Audio Taken Seriously The present and future of

  • Slides: 17
Download presentation
Audio Taken Seriously; The present and future of audio at Microsoft Ken Greenebaum kgreene@microsoft.

Audio Taken Seriously; The present and future of audio at Microsoft Ken Greenebaum [email protected] com Internet Platforms and tools Division Microsoft Corporation November 4 th, 1996 1 ICAD Industry Panel

Slides, other materials online: http: //www. research. microsoft. com/research/grap hics/kgreene/icad November 4 th, 1996

Slides, other materials online: http: //www. research. microsoft. com/research/grap hics/kgreene/icad November 4 th, 1996 2 ICAD Industry Panel

Overview ù Today ù Solid media foundations (Direct. X, Active. Movie) ù Soon ù

Overview ù Today ù Solid media foundations (Direct. X, Active. Movie) ù Soon ù Advanced media (Active. Animation, Whisper/Whistler) ù Tomorrow ù Conversational interfaces November 4 th, 1996 3 ICAD Industry Panel

Today: Direct. Sound http: //www. microsoft. com/mediadev/audio/iaud. htm ù ù ù Streaming audio Reasonable

Today: Direct. Sound http: //www. microsoft. com/mediadev/audio/iaud. htm ù ù ù Streaming audio Reasonable latency Input (soon) Device independence Multiple app’s audio mix DSound 3 D November 4 th, 1996 4 ICAD Industry Panel

Today: Active Movie ù ù ù Graph based media architecture Movie playback Movie record

Today: Active Movie ù ù ù Graph based media architecture Movie playback Movie record (soon!) Open filter API Audio plugin technology November 4 th, 1996 5 ICAD Industry Panel

Today: Netshow http: //www. microsoft. com/netshow/ ù ù Streaming network audio/video Multicast audio using

Today: Netshow http: //www. microsoft. com/netshow/ ù ù Streaming network audio/video Multicast audio using RTP (real-time protocol) ASF file format, conversion, editing tools NT server November 4 th, 1996 6 ICAD Industry Panel

Today: Interactive Music (Formerly Blue. Ribbon’s Audio. Active) ù ù ù Intelligent interactive music

Today: Interactive Music (Formerly Blue. Ribbon’s Audio. Active) ù ù ù Intelligent interactive music Composes/Delivers music Based on expert system Human composer ‘authors’ templates Music always sounds fresh and original Look for it: Power. Point ‘ 97, MSN Riff November 4 th, 1996 7 ICAD Industry Panel

Soon: Direct. Music Contact: craighs@microsoft. com ù ù ù Consistent Playback of MIDI Music

Soon: Direct. Music Contact: [email protected] com ù ù ù Consistent Playback of MIDI Music Internet support for Music DLS downloadable sample sets Optional software MIDI synth Internet MIDI jamming? November 4 th, 1996 8 ICAD Industry Panel

Soon: “Appelles” ù ù ù ù Expect an announcement soon! Animation Description Language Functional

Soon: “Appelles” ù ù ù ù Expect an announcement soon! Animation Description Language Functional Paradigm Media Integration Implicit Time Language Integration (Java) Enable sophisticated Web animation November 4 th, 1996 9 ICAD Industry Panel

Appelles Audio Capabilities: ù ù ù ù All audio types orthogonal Parametric Synthesis MIDI

Appelles Audio Capabilities: ù ù ù ù All audio types orthogonal Parametric Synthesis MIDI Audio Active Music Synthesis Streaming audio PCM Audio 3 D Spatialized sound embedded in geometry November 4 th, 1996 10 ICAD Industry Panel

Soon: “Talisman” Audio http: //www. microsoft. com/hwdev/devdes/talisman. htm/ ù Hardware acceleration of: ù ù

Soon: “Talisman” Audio http: //www. microsoft. com/hwdev/devdes/talisman. htm/ ù Hardware acceleration of: ù ù ù DSound/DSound 3 D Echo Cancellation Active Movie filter accelerator 32 bit mixer DLS compatible synthesizer MODEM/Telephony November 4 th, 1996 11 ICAD Industry Panel

Soon: “Whisper” http: //www. research. microsoft. com/research/srg/ ù ù ù Windows Highly Intelligent Speech

Soon: “Whisper” http: //www. research. microsoft. com/research/srg/ ù ù ù Windows Highly Intelligent Speech Recognizer Based on Sphinx. II Continuous speech recognition Speaker independent Context-free grammar decoding November 4 th, 1996 12 ICAD Industry Panel

Soon: “Whistler” http: //www. research. microsoft. com/research/srg/ ù ù Trainable Text to Speech Synthesizer

Soon: “Whistler” http: //www. research. microsoft. com/research/srg/ ù ù Trainable Text to Speech Synthesizer Training from human speech; maintains: ù Natural prosody ù Characteristics of original human ù ù Emotional control Uses NLP technology to parse text November 4 th, 1996 13 ICAD Industry Panel

Tomorrow: Conversational Interfaces ù Motivation: ù Given choice people communicate with speech ù People

Tomorrow: Conversational Interfaces ù Motivation: ù Given choice people communicate with speech ù People prefer natural language over ‘command languages’ ù anthropomorphism unavoidable w/spoken interaction November 4 th, 1996 14 ICAD Industry Panel

Persona Project http: //www. research. microsoft. com/ui/persona/home. htm/ ù Conversational Assistant as UI ù

Persona Project http: //www. research. microsoft. com/ui/persona/home. htm/ ù Conversational Assistant as UI ù ù Spoken conversation (voice recognition/synth) Natural Language (in limited domains) Assistant w/Rich visual presence Simulates verbal and non-verbal cues November 4 th, 1996 15 ICAD Industry Panel

Here’s Peedy and Gene: November 4 th, 1996 16 ICAD Industry Panel

Here’s Peedy and Gene: November 4 th, 1996 16 ICAD Industry Panel

Conclusion: ù Microsoft is: ù Taking media very seriously ù Offering a solid foundation

Conclusion: ù Microsoft is: ù Taking media very seriously ù Offering a solid foundation today ù Designing the future November 4 th, 1996 17 ICAD Industry Panel