Chris@117: #!/usr/bin/env python
Chris@117: 
Chris@117: #   Python Vamp Host
Chris@117: #   Copyright (c) 2008-2015 Queen Mary, University of London
Chris@117: #
Chris@117: #   Permission is hereby granted, free of charge, to any person
Chris@117: #   obtaining a copy of this software and associated documentation
Chris@117: #   files (the "Software"), to deal in the Software without
Chris@117: #   restriction, including without limitation the rights to use, copy,
Chris@117: #   modify, merge, publish, distribute, sublicense, and/or sell copies
Chris@117: #   of the Software, and to permit persons to whom the Software is
Chris@117: #   furnished to do so, subject to the following conditions:
Chris@117: #
Chris@117: #   The above copyright notice and this permission notice shall be
Chris@117: #   included in all copies or substantial portions of the Software.
Chris@117: #
Chris@117: #   THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
Chris@117: #   EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
Chris@117: #   MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
Chris@117: #   NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY
Chris@117: #   CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF
Chris@117: #   CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
Chris@117: #   WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
Chris@117: #
Chris@117: #   Except as contained in this notice, the names of the Centre for
Chris@117: #   Digital Music and Queen Mary, University of London shall not be
Chris@117: #   used in advertising or otherwise to promote the sale, use or other
Chris@117: #   dealings in this Software without prior written authorization.
Chris@117: 
Chris@56: '''A high-level interface to the vampyhost extension module, for quickly and easily running Vamp audio analysis plugins on audio files and buffers.'''
Chris@56: 
Chris@56: import vampyhost
Chris@112: import vamp.frames
Chris@112: import vamp.load
Chris@68: 
Chris@100: def process_with_initialised_plugin(ff, sample_rate, step_size, plugin, outputs):
Chris@89: 
Chris@98:     out_indices = dict([(id, plugin.get_output(id)["output_index"])
Chris@98:                         for id in outputs])
Chris@89:     plugin.reset()
Chris@89:     fi = 0
Chris@89: 
Chris@89:     for f in ff:
Chris@89:         timestamp = vampyhost.frame_to_realtime(fi, sample_rate)
Chris@89:         results = plugin.process_block(f, timestamp)
Chris@89:         # results is a dict mapping output number -> list of feature dicts
Chris@89:         for o in outputs:
Chris@89:             ix = out_indices[o]
Chris@89:             if ix in results:
Chris@89:                 for r in results[ix]:
Chris@89:                     yield { o: r }
Chris@89:         fi = fi + step_size
Chris@89: 
Chris@89:     results = plugin.get_remaining_features()
Chris@89:     for o in outputs:
Chris@89:         ix = out_indices[o]
Chris@89:         if ix in results:
Chris@89:             for r in results[ix]:
Chris@89:                 yield { o: r }
Chris@89: 
Chris@89: 
Chris@140: def process_audio(data, sample_rate, plugin_key, output = "", parameters = {}, **kwargs):
Chris@98:     """Process audio data with a Vamp plugin, and make the results from a
Chris@98:     single plugin output available as a generator.
Chris@98: 
Chris@98:     The provided data should be a 1- or 2-dimensional list or NumPy
Chris@98:     array of floats. If it is 2-dimensional, the first dimension is
Chris@98:     taken to be the channel count.
Chris@98: 
Chris@98:     The returned results will be those calculated by the plugin with
Chris@98:     the given key and returned through its output with the given
Chris@98:     output identifier. If the requested output is the empty string,
Chris@98:     the first output provided by the plugin will be used.
Chris@98: 
Chris@98:     If the parameters dict is non-empty, the plugin will be configured
Chris@98:     by setting its parameters according to the (string) key and
Chris@98:     (float) value data found in the dict.
Chris@98: 
Chris@98:     This function acts as a generator, yielding a sequence of result
Chris@98:     features as it obtains them. Each feature is represented as a
Chris@98:     dictionary containing, optionally, timestamp and duration
Chris@98:     (RealTime objects), label (string), and a 1-dimensional array of
Chris@98:     float values.
Chris@99: 
Chris@143:     If you wish to override the step size, block size, or process
Chris@143:     timestamp method to be used, you may supply them as keyword
Chris@143:     arguments with keywords step_size (int), block_size (int), and
Chris@143:     process_timestamp_method (choose from vamp.vampyhost.SHIFT_DATA,
Chris@143:     vamp.vampyhost.SHIFT_TIMESTAMP, or vamp.vampyhost.NO_SHIFT).
Chris@143: 
Chris@99:     If you would prefer to obtain all features in a single output
Chris@99:     structure, consider using vamp.collect() instead.
