feat(audio): add a family dictation tone picker (#707)
Settings → Audio can pick start/stop cues from the shared catalog (lift, flick, ember, step, voca, soft, chirp, scale, drop, glass, off). New installs and any missing or unknown sound_effects.tone value use voca. A saved id, including off, is left alone. The old start_recording/stop_recording files stay on disk. They are no longer the implicit default. Error cues still follow the existing enable switch.
This commit is contained in:
Jatin K Malik
+215
-83
@@ -1,33 +1,51 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
Sound generation utility for Vocalinux.
|
||||
"""Generate Vocalinux notification WAVs.
|
||||
|
||||
This script generates the notification sounds used by Vocalinux for:
|
||||
- Start recording (ascending pitch glide)
|
||||
- Stop recording (descending pitch glide)
|
||||
- Error notification (lower descending glide)
|
||||
Writes 16-bit PCM mono files at 44.1 kHz into resources/sounds/:
|
||||
|
||||
The sounds use smooth pitch glides with pure sine waves and soft envelopes
|
||||
to create pleasant, headphone-friendly audio feedback.
|
||||
- start_recording.wav / stop_recording.wav / error.wav: the original Linux
|
||||
pair. Kept on disk; not used as the implicit default anymore.
|
||||
- {id}_start.wav / {id}_stop.wav for each family tone except Off.
|
||||
|
||||
Usage:
|
||||
python scripts/generate_sounds.py
|
||||
|
||||
The generated sounds will be placed in resources/sounds/
|
||||
|
||||
Design Philosophy:
|
||||
- Pure sine waves (no harsh harmonics)
|
||||
- Smoothstep interpolation for pitch transitions
|
||||
- Soft attack/decay envelopes
|
||||
- Lower frequencies (F4-A4 range) for warmth
|
||||
- Consistent glide aesthetic across all sounds
|
||||
"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
import math
|
||||
import os
|
||||
import shutil
|
||||
import struct
|
||||
import wave
|
||||
|
||||
SAMPLE_RATE = 44100
|
||||
|
||||
# Named pitches used by the family catalog.
|
||||
G2 = 98.00
|
||||
C3 = 130.81
|
||||
D3 = 146.83
|
||||
F3 = 174.61
|
||||
G3 = 196.00
|
||||
C4 = 261.63
|
||||
E4 = 329.63
|
||||
F4 = 349.23
|
||||
G4 = 392.00
|
||||
A4 = 440.00
|
||||
C5 = 523.25
|
||||
A5 = 880.00
|
||||
C6 = 1046.50
|
||||
E7 = 2637.02
|
||||
C7 = 2093.00
|
||||
|
||||
|
||||
def _smoothstep(progress: float) -> float:
|
||||
return progress * progress * (3.0 - 2.0 * progress)
|
||||
|
||||
|
||||
def _sine_envelope(progress: float) -> float:
|
||||
return (math.sin(math.pi * progress) ** 2) * math.exp(-0.5 * progress)
|
||||
|
||||
|
||||
def generate_glide_tone(
|
||||
filename: str,
|
||||
@@ -35,99 +53,213 @@ def generate_glide_tone(
|
||||
freq_end: float,
|
||||
duration: float = 0.6,
|
||||
amplitude: float = 0.16,
|
||||
sample_rate: int = 44100,
|
||||
sample_rate: int = SAMPLE_RATE,
|
||||
) -> None:
|
||||
"""
|
||||
Generate a smooth pitch glide tone.
|
||||
|
||||
Args:
|
||||
filename: Output WAV file path
|
||||
freq_start: Starting frequency in Hz
|
||||
freq_end: Ending frequency in Hz
|
||||
duration: Duration in seconds
|
||||
amplitude: Volume (0.0 to 1.0)
|
||||
sample_rate: Audio sample rate
|
||||
"""
|
||||
"""Generate a smooth pitch glide tone (original Linux pair recipe)."""
