DEMONSTRATION - Singing Voice Synthesis

Speech synthesis systems usually use "reading out" speech database to training, but the proposed system uses "singing voice" speech database to training. It can constructs singing voice synthesis system.

• Song sample for training
  • "a me a me, fu re fu re, ka: sa N ga"
    Singing voice

• Synthesized song
  • "shi a wa se na ra te wo ta ta ko u"
    Singing voice

On this system, special tunings are not required, only needs music score(MIDI with lyrics information) to synthesis singing voice.