MEETING REPORT
Attendees:
Andy Hunt introduced the session by suggesting that PD was Great (flexible, free, open-source etc.) and Scary (few defaults, blank screen, diffuse tutorials mostly written by developers). The group joined in with their own experiences and thoughts about PD.
Sandra Pauletto described the basic structure of PD and its associated libraries.
She uses zexy (for matrices), GEM for graphical output, GriPD for user interface controls, a VST object for plug-ins, and an adcReader object for the Picotech interface. She showed us various objects which she had written (with a brief explanation of what is involved to do this), followed by a demonstration of a Vowel Synthesiser using a VST plug-in.
She also demonstrated a suite of interactive programs for the general public to allow them to modify sounds. This used a GriPD interface to hide the traditional programming screens, and to present a series of pictures, text boxes, buttons and sliders which were restricted to allow a certain amount of experimentation.
This was followed by an experiment to compare the audio and visual diplays of a complex data set. This used PD to gather and process the data, while a GriPD interface again handled the user interaction.
Sandra described the result of a COST STSM (Sandras visit to Bielefeld to work with Thomas Hermann), on a multidimensional representation of helicopter data. Finally she showed us the interactive sonification toolkit which is made from PD and GriPD.
Liz Bramley is a secondary school Music Teacher, currently doing a Music Technology Masters at York, and has seen how music technology has been used in schools over a decade. She is just starting a project to use PD creatively in secondary schools, not just for music, but for other subject areas such as physics. Personally she has found PD useful for experimenting with synthesis, random sound generation and graphical output via GEM etc. She has tried PD on an interactive whiteboard (SmartBoard). She has found the Help patches useful, but they will need re-writing. She sees the advantages as being the free availability and the fact that it can be used on a PC without any other special hardware and it can be used at home! She aims to develop workshops for the creative use of PD in secondary schools. Feedback from the workshop was that it would be great to have the results of this available to the community.
Roberto Bresin, Sofia Dahl and Kjetil Hansen described various projects using PD at KTH, Sweden. They described how they teach PD at KTH in the Musical Communication and Music Technology course. 3 hours of PD tutorials in a computer room (2 people at a computer), then a 3 hour lab with audio, MIDI controllers etc. 30 hour project, with some students using PD.
The lab exercise was shown (e.g. how to build a Theremin), with background, instructions, links to be read in advance of the tutorial. The teaching patches were very impressive encouraging students to construct solutions to sound problems by configuring some pre-provided objects.
They then demonstrated a range of patches from KTH TMH projects, including:
- an output of the Sounding Object project (allowing sensors including a mobile phone! to control a drumming model).
- the Ballancer a model of a rolling ball driving an impact sound model, and controlled by a tilt device.
- an interactive interface and control screen for DJ scratching (and how to map the inputs to various control parameters)
- Director Musices, which allows expressive rules to be applied to MIDI files, with real-time control inputs to shift the emotional state of the music as it plays.
- Expressiball - a visualisation of the performance parameters of a piece of music, using GEM
Thomas Hermann gave us an enlightening overview of the differences between PD and Supercollider. He gave the presentation in the Supercollider environment, which meant that we saw and heard all the examples as he went along, executing lines of text, as it is basically a text evaluation environment.
He summarised that while PD is advantageous in interaction with the mouse, the text philosophy of Supercollider allows powerful scripting, and is operable by blind people (unlike PD). Perhaps text-based programming and graphical interfaces appeal to different people and possibly use different parts of the brain.
The Workshop participants then had a chance to try out the presenters demonstrations for themselves, and to break into smaller groups for discussion.
Final Discussion:
This is free software which exists because people need it, and by sharing it with others a community has been built. So, in turn, we should make our work available too. Programs are not just successful because they are free- it needs organisation and dissemination.
How to introduce PD educationally.
We should start - Good web-links e.g. a good place to start for PD.
Nicola stressed the importance of licensing any work that we put on the web. (see www.gnu.org for more detail or www.creativecommons.org )
Andy Hunt offered to collect all the presentation materials and to make them available on the workshop website.
Idea: a list of problems that we encountered.
Manuals: Could we gather together the different materials that we have currently have within ConGAS.
Idea: How to move from one program to another (e.g.MAX-to-PD)
Generally we need to assess what is currently there, before we start writing new materials.
But after introducing PD in a general sense we need to describe how it relates to ConGAS, i.e. how PD can be used in the context of gestural control of audio systems.
So, how about linking to completed projects that have been done in PD.
The problem is getting the information you want when you need it (e.g. when youre developing does anything else exist out there before I write it).
In other words how can people find the information they need about PD at a level that is appropriate for their position on the learning curve.
Perhaps we need a PD for Dummies book, or at least a first few chapters of such a book. Asri explained that there seems to be missing a document which explains the abstract philosophy behind PD, which will aid a deep learning, rather than diving into complex usage examples.