| import glob, os, random |
| from concurrent.futures import ProcessPoolExecutor |
| import numpy as np |
| import soundfile as sf |
| SR = 16000 |
| NOISES = sorted(glob.glob("/scratch/noise/*.flac")) |
| os.makedirs("/scratch/rehearsal_deg/audio", exist_ok=True) |
|
|
| def degrade(src): |
| rng = random.Random(hash(os.path.basename(src)) & 0xffffffff) |
| dst = "/scratch/rehearsal_deg/audio/" + os.path.basename(src) |
| if os.path.exists(dst): |
| return |
| wav, _ = sf.read(src, dtype="float32") |
| if rng.random() < 0.7 and NOISES: |
| n, _ = sf.read(NOISES[rng.randrange(len(NOISES))], dtype="float32") |
| if len(n) < len(wav): |
| n = np.tile(n, len(wav) // len(n) + 1) |
| s = rng.randrange(max(len(n) - len(wav), 1)) |
| n = n[s:s + len(wav)] |
| snr = rng.uniform(3.0, 15.0) |
| pw, pn = float((wav ** 2).mean()) + 1e-9, float((n ** 2).mean()) + 1e-9 |
| wav = wav + n * np.sqrt(pw / (pn * 10 ** (snr / 10))) |
| if rng.random() < 0.5: |
| f = rng.uniform(0.9, 1.1) |
| idx = np.arange(0, len(wav) - 1, f) |
| wav = np.interp(idx, np.arange(len(wav)), wav).astype(np.float32) |
| if rng.random() < 0.5: |
| idx8 = np.arange(0, len(wav) - 1, 2.0) |
| down = np.interp(idx8, np.arange(len(wav)), wav) |
| idx16 = np.arange(0, len(down) - 1, 0.5) |
| wav = np.interp(idx16, np.arange(len(down)), down).astype(np.float32) |
| wav = np.clip(wav * (10 ** (rng.uniform(-6, 3) / 20)), -1, 1).astype(np.float32) |
| sf.write(dst, wav, SR) |
|
|
| files = sorted(glob.glob("/scratch/rehearsal/audio/*.flac")) |
| with ProcessPoolExecutor(max_workers=8) as ex: |
| list(ex.map(degrade, files, chunksize=64)) |
| print("DEGRADE_DONE", len(files)) |
|
|
