Skip to content

User guide

Installation

We highly recommend installing into a virtual environment.

Creating a virtual environment

To create a virtual environment we can use Python's inbuilt venv tool. Simply cd into the TAAT repository and execute the following command:

python -m venv .venv

This will create the hidden folder .venv at the top level of the repository.

You can call the virtual environment anything, but if you have cloned the repository we recommend only using the names venv, .venv, env or .env, as these will be ignored by Git.

To use your virtual environment you will need to activate it, using:

source .venv/bin/activate

Note the relative path to the activate command. When you are done with the virtual environment you should deactivate it using:

deactivate

Install using pip

To install taat using pip, from the top-level TAAT directory execute the following:

python -m pip install .

Install using conda

TBC

Basic Usage

The code below provides an example of basic TAAT usage. For more advanced examples see the Tutorials page.

from taat import query


FEATURES = ["melspectrogram", "tempogram", "rms", "spectral_centroid"]

# Run a query.
query_result = query(source_dir="path/to/audio/files/to/query/against",
                     query_filepath="path/to/file/to/query.wav",
                     features=FEATURES,
                     sr=16000,
                     k=7,
                     n_fft=2048,
                     hop_length=1024)

# Write matches to disk as audio files
query_result.write("path/to/output/folder")

The taat_interpreter Max patch

We also have a Max patch available, the taat_interpreter (created by Dr Sam Gillies), which allows to load exported JSON output from a query result and display the it as playable audio waveforms:

image