-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathMultiSpeech.py
262 lines (229 loc) · 10 KB
/
MultiSpeech.py
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
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
# -*- coding: utf-8 -*-
import subprocess
import requests
import json
import time
import os
import re
import numpy as np
import jaconv
import librosa
import soundfile as sf
from datetime import datetime
Voicevox_speakers = {"四国めたん（あまあま）":0, "ずんだもん（あまあま）":1,
"四国めたん（ノーマル）":2, "ずんだもん（ノーマル）":3,
"四国めたん（セクシー）":4, "ずんだもん（セクシー）":5,
"四国めたん（ツンツン）":6, "ずんだもん（ツンツン）":7,
"九州そら（ノーマル）":16, "九州そら（あまあま）":15,
"九州そら（ツンツン）":18, "九州そら（セクシー）":17,
"九州そら（セクシー）":19,
"春日部つむぎ":8, "波音リツ":9, "雨晴はう":10,
"玄野武宏":11,"白上虎太郎":12,"青山龍星":13,
"冥鳴ひまり":14}
Coeiroink_speakers = {"つくよみちゃん":0, "AI声優-朱花":50}
AIVOICE_speakers = {"紲星あかり":5209, "紲星あかり（蕾）":5210}
def wav_wait(file):
while True:
try:
if os.path.isfile("./wav/%s"%file):
break
except PermissionError: continue
def nchars(s, n):
assert n > 0
reg = re.compile("(.)\\1{%d,}" % (n - 1)) # カンマを取ると n 個ちょうどになる
while True:
m = reg.search(s)
if not m:
break
yield m.group(0)
s = s[m.end():]
def start():
"""Start up SofTalk.
"""
cmd = "Softalk.exe /NM:女性01 /O:120 /S:100 /V:50 /X:1 /W:"
v_start = subprocess.Popen(cmd, shell=True)
v_start.kill
def voice_only(v_name, interval, speed, text):
"""Speak sentences using SofTalk.\n
Japanese name -> (音声名, 音程, 速度, 発声文章)\n
UtaYomi -> True(重音テト, 東北きりたん)\n
return -> None
"""
if v_name in ["重音テト", "MOTRoid", "試聴用", "東北きりたん"]:
v_name = "唄詠：" + v_name
cmd = "Softalk /NM:%s /O:%s /S:%s /V:50 /X:0 /W:%s" % (v_name, interval, speed, text)
v_in = subprocess.Popen(cmd, shell=True)
time.sleep(1)
def finish():
"""Stop to softalk after "softalk.voice(v_name, interval, speed, text, emo)".
"""
cmd = "Softalk.exe /close"
v_f = subprocess.Popen(cmd, shell=True)
v_f.kill
# Voicevoxでのtext_to_speech
def Voicevox_voice(speaker, filename, interval, speed, intonation, text):
"""
Internal Server Error(500)が出ることがあるのでリトライする\n
（HTTPAdapterのretryはうまくいかなかったので独自実装）\n
connect timeoutは10秒、read timeoutは300秒に設定（処理が重いので長めにとっておく）\n
audio_query
＜音声引数＞\n
0：四国めたん（あまあま）\n
1：ずんだもん（あまあま）\n
2：四国めたん（ノーマル）\n
3：ずんだもん（ノーマル）\n
4：四国めたん（セクシー）\n
5：ずんだもん（セクシー）\n
6：四国めたん（ツンツン）\n
7：ずんだもん（ツンツン）\n
8：春日部つむぎ\n
9：波音リツ\n
10：雨晴はう
"""
wav_file = "%s\wav\%s.wav" % (os.getcwd(), filename)
# 音声名の決定
name = Voicevox_speakers[speaker]
if int(interval) > 115: interval = "115"
elif int(interval) < 85: interval = "85"
if int(speed) > 200 : speed = "200"
elif int(speed) < 50: speed = "50"
if text.endswith("!") or text.endswith("?"):
intonation = "150"
query_payload = {"text": text, "speaker": name}
r = requests.post("http://127.0.0.1:50021/audio_query",
params=query_payload)
query_data = r.json()
# 音質の変更
interval_scale = int(interval) - 100
query_data["speedScale"] = int(speed) /100
query_data["pitchScale"] = interval_scale /100
query_data["intonationScale"] = int(intonation) / 100
# 音素解析
mora_list = []
text = ""
for sequence in query_data['accent_phrases']:
for val in sequence:
if val == "moras":
for j in sequence[val]:
mora_list.append(j)
elif val == "pause_mora" and sequence[val]:
mora_list.append(sequence[val])
for i in mora_list:
for j in i:
if j in ["consonant", "vowel"] and i[j]:
text+="%s,%d,"%(i[j], int(i[j+"_length"] * 1000))
with open('./wav/%s.txt'%filename, 'w', encoding='UTF-8') as f:
f.write(text.rstrip(","))
# synthesis
synth_payload = {"speaker": name}
r = requests.post("http://127.0.0.1:50021/synthesis", params=synth_payload,
data=json.dumps(query_data))
with open(wav_file, "wb") as fp:
fp.write(r.content)
def Coeiroink_voice(speaker, filename, param, interval, speed, intonation, text, emo):
"""
Internal Server Error(500)が出ることがあるのでリトライする\n
（HTTPAdapterのretryはうまくいかなかったので独自実装）\n
connect timeoutは10秒、read timeoutは300秒に設定（処理が重いので長めにとっておく）\n
