1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
|
syntax = "proto3";
package speechkit.tts.v3;
option go_package = "github.com/yandex-cloud/go-genproto/yandex/cloud/ai/tts/v3;tts";
option java_package = "yandex.cloud.api.ai.tts.v3";
message AudioContent {
// The audio source to read the data from.
oneof AudioSource {
// Bytes with audio data.
bytes content = 1;
}
// Description of the audio format.
AudioFormatOptions audio_spec = 2;
}
message AudioFormatOptions {
oneof AudioFormat {
// The audio format specified in request parameters.
RawAudio raw_audio = 1;
// The audio format specified inside the container metadata.
ContainerAudio container_audio = 2;
}
}
message RawAudio {
enum AudioEncoding {
AUDIO_ENCODING_UNSPECIFIED = 0;
// Audio bit depth 16-bit signed little-endian (Linear PCM).
LINEAR16_PCM = 1;
}
// Encoding type.
AudioEncoding audio_encoding = 1;
// Sampling frequency of the signal.
int64 sample_rate_hertz = 2;
}
message ContainerAudio {
enum ContainerAudioType {
CONTAINER_AUDIO_TYPE_UNSPECIFIED = 0;
// Audio bit depth 16-bit signed little-endian (Linear PCM).
WAV = 1;
// Data is encoded using the OPUS audio codec and compressed using the OGG container format.
OGG_OPUS = 2;
// Data is encoded using MPEG-1/2 Layer III and compressed using the MP3 container format.
MP3 = 3;
}
ContainerAudioType container_audio_type = 1;
}
message TextVariable {
// The name of the variable.
string variable_name = 1;
// The text of the variable.
string variable_value = 2;
}
message AudioVariable {
// The name of the variable.
string variable_name = 1;
// Start time of the variable in milliseconds.
int64 variable_start_ms = 2;
// Length of the variable in milliseconds.
int64 variable_length_ms = 3;
}
message UtteranceSynthesisResponse {
// Part of synthesized audio.
AudioChunk audio_chunk = 1;
// Part of synthesized text.
TextChunk text_chunk = 2;
// Start time of the audio chunk in milliseconds.
int64 start_ms = 3;
// Length of the audio chunk in milliseconds.
int64 length_ms = 4;
}
message AudioTemplate {
// Audio file.
AudioContent audio = 1;
// Template and description of its variables.
TextTemplate text_template = 2;
// Describing variables in audio.
repeated AudioVariable variables = 3;
}
message AudioChunk {
// Sequence of bytes of the synthesized audio in format specified in output_audio_spec.
bytes data = 1;
}
message TextChunk {
// Synthesized text.
string text = 1;
}
message TextTemplate {
// Template text.
//
// Sample:`The {animal} goes to the {place}.`
string text_template = 1;
// Defining variables in template text.
//
// Sample: `{animal: cat, place: forest}`
repeated TextVariable variables = 2;
}
message DurationHint {
enum DurationHintPolicy {
DURATION_HINT_POLICY_UNSPECIFIED = 0;
// Limit audio duration to exact value.
EXACT_DURATION = 1;
// Limit the minimum audio duration.
MIN_DURATION = 2;
// Limit the maximum audio duration.
MAX_DURATION = 3;
}
// Type of duration constraint.
DurationHintPolicy policy = 1;
// Constraint on audio duration in milliseconds.
int64 duration_ms = 2;
}
message Hints {
// The hint for TTS engine to specify synthesised audio characteristics.
oneof Hint {
// Name of speaker to use.
string voice = 1;
// Template for synthesizing.
AudioTemplate audio_template = 2;
// Hint to change speed.
double speed = 3;
// Hint to regulate normalization level.
// * For `MAX_PEAK` loudness_normalization_type: volume changes in a range (0;1], default value is 0.7.
// * For `LUFS` loudness_normalization_type: volume changes in a range [-145;0), default value is -19.
double volume = 4;
// Hint to specify pronunciation character for the speaker.
string role = 5;
// Hint to increase (or decrease) speaker's pitch, measured in Hz. Valid values are in range [-1000;1000], default value is 0.
double pitch_shift = 6;
// Hint to limit both minimum and maximum audio duration.
DurationHint duration = 7;
}
}
message UtteranceSynthesisRequest {
// The name of the model.
// Specifies basic synthesis functionality. Currently should be empty. Do not use it.
string model = 1;
// Text to synthesis, one of text synthesis markups.
oneof Utterance {
// Raw text (e.g. "Hello, Alice").
string text = 2;
// Text template instance, e.g. `{"Hello, {username}" with username="Alice"}`.
TextTemplate text_template = 3;
}
// Optional hints for synthesis.
repeated Hints hints = 4;
// Optional. Default: 22050 Hz, linear 16-bit signed little-endian PCM, with WAV header
AudioFormatOptions output_audio_spec = 5;
enum LoudnessNormalizationType {
LOUDNESS_NORMALIZATION_TYPE_UNSPECIFIED = 0;
// The type of normalization, wherein the gain is changed to bring the highest PCM sample value or analog signal peak to a given level.
MAX_PEAK = 1;
// The type of normalization based on EBU R 128 recommendation.
LUFS = 2;
}
// Specifies type of loudness normalization.
// Optional. Default: `LUFS`.
LoudnessNormalizationType loudness_normalization_type = 6;
// Optional. Automatically split long text to several utterances and bill accordingly. Some degradation in service quality is possible.
bool unsafe_mode = 7;
}
|
