AudioTagger

A simple program to tag wav files by drawing bounding boxes on the spectrogram.

Quick Install

On most systems it should work to

Install Miniconda Python 2.7 distribution (http://conda.pydata.org/miniconda.html). Then type in the terminal:

conda config --add channels https://conda.binstar.org/groakat
conda install audiotagger-dependencies
pip install -U https://github.com/groakat/AudioTagger/archive/master.zip

Now you can start the AudioTagger by opening a terminal and simply typing

AudioTagger

If this does not work, please proceed reading the remainder of the README. Otherwise you are done. :)

Using the github client

On Windows and maybe Mac OSX git might not be installed. Then the easiest is to install the github client and synchronize thsi repository.

Once it the repository is synchronized, open a terminal and type

cd C:\path\to\github\AudioTagger

By replacing C:\path\to\github\AudioTagger with the path where the github client saved the AudioTagger repository. The folder you are looking for has a setup/py file. Now type

python setup.py develop

You should now be able to open the AudioTagger by simply typing AudioTagger in the terminal. (You might have to iopen a new terminal for it to work.

Dependencies

The program should run with an anaconda installation plus the following dependencies:

My fork of sound4python https://github.com/groakat/sound4python (please follow the readme of this project to install sox):

pip install https://github.com/groakat/sound4python/archive/master.zip

My fork of qimage2ndarray:

pip install https://github.com/groakat/qimage2ndarray/archive/master.zip

---

General dependencies (if not anaconda is used as python distro):

pip install numpy
pip install scipy
pip install PySide

Detailed Install Instructions

You need to

• download anaconda
• install the dependencies
• install the AudioTagger

Downloading Anaconda

On any system, install anaconda 64bit (if possible, otherwise use 32bit and install the dependencies by hand). If you have the Enthought distro remove it. Click the boxes that say

1. Add to system path and variables
2. Register anaconda as the system Python 2.7

Installing dependencies

If you have Windows 64 bit (all newer versions) or OSX, follow the instructions in the section "Installing dependencies using Conda (Windows 64bit or OSX)" and then continue with "Installing github and audioTagger".

If you have Windows 32 bit or Linux, follow the instructions in the section "Installing dependencies manually (Linux, Windows 32bit)" and then continue with "Installing github and audioTagger".

Installing dependencies using Conda (Windows 64bit or OSX)

If you have Windows 64 bit (all newer versions) or OSX, you can install all dependencies within anaconda by following the following steps (then continue with the github part below):

Open a command window (Click the start button and type cmd) In the command window type

conda config --add channels https://conda.binstar.org/groakat
conda install audiotagger-dependencies

Installing dependencies manually (Linux, Windows 32bit)

If you have an 32bit version of windows, you have currently to follow this steps and install the dependencies manually:

Open a command window (Click the start button and type cmd) In the command window type

conda install PySide

followed by

pip install https://github.com/groakat/sound4python/archive/master.zip

(please follow the instructions on https://github.com/groakat/sound4python to install sox as well.)

and finally do

pip install https://github.com/groakat/qimage2ndarray/archive/master.zip

Installing

stable install (standard)

Open a terminal and type:

pip install git+https://github.com/groakat/AudioTagger.git

github and audioTagger

Download GitHub for windows (https://windows.github.com/) or Max OSX (https://mac.github.com) After signing in hit plus in top left to add repository Click clone Select groakat/AudioTagger and specify a folder on your computer to save it to e.g. C:\Users\localadmin\projects\.

Now open a terminal (Windows: hit [start-key] and type cmd [return]) and navigate to the path you have downloaded the AudioTagger to using GitHub, e.g. cd C:\Users\localadmin\projects\AudioTagger. Now you can install the AudioTagger by typing

python setup.py develop

To run the AudioTagger, open a terminal (Windows: hit [start-key] and type cmd [return]), and navigate to the audioTagger.py file. E.g. cd C:\Users\localadmin\projects\AudioTagger\AudioTagger and run:

ipython audioTagger.py

You will then be asked for the folder containing your wav files and a folder to save you output annotations in.