soprano_to_rvc/__pycache__/api_server.cpython-310.pyc

o
<00>uei<65>O<00>@s
UdZd
dl
Z
d
dlZd
dlZd
dlZd
dlZd
dlZd
dlZd
dlZd
dl	m
Z

d
dlmZm
Z

d
dlZd
dlZd
dlmZmZmZ
d
dlmZmZ
d
dlmZmZ
d
dlZejejdd	<09>
e<04>e <20>
Z!d
d
l"m#Z#m$Z$
d
dl%m&Z&m'Z'
edd
dd<0F>Z(da)ee#e*d<e<05>+<2B>Z,e<08>-<2D>Z.da/Gdd<13>d<13>Z0Gdd<15>de<1A>Z1Gdd<17>de<1A>Z2e(<28>3d<18>
dd<1A><00>
Z4e(<28>3d<1B>
dd<1D><00>
Z5e(<28>6d<1E>
dd <20><00>
Z7e(<28>6d!<21>
d"d#<23><00>
Z8e(<28>6d$<24>
d%d&<26><00>
Z9e(<28>:d'<27>
d(e1fd)d*<2A><04>
Z;e(<28>:d+<2B>
d(e1fd,d-<2D><04>
Z<e(<28>:d.<2E>
d(e1fd/d0<64><04>
Z=d1d2<64>Z>e d3k<02>
re><3E>
dSdS)4zj
FastAPI Server for Soprano TTS + RVC Pipeline
Provides HTTP API for text-to-speech with voice conversion
<EFBFBD>N)
<01>Path)<02>Optional<61>List)<03>FastAPI<50>
HTTPException<6F>BackgroundTasks)<02>StreamingResponse<73>JSONResponse)<02>	BaseModel<65>Fieldz4%(asctime)s - %(name)s - %(levelname)s - %(message)s)<02>level<65>format)<02>UnifiedPipeline<6E>	RVCConfig)<02>HeadlessRVC<56>HeadlessRVCConfig<69>Soprano TTS + RVC APIz.Text-to-speech with real-time voice conversion<6F>1.0.0)<03>title<6C>description<6F>version<6F>pipelineFc@sHeZ
dZd
Zdejfdd<04>Zdd<06>Zdd<08>Zd	d
<EFBFBD>Z	dd<0C>Z
d
d<0E>ZdS)<10>AudioCaptureStreamzAStream that captures audio data to a buffer instead of playing it<69>buffercCs|
|_d
|_
d|_dS)N逻<00>)r<00>
samplerate<EFBFBD>channels)<02>selfr<00>r<00>'/data/NVMe/AI/soprano2RVC/api_server.py<70>__init__4s



zAudioCaptureStream.__init__cCs|j<00>
|
<01><02><00>

d
S)zCapture audio data to bufferN)r<00>put<75>copy)r<00>datarrr <00>write9szAudioCaptureStream.writec


C<00>d
S<00>zNo-op for compatibilityNr<00>
rrrr <00>start=<00>zAudioCaptureStream.startc


Cr&r'rr(rrr <00>stopAr*zAudioCaptureStream.stopc


Cr&r'rr(rrr <00>abortEr*zAudioCaptureStream.abortc


Cr&r'rr(rrr <00>closeIr*zAudioCaptureStream.closeN)<0C>__name__<5F>
__module__<EFBFBD>__qualname__<5F>__doc__<5F>queue<75>Queuer!r%r)r+r,r-rrrr r2s

rc@sNeZ
dZUed
dddd<05>Zeed<eddd	<09>Zeed
<eddd	<09>Ze	ed
<dS)<0F>
TTSRequest.zText to synthesize<7A>
i<>)r<00>
min_length<EFBFBD>
max_length<EFBFBD>text<78>wavzAudio format (wav, mp3, flac))<02>defaultrr
rzOutput sample rate in Hz<48>sample_rateN)
r.r/r0rr8<00>str<74>__annotations__r
r;<00>intrrrr r4Ns



