add_argument("source_uri", default=None, nargs='?',
-            help="input sound file to analyse", metavar = "<source_uri>")
-    parser.add_argument("-i", "--input", dest = "source_uri2",
-            help="input sound file to analyse", metavar = "<source_uri>")
-    parser.add_argument("-r", "--samplerate",
-            metavar = "<freq>", type=int,
-            action="store", dest="samplerate", default=0,
-            help="samplerate at which the file should be represented")
+def parser_add_subcommand_cut(subparsers):
+    # quiet subcommand
+    subparser = subparsers.add_parser('cut',
+            help='slice at timestamps')
+    subparser.add_input()
+    helpstr = "onset novelty function"
+    helpstr += " <default|energy|hfc|complex|phase|specdiff|kl|mkl|specflux>"
+    subparser.add_method(helpstr=helpstr)
+    subparser.add_buf_hop_size()
+    subparser.add_silence()
+    subparser.add_threshold(default=0.3)
+    subparser.add_minioi()
+    subparser.add_slicer_options()
+    subparser.add_time_format()
+    subparser.add_verbose_help()
+    subparser.set_defaults(process=process_cut)
 
-def parser_add_verbose_help(parser):
-    parser.add_argument("-v","--verbose",
-            action="count", dest="verbose", default=1,
-            help="make lots of noise [default]")
-    parser.add_argument("-q","--quiet",
-            action="store_const", dest="verbose", const=0,
-            help="be quiet")
+class AubioArgumentParser(argparse.ArgumentParser):
 
-def parser_add_buf_hop_size(parser, buf_size=512, hop_size=256):
-    parser.add_argument("-B","--bufsize",
-            action="store", dest="buf_size", default=buf_size,
-            metavar = "<size>", type=int,
-            help="buffer size [default=%d]" % buf_size)
-    parser.add_argument("-H","--hopsize",
-            metavar = "<size>", type=int,
-            action="store", dest="hop_size", default=hop_size,
-            help="overlap size [default=%d]" % hop_size)
+    def add_input(self):
+        self.add_argument("source_uri", default=None, nargs='?',
+                help="input sound file to analyse", metavar = "<source_uri>")
+        self.add_argument("-i", "--input", dest = "source_uri2",
+                help="input sound file to analyse", metavar = "<source_uri>")
+        self.add_argument("-r", "--samplerate",
+                metavar = "<freq>", type=int,
+                action="store", dest="samplerate", default=0,
+                help="samplerate at which the file should be represented")
 
-def parser_add_method(parser, method='default', helpstr='method'):
-    parser.add_argument("-m","--method",
-            metavar = "<method>", type=str,
-            action="store", dest="method", default=method,
-            help="%s [default=%s]" % (helpstr, method))
+    def add_verbose_help(self):
+        self.add_argument("-v","--verbose",
+                action="count", dest="verbose", default=1,
+                help="make lots of noise [default]")
+        self.add_argument("-q","--quiet",
+                action="store_const", dest="verbose", const=0,
+                help="be quiet")
 
-def parser_add_threshold(parser, default=None):
-    parser.add_argument("-t","--threshold",
-            metavar = "<threshold>", type=float,
-            action="store", dest="threshold", default=default,
-            help="threshold [default=%s]" % default)
+    def add_buf_hop_size(self, buf_size=512, hop_size=256):
+        self.add_buf_size(buf_size=buf_size)
+        self.add_hop_size(hop_size=hop_size)
+
+    def add_buf_size(self, buf_size=512):
+        self.add_argument("-B","--bufsize",
+                action="store", dest="buf_size", default=buf_size,
+                metavar = "<size>", type=int,
+                help="buffer size [default=%d]" % buf_size)
+
+    def add_hop_size(self, hop_size=256):
+        self.add_argument("-H","--hopsize",
+                metavar = "<size>", type=int,
+                action="store", dest="hop_size", default=hop_size,
+                help="overlap size [default=%d]" % hop_size)
+
+    def add_method(self, method='default', helpstr='method'):
+        self.add_argument("-m","--method",
+                metavar = "<method>", type=str,
+                action="store", dest="method", default=method,
+                help="%s [default=%s]" % (helpstr, method))
 
