454 J. Audio Eng. Soc., Vol. 61, No. 6, 2013 June

The AES 49th International Conference, Audiofor Games, was held in London this year fromFebruary 6th through 8th. This was the third

AES conference devoted to game audio, held everytwo years in London. The conference is one of thelargest conferences in the world focused specificallyon sound and technical art of interactive audio. Look-ing to bring together the working practices, discus-sions, innovations, and diversity in interactive audioacross professional game development and academicresearch, the AES Audio for Games conference helpsbridge the gap between works done across differentdisciplines in interactive audio. The inaugural AESAudio for Games conference in 2009 (the 35th) ush-ered in the beginning of the current video game con-sole generation. As the industry stands on the cusp ofanother transition to new technology, the importance

of knowledge sharing between theoretical researchinto interactive audio and the practical application oftechniques is more critical than ever.The facilities at 2 Carlton House Terrace, where the

conference was held, provided an intimate and flexi-ble space to accommodate the free flow of ideasthroughout the week. The main presentation hallplayed host to over fifteen speakers, and another audi-torium hosted the paper and poster sessions. Whenthis second auditorium was not in use, a live videofeed from the main presentation hall was projected,allowing for overflow and to help inspire furtherconversation. In addition to these spaces there werethree smaller rooms that ran short sponsored educa-tional sessions throughout the week.The central London location proved the perfect

stage for the three days of conference proceedings in

AES 49th International Conference

Audio for GamesLondon, UK6–8 February 2013

CONFERENCE REPORT

CONFERENCE REPORT

J. Audio Eng. Soc., Vol. 61, No. 6, 2013 June 455

Michael Kelly, Conference Chair

Sponsored by

CONFERENCE REPORT

an inspiring program of sessions, workshops, and tutorialscentered on game audio. Split between the main conferenceroom, an auxiliary auditorium, and several small breakoutrooms, the space was large enough in scope with added diversityto suit the different program materials. While all of the sessions,along with the keynote, were held in the main room, papers werepresented downstairs, along with sponsored tool tutorials anddemonstrations in the break-out rooms. The diversity of presen-tations allowed the opportunity to fine-tune the focus of one’sexperience by easily mixing between the main track, technicalpapers, and tools/tutorial discussions.The first day of the conference started with short introductions

that found conference-goers in a crystalline state of anticipationthat shaped the next day’s events. Attendees had come from allover the world to participate in focused presentations, technology,and spirited discussions centered around interactive, and specifi-cally, game audio. The palpable excitement of these first moments,punctuated by stories of commutes to the conference on the ubiq-uitous London Underground, found everyone quickly warming tothe idea of being surrounded by colleagues and old friends. Thebest of the game audio community was out representing legendaryLondon studios such as Sony Computer Entertainment Europe,Codemasters, Lionhead, and EA in addition to European favoritesGuerrilla, IO Interactive, Wargaming, and North America’s finestfrom Microsoft, Audiokinetic, and a contingent of developers fromJapan and Australia.One of the pervasive aspects apparent at this year’s conference

was the presence of dynamic mixing and loudness trends. Theseaspects were pervasive throughout several talks and panels overthe course of the three days. It would seem that with the escalat-ing voice count and increased complexity of dynamic systems thattechniques and tools are finally emerging to address this formida-ble task. From Garry Taylor’s talk on loudness standards, to apanel of game audio experts discussing the disparity between filmand game mixing, a thread of best practices was discussed thatfurther overflowed into a panel discussing the future of gameaudio. Meanwhile Xavier Buffoni’s talk on HDR Audio exposed atopic that has been circling around in proprietary toolsets foryears, and is now poised to become accessible to users in the forth-coming generation of middleware. Coupled with the Atmosdemonstration during Dolby’s sponsored event, dynamic mixing

was a topic of conversation for many. It became clear through theweek that it is a topic at the forefront of people’s thinking.