audio_query
＜音声引数＞\n
0：つくよみちゃん\n
1：AI声優-朱花
"""
if int(interval) > 115: interval = "115"
elif int(interval) < 85: interval = "85"
if int(speed) > 200 : speed = "200"
elif int(speed) < 50: speed = "50"
wav_file = "%s\wav\%s.wav" % (os.getcwd(), filename)
# 音声名の決定
name = Coeiroink_speakers[speaker]
query_payload = {"text": text, "speaker": name}
r = requests.post("http://127.0.0.1:50031/audio_query",
params=query_payload)
query_data = r.json()
# 音質の変更
interval_scale = int(interval) - 100
query_data["speedScale"] = int(speed) /100
query_data["pitchScale"] = interval_scale /100
query_data["intonationScale"] = int(intonation) / 100
# 音素解析
mora_list = []
text = ""
for sequence in query_data['accent_phrases']:
for val in sequence:
if val == "moras":
for j in sequence[val]:
mora_list.append(j)
elif val == "pause_mora" and sequence[val]:
mora_list.append(sequence[val])
for i in mora_list:
for j in i:
if j in ["consonant", "vowel"] and i[j]:
text+="%s,%d,"%(i[j], int(i[j+"_length"] * 1000))
with open('./wav/%s.txt'%filename, 'w', encoding='UTF-8') as f:
f.write(text.rstrip(","))
# synthesis
synth_payload = {"speaker": name}
r = requests.post("http://127.0.0.1:50031/synthesis", params=synth_payload,
data=json.dumps(query_data))
with open(wav_file, "wb") as fp:
fp.write(r.content)
def AIVOICE_voice(v_name:str, filename:str, interval:str, speed:str, intonation:str, sentence:str):
wav_file = "%s\wav\%s.wav" % (os. getcwd(), filename)
if int(interval) > 200: interval = "200"
elif int(interval) < 50: interval = "50"
if int(speed) > 400 : speed = "400"
elif int(speed) < 50: speed = "50"
AIVOICE_name = AIVOICE_speakers[v_name]
# 連続した半角記号を一文字化
match_filler = list(nchars(sentence, 2))
if match_filler:
for i in match_filler:
filter = re.search('[「『!?！？。、]+', i)
if filter:
sentence = sentence.replace(i, list(set(i))[0])
if sentence.endswith("!") or sentence.endswith("?"):
intonation = "150"
# 発声処理
cmd = "SeikaSay2.exe -cid %d -save %s -volume 1 -speed %f -pitch %f -intonation %f" % (AIVOICE_name, wav_file, int(speed) / 100, int(interval) / 100, int(intonation) / 100)
cmd += " -t %s" % sentence
subprocess.Popen(cmd, shell=True)
wav_wait("%s.wav"%filename)
wav_wait("%s.lab"%filename)
# 音素解析
l = [s.strip() for s in open("./wav/%s.lab"%filename, encoding = "utf-8_sig").readlines()]
text = ""
for i, val in enumerate(l):
temp = val.split()
_time = (float(temp[1]) - float(temp[0])) / 10000
text+= temp[2]+",%d,"%int(_time)
text = text.replace("q,", "cl,")
with open('./wav/%s.txt'%filename, 'w', encoding='UTF-8') as f:
f.write(text.rstrip(",")) #音素列を文字として保存
def softalk_voice(v_name, filename, interval, speed, sentence):
"""Make wav file used SofTalk.\n
Japanese name -> (音声名, 音程, 速度, 発声文章、感情)\n
UtaYomi -> True(重音テト, 東北きりたん)\n
return -> wav_file: path
"""
wav_file = "%s\wav\%s.wav" % (os.getcwd(), filename)
if v_name in ["重音テト", "MOTRoid", "試聴用", "東北きりたん"]:
v_name = "唄詠：" + v_name
cmd = "Softalk.exe /NM:%s /PS:True /O:%s /S:%s /V:50 /R:%s /X:1 /W:%s" % (v_name, interval, speed, wav_file, sentence)
_ = subprocess.Popen(cmd, shell=True)
def voice(v_name:str, interval:str, speed:str, intonation:str, text:str):
"""Make wav file used SofTalk or Voicevox.\n
Japanese name -> (音声名, 音程, 速度, 発声文章、感情)\n
UtaYomi -> True(重音テト, 東北きりたん)\n
return -> wav_file: path
"""
file_name = "%s" %(datetime.now().strftime("%d%H%M%S"))
if v_name in [j for j in Voicevox_speakers]:
Voicevox_voice(v_name, file_name, interval, speed, intonation, text)
elif v_name in [q for q in Coeiroink_speakers]:
Coeiroink_voice(v_name, file_name, interval, speed, intonation, text)
elif v_name in [i for i in AIVOICE_speakers]:
AIVOICE_voice(v_name, file_name, interval, speed, intonation, text)
else:
softalk_voice(v_name, file_name, interval, speed, text)
return file_name