SYNTHESIZED SINGING VOICE WAVEFORM GENERATOR
Various technologies for generating a synthesized singing voice waveform. In one implementation, the computer program may receive a request from a user to create a synthesized singing voice using the lyrics of a song and a digital file containing its melody as inputs. The computer program may then dissect the lyrics' text and its melody file into its corresponding sub-phonemic units and musical score respectively. The musical score may be further dissected into a sequence of musical notes and duration times for each musical note. The computer program may then determine a fundamental frequency (F 0 ), or pitch, of each musical note.
1 . A method comprising:
adapting, according to second models of a voice and a corresponding transcript of a second person, first models that are based on a voice and a corresponding transcript of a first person; and
synthesizing, by a computer, a voice based on the adapted first models, and based on a sequence of contextual parametric models of lyrics, and based on a sequence of notes and their duration times from a digital melody.
2 . The method of claim 1 wherein the sequence of contextual parametric models corresponds to speech units derived from the lyrics;
3 . The method of claim 1 wherein the first models comprise statistically trained parametric models that include sequences of representations of speech units of the voice of the first person.
4 . The method of claim 3 wherein the second models comprise statistically trained parametric models that include sequences of representations of speech units of the voice of the second person.
5 . The method of claim 4 wherein the adapting comprises matching the adapted first models to the sequence of contextual parametric models.
6 . The method of claim 5 wherein the synthesizing is based on the matching.
7 . The method of claim 1 wherein the synthesized voice sounds substantially the same as the second voice as opposed to the first voice.
8 . At least one computer storage media storing computer-executable instructions that, when executed by a computer, cause the computer to perform a method comprising:
adapting, according to second models of a voice and a corresponding transcript of a second person, first models that are based on a voice and a corresponding transcript of a first person; and
synthesizing a voice based on the adapted first models, and based on a sequence of contextual parametric models of lyrics, and based on a sequence of notes and their duration times from a digital melody.
9 . The method of claim 8 wherein the sequence of contextual parametric models corresponds to speech units derived from the lyrics;
10 . The method of claim 8 wherein the first models comprise statistically trained parametric models that include sequences of representations of speech units of the voice of the first person.
11 . The method of claim 10 wherein the second models comprise statistically trained parametric models that include sequences of representations of speech units of the voice of the second person.
12 . The method of claim 11 wherein the adapting comprises matching the adapted first models to the sequence of contextual parametric models.
13 . The method of claim 12 wherein the synthesizing is based on the matching.
14 . The method of claim 8 wherein the synthesized voice sounds substantially the same as the second voice as opposed to the first voice.
15 . A system comprising:
a computer;
an adaptation module implemented at least in part by the computer and configured for adapting, according to second models of a voice and a corresponding transcript of a second person, first models that are based on a voice and a corresponding transcript of a first person; and
synthesizing a voice based on the adapted first models, and based on a sequence of contextual parametric models of lyrics, and based on a sequence of notes and their duration times from a digital melody.
16 . The system of claim 15 wherein the sequence of contextual parametric models corresponds to speech units derived from the lyrics;
17 . The system of claim 15 wherein the first models comprise statistically trained parametric models that include sequences of representations of speech units of the voice of the first person.
18 . The system of claim 17 wherein the second models comprise statistically trained parametric models that include sequences of representations of speech units of the voice of the second person.
19 . The system of claim 18 wherein the adapting comprises matching the adapted first models to the sequence of contextual parametric models.
20 . The system of claim 19 wherein the synthesizing is based on the matching, wherein the synthesized voice sounds substantially the same as the second voice as opposed to the first voice.