-def parser_add_silence(parser):
-    parser.add_argument("-s", "--silence",
-            metavar = "<value>", type=float,
-            action="store", dest="silence", default=-70,
-            help="silence threshold")
+    def add_threshold(self, default=None):
+        self.add_argument("-t","--threshold",
+                metavar = "<threshold>", type=float,
+                action="store", dest="threshold", default=default,
+                help="threshold [default=%s]" % default)
 
-def parser_add_minioi(parser):
-    parser.add_argument("-M", "--minioi",
-            metavar = "<value>", type=str,
-            action="store", dest="minioi", default="12ms",
-            help="minimum Inter-Onset Interval")
+    def add_silence(self):
+        self.add_argument("-s", "--silence",
+                metavar = "<value>", type=float,
+                action="store", dest="silence", default=-70,
+                help="silence threshold")
 
-def parser_add_pitch_unit(parser, default="Hz"):
-    help_str = "frequency unit, should be one of Hz, midi, bin, cent"
-    help_str += " [default=%s]" % default
-    parser.add_argument("-u", "--pitch-unit",
-            metavar = "<value>", type=str,
-            action="store", dest="pitch_unit", default=default,
-            help=help_str)
+    def add_minioi(self, default="12ms"):
+        self.add_argument("-M", "--minioi",
+                metavar = "<value>", type=str,
+                action="store", dest="minioi", default=default,
+                help="minimum Inter-Onset Interval [default=%s]" % default)
 
-def parser_add_time_format(parser):
-    helpstr = "select time values output format (samples, ms, seconds)"
-    helpstr += " [default=seconds]"
-    parser.add_argument("-T", "--time-format",
-             metavar='format',
-             dest="time_format",
-             default=None,
-             help=helpstr)
+    def add_pitch_unit(self, default="Hz"):
+        help_str = "frequency unit, should be one of Hz, midi, bin, cent"
+        help_str += " [default=%s]" % default
+        self.add_argument("-u", "--pitch-unit",
+                metavar = "<value>", type=str,
+                action="store", dest="pitch_unit", default=default,
+                help=help_str)
 
+    def add_time_format(self):
+        helpstr = "select time values output format (samples, ms, seconds)"
+        helpstr += " [default=seconds]"
+        self.add_argument("-T", "--time-format",
+                 metavar='format',
+                 dest="time_format",
+                 default=None,
+                 help=helpstr)
+
+    def add_slicer_options(self):
+        self.add_argument("-o","--output", type = str,
+                metavar = "<outputdir>",
+                action="store", dest="output_directory", default=None,
+                help="specify path where slices of the original file should be created")
+        self.add_argument("--cut-until-nsamples", type = int,
+                metavar = "<samples>",
+                action = "store", dest = "cut_until_nsamples", default = None,
+                help="how many extra samples should be added at the end of each slice")
+        self.add_argument("--cut-every-nslices", type = int,
+                metavar = "<samples>",
+                action = "store", dest = "cut_every_nslices", default = None,
+                help="how many slices should be groupped together at each cut")
+        self.add_argument("--cut-until-nslices", type = int,
+                metavar = "<slices>",
+                action = "store", dest = "cut_until_nslices", default = None,
+                help="how many extra slices should be added at the end of each slice")
+
 # some utilities
 
 def samples2seconds(n_frames, samplerate):
@@ -176,7 +256,7 @@
 def samples2milliseconds(n_frames, samplerate):
     return "%f\t" % (1000. * n_frames / float(samplerate))
 
-def samples2samples(n_frames, samplerate):
+def samples2samples(n_frames, _samplerate):
     return "%d\t" % n_frames
 
 def timefunc(mode):
@@ -199,7 +279,7 @@
             name = type(self).__name__.split('_')[1]
             optstr = ' '.join(['running', name, 'with options', repr(self.options), '\n'])
             sys.stderr.write(optstr)
-    def flush(self, n_frames, samplerate):
+    def flush(self, frames_read, samplerate):
         # optionally called at the end of process
         pass
 
