engine_phoneme_spec Class Reference

#include <phone_estimate.h>

Detailed Description

This data class describes an engine phoneme.

As described above, in order to produce better spreading of phonemes within the word alignment, we can use heuristics, but in order to use them, we need some information about the phoneme, whether it is voiced, unvoiced, a dipthong, etc. These classes typically are the same duration. By classifying the phoneme, we are able to do a little better job. It is not required though.

Additionally, the spec provides a mapping. This is somewhat of a hack, I want to be able to map everything to the Annosoft Phoneme Labels so that they display better in the the lipsync tool. Again, this can be adjusted to taste

Todo:: provide a way to read these in from an external file.

Definition at line 124 of file phone_estimate.h.

Public Types

enum phoneme_type { silence, unvoiced, voiced, dipthong }

type of phoneme More...

Public Member Functions

engine_phoneme_spec (const wchar_t *_enginePhn, wchar_t *_outputPhn, wchar_t *_desc, phoneme_type _pt)

paramaterized constructor

Data Fields

std::wstring enginePhoneme

SAPI engine phoneme.

std::wstring outputPhoneme

output phoneme

std::wstring desc

description

phoneme_type m_type

the type of phoneme one of engine_phoneme_spec::phoneme_type

The documentation for this class was generated from the following file:

sapi_lipsync/phone_estimate.h