r4c@s>eZ
dZUeed
<eed<dZeeed<dZ	ee
ed<dS)<06>TTSResponse<73>success<73>messageN<65>audio_lengthr;)r.r/r0<00>boolr=r<rBr<00>floatr;r>rrrr r?Ts





r?<00>startupc<00>s<><00>
t<00>
d
<01>

tt<03>
jd}|<00><05>s t<00>d|<00><00><02>

td|<00><00><02>
<01>
t<00>
d|<00><00><02>

t<08>	t
|<00>
<01>
}
t<00>
d<05>

t|
<01>
at<0C>
<0A>
t<00>
d<06>

t<0C><0E>
t<00>
d<07>

t<0C><0F>
t<00>
d<08>

d	S)
z"Initialize the pipeline on startupzStarting API server...zrvc_config.jsonzConfiguration file not found: zLoading configuration from: z*Initializing Soprano TTS + RVC pipeline...zLoading Soprano TTS model...z Starting RVC voice conversion...u✓ API server ready!N)<10>logger<65>infor<00>__file__<5F>parent<6E>exists<74>error<6F>RuntimeErrorr<00>	from_filer<rr<00>ensure_virtual_sink<6E>initialize_soprano<6E>	start_rvc)<02>config_path<74>
rvc_configrrr <00>
startup_event[s <02>










rS<00>shutdownc<00>s&<00>
t<00>
d
<01>

trt<02><03>
t<00>
d<02>

dS)zCleanup on shutdownzShutting down API server...u✓ Shutdown completeN)rFrGr<00>cleanuprrrr <00>shutdown_event}s
<02>

rV<00>
/c	<00>s&<00>
d
dtrdn
dddddd	d
<EFBFBD>dd<0C>S)
zRoot endpoint - API inforr<00>running<6E>initializingu1/api/v1/synthesize (FULL PIPELINE: Soprano→RVC)z2/api/v1/synthesize/soprano-only (TTS only, no RVC)z./api/v1/synthesize/stream (Streaming with RVC)<29>/health<74>/config)<05>
synthesize<EFBFBD>synthesize_soprano_only<6C>synthesize_stream<61>health<74>configzMDefault /synthesize endpoint now uses FULL pipeline with RVC voice conversion)<05>servicer<00>status<75>	endpoints<74>note)
rrrrr <00>root<6F>s<02>






<04><06>rerZc<00>s8<00>
td
urt
ddd<04><02>
dtjd
u
tjd
u
otj<03><04>d<06>S)zHealth check endpointN<74><4E>
<00>Pipeline not initialized<65><02>status_code<64>detail<69>healthy)rb<00>soprano_loaded<65>rvc_running)rr<00>soprano<6E>
rvc_thread<EFBFBD>is_aliverrrr <00>health_check<63>s<02>


<06>rqr[c
<00>sR<00>
td
urt
ddd<04><02>
tjjtjjtjjtjjtjjtjjtjj	tjj
tjjd<05>	S)zGet current RVC configurationNrfrgrh)	<09>pth_path<74>
index_path<EFBFBD>pitch<63>
index_rate<EFBFBD>f0method<6F>input_device<63>
output_device<63>formant<6E>use_pv)rrrR<00>pth<74>indexrtrurvrwrxryrzrrrr <00>
get_config<EFBFBD>s<02>








<06>r}z/api/v1/synthesize<7A>requestc
<00>s\<00>
td
urt
ddd<04><02>
t4Id
H<00><02>
<00>
z<>t<03>d|jd
d<06><00>d<07><03>

t<06><07>s<zt<06><08>
Wn
t	j
y7


Ynwt<06><07>r'dat<03>d	<09>

id
tjj
<0A>
dtjj<0E>
dtjj<0F>
d
tjj<10>
dtjj<11>
dtjj<12>
dd
<01>
dd<12>
dd<14>
dd<16>
dd<18>
dd<1A>
dd<1C>
dtjj<13>
dtjj<14>
dtjj<15>
d tjj<16>
tjjtjjtjjd!d"<22><04>
}
t|
<01>
}t|<02>
}|<03><1C>
g}t<03>d#<23>

