ppp-svc-helper 0.1.0__tar.gz

ppp_svc_helper-0.1.0/LICENSE.md

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2025 therealvul
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

ppp_svc_helper-0.1.0/PKG-INFO

@@ -0,0 +1,26 @@
+Metadata-Version: 2.4
+Name: ppp_svc_helper
+Version: 0.1.0
+Summary: Utility library for singing voice conversion work
+Author: effusiveperiscope
+Author-email: therealvul <wowitsvul@gmail.com>
+License-Expression: MIT
+Project-URL: Homepage, https://github.com/pypa/sampleproject
+Project-URL: Issues, https://github.com/pypa/sampleproject/issues
+Classifier: Programming Language :: Python :: 3
+Classifier: Operating System :: OS Independent
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+License-File: LICENSE.md
+Dynamic: author
+Dynamic: license-file
+
+# What is this
+Contains various utilities for common SVC (singing voice conversion) data
+ tasks
+based on popular singing voice conversion libraries--interfaces for
+pitch extraction, speech feature extraction, singing voice conversion, data augmentation.
+
+# Credits
+* [RVC](https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI)
+* Black, Dawn A. A., Li, Ma and Tian, Mi. "Automatic Identification of Emotional Cues in Chinese Opera Singing", in Proc. of 13th Int. Conf. on Music Perception and Cognition and the 5th Conference for the Asian-Pacific Society for Cognitive Sciences of Music (ICMPC 13-APSC0M 5 2014), Seoul, South Korea, August 2014.

ppp_svc_helper-0.1.0/README.md

@@ -0,0 +1,9 @@
+# What is this
+Contains various utilities for common SVC (singing voice conversion) data
+ tasks
+based on popular singing voice conversion libraries--interfaces for
+pitch extraction, speech feature extraction, singing voice conversion, data augmentation.
+
+# Credits
+* [RVC](https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI)
+* Black, Dawn A. A., Li, Ma and Tian, Mi. "Automatic Identification of Emotional Cues in Chinese Opera Singing", in Proc. of 13th Int. Conf. on Music Perception and Cognition and the 5th Conference for the Asian-Pacific Society for Cognitive Sciences of Music (ICMPC 13-APSC0M 5 2014), Seoul, South Korea, August 2014.

ppp_svc_helper-0.1.0/ppp_svc_helper.egg-info/PKG-INFO

@@ -0,0 +1,26 @@
+Metadata-Version: 2.4
+Name: ppp_svc_helper
+Version: 0.1.0
+Summary: Utility library for singing voice conversion work
+Author: effusiveperiscope
+Author-email: therealvul <wowitsvul@gmail.com>
+License-Expression: MIT
+Project-URL: Homepage, https://github.com/pypa/sampleproject
+Project-URL: Issues, https://github.com/pypa/sampleproject/issues
+Classifier: Programming Language :: Python :: 3
+Classifier: Operating System :: OS Independent
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+License-File: LICENSE.md
+Dynamic: author
+Dynamic: license-file
+
+# What is this
+Contains various utilities for common SVC (singing voice conversion) data
+ tasks
+based on popular singing voice conversion libraries--interfaces for
+pitch extraction, speech feature extraction, singing voice conversion, data augmentation.
+
+# Credits
+* [RVC](https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI)
+* Black, Dawn A. A., Li, Ma and Tian, Mi. "Automatic Identification of Emotional Cues in Chinese Opera Singing", in Proc. of 13th Int. Conf. on Music Perception and Cognition and the 5th Conference for the Asian-Pacific Society for Cognitive Sciences of Music (ICMPC 13-APSC0M 5 2014), Seoul, South Korea, August 2014.