@@ -238,7 +318,7 @@
         super(process_onset, self).__init__(args)
     def __call__(self, block):
         return self.onset(block)
-    def repr_res(self, res, frames_read, samplerate):
+    def repr_res(self, res, _frames_read, samplerate):
         if res[0] != 0:
             outstr = self.time2string(self.onset.get_last(), samplerate)
             sys.stdout.write(outstr + '\n')
@@ -269,7 +349,7 @@
         super(process_beat, self).__init__(args)
     def __call__(self, block):
         return self.tempo(block)
-    def repr_res(self, res, frames_read, samplerate):
+    def repr_res(self, res, _frames_read, samplerate):
         if res[0] != 0:
             outstr = self.time2string(self.tempo.get_last(), samplerate)
             sys.stdout.write(outstr + '\n')
@@ -278,7 +358,7 @@
     def __init__(self, args):
         super(process_tempo, self).__init__(args)
         self.beat_locations = []
-    def repr_res(self, res, frames_read, samplerate):
+    def repr_res(self, res, _frames_read, samplerate):
         if res[0] != 0:
             self.beat_locations.append(self.tempo.get_last_s())
     def flush(self, frames_read, samplerate):
@@ -362,6 +442,61 @@
         fmt_out += ' '.join(["% 9.7f" % f for f in res.tolist()])
         sys.stdout.write(fmt_out + '\n')
 
+class process_quiet(default_process):
+    def __init__(self, args):
+        self.args = args
+        valid_opts = ['hop_size', 'silence']
+        self.parse_options(args, valid_opts)
+        self.wassilence = 1
+
+        if args.silence is not None:
+            self.silence = args.silence
+        super(process_quiet, self).__init__(args)
+
+    def __call__(self, block):
+        if aubio.silence_detection(block, self.silence) == 1:
+            if self.wassilence != 1:
+                self.wassilence = 1
+                return 2 # newly found silence
+            return 1 # silence again
+        else:
+            if self.wassilence != 0:
+                self.wassilence = 0
+                return -1 # newly found noise
+            return 0 # noise again
+
+    def repr_res(self, res, frames_read, samplerate):
+        fmt_out = None
+        if res == -1:
+            fmt_out = "NOISY: "
+        if res == 2:
+            fmt_out = "QUIET: "
+        if fmt_out is not None:
+            fmt_out += self.time2string(frames_read, samplerate)
+            sys.stdout.write(fmt_out + '\n')
+
+class process_cut(process_onset):
+    def __init__(self, args):
+        super(process_cut, self).__init__(args)
+        self.slices = []
+        self.options = args
+
+    def __call__(self, block):
+        ret = super(process_cut, self).__call__(block)
+        if ret: self.slices.append(self.onset.get_last())
+        return ret
+
+    def flush(self, frames_read, samplerate):
+        from aubio.cut import _cut_slice
+        _cut_slice(self.options, self.slices)
+        duration = float (frames_read) / float(samplerate)
+        base_info = '%(source_file)s' % {'source_file': self.options.source_uri}
+        base_info += ' (total %(duration).2fs at %(samplerate)dHz)\n' % \
+                {'duration': duration, 'samplerate': samplerate}
+        info = "created %d slices from " % len(self.slices)
+        info += base_info
+        sys.stderr.write(info)
+
 def main():
     parser = aubio_parser()
     args = parser.parse_args()
@@ -370,10 +505,13 @@
         sys.exit(0)
     elif 'verbose' in args and args.verbose > 3:
         sys.stderr.write('aubio version ' + aubio.version + '\n')
-    if 'command' not in args or args.command is None:
+    if 'command' not in args or args.command is None or args.command in ['help']:
         # no command given, print help and return 1
         parser.print_help()
-        sys.exit(1)
+        if args.command and args.command in ['help']:
+            sys.exit(0)
+        else:
+            sys.exit(1)
     elif not args.source_uri and not args.source_uri2:
         sys.stderr.write("Error: a source is required\n")
         parser.print_help()
--- a/python/lib/aubio/cut.py
+++ b/python/lib/aubio/cut.py
@@ -5,16 +5,11 @@
 """
 
