Labyricorn/voicebox
1
0
Fork 0
mirror of https://github.com/jamiepine/voicebox.git synced 2026-10-02 16:45:15 -07:00
Code Issues Packages Projects Releases Wiki Activity

comment cleanup

Browse Source
This commit is contained in:
James Pine
2026-03-16 01:46:19 -07:00
parent fe19a9ca47
commit b7781951df
10 changed files with 738 additions and 694 deletions
Download Patch File Download Diff File Expand all files Collapse all files
backend/utils/chunked_tts.py
+16 -19
View File
@@ -58,11 +58,6 @@ _ABBREVIATIONS = frozenset(
_PARA_TAG_RE = re.compile(r"\[[^\]]*\]")
# ---------------------------------------------------------------------------
# Text splitting
# ---------------------------------------------------------------------------
def split_text_into_chunks(text: str, max_chars: int = DEFAULT_MAX_CHUNK_CHARS) -> List[str]:
"""Split *text* at natural boundaries into chunks of at most *max_chars*.
@@ -174,11 +169,6 @@ def _safe_hard_cut(segment: str, max_chars: int) -> int:
return cut
# ---------------------------------------------------------------------------
# Audio concatenation
# ---------------------------------------------------------------------------
def concatenate_audio_chunks(
chunks: List[np.ndarray],
sample_rate: int,
@@ -211,11 +201,6 @@ def concatenate_audio_chunks(
return result
# ---------------------------------------------------------------------------
# Engine-agnostic chunked generation
# ---------------------------------------------------------------------------
async def generate_chunked(
backend,
text: str,
@@ -264,7 +249,11 @@ async def generate_chunked(
if len(chunks) <= 1:
# Short text — single-shot fast path
audio, sample_rate = await backend.generate(
text, voice_prompt, language, seed, instruct,
text,
voice_prompt,
language,
seed,
instruct,
)
if trim_fn is not None:
audio = trim_fn(audio, sample_rate)
@@ -273,7 +262,9 @@ async def generate_chunked(
# Long text — chunked generation
logger.info(
"Splitting %d chars into %d chunks (max %d chars each)",
len(text), len(chunks), max_chunk_chars,
len(text),
len(chunks),
max_chunk_chars,
)
audio_chunks: List[np.ndarray] = []
sample_rate: int | None = None
@@ -281,7 +272,9 @@ async def generate_chunked(
for i, chunk_text in enumerate(chunks):
logger.info(
"Generating chunk %d/%d (%d chars)",
i + 1, len(chunks), len(chunk_text),
i + 1,
len(chunks),
len(chunk_text),
)
# Vary the seed per chunk to avoid correlated RNG artefacts,
# but keep it deterministic so the same (text, seed) pair
@@ -289,7 +282,11 @@ async def generate_chunked(
chunk_seed = (seed + i) if seed is not None else None
chunk_audio, chunk_sr = await backend.generate(
chunk_text, voice_prompt, language, chunk_seed, instruct,
chunk_text,
voice_prompt,
language,
chunk_seed,
instruct,
)
if trim_fn is not None:
chunk_audio = trim_fn(chunk_audio, chunk_sr)
backend/utils/effects.py
+57 -40
View File
@@ -35,10 +35,6 @@ from pedalboard import (
)
# ---------------------------------------------------------------------------
# Effect registry: maps type names -> (pedalboard class, param definitions)
# ---------------------------------------------------------------------------
# Each param definition: (default, min, max, description)
EFFECT_REGISTRY: Dict[str, Dict[str, Any]] = {
"chorus": {
@@ -46,11 +42,17 @@ EFFECT_REGISTRY: Dict[str, Dict[str, Any]] = {
"label": "Chorus / Flanger",
"description": "Modulated delay for flanging or chorus effects. Short centre_delay_ms (<10) gives flanger; longer gives chorus.",
"params": {
"rate_hz": {"default": 1.0, "min": 0.01, "max": 20.0, "step": 0.01, "description": "LFO speed (Hz)"},
"depth": {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01, "description": "Modulation depth"},
"feedback": {"default": 0.0, "min": 0.0, "max": 0.95, "step": 0.01, "description": "Feedback amount"},
"centre_delay_ms": {"default": 7.0, "min": 0.5, "max": 50.0, "step": 0.1, "description": "Centre delay (ms)"},
"mix": {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01, "description": "Wet/dry mix"},
"rate_hz": {"default": 1.0, "min": 0.01, "max": 20.0, "step": 0.01, "description": "LFO speed (Hz)"},
"depth": {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01, "description": "Modulation depth"},
"feedback": {"default": 0.0, "min": 0.0, "max": 0.95, "step": 0.01, "description": "Feedback amount"},
"centre_delay_ms": {
"default": 7.0,
"min": 0.5,
"max": 50.0,
"step": 0.1,
"description": "Centre delay (ms)",
},
"mix": {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01, "description": "Wet/dry mix"},
},
},
"reverb": {
@@ -58,11 +60,11 @@ EFFECT_REGISTRY: Dict[str, Dict[str, Any]] = {
"label": "Reverb",
"description": "Room reverb effect.",
"params": {
"room_size": {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01, "description": "Room size"},