|
||||
num_samples = int(sample_rate * duration)
|
||||
samples = []
|
||||
|
||||
for i in range(num_samples):
|
||||
t = i / sample_rate
|
||||
progress = i / num_samples
|
||||
|
||||
# Ultra-soft envelope: sine-squared with exponential decay
|
||||
envelope = (math.sin(math.pi * progress) ** 2) * math.exp(-0.5 * progress)
|
||||
|
||||
# Smooth pitch glide using smoothstep interpolation
|
||||
# This creates buttery smooth transitions with no abrupt changes
|
||||
glide = progress ** 2 * (3 - 2 * progress)
|
||||
envelope = _sine_envelope(progress)
|
||||
glide = _smoothstep(progress)
|
||||
freq_current = freq_start + (freq_end - freq_start) * glide
|
||||
|
||||
# Generate pure sine wave
|
||||
phase = 2 * math.pi * freq_current * t
|
||||
sample = amplitude * envelope * math.sin(phase)
|
||||
samples.append(sample)
|
||||
samples.append(amplitude * envelope * math.sin(phase))
|
||||
|
||||
# Convert to 16-bit PCM
|
||||
wav_samples = [int(sample * 32767) for sample in samples]
|
||||
_write_wav(filename, samples, sample_rate)
|
||||
|
||||
# Write WAV file
|
||||
|
||||
def _write_wav(filename: str, samples: list[float], sample_rate: int = SAMPLE_RATE) -> None:
|
||||
wav_samples = [int(max(-1.0, min(1.0, sample)) * 32767) for sample in samples]
|
||||
os.makedirs(os.path.dirname(filename) or ".", exist_ok=True)
|
||||
with wave.open(filename, "w") as wav_file:
|
||||
wav_file.setnchannels(1) # Mono
|
||||
wav_file.setsampwidth(2) # 16-bit
|
||||
wav_file.setnchannels(1)
|
||||
wav_file.setsampwidth(2)
|
||||
wav_file.setframerate(sample_rate)
|
||||
wav_file.writeframes(struct.pack("<" + "h" * len(wav_samples), *wav_samples))
|
||||
|
||||
print(f"Generated: {filename}")
|
||||
|
||||
|
||||
def main():
|
||||
"""Generate all Vocalinux notification sounds."""
|
||||
# Determine output directory
|
||||
def _tone_burst(
|
||||
freq: float,
|
||||
duration: float,
|
||||
amplitude: float = 0.12,
|
||||
decay: float = 8.0,
|
||||
) -> list[float]:
|
||||
n = int(SAMPLE_RATE * duration)
|
||||
samples = []
|
||||
for i in range(n):
|
||||
t = i / SAMPLE_RATE
|
||||
progress = i / max(n - 1, 1)
|
||||
attack = min(1.0, i / max(int(0.004 * SAMPLE_RATE), 1))
|
||||
env = attack * math.exp(-decay * t) * (1.0 - 0.15 * progress)
|
||||
samples.append(amplitude * env * math.sin(2 * math.pi * freq * t))
|
||||
return samples
|
||||
|
||||
|
||||
def _silence(duration: float) -> list[float]:
|
||||
return [0.0] * int(SAMPLE_RATE * duration)
|
||||
|
||||
|
||||
def generate_ticks(
|
||||
filename: str,
|
||||
freqs: tuple[float, ...],
|
||||
tick_duration: float = 0.07,
|
||||
gap: float = 0.045,
|
||||
amplitude: float = 0.11,
|
||||
decay: float = 14.0,
|
||||
) -> None:
|
||||
samples: list[float] = []
|
||||
for index, freq in enumerate(freqs):
|
||||
if index:
|
||||
samples.extend(_silence(gap))
|
||||
samples.extend(_tone_burst(freq, tick_duration, amplitude, decay))
|
||||
_write_wav(filename, samples)
|
||||
|
||||
|
||||
def generate_fifth_swell(
|
||||
filename: str,
|
||||
low: float,
|
||||
high: float,
|
||||
duration: float = 0.42,
|
||||
amplitude: float = 0.13,
|
||||
opening: bool = True,
|
||||
) -> None:
|
||||
"""Open-fifth swell. Start opens C4→G4; stop closes G4→C4."""
