Thursday 12th February

Paper Session 2: Audio Codecs

Session Chair:
Steve Martz, THX, USA

2-1: SAOC for Gaming - The Upcoming MPEG Standard on Parametric Object Based Audio Coding

Leonid Terentiev, Oliver Hellmuth, Johannes Hilpert, Cornelia Falch, Fraunhofer Institute for Integrated Circuits IIS, Erlangen, Germany; Werner Oomen, Philips Applied Technologies, Eindhoven, The Netherlands; Jonas Engdegard, Dolby Sweden AB

Following the current trend of employing parametric enhancement tools for increasing coding and spatial rendering efficiency, the MPEG audio group performs the standardization activities on Spatial Audio Object Coding (SAOC) technology. The SAOC system extends the MPEG Surround standard by exploiting its rendering capability and decoding efficiency. Moreover, the SAOC technology introduces user-controllable rendering functionality together with flexible choice of various playback configurations. These aspects are of potential interest for a large range of gaming applications that will benefit from the efficient coding and interactive rendering. Although SAOC targets many different application scenarios, this paper will describe the basic SAOC architecture and the manifold of enhancement tools, with a specific focus on the relevance for gaming applications.

2-2: Viable Distribution of Multichannel Audio-over-IP for Live and Interactive "Voice Talent"-Based Gaming Using High-Quality, Low-latency Audio Codec Technology

Gregory Massey, APT Ltd., Belfast, UK

The delivery of multichannel audio - from monophonic to surround-sound - in real-time over public IP networks for the purpose of interactive, multi-participant, on-line game entertainment presents a serious design engineering challenge to game developers, console architects, content distribution networks, and ISPs. Leveraging audio engineering expertise gained in professional broadcasting and recording studio postproduction, APT has developed a robust and scalable audio codec technology that enables on-line gaming systems to realize real-time distribution of high-quality audio for immersive, instantaneous audio experiences in multi-player scenarios of the kind where mass audiences engage with voice talent: talking and singing.

2-3: An Analysis of Differences in Playback Levels Among In-Game Sound Levels, Game Start-up Sounds, and Reference Levels of Game Production Studios

Masataka Nakahara, SONA Corp./ONFUTURE Ltd., Tokyo, Japan; Kazutaka Someya, DiMAGIC Co., Ltd., Tokyo, Japan; Steven P. Martz, THX Ltd., San Rafael, CA, USA

The audio experience can vary greatly when playing a video game. After establishing a playback level for a reproduction device, the game may still be too loud or too soft. Average reproduction levels can fluctuate from title to title and even within a game. It is important for game production studios to carefully consider how they configure their systems and the effects it may have on consistency in their games. In this paper, the authors survey; 1) the in-game recorded levels of 55 opening cut-scenes from video game titles, 2) the output levels of start-up sounds for four kinds of game consoles, and 3) the current reference levels of sound production facilities for video games. As a result, we conclude; 1) that -20.4dBFS is the average RMS value per channel of in-game sound, so that -20dBFS is considered to be the useful reference level for video game production, 2) that -16.4dBFS is the average Leq(A) value from the summed output of in-game sound, and that it has a good relationship with the RMS value, 3) that there is little difference among the average levels in games across genres and across game consoles, and that there are large differences among production companies, and 4) that the level of start-up sounds which are implemented into game consoles are approximately 10dB lower than the level of in-game sounds in opening cut-scenes.