 import sys
-import argparse
+from aubio.cmd import AubioArgumentParser
 
-def parse_args():
-    usage = "usage: %s [options] -i soundfile" % sys.argv[0]
-    usage += "\n help: %s -h" % sys.argv[0]
-    parser = argparse.ArgumentParser()
-    parser.add_argument("source_file", default=None, nargs='?',
-            help="input sound file to analyse", metavar = "<source_file>")
-    parser.add_argument("-i", "--input", action = "store", dest = "source_file2",
-            help="input sound file to analyse", metavar = "<source_file>")
+def aubio_cut_parser():
+    parser = AubioArgumentParser()
+    parser.add_input()
     parser.add_argument("-O","--onset-method",
             action="store", dest="onset_method", default='default',
             metavar = "<onset_method>",
@@ -23,7 +18,7 @@
     # cutting methods
     parser.add_argument("-b","--beat",
             action="store_true", dest="beat", default=False,
-            help="use beat locations")
+            help="slice at beat locations")
     """
     parser.add_argument("-S","--silencecut",
             action="store_true", dest="silencecut", default=False,
@@ -34,33 +29,16 @@
             help="silence threshold [default=-70]")
             """
     # algorithm parameters
-    parser.add_argument("-r", "--samplerate",
-            metavar = "<freq>", type=int,
-            action="store", dest="samplerate", default=0,
-            help="samplerate at which the file should be represented")
-    parser.add_argument("-B","--bufsize",
-            action="store", dest="bufsize", default=512,
-            metavar = "<size>", type=int,
-            help="buffer size [default=512]")
-    parser.add_argument("-H","--hopsize",
-            metavar = "<size>", type=int,
-            action="store", dest="hopsize", default=256,
-            help="overlap size [default=256]")
-    parser.add_argument("-t","--onset-threshold",
-            metavar = "<value>", type=float,
+    parser.add_buf_hop_size()
+    parser.add_argument("-t","--threshold", "--onset-threshold",
+            metavar = "<threshold>", type=float,
             action="store", dest="threshold", default=0.3,
             help="onset peak picking threshold [default=0.3]")
     parser.add_argument("-c","--cut",
             action="store_true", dest="cut", default=False,
-            help="cut input sound file at detected labels \
-                    best used with option -L")
+            help="cut input sound file at detected labels")
+    parser.add_minioi()
 
-    # minioi
-    parser.add_argument("-M","--minioi",
-            metavar = "<value>", type=str,
-            action="store", dest="minioi", default="12ms",
-            help="minimum inter onset interval [default=12ms]")
-
     """
     parser.add_argument("-D","--delay",
             action = "store", dest = "delay", type = float,
@@ -81,9 +59,7 @@
             metavar = "<value>",
             action="store", dest="zerothres", default=0.008,
             help="zero-crossing threshold for slicing [default=0.00008]")
-            """
     # plotting functions
-    """
     parser.add_argument("-p","--plot",
             action="store_true", dest="plot", default=False,
             help="draw plot")
@@ -109,51 +85,24 @@
             action="store_true", dest="spectro", default=False,
             help="add spectrogram to the plot")
     """
-    parser.add_argument("-o","--output", type = str,
-            metavar = "<outputdir>",
-            action="store", dest="output_directory", default=None,
-            help="specify path where slices of the original file should be created")
-    parser.add_argument("--cut-until-nsamples", type = int,
-            metavar = "<samples>",
-            action = "store", dest = "cut_until_nsamples", default = None,
-            help="how many extra samples should be added at the end of each slice")
-    parser.add_argument("--cut-every-nslices", type = int,
-            metavar = "<samples>",
-            action = "store", dest = "cut_every_nslices", default = None,
-            help="how many slices should be groupped together at each cut")
-    parser.add_argument("--cut-until-nslices", type = int,
-            metavar = "<slices>",
-            action = "store", dest = "cut_until_nslices", default = None,
-            help="how many extra slices should be added at the end of each slice")
+    parser.add_slicer_options()
+    parser.add_verbose_help()
+    return parser
 
-    parser.add_argument("-v","--verbose",
-            action="store_true", dest="verbose", default=True,
-            help="make lots of noise [default]")
-    parser.add_argument("-q","--quiet",
-            action="store_false", dest="verbose", default=True,
-            help="be quiet")
-    args = parser.parse_args()
-    if not args.source_file and not args.source_file2:
-        sys.stderr.write("Error: no file name given\n")
-        parser.print_help()
-        sys.exit(1)
-    elif args.source_file2 is not None:
-        args.source_file = args.source_file2
-    return args
 
-def main():
-    options = parse_args()
-
-    source_file = options.source_file
-    hopsize = options.hopsize
-    bufsize = options.bufsize
+def _cut_analyze(options):
+    hopsize = options.hop_size
+    bufsize = options.buf_size
     samplerate = options.samplerate
-    source_file = options.source_file
+    source_uri = options.source_uri
 
