SPPySound

SPPySound is a python project that acts as a wrapper to the pysndfile library, adding aditional functionality needed for audio and DSP related projects. Currently the project is focused on the generation of audio descriptor meta data for sound databases as part of my final year project. The project is currently in development but aims to generate a range of useful functions for sound manipulation. The project is designed with flexability in mind so that new functions and features can be added as needed. To address this the project has a modular nature and currently consists of 3 main classes for the encapsulation of audio files:

Main audiofile Module

AudioFile

This is the base class for creating an audio file object and has methods for simple processed on the audio file. Functionality such as converting the file to mono and convenience functions for reading and writing grains of audio are included.

AnalysedAudioFile

Inheriting from the AudioFile class, this encapsulates the audio file and the analyses that can be generated through the seperate audio analysis modules. This will be used for the processing of analyses of audio files and to group audio files with their analyses.

AudioDatabase

An audio database is a class used to group AnalysedAudioFiles for batch processing. This class will be used for comparison methods to compare audiofiles and their meta data.

Analysis Modules - Under development

RMSAnalysis

Generates an RMS analysis file. Currently stored as a plain text file. Values can be accessed through RMSAnalysis class methods.

ZeroXAnalysis

Generates a windowed zero-crossing anaysis of the audio file Currently stored as a plain text file Values can be accessed through ZeroXAnalysis class methods.

AttackAnalysis

Generates an estimate for the attack start and end of a monophonic audio file. Will only work on single event audio files Estimation algorithm needs improvments.

Project Roadmap:

Choose best audio descriptors for implementation [ ] Create a few basic descriptors to test matching on [x] Create database object matching algorithm [ ]

• Starting with a brute force algorithm that tests every entry against every other entry [ ] Create simple synthesis algorithm [ ] Create all descriptors required for project [ ] Work on improving matching and synthesis algorithms [ ] Create front-end? [ ]

Package Installation

Simply run: pip install . from the package folder Alternatively: pip install -e . will install the package keeping it's contents in place in the folder so that they can be easily developed and tested.