KEYNOTE

This year’s keynote featured British composer Robin Rimbaud(aka Scanner) who chronicled his career as a fearless

designer of sounds across many media. While he’s not beeninvolved with video games directly, his work has often touchedon real-time collaborative interactions, be it with people directlyor within a space. His rapid-fire explanation of various worksassured that there was something that everyone could connectwith—an understanding of the overlap and a feeling of inspira-tion from his work creating unique audio experiences. In one example he told of a bus ride through London where

microphones attached to the outside and speakers on the insideprovided a dynamic and randomized sample-set that he mixed inreal-time during the course of each sojourn. Guaranteed to bedifferent every time, the audience was treated to an aural view ofLondon outside their usual experience (and in one case, heexplained, a very unusual experience!). Another project thatprovided direct interaction between participants was an installa-tion that consisted of a large geometric cube housed outdoors inBrooklyn, NY, that could be walked through. A series of triggerscontrolled sample playback and lighting effects throughout thestructure, providing an interactive soundtrack based on move-ment throughout the space. As Michael Kelly explained in his introduction, Robin was a

risky choice as a keynote speaker because it was unknown howwell his career would translate to the discipline of interactiveaudio and specifically audio for games. Throughout his presenta-tion he was able to cite examples that bridged disciplines andsome projects which would be equally comfortable within aninteractive gaming experience. It was clear through conversa-tions with conference-goers after his presentation that his workwas inspirational and helped set a tone of exploration outside theboundaries of typical interactive sound.

456 J. Audio Eng. Soc., Vol. 61, No. 6, 2013 June

Robin Rimbaud (aka “Scanner”) delivers an inspirational keynote.

The impressive Mall to the rear of Number Two Carlton House Terrace,where the conference took place.

CONFERENCE REPORT

DAY 1

After introductions on Wednesday morning, sessions beganwith Scott Selfon (Microsoft Advanced Technology Group)

presenting a talk entitled “Crossing the Streams.” His talk aimedto increase awareness of areas where audio has an ability tointeroperate between a game and an audio engine. Kickingthings off with a series of comedic videos, the presentationhelped to communicate a common language of the subtle intri-cacies at work in games that are harnessing dynamic audio.Some of the specific examples that broke out of this paradigmincluded “You Don’t Know Jacks” implementation of specificcallbacks to audio regarding day-of-the-week information today/time-specific VO and a comparison of the PS2 version ofSSX, which modulated a low-pass filter when in air, versus themore recent implementation of EA’s RUMOR technology, whichenabled on-beat landing (where an animation will wait formusic-specific timing before finishing a jump). Throughout thepresentation, Scott showed the power of two-way communica-tion between the game and audio engine and encouraged devel-opers to leverage the inherent interactivity to extend the capabil-ities of audio to reflect gameplay.Xavier Buffoni dug in deep to expose the “Myths, Facts, and

Techniques behind HDR Audio” in a presentation that providedan overview of the known facts of HDR from DICE’s publicpresentations on HDR Audio using their Frostbite engine. Hewent on to describe the process of Audiokinetics’ own researchinto the application of HDR Audio and shared some of their find-ings. Fundamental to the presentation was the question: “Whymap sounds to HDR audio?” His simple answer being: in order torepresent real li fe dynamics and greater dynamic range.Referencing HDR photography, he went on to compare this tech-nique with its use in HDR mixing with the focus on loudness asan approximation of the importance of sound in time. He thenpresented examples that helped to illustrate and frame decisionsand workflow choices involved with implementing HDR Audio

within a game audio pipeline using the Wwise authoring applica-tion. Key to these examples was the ability to visualize the appli-cation of an HDR Audio system working in different ways withinWwise to illustrate the resulting affect. Martin Walsh overviewed a “A (Fairly) Brief History of 3D