tj<1D>|j<05>
D]<5D>}t|d$<24>r<>|<05> <20><00>!<21>}|j"d%kr<>|<05>#<23>}d&d'l$m%}
t&t'|<05>
dd(<00>
}|<06>(||<07>}t)j*|d)d*d+d,<2C>}t)<29>+|d+d*<2A>}t)<29>,||f<02>
<01>-t)j.<2E>
}	t)<29>/|	<09>
}
z|<03>0|	|
t'|	<09>
d
d
<01>
|<04>1|
d
d
<01>d&f<00>

Wq<>t2<74>
y7
}
zt<03>3d-|<0B><00><02>

|<04>1|<08>

WYd
}~q<>d
}~w
w|<04>
sAt
d.d/d<04><02>
t)<29>4|<04>
}t<03>d0t'|<0C>
dd1<64>d2<64><03>

|j5dk<03>
rkt&t'|<0C>
|j5d<00>
}|<06>(||<07>}t)j*|d)d*d+d,<2C>}t)<29>+|d+d*<2A>}t6<74>7<EFBFBD>}
|j8d3k<02>
r<>t9j:|
||j5d4d5<64>
d6}n0|j8d7k<02>
r<>t9j:|
||j5d8d5<64>
d9}n|j8d:k<02>
r<>t9j:|
||j5d;d5<64>
d<}n
t
d.d=|j8<6A><00>d<04><02>
|
<0A>;d&<26>

t'|<0C>
|j5}t<03>d>|d1<64>d?<3F><03>

d@at<|
|dA|j8<6A><00>t=|<0F>
t=|j5<6A>
dBdC<64>dD<64>WWd
<00>Id
H
St2<74>y
}
zd@atj>dEt=|<0B>
<01><00>ddF<64>
t
dGdHt=|<0B>
<01><00>d<04><02>
d
}~w
w1Id
H<00>s'w



Y
d
S)Iuz
    Synthesize speech from text through the FULL pipeline (Soprano TTS → RVC)
    Returns RVC-processed audio file
    NrfrgrhzProcessing synthesis request: '<27>2u*...' through FULL pipeline (Soprano→RVC)Tz$Setting up RVC with audio capture...rrrsrtrurvrwrxrr<00>
block_time<EFBFBD><00>?<3F>crossfade_time<6D>{<14>G<EFBFBD>z<EFBFBD>?<3F>
extra_time<EFBFBD>@<40>n_cpu<70><00>	threshold<6C>N<>ry<00>
filter_radius<75>rms_mix_rate<74>protectr<00><04>I_noise_reduce<63>O_noise_reducerzrz%Generating speech with Soprano TTS...<2E>cpur5r
<00>
<01>signal<61>}<00><00><00>?<3F><00><><EFBFBD><03>nan<61>posinf<6E>neginfzRVC processing error: <20><>
<00>No audio generatedz'RVC processing complete. Audio length: <20>.2f<EFBFBD>
sr9<00>WAV<41>
r
<00>	audio/wav<61>mp3<70>MP3<50>
audio/mpeg<65>flac<61>FLAC<41>
audio/flac<61>Unsupported format: <20>
Generated zs of RVC-processed audioFz#attachment; filename=synthesis_rvc.zSoprano-TTS-RVC<56><04>Content-DispositionzX-Audio-Duration<6F>
X-Sample-Rate<74>X-Processed-By<42><02>
media_type<EFBFBD>headers<72>Synthesis error: <20>
<01>exc_info<66><6F>
<00>Synthesis failed: )?rr<00>
pipeline_lockrFrGr8<00>audio_capture_buffer<65>empty<74>
get_nowaitr2<00>Empty<74>capture_enabledrRr{r|rtrurvrwryr<>r<>r<>r<>r<>rzrr<00>initialize_rvcrn<00>infer_stream<61>hasattrr<72><00>numpy<70>ndim<69>squeeze<7A>scipyr<79>r><00>len<65>resample<6C>np<6E>
nan_to_num<EFBFBD>clip<69>column_stack<63>astype<70>float32<33>
zeros_like<EFBFBD>audio_callback<63>append<6E>	Exception<6F>warning<6E>concatenater;<00>io<69>BytesIOr
<00>sfr%<00>seekrr<rK)r~<00>config_dictrR<00>headless_rvc<76>collected_audio<69>chunk<6E>scipy_signal<61>num_samples<65>	chunk_48k<38>chunk_stereo<65>output_chunk<6E>
e<>
audio_data<EFBFBD>audio_bufferr<72><00>durationrrr r\<00>s<><02>






