Introduction

Welcome

My plan is to build on the work of the Creative Speech Technology Network by providing a platform for discussion on the creative use of speech technology. I have chosen to call this “The Art of Artificial Voices.”

If your work with speech technology has an interdisciplinary flavour, touching upon both art and science, you may find it valuable to register and post. A significant additional resource may also be found by going to the archive of the CreST Network project (linked above) – However this is now a few years old and some of the material and links may no longer exists or not  be as relevant as originally intended.

The plan will take time to realise so please bear with me.

Chris Newell c.newell@hull.ac.uk

I became interested in the potential of artificial or computer generated voices as a consequence of working with actors. At the time I was assisting Peter Hall on a season of Shakespeare plays and he asked me to train the actors in a method of speaking verse he and John Barton had evolved at the Royal Shakespeare Company. The method was very procedural and structured, producing spectacular results very quickly. It gave an impression of intelligence when none was necessarily present. This experience was a big influence on my subsequent work. I became more and more interested in disconnecting human intelligence from acting and finding ways of acting that were rule based rather than intuition based. Eventually this evolved into the subject of a PhD in Computer Science at the University of York. I was lucky enough to find a supervisor in Alistair Edwards brave enough to take the risk and foolish enough to let me add a significant amount of theatre practice to a traditional science PhD. I am also indebted to Roger Moore. He was prepared to listen to my ideas, these are too naïve to report in detail but they were primarily focused on how I would revolutionise speech synthesis by getting voices to pause more. As you can imagine these ‘innovations’ came and went.

Having served my time as a not particularly good director prior to my new career as an academic, I have found the freedom from human actors and their demands particularly pleasing. Computers do not look for motivation or a back-story, they usually do what they are told. This megalomanic status means I am able to create work (be it without the normal visual accoutrements) in the privacy of my studio. My actors are the voices that cleverer scientists than me have created and my medium is either melodrama (let’s call that speech and music) or radio drama. I have had several goes (including several failures) at creating interesting work and there are several more on the go. Meanwhile I hope to stockpile everything there is to know about this very particular art on this site. I hope others prepared to cross the line between art and science will help me out.

A few thoughts on contributions. Initially I will be moderating them all until we have begun to identify our niche and our style. If we write posts in a highly academic or arty style we may be in danger of becoming just another self serving resource we can all claim on our REF return (a silly way of scoring academic credibility (explanation of annoying acronym for the non academic community)). It may be better to write in a chattier style – for example: please don’t link to academic papers because I for one won’t read them. Risk a bit of stream of consciousness and don’t get hung up on spelling and grammar. Perhaps we should remind ourselves that we are dealing with a subject (the voice) which has great emotional, possibly spiritual and certainly philosophical significance when embodied by a machine and also this is a blog, not a conference platform and certainly not a journal.

I particularly encourage contributions from people who use speaking machines. Many of these may have a disability and many of those may be frustrated by the current absence of great creative potential demonstrated by their machines. It would be nice to get together and do something creative.