AudioIO Objects

Provide real time audio playback and recording, with special classes to concurrently read input audio.

class pytta.Streaming(IO: str, samplingRate: int, device: int, datatype: str = 'float32', blocksize: int = 0, inChannels: Optional[ChannelsList] = None, outChannels: Optional[ChannelsList] = None, excitation: Optional[SignalObj] = None, duration: Optional[float] = None, numSamples: Optional[int] = None, monitor: Optional[Monitor] = None, *args, **kwargs)

Stream control.

__init__(IO: str, samplingRate: int, device: int, datatype: str = 'float32', blocksize: int = 0, inChannels: Optional[ChannelsList] = None, outChannels: Optional[ChannelsList] = None, excitation: Optional[SignalObj] = None, duration: Optional[float] = None, numSamples: Optional[int] = None, monitor: Optional[Monitor] = None, *args, **kwargs)

Manage input and output of audio.

Parameters:

• IO (str) – DESCRIPTION.

• msmnt (Measurement) – DESCRIPTION.

• datatype (str, optional) – DESCRIPTION. Defaults to ‘float32’.

• blocksize (int, optional) – DESCRIPTION. Defaults to 0.

• duration (Optional[float], optional) – DESCRIPTION. Defaults to 5.

• monitor (Optional[Monitor], optional) – DESCRIPTION. Defaults to None.

• *args (TYPE) – DESCRIPTION.

• **kwargs (TYPE) – DESCRIPTION.

Returns:

None.

__enter__()

Provide context functionality, the with keyword, e.g.

>>> with Streaming(*args, **kwargs) as strm:  # <-- called here
...     strm.playrec()
...
>>>

__exit__(exc_type: Type, exc_val: Exception, exc_tb: Type)

Provide context functionality, the with keyword, e.g.

>>> with Streaming(*args, **kwargs) as strm:
...     strm.playrec()
...                             # <-- called here
>>>

set_io_properties(io: str, channels: ChannelsList)

Allocate memory for input and output of data, set counter.

Parameters:

msmnt (TYPE) – DESCRIPTION.

Returns:

None.

set_monitoring(monitor: Optional[Monitor] = None)

Set up the class used as monitor. It must have the following methods with these names.

def setup(None) -> None:

_Call any function and other object configuration needed for the monitoring_ return

def callback(indata: np.ndarray, outdata: np.ndarray,

frames: int, status: sd.CallbackFlags) -> None:

_Process the data gathered from the stream_ return

It will be called from within a parallel process that the Recorder starts and terminates during it’s .run() call.

Parameters:

monitor (object) – Object or class that will be used to monitor the stream data flow.

runner(StreamType: Type, stream_callback: Callable, numchannels: Union[List[int], int])

Do the work.

Instantiates a sounddevice.*Stream and calls for a threading.Thread if any Monitor is set up. Then turn on the monitorCheck Event, and starts the stream. Waits for it to finish, unset the event And terminates the process

Returns:

Return type:

input_callback(indata: ndarray, frames: int, times: type, status: CallbackFlags)

This method will be called from the stream, as stated on sounddevice’s documentation.

output_callback(outdata: ndarray, frames: int, times: type, status: CallbackFlags)

This method will be called from the stream, as stated on sounddevice’s documentation.

stream_callback(indata: ndarray, outdata: ndarray, frames: int, time: type, status: CallbackFlags)

This method will be called from the stream, as stated on sounddevice’s documentation.

class pytta.Monitor(numsamples: int, samplingrate: int = 44100, numchannels: List[int] = [1, 1], datatype: str = 'float32')

PyTTa default Monitor base class.

__init__(numsamples: int, samplingrate: int = 44100, numchannels: List[int] = [1, 1], datatype: str = 'float32')

Default Monitor class.

Subclasses must override setup, callback and tear_down methods.

Parameters:

• numsamples (int) – DESCRIPTION.

• samplingrate (int, optional) – DESCRIPTION. The default is default.samplingRate.

• numchannels (List[int], optional) – DESCRIPTION. The default is [len(default.inChannel), len(default.outChannel)].

• datatype (str, optional) – DESCRIPTION. The default is ‘float32’.

Return type:

None.

setup()

Start up widgets, threads, anything that will be used during audio processing.

reset()

Reset write counter.

callback(frames: int, indata: ndarray, outdata: Optional[ndarray] = None)

The audio processing itself, will be called for every chunk of data taken from the queue.

tear_down()

Finish any started object here, like GUI members, to allow the Monitor parallel process be joined.