ppp_svc_helper-0.1.0/ppp_svc_helper.egg-info/SOURCES.txt

@@ -0,0 +1,55 @@
+LICENSE.md
+README.md
+pyproject.toml
+setup.py
+ppp_svc_helper.egg-info/PKG-INFO
+ppp_svc_helper.egg-info/SOURCES.txt
+ppp_svc_helper.egg-info/dependency_links.txt
+ppp_svc_helper.egg-info/top_level.txt
+svc_helper/__init__.py
+svc_helper/augmentation/__init__.py
+svc_helper/augmentation/pedalboard.py
+svc_helper/pitch/__init__.py
+svc_helper/pitch/rmvpe.py
+svc_helper/pitch/utils.py
+svc_helper/sfeatures/__init__.py
+svc_helper/sfeatures/models.py
+svc_helper/sfeatures/whisper/__init__.py
+svc_helper/sfeatures/whisper/audio.py
+svc_helper/sfeatures/whisper/decoding.py
+svc_helper/sfeatures/whisper/model.py
+svc_helper/sfeatures/whisper/tokenizer.py
+svc_helper/sfeatures/whisper/utils.py
+svc_helper/svc/__init__.py
+svc_helper/svc/rvc/__init__.py
+svc_helper/svc/rvc/configs/__init__.py
+svc_helper/svc/rvc/configs/config.py
+svc_helper/svc/rvc/configs/v1/32k.json
+svc_helper/svc/rvc/configs/v1/40k.json
+svc_helper/svc/rvc/configs/v1/48k.json
+svc_helper/svc/rvc/configs/v2/32k.json
+svc_helper/svc/rvc/configs/v2/48k.json
+svc_helper/svc/rvc/lib/__init__.py
+svc_helper/svc/rvc/lib/audio.py
+svc_helper/svc/rvc/lib/infer_pack/__init__.py
+svc_helper/svc/rvc/lib/infer_pack/attentions.py
+svc_helper/svc/rvc/lib/infer_pack/commons.py
+svc_helper/svc/rvc/lib/infer_pack/models.py
+svc_helper/svc/rvc/lib/infer_pack/models_onnx.py
+svc_helper/svc/rvc/lib/infer_pack/modules.py
+svc_helper/svc/rvc/lib/infer_pack/onnx_inference.py
+svc_helper/svc/rvc/lib/infer_pack/transforms.py
+svc_helper/svc/rvc/lib/train/__init__.py
+svc_helper/svc/rvc/lib/train/data_utils.py
+svc_helper/svc/rvc/lib/train/losses.py
+svc_helper/svc/rvc/lib/train/mel_processing.py
+svc_helper/svc/rvc/lib/train/process_ckpt.py
+svc_helper/svc/rvc/lib/train/utils.py
+svc_helper/svc/rvc/modules/__init__.py
+svc_helper/svc/rvc/modules/vc/__init__.py
+svc_helper/svc/rvc/modules/vc/modules.py
+svc_helper/svc/rvc/modules/vc/pipeline.py
+svc_helper/svc/rvc/modules/vc/utils.py
+tests/test_pitch.py
+tests/test_sfeatures.py
+tests/test_svc.py

ppp_svc_helper-0.1.0/ppp_svc_helper.egg-info/dependency_links.txt

@@ -0,0 +1 @@
+

ppp_svc_helper-0.1.0/ppp_svc_helper.egg-info/top_level.txt

@@ -0,0 +1 @@
+svc_helper

ppp_svc_helper-0.1.0/pyproject.toml

@@ -0,0 +1,23 @@
+[build-system]
+requires = ["setuptools >= 77.0.3"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "ppp_svc_helper"
+version = "0.1.0"
+authors = [
+  { name="therealvul", email="wowitsvul@gmail.com" },
+]
+description = "Utility library for singing voice conversion work"
+readme = "README.md"
+requires-python = ">=3.9"
+classifiers = [
+    "Programming Language :: Python :: 3",
+    "Operating System :: OS Independent",
+]
+license = "MIT"
+license-files = ["LICEN[CS]E*"]
+
+[project.urls]
+Homepage = "https://github.com/pypa/sampleproject"
+Issues = "https://github.com/pypa/sampleproject/issues"

ppp_svc_helper-0.1.0/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+

ppp_svc_helper-0.1.0/setup.py

@@ -0,0 +1,14 @@
+import os
+from setuptools import setup, find_packages
+
+setup(
+    name="svc_helper",
+    version="0.1.0",
+    author="effusiveperiscope",
+    description = (""),
+    license="MIT",
+    packages=find_packages(exclude=['tests']),
+    package_data={
+        'svc_helper': ['svc/rvc/configs/**/*.json']
+    }
+)