"damping": {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01, "description": "High frequency damping"},
"wet_level": {"default": 0.33, "min": 0.0, "max": 1.0, "step": 0.01, "description": "Wet level"},
"dry_level": {"default": 0.4, "min": 0.0, "max": 1.0, "step": 0.01, "description": "Dry level"},
"width": {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01, "description": "Stereo width"},
"room_size": {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01, "description": "Room size"},
"damping": {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01, "description": "High frequency damping"},
"wet_level": {"default": 0.33, "min": 0.0, "max": 1.0, "step": 0.01, "description": "Wet level"},
"dry_level": {"default": 0.4, "min": 0.0, "max": 1.0, "step": 0.01, "description": "Dry level"},
"width": {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01, "description": "Stereo width"},
},
},
"delay": {
@@ -70,9 +72,15 @@ EFFECT_REGISTRY: Dict[str, Dict[str, Any]] = {
"label": "Delay",
"description": "Echo / delay line.",
"params": {
"delay_seconds": {"default": 0.3, "min": 0.01, "max": 2.0, "step": 0.01, "description": "Delay time (seconds)"},
"feedback": {"default": 0.3, "min": 0.0, "max": 0.95, "step": 0.01, "description": "Feedback amount"},
"mix": {"default": 0.3, "min": 0.0, "max": 1.0, "step": 0.01, "description": "Wet/dry mix"},
"delay_seconds": {
"default": 0.3,
"min": 0.01,
"max": 2.0,
"step": 0.01,
"description": "Delay time (seconds)",
},
"feedback": {"default": 0.3, "min": 0.0, "max": 0.95, "step": 0.01, "description": "Feedback amount"},
"mix": {"default": 0.3, "min": 0.0, "max": 1.0, "step": 0.01, "description": "Wet/dry mix"},
},
},
"compressor": {
@@ -80,10 +88,16 @@ EFFECT_REGISTRY: Dict[str, Dict[str, Any]] = {
"label": "Compressor",
"description": "Dynamic range compression for consistent loudness.",
"params": {
"threshold_db": {"default": -20.0, "min": -60.0, "max": 0.0, "step": 0.5, "description": "Threshold (dB)"},
"ratio": {"default": 4.0, "min": 1.0, "max": 20.0, "step": 0.1, "description": "Compression ratio"},
"attack_ms": {"default": 10.0, "min": 0.1, "max": 100.0, "step": 0.1, "description": "Attack time (ms)"},
"release_ms": {"default": 100.0, "min": 10.0, "max": 1000.0,"step": 1.0, "description": "Release time (ms)"},
"threshold_db": {"default": -20.0, "min": -60.0, "max": 0.0, "step": 0.5, "description": "Threshold (dB)"},
"ratio": {"default": 4.0, "min": 1.0, "max": 20.0, "step": 0.1, "description": "Compression ratio"},
"attack_ms": {"default": 10.0, "min": 0.1, "max": 100.0, "step": 0.1, "description": "Attack time (ms)"},
"release_ms": {
"default": 100.0,
"min": 10.0,
"max": 1000.0,
"step": 1.0,
"description": "Release time (ms)",
},
},
},
"gain": {
@@ -99,7 +113,13 @@ EFFECT_REGISTRY: Dict[str, Dict[str, Any]] = {
"label": "High-Pass Filter",
"description": "Removes frequencies below the cutoff.",
"params": {
"cutoff_frequency_hz": {"default": 80.0, "min": 20.0, "max": 8000.0, "step": 1.0, "description": "Cutoff frequency (Hz)"},
"cutoff_frequency_hz": {
"default": 80.0,
"min": 20.0,
"max": 8000.0,
"step": 1.0,
"description": "Cutoff frequency (Hz)",
},
},
},
"lowpass": {
@@ -107,7 +127,13 @@ EFFECT_REGISTRY: Dict[str, Dict[str, Any]] = {
"label": "Low-Pass Filter",
"description": "Removes frequencies above the cutoff.",
"params": {
"cutoff_frequency_hz": {"default": 8000.0, "min": 200.0, "max": 20000.0, "step": 1.0, "description": "Cutoff frequency (Hz)"},
"cutoff_frequency_hz": {
"default": 8000.0,
"min": 200.0,
"max": 20000.0,
"step": 1.0,
"description": "Cutoff frequency (Hz)",
},
},
},
"pitch_shift": {
@@ -121,10 +147,6 @@ EFFECT_REGISTRY: Dict[str, Dict[str, Any]] = {
}
# ---------------------------------------------------------------------------
# Built-in presets
# ---------------------------------------------------------------------------
BUILTIN_PRESETS: Dict[str, Dict[str, Any]] = {
"robotic": {
"name": "Robotic",
@@ -233,10 +255,6 @@ BUILTIN_PRESETS: Dict[str, Dict[str, Any]] = {
}
# ---------------------------------------------------------------------------
# Public API
# ---------------------------------------------------------------------------
def get_available_effects() -> List[Dict[str, Any]]:
"""Return the list of available effect types with their parameter definitions.
@@ -244,15 +262,14 @@ def get_available_effects() -> List[Dict[str, Any]]:
"""
result = []
for effect_type, info in EFFECT_REGISTRY.items():
result.append({
"type": effect_type,
"label": info["label"],
"description": info["description"],
"params": {
name: {k: v for k, v in pdef.items()}
for name, pdef in info["params"].items()
},
})
result.append(
{
"type": effect_type,
"label": info["label"],
"description": info["description"],
"params": {name: {k: v for k, v in pdef.items()} for name, pdef in info["params"].items()},
}
)
return result