Audio in Games” pulling from his past experience. As someonewho has been working in the industry for the last 20 years, push-ing the forefront of 3D audio, he provided a valuable perspectiveon the development of technology that has lived on the cuttingedge during his time at Creative Labs and other associatedcompanies. His personal history has been skewed toward the PCplatform, and his experience stems back to the definition of 3Daudio. This included how spatialization was approached throughtechnology, including positional audio, dynamic reverb “morph-ing,” and HRTFs. His history covered the inception of audio APIsfor handling 3D audio, through DirectSound 3D, A3D, andbeyond. His deep involvement as one of the major players in thedevelopment of early 3D audio lent a “fly-on-the-wall” perspec-tive that resulted in many inside anecdotes and stories from thebygone days when the battle for 3D audio raged between differentdevelopers. His presentation provided more than just a historyand helped shed light on the behind-the-scenes machinations atwork during the upheaval in hardware versus. software interac-tive audio development on the PC.Tireless audio advocate Stephan Schutze took the stage with a

confidence that was infectious. His talk on “How Sound AffectsRealities: Enhancing Narrative Through Audio” was a borderlinesermon on the ability of sound to communicate to the player in away that visuals are sometimes unable. He used a sound sampleof an airplane and asked conference attendees to listen andconnect with the sound through their own experiences. He thenwent on to detail an instance where the authenticity of thissound was more than just an airplane; for him it held veryspecific connotations that transcend the simple definition of thesound. As creators of experiences, he urged the audience torespect the use of sound and what it can convey in a given situa-tion, outside of the direct need to reflect visuals. His example ofthe use of airplane sounds, detached from any physical represen-tation in a game, showed the value that contextual sound canbring to an experience.

DAY 2

Garry Taylor kicked off Day 2 with his talk on “MeasuringLoudness in Interactive Entertainment.” Garry has worked

tirelessly at SCEE to assemble and analyze data related to loud-ness in games as part of an initiative within Sony, and he sharedthis experience withattendees. As part ofhis f indings he wasfinally able to sharethe standard that Sonyhas adopted, based onthe ITU-R BS.1770.The process they haveadopted includes audiofrom a title measuredas a whole; not sepa-rated by music, sound,or dialog, for a mini-mum of 30 minutes.This measured contentshould be representa-

J. Audio Eng. Soc., Vol. 61, No. 6, 2013 June 457

SOME DAY 1 SPEAKERSScott Selfon (top left), Xavier Buffoni (top right), Martin Walsh(bottom left), and Stephan Schutze (bottom right)

Olivier Thereaux and Jory Prum after theirpresentation on a new standard for webaudio.

CONFERENCE REPORT

458 J. Audio Eng. Soc., Vol. 61, No. 6, 2013 June

tive of the dif ferenttypes of gameplay. Theirstandard is beingenforced through qual-ity assurance testing aspart of platform require-ments for their f irst-party Sony titles. Thisstandard also forms the basis for a recom-mendation for all titlesreleased on Sony platforms. Jory Prum and Olivier

Thereaux gave anoverview for a “NewStandard for Web

Audio” by educating conference attendees in the history of webaudio, current activity in W3C Audio Working Group andWebAudioAPI, as well as providing examples of the currentimplementations and applications. The general feeling was thatweb audio has established a substantial foundation for the futureof interactive audio in the browser. The crowd response wasfavorable to this presentation; many spoke of the potential beingestablished by web audio. While the effort is still relatively new,it also affords great possibilities for crosscollaboration betweengame and web audio in order to educate and inform the directionof future efforts. As such, the potential for game audio profes-

sionals to assist in establishing these standards and best prac-tices is greatly needed in order to leverage the experience of thegame audio community.Adele Cutting moderated a panel called “The Future of Game

Audio: A Retrospective” that took predictions of the future of gameaudio from the AES 35th Audio for Games conference and reflectedon where the industry at the time thought we would be by now.The first topic addressed was the use and development of the mixprocess and comments on the rise of dynamic mixing and readilyavailable tools to accomplish this. Tom Colvin (Ninja Theory) advo-cated the continued rise of standardization in game audio and theaddition of a postproduction phase for mixing. This postproductionmix was seconded by Thomas Bärtschi (IO Interactive) who wasable to achieve this nearing the end of Hitman: Absolution.Storytelling and narrative were seen as an area in games thatrequired work to achieve a greater quality, and the panel felt like inthe interim years there had been good progress across all disci-plines to increase the ability of games to tell stories with emotionswith the help of sound. A question about how to accomplish betterdialog in games turned into a general advocacy for the increasedcommunication between different departments as a way to improvethe quality of a title overall.