<02><08>

<02><02><02><02><02><02><02><02>	<02>
<02><02><02>
<02><02><02><02><02>


<08>















<08><02>




















<04><06><16>


<08><02><00><1C>r\z/api/v1/synthesize/soprano-onlyc

<00>sX<00>
td
urt
ddd<04><02>
t4Id
H<00>
<01>
z<>t<03>d|jd
d<06><00>d<07><03>

g}
tj<06>|j<05>
D]7}t|d<08>r8|<02>	<09><00>
<EFBFBD>}|jd	krA|<02><0C>}|j
d
kr]ddlm}
tt|<02>
|j
d
<00>
}|<03>||<04>}|
<01>|<02>

q+|
skt
d
dd<04><02>
t<14>|
<01>
}tj|dddd<12>}t<14>|dd<10>}t<18><19>}|jdkr<>tj|||j
dd<15>
d}n.|jdkr<>tj|||j
dd<15>
d}n|jdkr<>tj|||j
dd<15>
d}n
t
d
d|j<1A><00>d<04><02>
|<06>d<0B>

t|<05>
|j
}t<03>d|d<1F>d <20><03>

t||d!|j<1A><00>t|<08>
t|j
<0A>
d"d#<23>d$<24>WWd
<00>Id
H
St <20>
y
}	
ztj!d%t|	<09>
<01><00>d&d'<27>
t
d(d)t|	<09>
<01><00>d<04><02>
d
}	~	w
w1Id
H<00>
s%w



Y
d
S)*z<>
    Synthesize speech using ONLY Soprano TTS (no RVC voice conversion)
    Use this for comparison or if you want raw TTS output
    Nrfrgrhz$Processing Soprano-only synthesis: 'r<00>...'r<>r5r<>r
r<>r<>r<>r<>r<>r<>r<>r9r<>r<>r<>r<>r<>r<>r<>r<>r<>r<>r<>r<>zs of Soprano-only audioz'attachment; filename=synthesis_soprano.zSoprano-TTS-Onlyr<79>r<>r<>Tr<54>r<>r<>)"rrr<>rFrGr8rnr<>r<>r<>r<>r<>r<>r;r<>r<>r>r<>r<>r<>r<>r<>r<>r<>r<>r<>r
r<>r%r<>rr<r<>rK)
r~<00>audio_chunksr<73>r<>r<>r<>r<>r<>r<>r<>rrr r]R
sh<02>
































<04><06><16>B

<08><02><1C>r]z/api/v1/synthesize/streamc
<00>s@<00>
td
urt
ddd<04><02>
<01>f
dd<06>}
t|
<01>ddt<03>j<04>
d	d
<EFBFBD>d<0B>S)u<>
    Stream audio chunks through FULL pipeline (Soprano TTS → RVC) as they're generated
    Note: Streaming with RVC may have latency due to processing
    Nrfrgrhcs~<00><02>
z"t<00>
d
<01>jdd<03><00>d<04><03>