+    # analyze pass
     from aubio import onset, tempo, source
 
-    s = source(source_file, samplerate, hopsize)
-    if samplerate == 0: samplerate = s.get_samplerate()
+    s = source(source_uri, samplerate, hopsize)
+    if samplerate == 0:
+        samplerate = s.get_samplerate()
+        options.samplerate = samplerate
 
     if options.beat:
         o = tempo(options.onset_method, bufsize, hopsize, samplerate=samplerate)
@@ -170,7 +119,6 @@
 
     timestamps = []
     total_frames = 0
-    # analyze pass
     while True:
         samples, read = s()
         if o(samples):
@@ -179,15 +127,12 @@
         total_frames += read
         if read < hopsize: break
     del s
-    # print some info
-    nstamps = len(timestamps)
-    duration = float (total_frames) / float(samplerate)
-    info = 'found %(nstamps)d timestamps in %(source_file)s' % locals()
-    info += ' (total %(duration).2fs at %(samplerate)dHz)\n' % locals()
-    sys.stderr.write(info)
+    return timestamps, total_frames
 
+def _cut_slice(options, timestamps):
     # cutting pass
-    if options.cut and nstamps > 0:
+    nstamps = len(timestamps)
+    if nstamps > 0:
         # generate output files
         from aubio.slicing import slice_source_at_stamps
         timestamps_end = None
@@ -202,12 +147,36 @@
         if options.cut_until_nslices:
             timestamps_end = [t for t in timestamps[1 + options.cut_until_nslices:]]
             timestamps_end += [ 1e120 ] * (options.cut_until_nslices + 1)
-        slice_source_at_stamps(source_file, timestamps, timestamps_end = timestamps_end,
+        slice_source_at_stamps(options.source_uri,
+                timestamps, timestamps_end = timestamps_end,
                 output_dir = options.output_directory,
-                samplerate = samplerate)
+                samplerate = options.samplerate)
 
-        # print some info
-        duration = float (total_frames) / float(samplerate)
-        info = 'created %(nstamps)d slices from %(source_file)s' % locals()
-        info += ' (total %(duration).2fs at %(samplerate)dHz)\n' % locals()
+def main():
+    parser = aubio_cut_parser()
+    options = parser.parse_args()
+    if not options.source_uri and not options.source_uri2:
+        sys.stderr.write("Error: no file name given\n")
+        parser.print_help()
+        sys.exit(1)
+    elif options.source_uri2 is not None:
+        options.source_uri = options.source_uri2
+
+    # analysis
+    timestamps, total_frames = _cut_analyze(options)
+
+    # print some info
+    duration = float (total_frames) / float(options.samplerate)
+    base_info = '%(source_uri)s' % {'source_uri': options.source_uri}
+    base_info += ' (total %(duration).2fs at %(samplerate)dHz)\n' % \
+            {'duration': duration, 'samplerate': options.samplerate}
+
+    info = "found %d timestamps in " % len(timestamps)
+    info += base_info
+    sys.stderr.write(info)
+
+    if options.cut:
+        _cut_slice(options, timestamps)
+        info = "created %d slices from " % len(timestamps)
+        info += base_info
         sys.stderr.write(info)
--- /dev/null
+++ b/python/tests/test_aubio_cmd.py
@@ -1,0 +1,27 @@
+#! /usr/bin/env python
+
+import aubio.cmd
+from nose2 import main
+from numpy.testing import TestCase
+
+class aubio_cmd(TestCase):
+
+    def setUp(self):
+        self.a_parser = aubio.cmd.aubio_parser()
+
+    def test_default_creation(self):
+        assert self.a_parser.parse_args(['-V']).show_version
+
+class aubio_cmd_utils(TestCase):
+
+    def test_samples2seconds(self):
+        self.assertEqual(aubio.cmd.samples2seconds(3200, 32000), "0.100000\t")
+
+    def test_samples2milliseconds(self):
+        self.assertEqual(aubio.cmd.samples2milliseconds(3200, 32000), "100.000000\t")
+
+    def test_samples2samples(self):
+        self.assertEqual(aubio.cmd.samples2samples(3200, 32000), "3200\t")
+
+if __name__ == '__main__':
+    main()
--- /dev/null
+++ b/python/tests/test_aubio_cut.py
@@ -1,0 +1,16 @@
+#! /usr/bin/env python
+
+import aubio.cut
+from nose2 import main
+from numpy.testing import TestCase
+
+class aubio_cut(TestCase):
+
+    def setUp(self):
+        self.a_parser = aubio.cut.aubio_cut_parser()
+
+    def test_default_creation(self):
+        assert self.a_parser.parse_args(['-v']).verbose
+
+if __name__ == '__main__':
+    main()
--- a/python/tests/test_filterbank.py
+++ b/python/tests/test_filterbank.py
@@ -1,9 +1,8 @@
 #! /usr/bin/env python
 