DAY 3Mike Caviezel, long time racing game sound specialist and nowMicrosoft SoundLab manager, addressed the technical and emo-tional task of implementing vehicle sound systems for games in apresentation titled “Planes, Trains, and Automobiles: Creatingand Implementing Vehicle Sound Systems for Games.” One partscience and another part rock and roll, the presentation walked afine line between automotive specifics and the passionate deliv-ery of the speaker. The components’ combustible engines werediscussed in an accessible way, allowing for the presentation ofinteractive audio systems that are able to support both the accu-racy of today’s physics simulations, as well as the emotive needsspecific to games. Mike’s experience and personality shed lighton the complexities involved with recreating these systems bothtechnically and in a way that appropriately amplifies the dramaexpected from games as opposed to strict simulations.Mikkel Christiansen and Frans Galschiøt Quaade from IO

Interactive gave an overview of “Behind the Mix—An In-DepthLook at the Audio Engine in Hitman: Absolution.” Their overviewfocused on the music and dynamic crowd systems and their imple-mentation using their internally developed Glacier 2 engine, whichextended the functionality of the FMOD Ex audio engine. Their

SOME DAY 2SPEAKERSFrom top left: Garry Taylor,Adele Cutting, Mike Caviezel,Mikkel Christiansen and Franz Galschiøt Quaade

John Broomhall chairs a workshop on“Theoretical, Technical, and PracticalFrameworks for Interactive Mixing: AModerated Panel Discussion.”

A hands-on session about FMOD Studio on Thursday afternoon

CONFERENCE REPORT

J. Audio Eng. Soc., Vol. 61, No. 6, 2013 June 459

tool-side integration built upon the framework implementedthroughout their game engine pipeline and enabled, through theuse of a node graph interface, the complex scripting of audio behav-ior based on other aspects of the game development pipeline. Apartfrom establishing the building blocks of their audio pipeline withinthe game editor, on most occasions the team was able to makedecisions about the behavior of sound without involving program-mers. Different systems can be used as templates containingcomplex structures that can be reused throughout the project. Thedynamic crowd system in Hitman: Absolution constantly adjuststhe vocal reactions of thenearby crowd based on game-state information about theplayer’s action. The four statesderived from the behavior of theHitman that modify the vocalreactions of the crowd are:Ambient, Alerted, Scared,GetDown. These states are usedto change volume betweendifferent layers of crowd ambi-ence, which correspond to themood that should be expressedby the crowd. Additionally, thesize of the crowd and otherfactors can contribute to deter-mining which ambient soundsare played.

PAPERS

The true success of thepapers presented at this

year’s conference was not onlytheir diversity in subject mat-ter and depth of research, butin their relevance to development of audio for games. A diverseselection of topics assembled by Damian Murphy illustrated thesignificance of educational research into the reproduction ofreal-time audio. It could be said that by looking to the workbeing done within the educational incubators, one can catch aglimpse of the future. Some ideas and methodologies presentedrelied heavily upon deep-thinking, theory, and shared a utopianvision for interactive audio, while others sought answers to ques-tions that can better enable the greater understanding of ourrelationship to audio in games.“Modeling and Real-Time Generation of Pen Stroke Sounds for

Tactile Devices…” approached the task of delivering dynamicaudio feedback through touch interfaces by breaking downgestural interface interactions and modeling them in conjunc-tion with real-time resampling of prerecorded audio content. Thepaper identified the main sound components of the physicality ofpen stroke interaction and built an audio system to support thesecomponents for dynamic interaction with tactile devices. Seeking to expose the effectiveness of player interaction with

specific types of sound, Niels Bøttcher presented on Fridaymorning the findings of his paper titled “Can InteractiveProcedural Audio Affect the Motorical Behavior of Players inComputer Games with Motion Controllers?” Detailing his use ofprocedurally generated swordswings, in conjunction with theNintendo Wii remote, the presenter established a test to deter-mine whether the procedural and dynamic nature of synthesizedsound would be perceived differently from prerecorded samples

using a simulated swordfighting game. The test findings showedthat 16 of the 40 participants showed a difference in physicalinteraction between procedural and prerecorded swordswingtechniques.Also of note was a paper from David M. Young titled “The

