|
|
From text to speech: The MITalk system
|
|
|
|
|
|
The old man sat in a rocker.
|
|
|
PHONO2: Function word: DH IY
|
|
|
PHONO2: Content word: 'OW LX DD
|
|
|
|
|
|
PHONO2: Content word: MM ’'AE NN [End NOUN phrase]
|
|
|
PHONO2: Content word: SS '"AE DX
|
|
|
PHONO2: Function word: IH NN
|
|
|
PHONO2: Function word: AX
|
|
|
PHONQO2: Content word: RR ‘AA KK * - ER
|
|
|
PHONO2: Punctuation:
|
|
|
PHONQ2: <EOF>
|
|
|
|
|
|
PROSOD: (Silence] 30ms. 133.4Hz.
|
|
|
|
|
|
PROSOD: Function word:
|
|
|
|
|
|
PROSOD: DH 50ms. 123.4Hz.
|
|
|
|
|
|
PROSOD: IY 105ms. 131.4Hz.
|
|
|
|
|
|
PROSOD: Content word:
|
|
|
|
|
|
PROSOD: "OW 170ms. 174.5Hz. Stressed
|
|
|
PROSOD: LX 75ms. 151.0Hz.
|
|
|
|
|
|
PROSOD : DD 50ms. 146.0Hz.
|
|
|
|
|
|
PROSOD: Content word:
|
|
|
|
|
|
PROSOD: MM 70ms. 151.0Hz. Stressed
|
|
|
PROSOD: 'AE 210ms. 157.0Hz. Stressed
|
|
|
PROSOD: NN 55ms. 117.9Hz.
|
|
|
|
|
|
PROSOD: [End NOUN phrase]
|
|
|
|
|
|
PROSOD: Content word:
|
|
|
|
|
|
PROSOD: SS 100ms. 122.9Hz. Stressed
|
|
|
PROSOD: 'AE 175ms. 153.9Hz. Stressed
|
|
|
PROSOD: DX 20ms. 140.1Hz.
|
|
|
|
|
|
PROSOD: Function word:
|
|
|
|
|
|
PROSQOD: IH 55ms. 148.1Hz.
|
|
|
|
|
|
PROSOD: NN 50ms. 142.5Hz.
|
|
|
|
|
|
PROSOD: Function word:
|
|
|
|
|
|
PROSOD: AX 60ms. 142.5Hz.
|
|
|
|
|
|
PROSOD: Content word:
|
|
|
|
|
|
PROSOD: RR 80ms. 140.2Hz. Stressed
|
|
|
PROSOD: "AA 160ms. 146.2Hz. Stressed
|
|
|
PROSOD: KK 65ms. 113.1Hz.
|
|
|
|
|
|
PROSOD: *
|
|
|
|
|
|
PROSOD: -
|
|
|
|
|
|
PROSOD: ER 170ms. 108.1Hz.
|
|
|
|
|
|
PROSOD: Punctuation: .
|
|
|
|
|
|
PROSOD: [Silence] 400ms. 111.2Hz.
|
|
|
|
|
|
PROSOD: [End sentence]
|
|
|
|
|
|
PROSOD: <EOF>
|
|
|
|
|
|
Figure 9-1: Example of the processing performed by PROSOD
|
|
|
|
|
|
DUR=((INHDUR-MINDUR)xPRCNT)/100+MINDUR (1)
|
|
|
|
|
|
where INHDUR is the inherent duration of a segment in msec, MINDUR is the
|
|
|
minimum duration of a segment in msec, and PRCNT is the percentage shortening
|
|
|
determined by applying rules 1 to 10 below. The program begins by obtaining
|
|
|
values for INHDUR and MINDUR for the current segment from Table 9-1, and by
|
|
|
setting PRCNT to 100. The inherent duration has no special status other than a
|
|
|
starting point for rule application; it is roughly the duration to be expected in non-
|
|
|
sense CVCs spoken in the carrier phrase “Say bVb again” or “Say Cab again”.
|
|
|
The following ten rules are then applied, where each rule modifies the PRCNT
|
|
|
|
|
|
94
|