ppp_svc_helper-0.1.0/svc_helper/augmentation/pedalboard.py

@@ -0,0 +1,104 @@
+import pedalboard
+import random
+import numpy as np
+
+def _randp(prob):
+    return (random.random() < prob)
+
+# Randomly applies various pedalboard effects to audio
+class PedalboardRandomAugmentor:
+    def __init__(self, probs : dict = {}):
+        """
+        Accepts a dictionary of probabilities for various augmentations (see code for details)
+        """
+        default_probs = {
+            'base': 0.3, # probability of activating augmentation
+            'clipping': 0.05,
+            'comp_gentle': 0.2,
+            'comp_hard': 0.2,
+            'quiet': 0.2,
+            'limit': 0.1,
+            'resample_8k': 0.05,
+            'resample_16k': 0.05,
+            'resample_22k': 0.05,
+            'resample_24k': 0.05,
+            'reverb_short': 0.05,
+            'reverb_long': 0.05,
+            'delay_short': 0.05,
+            'bitcrush_8': 0.1,
+            'mp3_vbr2': 0.1,
+            'mp3_vbr0': 0.1,
+        }
+        self.probs = default_probs
+        self.probs.update(probs)
+        self.clipper = pedalboard.Clipping()
+        self.comp_gentle = pedalboard.Compressor(
+            threshold_db=-6.0, ratio=1.5, attack_ms=60.0)
+        self.comp_hard = pedalboard.Compressor(
+            threshold_db=-24.0, ratio=3)
+        self.limit = pedalboard.Limiter(threshold_db=-6.0)
+        self.quiet = pedalboard.Gain(gain_db=-6.0)
+        self.resample_8k = pedalboard.Resample(target_sample_rate=8000.0,
+            quality=pedalboard.Resample.Quality.Linear)
+        self.resample_16k = pedalboard.Resample(target_sample_rate=16000.0,
+            quality=pedalboard.Resample.Quality.Linear)
+        self.resample_22k = pedalboard.Resample(target_sample_rate=22050.0,
+            quality=pedalboard.Resample.Quality.Linear)
+        self.resample_24k = pedalboard.Resample(target_sample_rate=24000.0,
+            quality=pedalboard.Resample.Quality.Linear)
+        self.reverb_short = pedalboard.Reverb(room_size = 0.1, wet_level = 0.1)
+        self.reverb_long = pedalboard.Reverb(room_size = 0.9, wet_level = 0.3)
+        self.delay_short = pedalboard.Delay(delay_seconds=0.05, feedback=0.3)
+        self.bitcrush_8 = pedalboard.Bitcrush(bit_depth=8)
+        self.mp3_vbr2 = pedalboard.MP3Compressor(vbr_quality=2.0)
+        self.mp3_vbr0 = pedalboard.MP3Compressor(vbr_quality=0.1)
+
+    def process(self, audio : np.ndarray, sr : float):
+        self.clipper.reset()
+        self.comp_gentle.reset()
+        self.comp_hard.reset()
+        self.limit.reset()
+        self.delay_short.reset()
+        self.reverb_short.reset()
+        self.reverb_long.reset()
+
+        if not _randp(self.probs['base']):
+            return audio
+
+        if _randp(self.probs['reverb_short']):
+            audio = self.reverb_short.process(audio, sr)
+        if _randp(self.probs['reverb_long']):
+            audio = self.reverb_long.process(audio, sr)
+
+        if _randp(self.probs['delay_short']):
+            audio = self.delay_short.process(audio, sr)
+
+        if _randp(self.probs['clipping']):
+            audio = self.clipper.process(audio, sr)
+        if _randp(self.probs['comp_gentle']):
+            audio = self.comp_gentle.process(audio, sr)
+        if _randp(self.probs['comp_hard']):
+            audio = self.comp_hard.process(audio, sr)
+        if _randp(self.probs['limit']):
+            audio = self.limit.process(audio, sr)
+
+        if _randp(self.probs['resample_8k']):
+            audio = self.resample_8k.process(audio, sr)
+        elif _randp(self.probs['resample_16k']):
+            audio = self.resample_16k.process(audio, sr)
+        elif _randp(self.probs['resample_22k']):
+            audio = self.resample_22k.process(audio, sr)
+        elif _randp(self.probs['resample_24k']):
+            audio = self.resample_24k.process(audio, sr)
+
+        if _randp(self.probs['bitcrush_8']):
+            audio = self.bitcrush_8.process(audio, sr)
+
+        if _randp(self.probs['mp3_vbr2']) and sr >= 32000 and sr <= 48000:
+            audio = self.mp3_vbr2.process(audio, sr)
+        if _randp(self.probs['mp3_vbr0']) and sr >= 32000 and sr <= 48000:
+            audio = self.mp3_vbr0.process(audio, sr)
+        if _randp(self.probs['quiet']):
+            audio = self.quiet.process(audio, sr)
+
+        return audio