Future of Adaptive Game Music: The Continuing Evolution ofDynamic Music Systems,” which established the current state ofmusic for games and speculates on what is yet to come. In addi-tion to the technological advancements allowing for greatercontrol of authoring musical content for games, the author envi-

sions a future where biofeed-back is used to adapt thesoundtrack based on theplayer-state awareness enabledby future “emotion detectors.”

POSTERS

Posters were displayed inthe second-floor atrium

during the final day of presen-tations and served as a coun-terpoint to the discussionssurrounding the last day’sspeakers. Attendees were ableto speak directly with repre-sentatives and ask questionsbased on the information displayed.In a poster presented by

Tapani Pihlajamaki entitled“Plausible Mono-to-SurroundSynthesis in Virtual-WorldParametric Spatial Audio,” aprocess was presented to allowfor the synthesizing of multi-

channel output from a mono signal based on an estimation ofreverberant energy in a signal. Simply put, the paper suggests away of estimating the positional spread of a sound based on vari-ous factors as a way to “fill in the gaps in the sound scene to makeit sound surrounding and immersive to the listener.” This paperwas presented in conjunction with “Modular Architecture ofVirtual-World Parametric Spatial Audio Synthesis,” which

Intense discussion about posters in the library

SELECTED PAPER PRESENTERSNiels Bøttcher (top left), Erik Sikström (top right), Søren K. Jacobsen (bottom left), Tapani Pihlajamaki (bottom right)

restructures some ofDirectional Audio Coding(DirAC) work that is at the coreof their proposed system forparametric spatial audiocoding.A paper from Philip Jackson

and Naveen Desiraju attemptedthe “Use of 3D Head Shape forPersonalized Binaural Audio”in order to custom tailor thespatialization based on thehead shape of the l istener.Using a stereo-vision face-capture system comprised ofKinect camera hardware andproprietary capture system, amodel of the face and head wascaptured for use in creatingthe HRTF (or HRIR) dataset.Listening tests for subjectswho participated in personalface and head modeling wereable to better locate soundsources than participants thatused a dummy-head HRIR.A tool was suggested for

“Geometric and Wave-BasedAcoustic Modeling usingBlender” by Jelle van Mourikand Damien Murphy as a wayto automate reverberation based on 3D-modeled spaces.Presented as a “means of real-time sound rendering” the papergoes on to detail the process of developing acoustical modelsusing Blender and comparison between other models of reverber-ation. Findings have shown that the performance of Blender as a“sound scene” renderer is a viable way to generate impulseresponses that show similarities to other methods of modelingthe acoustics of spaces.While there exist other opportunities to submerge oneself in

game audio at other events, the AES Audio for Games conferenceallows a rare opportunity to rub shoulders with creatives andacademics alike. It’s this interaction between researchers, scien-tists, professionals, and students working with interactive audiothat makes the intersection of knowledge sharing at the confer-

ence such a unique experience. The exposure to the differentdisciplines and approaches within audio leads to rich dialogs,smart questions, and opportunities to educate across all levels ofexpertise. It’s hard to see what the direct result of this may be inthe future, but the opportunity to be able to get everyone on thesame page inspires new ways of thinking. It is this inspirationthat AES Chair Michael Kelly was hoping to cultivate when heset it as the theme forthis year’s conference,and judging by theacclamation fromcolleagues during theevent it can safely besaid that he succeeded.

CONFERENCE REPORT

460 J. Audio Eng. Soc., Vol. 61, No. 6, 2013 June

Roberto Pomoni takes a look at“Global Game Audio Production:Where Are We Going?”

The gathered conference congregation, led by secretary Becky Stewart, chair Michael Kelly, and paperschair Damian Murphy.

Editor’s note: You can downloadpapers from this conference fromthe AES e-library at:http://www.aes.org/e-lib/