idtjj<05>
dtjj<06>
dtjj<07>
dtjj<08>
d	tjj	<09>
d
tjj
<EFBFBD>
dd<02>
dd
<0A>
dd<0F>
dd<11>
dd<13>
dd<15>
dd<17>
dtjj<0B>
dtjj<0C>
dtjj
<0A>
dtjj<0E>
tjjtjjtjjdd<1D><04>
}t|<00>
}
t|
<01>
}|<02><14>
tj<15><16>j<02>
D]<5D>}t|d<1E>r<>|<03><18><00><19>}|jdkr<>|<03><1B>}d d!lm}
tt|<03>
d
d"<00>
}|<04> ||<05>}t!j"|d#d$d%d&<26>}t!<21>#|d%d$<24>}t!<21>$||f<02>
<01>%t!j&<26>
}t!<21>'|<07>
}z|<02>(||t|<07>
dd<02>
|dd<02>d f}	Wnt)y<>
}

zt<00>*d'|
<EFBFBD><00><02>

|}	WYd}
~
nd}
~
w
w<00>j+d
k<03>
rtt|	<09>
<01>j+d
<00>
}|<04> |	|<05>}	|	d(<00>%t!j,<2C>
}|<0B>-<2D>V
qzt<00>
d)<29>

WdSt)<29>
y>
}

z
tj.d*t/|
<EFBFBD>
<01><00>d+d,<2C>
<00>d}
~
w
w)-z#Generate RVC-processed audio chunkszStreaming synthesis with RVC: 'Nrr<>rrrsrtrurvrwrxrrr<>r<>r<>r<>r<>r<>r<>r<>r<>r<>ryr<>r<>r<>rr<>r<>r5r
r<>r<>r<>r<>r<>r<>zRVC streaming error: i<>zStreaming completezStreaming error: Tr<54>)0rFrGr8rrRr{r|rtrurvrwryr<>r<>r<>r<>r<>rzrrr<>rnr<>r<>r<>r<>r<>r<>r<>r<>r>r<>r<>r<>r<>r<>r<>r<>r<>r<>r<>r<>r<>r;<00>int16<31>tobytesrKr<)r<>rRr<>r<>r<>r<>r<>r<>r<><00>processed_audior<6F><00>chunk_int16<31>
r~rr <00>audio_generator<6F>
s<><02>

<02><02><02><02><02><02><02><02>	<02>
<02><02><02>
<02><02><02><02><02>


<08>















<08><02>




<08><02>z*synthesize_stream.<locals>.audio_generatorr<72>z-attachment; filename=synthesis_rvc_stream.wavzSoprano-TTS-RVC-Stream)r<>r<>r<>r<>)rrrr<r;)r~r<>rr<>r r^<00>
s<02>
O



<04><06>r^cCs<>d
dl}|j
dd<04>
}
|
jdtddd<08>
|
jd	td
dd<08>
|
jdd
dd<0F>
|
jdtddd<08>
|
<01><05>}tjd|j|j	|j
|jdd<15>
dS)zRun the API serverr
NzSoprano TTS + RVC API Server)
rz--hostz0.0.0.0zHost to bind to)<03>typer:<00>helpz--porti@zPort to bind toz--reload<61>
store_truezEnable auto-reload)<02>actionr<6E>z	--workersr5zNumber of worker processeszapi_server:apprG)<05>host<73>port<72>reload<61>workers<72>	log_level)<0C>argparse<73>ArgumentParser<65>add_argumentr<r><00>
parse_args<EFBFBD>uvicorn<72>runr<6E>r<>r<>r<>)r<><00>parser<65>argsrrr <00>mains










<EFBFBD>r<EFBFBD><00>__main__)?r1<00>osr<73><00>json<6F>logging<6E>asyncio<69>	threading<6E>timer2<00>pathlibr<00>typingrrr<>r<><00>	soundfiler<65><00>fastapirrr<00>fastapi.responsesrr	<00>pydanticr
rr<><00>basicConfig<69>INFO<46>	getLoggerr.rF<00>unified_soprano_rvcrrr<>rr<00>apprr=<00>Lockr<6B>r3r<>r<>rr4r?<00>on_eventrSrV<00>getrerqr}<00>postr\r]r^r<>rrrr <00><module>
sr

















<06>



<06>




!









O
b


<04>