next up previous
Next: Unit size Up: Unit Selection Synthesis Previous: The right domain

Synthesizing in style

It is possible to explicitly record different styles of speech and different prosodic contexts. For example we build a databases from a smaller 500 utterance prompt list where every second word was read with emphasis. For example from a database recorded as this

_Allow me _to interpret _this interesting _silence.
_Tarzan and _Jane raised _their heads.
Then each segment in each emphasized word is marked with an emphasis feature. During synthesis each word desired to have emphasis is constructed from the only those segment with that emphasis feature. We can synthesize emphasized examples like this
This is a short example.
_This is a short example.
This _is a short example.
This is _a short example.
This is a _short example.
This is a short _example.
This is a crude, but adequate, way to get explicit style. But without methods to modify the selected units, such explicit techniques are required.

Alan W Black 2002-09-30