-from unittest import main
-from numpy.testing import TestCase
-from numpy.testing import assert_equal, assert_almost_equal
 import numpy as np
+from numpy.testing import TestCase, assert_equal, assert_almost_equal
+
 from aubio import cvec, filterbank, float_type
 from .utils import array_from_text_file
 
@@ -62,6 +61,13 @@
         f.set_mel_coeffs_slaney(16000)
         assert_almost_equal ( expected, f.get_coeffs() )
 
+    def test_mfcc_coeffs_get_coeffs(self):
+        f = filterbank(40, 512)
+        coeffs = f.get_coeffs()
+        self.assertIsInstance(coeffs, np.ndarray)
+        assert_equal (coeffs, 0)
+        assert_equal (np.shape(coeffs), (40, 512 / 2 + 1))
+
 class aubio_filterbank_wrong_values(TestCase):
 
     def test_negative_window(self):
@@ -81,4 +87,5 @@
             f(cvec(256))
 
 if __name__ == '__main__':
-    main()
+    import nose2
+    nose2.main()
--- a/python/tests/test_filterbank_mel.py
+++ b/python/tests/test_filterbank_mel.py
@@ -1,11 +1,13 @@
 #! /usr/bin/env python
 
-from unittest import main
-from numpy.testing import TestCase
-from numpy.testing import assert_equal, assert_almost_equal
-from numpy import array, shape
+import numpy as np
+from numpy.testing import TestCase, assert_equal, assert_almost_equal
+
 from aubio import cvec, filterbank, float_type
 
+import warnings
+warnings.filterwarnings('ignore', category=UserWarning, append=True)
+
 class aubio_filterbank_mel_test_case(TestCase):
 
     def test_slaney(self):
@@ -12,33 +14,33 @@
         f = filterbank(40, 512)
         f.set_mel_coeffs_slaney(16000)
         a = f.get_coeffs()
-        assert_equal(shape (a), (40, 512/2 + 1) )
+        assert_equal(np.shape (a), (40, 512/2 + 1) )
 
     def test_other_slaney(self):
         f = filterbank(40, 512*2)
         f.set_mel_coeffs_slaney(44100)
-        _ = f.get_coeffs()
+        self.assertIsInstance(f.get_coeffs(), np.ndarray)
         #print "sum is", sum(sum(a))
         for win_s in [256, 512, 1024, 2048, 4096]:
             f = filterbank(40, win_s)
             f.set_mel_coeffs_slaney(32000)
-            _ = f.get_coeffs()
             #print "sum is", sum(sum(a))
+            self.assertIsInstance(f.get_coeffs(), np.ndarray)
 
     def test_triangle_freqs_zeros(self):
         f = filterbank(9, 1024)
         freq_list = [40, 80, 200, 400, 800, 1600, 3200, 6400, 12800, 15000, 24000]
-        freqs = array(freq_list, dtype = float_type)
+        freqs = np.array(freq_list, dtype = float_type)
         f.set_triangle_bands(freqs, 48000)
-        _ = f.get_coeffs().T
         assert_equal ( f(cvec(1024)), 0)
+        self.assertIsInstance(f.get_coeffs(), np.ndarray)
 
     def test_triangle_freqs_ones(self):
         f = filterbank(9, 1024)
         freq_list = [40, 80, 200, 400, 800, 1600, 3200, 6400, 12800, 15000, 24000]
-        freqs = array(freq_list, dtype = float_type)
+        freqs = np.array(freq_list, dtype = float_type)
         f.set_triangle_bands(freqs, 48000)
-        _ = f.get_coeffs().T
+        self.assertIsInstance(f.get_coeffs(), np.ndarray)
         spec = cvec(1024)
         spec.norm[:] = 1
         assert_almost_equal ( f(spec),
@@ -46,4 +48,5 @@
                     0.02133301, 0.02133301, 0.02133311, 0.02133334, 0.02133345])
 
