diffsptk

diffsptk is a differentiable version of SPTK based on the PyTorch framework.

Requirements

• Python 3.8+
• PyTorch 1.11.0+

Documentation

• See this page for a reference manual.
• Our paper is available on the ISCA Archive.

Installation

The latest stable release can be installed through PyPI by running

pip install diffsptk

The development release can be installed from the master branch:

pip install git+https://github.com/sp-nitech/diffsptk.git@master

Examples

Mel-cepstral analysis and synthesis

import diffsptk

fl = 400     # Frame length.
fp = 80      # Frame period.
n_fft = 512  # FFT length.
M = 24       # Mel-cepstrum dimensions.

# Read waveform.
x, sr = diffsptk.read("assets/data.wav")

# Compute STFT amplitude of x.
stft = diffsptk.STFT(frame_length=fl, frame_period=fp, fft_length=n_fft)
X = stft(x)

# Estimate mel-cepstrum of x.
alpha = diffsptk.get_alpha(sr)
mcep = diffsptk.MelCepstralAnalysis(cep_order=M, fft_length=n_fft, alpha=alpha, n_iter=10)
mc = mcep(X)

# Reconstruct x.
mlsa = diffsptk.MLSA(filter_order=M, frame_period=fp, alpha=alpha, taylor_order=30)
x_hat = mlsa(mlsa(x, -mc), mc)

# Write reconstructed waveform.
diffsptk.write("reconst.wav", x_hat, sr)

# Compute error.
error = (x_hat - x).abs().sum()
print(error)

# Extract pitch of x.
pitch = diffsptk.Pitch(frame_period=fp, sample_rate=sr, f_min=80, f_max=180)
p = pitch(x)

# Generate excitation signal.
excite = diffsptk.ExcitationGeneration(frame_period=fp)
e = excite(p)
n = diffsptk.nrand(x.size(0) - 1)

# Synthesize waveform.
x_voiced = mlsa(e, mc)
x_unvoiced = mlsa(n, mc)

# Output analysis-synthesis result.
diffsptk.write("voiced.wav", x_voiced, sr)
diffsptk.write("unvoiced.wav", x_unvoiced, sr)

Mel-spectrogram, MFCC, and PLP extraction

import diffsptk

fl = 400        # Frame length
fp = 80         # Frame period
n_fft = 512     # FFT length
n_channel = 80  # Number of channels
M = 12          # MFCC/PLP dimensions

# Read waveform.
x, sr = diffsptk.read("assets/data.wav")

# Compute STFT amplitude of x.
stft = diffsptk.STFT(frame_length=fl, frame_period=fp, fft_length=n_fft)
X = stft(x)

# Extract mel-spectrogram.
fbank = diffsptk.MelFilterBankAnalysis(
    n_channel=n_channel,
    fft_length=n_fft,
    sample_rate=sr,
)
Y = fbank(X)
print(Y.shape)

# Extract MFCC.
mfcc = diffsptk.MFCC(
    mfcc_order=M,
    n_channel=n_channel,
    fft_length=n_fft,
    sample_rate=sr,
)
Y = mfcc(X)
print(Y.shape)

# Extract PLP.
plp = diffsptk.PLP(
    plp_order=M,
    n_channel=n_channel,
    fft_length=n_fft,
    sample_rate=sr,
)
Y = plp(X)
print(Y.shape)

Subband decomposition

import diffsptk

K = 4   # Number of subbands.
M = 40  # Order of filter.

# Read waveform.
x, sr = diffsptk.read("assets/data.wav")

# Decompose x.
pqmf = diffsptk.PQMF(K, M)
decimate = diffsptk.Decimation(K)
y = decimate(pqmf(x), dim=-1)

# Reconstruct x.
interpolate = diffsptk.Interpolation(K)
ipqmf = diffsptk.IPQMF(K, M)
x_hat = ipqmf(interpolate(K * y, dim=-1)).reshape(-1)

# Write reconstructed waveform.
diffsptk.write("reconst.wav", x_hat, sr)

# Compute error.
error = (x_hat - x).abs().sum()
print(error)

Constant-Q transform

import diffsptk
import librosa  # This is to get sample audio.

fp = 128  # Frame period.
K = 252   # Number of CQ-bins.
B = 36    # Number of bins per octave.

# Read waveform.
x, sr = diffsptk.read(librosa.ex("trumpet"))

# Transform x.
cqt = diffsptk.CQT(fp, sr, n_bin=K, n_bin_per_octave=B)
c = cqt(x)

# Reconstruct x.
icqt = diffsptk.ICQT(fp, sr, n_bin=K, n_bin_per_octave=B)
x_hat = icqt(c, out_length=x.size(0))

# Write reconstructed waveform.
diffsptk.write("reconst.wav", x_hat, sr)

# Compute error.
error = (x_hat - x).abs().sum()
print(error)

Vector quantization

import diffsptk

K = 2  # Codebook size.
M = 4  # Order of vector.

# Prepare input.
x = diffsptk.nrand(M)

# Quantize x.
vq = diffsptk.VectorQuantization(M, K)
x_hat, indices, commitment_loss = vq(x)

# Compute error.
error = (x_hat - x).abs().sum()
print(error)

License

This software is released under the Apache License 2.0.

Citation

@InProceedings{sp-nitech2023sptk,
  author = {Takenori Yoshimura and Takato Fujimoto and Keiichiro Oura and Keiichi Tokuda},
  title = {{SPTK4}: An open-source software toolkit for speech signal processing},
  booktitle = {12th ISCA Speech Synthesis Workshop (SSW 2023)},
  pages = {211--217},
  year = {2023},
}