tuna2134/sbv2-v2
1
0
Fork 0
Code Issues Pull Requests 1 Actions Packages Projects Releases Wiki Activity

[Fix] Fix Volume inconsistent (#144)

Browse Source 
* [Fix] Fix Volume inconsistent

Fixed the problem of volume inconsistency between sentence segments caused by Gradio volume normalization

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
This commit is contained in:
vicky Lin
2023-11-02 17:27:52 +08:00
committed by GitHub
parent 74467f5f73
commit 6227012b9e
1 changed files with 14 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

21
webui.py
View File

@@ -40,7 +40,7 @@ def generate_audio(
language, language,
): ):
audio_list = [] audio_list = []
silence = np.zeros(hps.data.sampling_rate // 2) silence = np.zeros(hps.data.sampling_rate // 2, dtype=np.int16)
with torch.no_grad(): with torch.no_grad():
for piece in slices: for piece in slices:
audio = infer( audio = infer(
@@ -55,7 +55,8 @@ def generate_audio(
net_g=net_g, net_g=net_g,
device=device, device=device,
) )
audio_list.append(audio) audio16bit = gr.processing_utils.convert_to_16_bit_wav(audio)
audio_list.append(audio16bit)
audio_list.append(silence) # 将静音添加到列表中 audio_list.append(silence) # 将静音添加到列表中
return audio_list return audio_list
@@ -92,11 +93,13 @@ def tts_split(
net_g=net_g, net_g=net_g,
device=device, device=device,
) )
audio_list.append(audio) audio16bit = gr.processing_utils.convert_to_16_bit_wav(audio)
silence = np.zeros((int)(44100 * interval_between_para)) audio_list.append(audio16bit)
silence = np.zeros((int)(44100 * interval_between_para), dtype=np.int16)
audio_list.append(silence) audio_list.append(silence)
else: else:
for p in para_list: for p in para_list:
audio_list_sent = []
sent_list = re_matching.cut_sent(p) sent_list = re_matching.cut_sent(p)
for s in sent_list: for s in sent_list:
audio = infer( audio = infer(
@@ -111,14 +114,18 @@ def tts_split(
net_g=net_g, net_g=net_g,
device=device, device=device,
) )
audio_list.append(audio) audio_list_sent.append(audio)
silence = np.zeros((int)(44100 * interval_between_sent)) silence = np.zeros((int)(44100 * interval_between_sent))
audio_list.append(silence) audio_list_sent.append(silence)
if (interval_between_para - interval_between_sent) > 0: if (interval_between_para - interval_between_sent) > 0:
silence = np.zeros( silence = np.zeros(
(int)(44100 * (interval_between_para - interval_between_sent)) (int)(44100 * (interval_between_para - interval_between_sent))
) )
audio_list.append(silence) audio_list_sent.append(silence)
audio16bit = gr.processing_utils.convert_to_16_bit_wav(
np.concatenate(audio_list_sent)
) # 对完整句子做音量归一
audio_list.append(audio16bit)
audio_concat = np.concatenate(audio_list) audio_concat = np.concatenate(audio_list)
return ("Success", (44100, audio_concat)) return ("Success", (44100, audio_concat))