DAFx-12 / Program 15th International Conference on Digital Audio Effects DAFx-12 September 17-21, 2012 York, UK
Home Register Authors Venues Program Proceedings Pictures Committees Travel Guide Links

DAFx-12 Program

Keynote Speakers:

We are very happy to introduce our three keynote speakers for DAFx-12, all leaders in their respective fields:

(Tues 18) Trevor Wishart - The Music of Speech

Tuesday evening's social event is a performance of Trevor's latest work, Encounters in the Republic of Heaven. This is an 8-channel surround-sound piece based on speaking voices recorded in the North East of England where Trevor Wishart was Arts Council England Composer-in-Residence from October 1st, 2006 to September 2009, based at the University of Durham. The piece is constructed in 4 Acts of approximately 20 minutes each, combining portraits of individual speakers (accompanied by sounds and imaginary instruments derived from the voices themselves) - speech that waltzes, speech that locks in harmony, clouds of speech that circle the audience, culminating with speech that transforms into song.

The piece was completed on January 1st 2011 and has been performed around the world, but this will be its first performance in York.

(Weds 19) Trevor Cox - Back to Nature: Fantastical Echoes and Reverberations

Abstract:

Echo and reverb have been widely used as audio effects for decades. This paper explores unusual places, both ancient and modern, with remarkable natural sound effects including staircases that chirp, culverts that whistle and the most reverberant place in the world. The physics of the phenomena will be revealed, sometimes drawing on historical explanations, at other times using predictions from modern numerical models to visualise and better understand what is going on. The perception of listeners will also be explored, including what these spaces reveal about our hearing and our emotional response to architectural spaces.

Bio:

Trevor Cox is Professor of Acoustic Engineering at the University of Salford, a former Senior Media Fellow funded by the Engineering and Physical Sciences Research Council, and immediate past president of the Institute of Acoustics (IOA). One major strand of his research is room acoustics for intelligible speech and quality music production and reproduction. Trevor's diffuser designs can be found in rooms around the world. He has co-authored a research book entitled "acoustic absorbers and diffusers". He was award the IOA's Tyndall Medal in 2004. Trevor has a long track record of communicating acoustic engineering to the public and has been involved in engagement projects worth over £1M. He was given the IOA award for promoting acoustics to the public in 2009. He has developed and presented science shows to 15,000 pupils including performing at the Royal Albert Hall, Purcell Rooms and the Royal Institution.

Trevor has presented sixteen documentaries for BBC radio including: Life's soundtrack, Save our Sounds and Science vs the Strad. And is currently writing The Sound Book for Bodley Head.

(Thurs 20) Frank Melchior - Applications of Digital Audio Effects in Broadcast: Yesterday, Today and Tomorrow

Abstract:

The developments of broadcast technology are very related to the field of audio effects. In the early days a broadcast chain was more or less linear, only including different types of amplifiers. From there it has developed to a signal path with more and more signal processing. With the transformation into the digital domain digital audio effects have become important for the broadcaster on various levels.

Radio programmes like radio dramas or the BBC radiophonic workshop rely on audio signal processing and have transformed the use of audio effects into an art form. On the other side, processing driven by the needs of FM transmission like multiband dynamic processing are important for broadcasters to define the station sound and optimize coverage.

With the transformation of broadcasting systems in the future it is expected that also requirements and types of digital audio effects will change and develop further.

This talk will present some highlights on the history of audio processing in broadcasting and will try to highlight some potential options and needs for future developments in digital audio effects. Important aspects that will be discussed are the developments towards spatial audio effects, perceptual audio effects and personalized audio effects.

Bio:

Frank Melchior is head of the audio research team at BBC R&D. He received a Dipl.-Ing. (equiv. to Master of Science) in Media Technology from the Technical University Ilmenau and a doctoral degree on the topic of spatial sound design and acoustics from University of Technology, Delft, The Netherlands. Frank worked as a researcher and project manager at the Fraunhofer Institute for Digital Media Technology. He has spent his professional career developing audio systems including wave field synthesis based systems and leading R&D teams on various topics including authoring, live sound applications, automotive audio and motion picture sound. Before he joined the BBC he was the Chief Technical Officer at IOSONO, Erfurt, Germany, who develop spatial audio solutions. Frank's research interests include spatial array signal processing, spatial audio reproduction, user interface technology for audio systems and listening experiments.

Dr Frank Melchior is very passionate about finding the best way to bring the emotion and creative ideas embedded in sound to the listener.