|
||||
n = int(SAMPLE_RATE * duration)
|
||||
samples = []
|
||||
phase_low = 0.0
|
||||
phase_high = 0.0
|
||||
dt = 1.0 / SAMPLE_RATE
|
||||
for i in range(n):
|
||||
progress = i / max(n - 1, 1)
|
||||
env = math.sin(math.pi * progress) ** 2
|
||||
blend = _smoothstep(progress)
|
||||
if opening:
|
||||
low_amt = 1.0
|
||||
high_amt = blend
|
||||
else:
|
||||
low_amt = 1.0
|
||||
high_amt = 1.0 - blend
|
||||
sample = (
|
||||
amplitude
|
||||
* env
|
||||
* (low_amt * math.sin(phase_low) + 0.85 * high_amt * math.sin(phase_high))
|
||||
)
|
||||
samples.append(sample)
|
||||
phase_low += 2 * math.pi * low * dt
|
||||
phase_high += 2 * math.pi * high * dt
|
||||
_write_wav(filename, samples)
|
||||
|
||||
|
||||
def generate_ping(
|
||||
filename: str, freq: float, duration: float = 0.32, amplitude: float = 0.12
|
||||
) -> None:
|
||||
n = int(SAMPLE_RATE * duration)
|
||||
samples = []
|
||||
for i in range(n):
|
||||
t = i / SAMPLE_RATE
|
||||
env = math.exp(-14.0 * t)
|
||||
samples.append(amplitude * env * math.sin(2 * math.pi * freq * t))
|
||||
_write_wav(filename, samples)
|
||||
|
||||
|
||||
def generate_scale(
|
||||
filename: str,
|
||||
freqs: tuple[float, ...],
|
||||
note_duration: float = 0.085,
|
||||
gap: float = 0.02,
|
||||
amplitude: float = 0.12,
|
||||
) -> None:
|
||||
generate_ticks(filename, freqs, note_duration, gap, amplitude, decay=9.0)
|
||||
|
||||
|
||||
def _maybe_write_original(path: str, writer) -> None:
|
||||
if os.path.exists(path):
|
||||
print(f"Keeping existing: {path}")
|
||||
return
|
||||
writer()
|
||||
|
||||
|
||||
def main() -> None:
|
||||
script_dir = os.path.dirname(os.path.abspath(__file__))
|
||||
repo_root = os.path.dirname(script_dir)
|
||||
sounds_dir = os.path.join(repo_root, "resources", "sounds")
|
||||
|
||||
os.makedirs(sounds_dir, exist_ok=True)
|
||||
print(f"Generating sounds in: {sounds_dir}\n")
|
||||
|
||||
# START: Ascending F4→A4 (positive, uplifting)
|
||||
print("Generating start_recording.wav (ascending F4→A4)...")
|
||||
generate_glide_tone(
|
||||
os.path.join(sounds_dir, "start_recording.wav"),
|
||||
freq_start=349.23, # F4
|
||||
freq_end=440.00, # A4
|
||||
duration=0.6,
|
||||
amplitude=0.16,
|
||||
start_path = os.path.join(sounds_dir, "start_recording.wav")
|
||||
stop_path = os.path.join(sounds_dir, "stop_recording.wav")
|
||||
error_path = os.path.join(sounds_dir, "error.wav")
|
||||
|
||||
_maybe_write_original(
|
||||
start_path,
|
||||
lambda: generate_glide_tone(start_path, F4, A4, duration=0.6, amplitude=0.16),
|
||||
)
|
||||
_maybe_write_original(
|
||||
stop_path,
|
||||
lambda: generate_glide_tone(stop_path, A4, F4, duration=0.6, amplitude=0.16),
|
||||
)
|
||||
_maybe_write_original(
|
||||
error_path,
|
||||
lambda: generate_glide_tone(error_path, E4, C4, duration=0.7, amplitude=0.14),
|
||||
)
|
||||
|
||||
# STOP: Descending A4→F4 (resolution, completion)
|
||||
print("Generating stop_recording.wav (descending A4→F4)...")
|
||||
generate_glide_tone(
|
||||
os.path.join(sounds_dir, "stop_recording.wav"),
|
||||
freq_start=440.00, # A4
|
||||
freq_end=349.23, # F4
|
||||
duration=0.6,
|
||||
amplitude=0.16,
|
||||
)
|
||||
# lift: Linux-family F4→A4 sine glide (own files; does not replace the originals)
|
||||
generate_glide_tone(os.path.join(sounds_dir, "lift_start.wav"), F4, A4, 0.48, 0.16)
|
||||
generate_glide_tone(os.path.join(sounds_dir, "lift_stop.wav"), A4, F4, 0.48, 0.16)
|
||||
|
||||
# ERROR: Lower descending E4→C4 (sad, concerning but gentle)
|
||||
print("Generating error.wav (descending E4→C4)...")
