PDCON:Conference/Electro-Acoustic Tools (EAT): High-Level Abstractions for Audio Manipulation and Spatialisation

Electro-acoustic Tools (EAT): High-Level Abstractions for Audio Manipulation and Spatialisation

Author: Richard Thomas

This paper introduces Electro-Acoustic Tools (EAT), a new set of accessible abstractions for composition and diffusion that are currently under development in Pure Data Extended. The aim of this project is to create well documented intermediate level tools that embrace the advanced functionality that Pure Data can provide for digital signal processing effects and spatialisation. The modules will include those for sound file playback, global transport with MIDI and parameter preset functions, elementary synthesis, granular synthesis, chorus, distortion, dynamic routing, spatialisation behaviours, MIDI control, phase shift, pitch shift, delay, reverb, and data analysis amongst others. The spatialisation modules will feature customisable semi-automatic metachronal wave and pseudo-random amplitude-based panning behaviours that can be added sequentially. Analytical modules will provide real-time feedback and facilitate the reflexive control of other module parameters using the harvested data.

The abstractions are being designed partly for use as educational tools with intuitive, homogenous, and dynamic graphical user interfaces to encourage inexperienced composers to experiment with electronic music. First time Pd users can be put off by the basic interface and idiosyncrasies of the program; EAT will provide an enticing segue into Pd for users of mid-level DSP applications or proprietary graphical programming environments. The importance of palatable data representation to the user will be explained and the benefits of the graphical and text-based EAT help files. The EAT tools will hopefully prove useful for experienced users seeking to use pre-built modules in their own work, or through the use of EAT’s lower- level development abstractions involving MIDI or dynamic object management, for instance.

After an overview of EAT’s functionality and the motivations behind the project, this paper will focus closely upon the use of dynamic patching, natively in Pd and using Thomas Grill’s [dyn~] external. Dynamic object management in Pd has hitherto been a subject area lacking thorough documentation, which this paper will alleviate through discussion of patching strategies and current workarounds. Dynamic patching is used in EAT to produce multi-monophonic effects, which can provide up to thirteen audio channels per abstraction instance and ten instance-specific MIDI and parameter presets. This paper will also cover the development of EAT’s dynamic multichannel spatialisation modules, with flexible input-to-output ratios and multi-point panning behaviours. The spatialisation modules will offer greater affordances than live diffusion using basic software or hardware mixing consoles with individual or grouped fader control.

Potential EAT user groups:

• Novice users with no prior experience of Pd.
• Intermediate users of proprietary graphical programming environments seeking an alternative.
• Experienced users wishing to experiment with pre-built DSP tools, dynamic patching, or spatialisation.
• Educators wishing to teach the principles of DSP using an applied approach without proprietary software.