Workshop/Tutorial Sessions:

The main DAFx program is supported by invited workshop/tutorial sessions in key topics of interest to the DAFx audience on the Monday directly before, and Friday after the three days of oral presentations and poster sessions. These workshops are open to all DAFx12 delegates:

[Mon 17 - #1] GPUs for DAFx - Lauri Savioja, Department of Media Technology, Aalto University School of Science, Finland.

Abstract:

Current graphics processing units (GPUs) are massively parallel computation engines, and they have gained lots of popularity in high-performance computing in the recent years. This workshop aims to give a general understanding on the capabilities of GPUs, how the parallel computation is performed on a GPU, and what kind of problems are suitable for parallelization. Two commonly applied ways to utilize GPUs are introduced. While CUDA is the most typically applied programmer-level API, there is a good interface to use a GPU directly from Matlab, as well. The workshop will have several examples on utilization of GPUs on practical real-time audio signal processing tasks.

Bio:

Lauri Savioja received the degree of Doctor of Science in Technology, from the Helsinki University of Technology (TKK), Espoo, Finland, in 1999. He majored in computer science, and the topic of his doctoral thesis was room acoustic modeling. He worked at the TKK Laboratory of Telecommunications Software and Multimedia as a researcher, lecturer, and professor from 1995 until the formation of  Aalto University where he currently works as a professor in the Department of Media Technology at the School of Science. The academic year 2009-2010, he spent as a visiting researcher at NVIDIA Research.  His research interests include room acoustics, virtual reality, and parallel computation.

Prof. Savioja is a member of the Association for Computing Machinery (ACM), Audio Engineering Society (AES), and a life member of the Acoustical Society of Finland. From 2010 he has been an Associate Editor of the IEEE Transactions on Audio, Speech, and Language Processing.

[Mon 17 - 2#] Pointers to More Effective Software and Data in Audio Research - Mark Plumbley, Director of Centre for Digital Music, Queen Mary University of London, UK.

Abstract:

Sustainable and reusable software and data are becoming increasingly important in today's research environment. Methods for processing audio have become so complex they cannot fully be described in the printed paper. Researchers are becoming increasingly aware of the benefits of open access publication and reusable software and data in helping others to build on their work. However, many researchers find this difficult to achieve in practice. In this workshop we will explore ways to maximise the accessibility and impact of research through publication of research software, availability of research data, and open access paper publication. Topics will include: software version control; unit testing; writing maintainable code; code reviews; choice of language and platform; software licensing; data repositories and data management; levels of open access publication, negotiating publication agreements, and associating papers with software and data publications. The workshop will be of interest to researchers who wish to make immediate improvements to the effectiveness, sustainability and impact of their work, as well as to research group leaders with an interest in policy and management.

Bio:

Mark Plumbley is Director of the Centre for Digital Music (C4DM) at Queen Mary University of London, and leads the EPSRC-funded soundsoftware.ac.uk initiative. His work in audio signal analysis includes beat tracking, music transcription, source separation and object coding, using techniques such as neural networks, independent component analysis, sparse representations and Bayesian modelling. Prof. Plumbley is a member of the IEEE Machine Learning in Signal Processing Technical Committee, and leads the UK Digital Music Research Network.

Mark will be leading this workshop accompanied by his colleagues:

• Chris Cannam is a software developer with 15 years commercial and extensive open-source development experience. While at the C4DM he has worked on the widely-used Sonic Visualiser cross-platform audio analysis and visualisation application; Sonic Annotator, a tool for batch extraction of meaningful features from audio files; the Vamp plugin API for audio feature extraction, and many plugins using this API; and tools and ontologies for music description using RDF within the Semantic Web.


• Luis Figueira is a software developer with more than 5 years of experience with C/C++, Scheme, Web technologies and databases. He has an MSc in Electrotechnical and Computers Engineering from Instituto Superior Técnico in Lisbon, where he specialized in digital signal processing with a focus on speech synthesis. Luis has recently worked in a speech technology spin-off and an open-source web development company.


• Steve Welburn returned to education in 2005 for an MSc in Intelligent Systems at University College London, after 18 years in commercial IT. Following his MSc, he studied for a PhD at the Centre for Digital Music. Since December 2011 he has been working as a research assistant on research data management projects at the C4DM.

[Mon 17 - #3] FDTD Methods for Physical Modeling Sound Synthesis and Audio Effects Processing - Stefan Bilbao, Senior Lecturer in Acoustics, University of Edinburgh, UK.

Abstract:

This tutorial is intended as an introduction to time stepping methods for physical audio synthesis and effects processing, with some emphasis on how they fit into the larger picture of more standard audio signal processing/synthesis techniques. Various examples will be discussed and worked through in detail, spanning the range of acoustic musical instruments, electromechanical instruments and reverberation devices, and full 3D room acoustics simulation. Audio and video examples will be presented, along with Matlab code walk-throughs.