 if __name__ == '__main__':
-    main()
+    import nose2
+    nose2.main()
--- a/src/io/source_avcodec.c
+++ b/src/io/source_avcodec.c
@@ -313,15 +313,15 @@
     // TODO: use planar?
     //av_opt_set_int(avr, "out_sample_fmt",     AV_SAMPLE_FMT_FLTP,      0);
 #ifdef HAVE_AVRESAMPLE
-    if ( ( err = avresample_open(avr) ) < 0) {
+    if ( ( err = avresample_open(avr) ) < 0)
 #elif defined(HAVE_SWRESAMPLE)
-    if ( ( err = swr_init(avr) ) < 0) {
+    if ( ( err = swr_init(avr) ) < 0)
 #endif /* HAVE_AVRESAMPLE || HAVE_SWRESAMPLE */
+    {
       char errorstr[256];
       av_strerror (err, errorstr, sizeof(errorstr));
-      AUBIO_ERR("source_avcodec: Could not open AVAudioResampleContext for %s (%s)\n",
+      AUBIO_ERR("source_avcodec: Could not open resampling context for %s (%s)\n",
           s->path, errorstr);
-      //goto beach;
       return;
     }
     s->avr = avr;
--- a/src/mathutils.c
+++ b/src/mathutils.c
@@ -28,6 +28,7 @@
 /** Window types */
 typedef enum
 {
+  aubio_win_ones,
   aubio_win_rectangle,
   aubio_win_hamming,
   aubio_win_hanning,
@@ -63,7 +64,9 @@
   if (window_type == NULL) {
       AUBIO_ERR ("window type can not be null.\n");
       return 1;
-  } else if (strcmp (window_type, "rectangle") == 0)
+  } else if (strcmp (window_type, "ones") == 0)
+      wintype = aubio_win_ones;
+  else if (strcmp (window_type, "rectangle") == 0)
       wintype = aubio_win_rectangle;
   else if (strcmp (window_type, "hamming") == 0)
       wintype = aubio_win_hamming;
@@ -88,9 +91,11 @@
       return 1;
   }
   switch(wintype) {
+    case aubio_win_ones:
+      fvec_ones(win);
+      break;
     case aubio_win_rectangle:
-      for (i=0;i<size;i++)
-        w[i] = 0.5;
+      fvec_set_all(win, .5);
       break;
     case aubio_win_hamming:
       for (i=0;i<size;i++)
--- a/src/spectral/mfcc.c
+++ b/src/spectral/mfcc.c
@@ -36,7 +36,7 @@
 {
   uint_t win_s;             /** grain length */
   uint_t samplerate;        /** sample rate (needed?) */
-  uint_t n_filters;         /** number of  *filters */
+  uint_t n_filters;         /** number of filters */
   uint_t n_coefs;           /** number of coefficients (<= n_filters/2 +1) */
   aubio_filterbank_t *fb;   /** filter bank */
   fvec_t *in_dct;           /** input buffer for dct * [fb->n_filters] */
--- a/src/spectral/phasevoc.c
+++ b/src/spectral/phasevoc.c
@@ -143,6 +143,10 @@
   return NULL;
 }
 
+uint_t aubio_pvoc_set_window(aubio_pvoc_t *pv, const char_t *window) {
+  return fvec_set_window(pv->w, (char_t*)window);
+}
+
 void del_aubio_pvoc(aubio_pvoc_t *pv) {
   del_fvec(pv->data);
   del_fvec(pv->synth);
--- a/src/spectral/phasevoc.h
+++ b/src/spectral/phasevoc.h
@@ -95,6 +95,16 @@
 */
 uint_t aubio_pvoc_get_hop(aubio_pvoc_t* pv);
 
+/** set window type
+
+  \param pv phase vocoder to set the window type
+  \param window_type a string representing a window
+
+  \return 0 if successful, non-zero otherwise
+
+ */
+uint_t aubio_pvoc_set_window(aubio_pvoc_t *pv, const char_t *window_type);
+
 #ifdef __cplusplus
 }
 #endif