Chris@98:     """
Chris@89: 
Chris@140:     plugin, step_size, block_size = vamp.load.load_and_configure(data, sample_rate, plugin_key, parameters, **kwargs)
Chris@89: 
Chris@89:     if output == "":
Chris@89:         output = plugin.get_output(0)["identifier"]
Chris@89: 
Chris@112:     ff = vamp.frames.frames_from_array(data, step_size, block_size)
Chris@89: 
Chris@100:     for r in process_with_initialised_plugin(ff, sample_rate, step_size, plugin, [output]):
Chris@89:         yield r[output]
Chris@76:     
Chris@89:     plugin.unload()
Chris@89: 
Chris@76: 
Chris@129: def process_frames(ff, sample_rate, step_size, plugin_key, output = "", parameters = {}):
Chris@98:     """Process audio data with a Vamp plugin, and make the results from a
Chris@98:     single plugin output available as a generator.
Chris@95: 
Chris@98:     The provided data should be an enumerable sequence of time-domain
Chris@98:     audio frames, of which each frame is 2-dimensional list or NumPy
Chris@98:     array of floats. The first dimension is taken to be the channel
Chris@98:     count, and the second dimension the frame or block size. The
Chris@98:     step_size argument gives the increment in audio samples from one
Chris@98:     frame to the next. Each frame must have the same size.
Chris@95: 
Chris@98:     The returned results will be those calculated by the plugin with
Chris@98:     the given key and returned through its output with the given
Chris@98:     output identifier. If the requested output is the empty string,
Chris@98:     the first output provided by the plugin will be used.
Chris@95: 
Chris@98:     If the parameters dict is non-empty, the plugin will be configured
Chris@98:     by setting its parameters according to the (string) key and
Chris@98:     (float) value data found in the dict.
Chris@98: 
Chris@98:     This function acts as a generator, yielding a sequence of result
Chris@98:     features as it obtains them. Each feature is represented as a
Chris@98:     dictionary containing, optionally, timestamp and duration
Chris@98:     (RealTime objects), label (string), and a 1-dimensional array of
Chris@98:     float values.
Chris@99: 
Chris@99:     If you would prefer to obtain all features in a single output
Chris@99:     structure, consider using vamp.collect() instead.
Chris@98:     """
Chris@129:     plugin = vampyhost.load_plugin(plugin_key, sample_rate,
Chris@98:                                    vampyhost.ADAPT_INPUT_DOMAIN +
Chris@98:                                    vampyhost.ADAPT_BUFFER_SIZE +
Chris@98:                                    vampyhost.ADAPT_CHANNEL_COUNT)
Chris@98: 
Chris@98:     fi = 0
Chris@98:     channels = 0
Chris@98:     block_size = 0
Chris@98: 
Chris@98:     if output == "":
Chris@98:         out_index = 0
Chris@98:     else:
Chris@98:         out_index = plugin.get_output(output)["output_index"]
Chris@95:     
Chris@98:     for f in ff:
Chris@95: 
Chris@98:         if fi == 0:
Chris@98:             channels = f.shape[0]
Chris@98:             block_size = f.shape[1]
Chris@98:             plugin.set_parameter_values(parameters)
Chris@98:             if not plugin.initialise(channels, step_size, block_size):
Chris@98:                 raise "Failed to initialise plugin"
Chris@98: 
Chris@98:         timestamp = vampyhost.frame_to_realtime(fi, sample_rate)
Chris@98:         results = plugin.process_block(f, timestamp)
Chris@98:         # results is a dict mapping output number -> list of feature dicts
Chris@98:         if out_index in results:
Chris@98:             for r in results[out_index]:
Chris@98:                 yield r
Chris@98: 
Chris@98:         fi = fi + step_size
Chris@98: 
Chris@98:     if fi > 0:
Chris@98:         results = plugin.get_remaining_features()
Chris@98:         if out_index in results:
Chris@98:             for r in results[out_index]:
Chris@98:                 yield r
Chris@98:         
Chris@95:     plugin.unload()
Chris@95:     
Chris@95: 
Chris@140: def process_audio_multiple_outputs(data, sample_rate, plugin_key, outputs, parameters = {}, **kwargs):
Chris@99:     """Process audio data with a Vamp plugin, and make the results from a
Chris@99:     set of plugin outputs available as a generator.
Chris@99: 
Chris@99:     The provided data should be a 1- or 2-dimensional list or NumPy
Chris@99:     array of floats. If it is 2-dimensional, the first dimension is
Chris@99:     taken to be the channel count.
Chris@99: 
Chris@99:     The returned results will be those calculated by the plugin with
Chris@99:     the given key and returned through its outputs whose identifiers
Chris@99:     are given in the outputs argument.
Chris@99: 
Chris@99:     If the parameters dict is non-empty, the plugin will be configured
Chris@99:     by setting its parameters according to the (string) key and
Chris@99:     (float) value data found in the dict.