|
||||
generate_glide_tone(
|
||||
os.path.join(sounds_dir, "error.wav"),
|
||||
freq_start=329.63, # E4
|
||||
freq_end=261.63, # C4
|
||||
duration=0.7,
|
||||
amplitude=0.14, # Slightly quieter
|
||||
)
|
||||
# flick: same interval, shorter
|
||||
generate_glide_tone(os.path.join(sounds_dir, "flick_start.wav"), F4, A4, 0.18, 0.15)
|
||||
generate_glide_tone(os.path.join(sounds_dir, "flick_stop.wav"), A4, F4, 0.18, 0.15)
|
||||
|
||||
print("\n✓ All sounds generated successfully!")
|
||||
print("\nSound characteristics:")
|
||||
print("- Pure sine waves (headphone-friendly)")
|
||||
print("- Smoothstep pitch interpolation (no jumps)")
|
||||
print("- Soft envelopes (sine-squared with decay)")
|
||||
print("- Lower frequencies (F4-A4 range, warm and pleasant)")
|
||||
print("- Consistent glide aesthetic")
|
||||
# ember: low warm G3→C4
|
||||
generate_glide_tone(os.path.join(sounds_dir, "ember_start.wav"), G3, C4, 0.55, 0.15)
|
||||
generate_glide_tone(os.path.join(sounds_dir, "ember_stop.wav"), C4, G3, 0.55, 0.15)
|
||||
|
||||
# step: two soft C/E ticks
|
||||
generate_ticks(os.path.join(sounds_dir, "step_start.wav"), (C4, E4), 0.07, 0.05, 0.11)
|
||||
generate_ticks(os.path.join(sounds_dir, "step_stop.wav"), (E4, C4), 0.07, 0.05, 0.11)
|
||||
|
||||
# voca: open fifth swell (id is voca, not fifth)
|
||||
generate_fifth_swell(os.path.join(sounds_dir, "voca_start.wav"), C4, G4, 0.44, 0.14, True)
|
||||
generate_fifth_swell(os.path.join(sounds_dir, "voca_stop.wav"), C4, G4, 0.40, 0.13, False)
|
||||
|
||||
# soft: very short muted ticks
|
||||
generate_ticks(os.path.join(sounds_dir, "soft_start.wav"), (C5,), 0.05, 0.0, 0.07, decay=22.0)
|
||||
generate_ticks(os.path.join(sounds_dir, "soft_stop.wav"), (A4,), 0.055, 0.0, 0.07, decay=22.0)
|
||||
|
||||
# chirp: light high chirp
|
||||
generate_glide_tone(os.path.join(sounds_dir, "chirp_start.wav"), A5, C6, 0.09, 0.10)
|
||||
generate_glide_tone(os.path.join(sounds_dir, "chirp_stop.wav"), C6, A5, 0.10, 0.10)
|
||||
|
||||
# scale: pentatonic A4→C5
|
||||
generate_scale(os.path.join(sounds_dir, "scale_start.wav"), (A4, C5), 0.09, 0.025, 0.12)
|
||||
generate_scale(os.path.join(sounds_dir, "scale_stop.wav"), (C5, A4), 0.09, 0.025, 0.12)
|
||||
|
||||
# drop: low start, deeper stop
|
||||
generate_glide_tone(os.path.join(sounds_dir, "drop_start.wav"), D3, F3, 0.40, 0.16)
|
||||
generate_glide_tone(os.path.join(sounds_dir, "drop_stop.wav"), C3, G2, 0.48, 0.16)
|
||||
|
||||
# glass: thin glass ping
|
||||
generate_ping(os.path.join(sounds_dir, "glass_start.wav"), E7, 0.30, 0.10)
|
||||
generate_ping(os.path.join(sounds_dir, "glass_stop.wav"), C7, 0.34, 0.09)
|
||||
|
||||
# Package data lives in src/vocalinux/resources and is what ResourceManager
|
||||
# prefers when running from a checkout. Keep both trees in sync.
|
||||
packaged_dir = os.path.join(repo_root, "src", "vocalinux", "resources", "sounds")
|
||||
os.makedirs(packaged_dir, exist_ok=True)
|
||||
for name in os.listdir(sounds_dir):
|
||||
if not name.endswith(".wav"):
|
||||
continue
|
||||
if name in {"start_recording.wav", "stop_recording.wav", "error.wav"}:
|
||||
continue
|
||||
shutil.copy2(os.path.join(sounds_dir, name), os.path.join(packaged_dir, name))
|
||||
print(f"Copied: {os.path.join(packaged_dir, name)}")
|
||||
|
||||
print("\nCatalog pairs written. Original start/stop/error files were left in place.")
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
|
||||
Reference in New Issue
Block a user