sonic surrounds

Some more juicy sonic cherries picked from ICAD Day 2 …

Scott Brewer from the SIAL Sound Studio, RMIT, Melbourne, presented a poster, “CREATING A VIRTUAL SUIKINKUTSU”, describing the SuperCollider physically modeled synthesis of the Suikinkutsu - a Japanese garden feature that implements a below-ground chamber that resonates the sounds of water drips running off from a hand-wash bowl - kind of like a miniature of Jem Finer’s Score for a Hole in the Ground. Here’s a diagram (taken from a book by Morozumi) in Brewer’s paper:

K. Vogt presented a poster “SONIFICATION OF SPIN MODELS. LISTENING TO PHASE TRANSITIONS IN THE ISING AND POTTS MODEL.” which utilized two personal audio processing favourites of mine - spatialization and granulation to model particle spin theory. The result was burbley sound textures that morphed out of strongly pitched textures.

Peter Lennox (of University of Derby’s SPARG), in his paper with Tony Myatt, titled “CONCEPTS OF PERCEPTUAL SIGNIFICANCE FOR COMPOSITION AND REPRODUCTION OF EXPLORABLE SURROUND SOUND FIELDS”, mentioned the idea of cartoonification of spatial sound mixes - as a concept to economically increase the realism of virtual audio environments by focusing on creating the right *impression*, rather than painstakingly recreating exact features that may not be perceived for some reason, or that distract or diminish the desired impression. Of course, this cartoonification is exactly what film sound designers already do, though Lennox talks of taking this concept into the world of spatial sound rendering. He also made a strong argument for returning space to music, through applications such as spatial DJing - as a more sophisticated version of sound diffusion performance capability.

Another item from Day 1 that I forgot to mention in the previous post - the poster presentation titled “USER-SPECIFIC AUDIO RENDERING AND STEERABLE SOUND FOR DISTRIBUTED VIRTUAL ENVIRONMENTS”, by Mike Wozniewski, Zack Settel and Jeremy R. Cooperstock.

This work created a 3d-game like graphical environment with visible objects that also represent auditory sources and sinks. The result is kind of like a cross between Second Life and Native Instruments’ Reaktor, or some other audio processing patcher software. The user can take a first-person or second-person perspective of a world that contains sounding objects, other users, and sound-modifying objects. For example, the demonstrator wore a microphone headset that fed audio into the environment (which itself contained virtual objects that generated spatial audio), and was able to effect his voice by aiming a wireframe, cone-shaped sound radiation pattern towards the effector objects - such as a flanger. Other sounding objects contained sound loops that could be similarly triggered by the viewer’s site lines. Overall, the environment enabled an enjoyable, playful collaborative, networked, immersive experimental audio experience. Nice work. Here’s a quote from the paper abstract:

We present a method for user-specific audio rendering of a virtual environment that is shared by multiple participants. … Spatialization of sound sources is accomplished via acoustic physical modelling, yet our approach also allows for localized signal processing within the scene. In order to control the flow of sound within the scene, the user has the ability to steer audio in specific directions. This paradigm leads to many novel applications where groups of individuals can share one continuous interactive sonic space.

And an image from the paper:

On a separate note, Somaya Langley has an ICAD photoset on her Flickr account (hopefully not censored!), including the photo below of a slide showing the geodesic head-cage for HRTF capture via reciprocity technique from a paper I mentioned on the previous post.

Today is the first day of my third ICAD in Montréal (program available here), after Sydney 2004 (as audience only) and Limerick, Ireland, 2005 (when I presented at the student thinktank). This morning I presented my paper, “Mitigation of Binaural Front Back Confusions by Body Motion in Audio Augmented Reality”, on one of my perceptual experiments characterizing mobile augmented reality (AR) performance - a part of my PhD research.

Being preoccupied with final preparations, I missed the details of early presentations, though it was impressive to see the sea of wireless headphone-clad ICAD audience listening to binaural examples. Below are some memorable observations of the day.

One notable idea that I’m surprised not to have come across before was a process to capture 128 simultaneous Head Related Impulse Responses (HRIRs) from a single subject in a matter of seconds, using the reciprocity technique, where 128 microphones mounted on a very sci-fi polyhedral head cage capture signals emitted by tiny ear-mounted loudspeakers - a reversal of the usual microphones-in-ears technique. This idea and the below image are from the paper “PERCEIVED NATURALNESS OF SPEECH SOUNDS PRESENTED USING PERSONALIZED VERSUS NON-PERSONALIZED HRTFS” by John Usher and William L. Martens.

Another noteworthy paper was titled “SPATIAL AUDIO QUALITY EVALUATION: COMPARING TRANSAURAL, AMBISONICS AND STEREO” by Catherine Guastavino, Véronique Larcher, Guillaume Catusseau and Patrick Boussard. The paper described two experiments evaluating subjective spatial quality and localization errors for the two multichannel spatial audio presentation methods on a 6 speaker array vs 2 and 4 speaker transaural, and showed that Ambisonics provides best sense of immersion, while pairwise amplitude panning provides best llocalization errors.

ICAD afternoon 1 presented a series of sound-art and music tape/(actually, hard disk) performances, including various performances by Pauline Oliveros (presented in person), with comments on the acoustic and artificial spatial environments as co-performer. The day finished with a networked performance by Tintinnabulate Ensemble, involving Oliveros, Jonas Braash, and others on several acoustic instruments and computers, with video and audio link between two internet-joined sites, multichannel speaker array diffusion and simultaneous binaural headphone channels involving binaural signals captured via dummy head in the first row of the theatre… a swagful of technology and complication that reached some interesting moments, but also some limitations of the delayed interaction and tendency towards continual playing by all musicians… nevertheless an interesting musical / technological experiment in line with the expansive mindset of the conference. Now to leave the wifi and get out before day two arrives.

here’s a sketch of an improv performance I’m posting as a reminder to resume this particular experiment when I have more free time after finishing the thesis.

this piece is done in Pd, using field recordings of a street cleaner machine; also some granulation controlled by a particle swarm simulation, providing some great organic bass sounds in the later stages (and mids earlier on); a glitch/pop generator run through a loop catcher; and an inertial-physical-model-controlled drone providing some slowly descending glue tones, though occasionally damaged by very slight dropouts serving as a reminder to use Pd with Jack for better reliability.

as i said, a sketch only, but with elements well worth a revisit. the swarm-granular-bass might be an interesting anchor for some warped dsp dubstep hybrid (!?).

 improv-pd-gran-swarm-etc: Play Now | Play in Popup | Download

French-born, Barcelona-based sound designer and artist, Philippe Faujas presents Kiln, a multi-form sound artwork presented as downloadable software, online flash composition and presumably as a physical installation (?), composed with sounds from SoundTransit.nl, where selected cycles of audio interspersed with silence, play back from multiple speakers in a darkened room with concert seating. The sounds are all creative commons licensed, including my binaural wisdom tooth extraction. I wish I could visit the actual installation, though the online flash version is a nice piece itself. Great to see another example of how well Creative Commons can work to foster reuse of media materials.

* Derek Holzer (Binaural Tunnel Study, Binaural rainstorm, Seto song)
* Yannick Dauby (Bats-Echolocation)
* Planktone (Industry 2, Windmill)
* Cedric Peyronnet (Fences And Wind)
* Dallas Simpson (Binaural Environmental)
* Nick Mariette ( Wisdom Tooth Extraction In Binaural)
* John Tenny (Desert Wind In The Hall)