Bio:

Stefan Bilbao is currently a Senior Lecturer in Acoustics at the University of Edinburgh. He was formerly a lecturer at the Sonic Arts Research Centre, at the Queen's University Belfast, and a postdoctoral research associate at the Stanford Space Telecommunications and Radioscience Laboratory. Prior to that, he received the MSc and PhD degrees in Electrical Engineering at Stanford University while working at the Centre for Computer Research in Music and Acoustics, and a BA in Physics at Harvard University, and in between spent two years in the real-time group at the Institut de Recherche et Coordination Acoustique/Musique in Paris.

His main interests are in sound synthesis, numerical method design, and parallel processing for audio on GPUs.

[Fri 21 - #1] The Acoustics of the Singing Voice - David Howard, Head of Department of Electronics, University of York, UK.

Abstract:

Singing is a way humans have expressed themselves from earliest times, and the range of sounds that they can achieve from what is basically a variable-pitch buzzer and a flexible short tube is immense. This workshop will explore the physiology of the human singing voice and relate its elements to the nature of the acoustic output and how it can be varied. Shape changes that are made to the flexible tube or 'vocal tract' will be explored based on magnetic resonance images (MRI) from which the airway itself can be reconstructed. This enables the shape changes to be investigated in terms of the physiology as well as the acoustics by means of digital waveguide synthesis. Reference will be made to choral singing; in particular tuning in unaccompanied or 'a cappella' singing.

Bio:

David Howard is Head of the Department of Electronics and Head of the Audio Laboratory at the University of York. His research centres around the analysis and synthesis of singing, speech and music and the use of real-time visual displays in vocal pedagogy. David has been an Engineering and Physical Sciences Research Council Senior Media Fellow and he has been involved in a number of radio and TV programmes mainly about the science of sound, particularly singing and music. In the BBC4-TV programme 'Voice' he worked with Jeremy Hardy on his singing(!) and Rory Bremner on his impression of Tony Blair. Recently he was involved in Channel 4's 'Hidden Talent Show'.

David conducts the 'Vale of York Voice'; a local choir that sings evensong in York Minster about once a month.

[Fri 21 - #2] Practical Ambisonics - Bruce Wiggins, Lecturer in Electronics and Sound, University of Derby, UK.

Abstract:

Ambisonics, with its channel based, speaker agnostic approach to surround sound encoding and decoding, has always been problematic to use in speaker layout obsessed Digital Audio Workstations. This session will explain what Ambisonics is and demonstrate how to setup and use the WigWare Ambisonic Plug-in Suite with the hierarchical routing structure presented by Reaper paying particular attention to the differences between first and higher order Ambisonics, and the lesser used near field compensation and distance filters. Example mixes created by 2nd year undergraduate students at the University of Derby will also be presented, along with newly developed teaching aids and visualisations that are soon to be released as open educational material for the dissemination of Ambisonics to an audience new to the subject.

Bio:

Dr Bruce Wiggins graduated with a 1st class honours in Music Technology and Audio System Design from the University of Derby in 1999. His interest in audio signal processing spurred him to continue at Derby completing his PhD entitled "An Investigation into the Real-time Manipulation and Control of 3D Sound Fields" in 2004 where he solved the problem of generating Ambisonic decoders for irregular speaker arrays and also carried out work on binaural/transaural reproduction systems.

He is now a lecturer in the Electronics and Sound subject group at the University where he teaches electronics, audio programming and digital signal processing which are all fed from his continuing research interests in Ambisonic surround sound systems earning him excellence awards for his promising research in 2005/6, his application of technology in 2006/7 and excellence in learning, teaching and assessment in 2007/8.

Satellite Workshop:

The SoundSoftware and Software Carpentry projects are organising a three-day bootcamp on software development, during the week preceding the main DAFx-12 conference. The bootcamp will run from the 13th to 15th of September 2012 at the University of York.

The goal of the bootcamp is to help scientists and engineers become more productive by teaching them basic software development skills like program design, version control, testing, and task automation. In this three-day boot camp, short tutorials will alternate with hands-on practical exercises. Participants will be encouraged both to help one another, and to apply what they have learned to their own research problems during and between sessions. There will be online follow up sessions for 6 to 8 weeks extending the material from the boot camp.

More information and separate registration for this event is available at the York Bootcamp Website. Note that this workshop is not included in the main DAFx-12 registration, but delegates may wish to take advantage of this event alongside their attendance at the DAFx conference.

Main DAFx-12 Program:

The final program below is also available as a PDF here.

DAFx-12 Complete Program: