Home   About   News   MusicTech   AudioEngine   Audio Compression & Mobile Apps

MusicTech is a set of music signal processing algorithms to serve as main functions or to add great entertainment value to a wide array of interactive music devices - toy, learning device, educational aid, OID pen, e.dictionary, picture frame, digital camera or camcoder, mobile phone, tablet PC, soundcard, and so on.

Solo and Duet

Two new digital music technologies to be available soon ...

 

AnyTune (AT)

Function: AT converts spoken words into a song with accompaniment music. AT features include the following:

- Speech maps onto the melody (of any song!) with or without accompaniment music.

- Automatic repeating of speech when the accompaniment music is longer than speech.

- Automatic time-synchronization (onset matching) of speech and music.

- Automatic segmentation of speech for preset or random playback of speech segments.

- Optional slowing-down of speech to sound more like singing.

- Timbre/tone control to convert a person's speech into another person's singing.

                                           

Audio Demo:   Original Speech       AT-Processed Song       Original Speech       AT-Processed Song

The speech/lyric, mapped onto selected melody, can be anything - a message for loved ones, poem, joke, rap, kid's talking, baby's voice, and so on. The melody can be classical music, pop, opera, gospel, folk, and so on. By creative matching of lyric and melody, AT, a "talking karaoke", provides endless fun!

Applications: For interactive toy, lyric creation/music learning device, mobile app and many other applications.

Specifications:
- 16-bit integer C code (bit-exact performance for HW & SW platforms)
- Audio data: 16-bit linear PCM (.wav or raw data)
- Sample rate: 8/10/11.025/12 KHz
- Audio I/O: 1-ch (mono) input and output

Resources Required: For "record-then-playback" mode at 8KHz sample rate - 5.2MIPS, 2.8KB RAM, 5.2KB ROM (data).

AnyTune/Voice Morphing (ATVM)

Function: ATVM morphs any voice into the voice of a male, female, kid, robot, monster, cartoon character; plus slower voice, faster voice, whisper voice and singing (spoken words mapped on select melody, such as a mother's loving words on Brahms' Lullaby).

                                   

Audio Demo:   Original Voice       Cartoon Voice       Original Voice       Monster Voice       Original Voice       Robot Voice

 

ATVM features include:

 

 

 

 

Applications: For applications from serious (such as to hide personal identity) to fun (such as to play tricks on friends), including for VoIP phone conversation (Skype, ..), Internet voice chat, on-line games, voice messaging, toys/robots, learning devices, children books, text-to-speech and many others.

Specifications:
- 16-bit integer C code (bit-exact performance for HW & SW platforms)
- Audio data: 16-bit linear PCM (.wav or raw data)
- Sample rate: 8/10/11.025/12 KHz (extendable to 16/22.05/32/44.1/48 KHz)
- Audio I/O: 1-ch (mono) input and output

Resources Required: For "record-then-playback" mode at 8KHz sample rate - 5.8MIPS, 4KB RAM, 5.6KB ROM (data).

Beat Tracking (BT)

Function: BT tracks music beats accurately in real time for various music-interactive, music-synchronous applications. BT picks up beat quickly when music starts or music changes. BT stops beat tracking immediately when music stops. BT works well even for music without strong drum sound.

Applications: Music-sync LED light show for any mobile device that plays music -- mobile phone, notebook/netbook PC, MP4/MP3 players, iPod docking station, children books. Also dancing robots/toys, etc.

Audio Demo:   Original Music       BT       Original Music       BT

Specifications:
- 16-bit integer C code (bit-exact performance for HW & SW platforms)
- Audio data: 16-bit linear PCM (.wav or raw data)
- Sample rate: 8/10/11.025/12/16/22.05/32/44.1/48 KHz
- Audio I/P: 1-ch (mono) microphone or line-in input

Resources Required: With mono audio input at 8KHz sample rate - MIPS=4.8, RAM=1.5KB, ROM (data)=0.64KB.