Chris@99: 
Chris@99:     This function acts as a generator, yielding a sequence of result
Chris@99:     feature sets as it obtains them. Each feature set is a dictionary
Chris@99:     mapping from output identifier to a list of features, each
Chris@99:     represented as a dictionary containing, optionally, timestamp and
Chris@99:     duration (RealTime objects), label (string), and a 1-dimensional
Chris@99:     array of float values.
Chris@143: 
Chris@143:     If you wish to override the step size, block size, or process
Chris@143:     timestamp method to be used, you may supply them as keyword
Chris@143:     arguments with keywords step_size (int), block_size (int), and
Chris@143:     process_timestamp_method (choose from vamp.vampyhost.SHIFT_DATA,
Chris@143:     vamp.vampyhost.SHIFT_TIMESTAMP, or vamp.vampyhost.NO_SHIFT).
Chris@99:     """
Chris@68: 
Chris@140:     plugin, step_size, block_size = vamp.load.load_and_configure(data, sample_rate, plugin_key, parameters, **kwargs)
Chris@64: 
Chris@112:     ff = vamp.frames.frames_from_array(data, step_size, block_size)
Chris@64: 
Chris@100:     for r in process_with_initialised_plugin(ff, sample_rate, step_size, plugin, outputs):
Chris@89:         yield r
Chris@61: 
Chris@89:     plugin.unload()
Chris@99: 
Chris@100: 
Chris@129: def process_frames_multiple_outputs(ff, sample_rate, step_size, plugin_key, outputs, parameters = {}):
Chris@100:     """Process audio data with a Vamp plugin, and make the results from a
Chris@100:     set of plugin outputs available as a generator.
Chris@100: 
Chris@100:     The provided data should be an enumerable sequence of time-domain
Chris@100:     audio frames, of which each frame is 2-dimensional list or NumPy
Chris@100:     array of floats. The first dimension is taken to be the channel
Chris@100:     count, and the second dimension the frame or block size. The
Chris@100:     step_size argument gives the increment in audio samples from one
Chris@100:     frame to the next. Each frame must have the same size.
Chris@100: 
Chris@100:     The returned results will be those calculated by the plugin with
Chris@100:     the given key and returned through its outputs whose identifiers
Chris@100:     are given in the outputs argument.
Chris@100: 
Chris@100:     If the parameters dict is non-empty, the plugin will be configured
Chris@100:     by setting its parameters according to the (string) key and
Chris@100:     (float) value data found in the dict.
Chris@100: 
Chris@100:     This function acts as a generator, yielding a sequence of result
Chris@100:     feature sets as it obtains them. Each feature set is a dictionary
Chris@100:     mapping from output identifier to a list of features, each
Chris@100:     represented as a dictionary containing, optionally, timestamp and
Chris@100:     duration (RealTime objects), label (string), and a 1-dimensional
Chris@100:     array of float values.
Chris@100:     """
Chris@129:     plugin = vampyhost.load_plugin(plugin_key, sample_rate,
Chris@100:                                    vampyhost.ADAPT_INPUT_DOMAIN +
Chris@100:                                    vampyhost.ADAPT_BUFFER_SIZE +
Chris@100:                                    vampyhost.ADAPT_CHANNEL_COUNT)
Chris@100: 
Chris@100:     out_indices = dict([(id, plugin.get_output(id)["output_index"])
Chris@100:                         for id in outputs])
Chris@100:     
Chris@100:     fi = 0
Chris@100:     channels = 0
Chris@100:     block_size = 0
Chris@100: 
Chris@100:     for f in ff:
Chris@100: 
Chris@100:         if fi == 0:
Chris@100:             channels = f.shape[0]
Chris@100:             block_size = f.shape[1]
Chris@100:             plugin.set_parameter_values(parameters)
Chris@100:             if not plugin.initialise(channels, step_size, block_size):
Chris@100:                 raise "Failed to initialise plugin"
Chris@100: 
Chris@100:         timestamp = vampyhost.frame_to_realtime(fi, sample_rate)
Chris@100:         results = plugin.process_block(f, timestamp)
Chris@100:         # results is a dict mapping output number -> list of feature dicts
Chris@100:         for o in outputs:
Chris@100:             ix = out_indices[o]
Chris@100:             if ix in results:
Chris@100:                 for r in results[ix]:
Chris@100:                     yield { o: r }
Chris@100:         fi = fi + step_size
Chris@100: 
Chris@100:     if fi > 0:
Chris@100:         results = plugin.get_remaining_features()
Chris@100:         for o in outputs:
Chris@100:             ix = out_indices[o]
Chris@100:             if ix in results:
Chris@100:                 for r in results[ix]:
Chris@100:                     yield { o: r }
Chris@100:         
Chris@100:     plugin.unload()
Chris@100: 
Chris@100: