def convert_loop(self, in_feature: AcousticFeature, n_len: int = 512, n_wrap: int = 128):
out_feature_list: List[AcousticFeature] = []
N = len(in_feature.f0)
for i in numpy.arange(0, int(numpy.ceil(N / n_len))):
# convert with overwrapped
start = i * n_len
mi = max(start - n_wrap, 0)
ma = min(start + n_len + n_wrap, N)
f = in_feature.indexing(numpy.arange(mi, ma))
o_warp = self.convert(f)
# eliminate overwrap
ex_mi = start - mi
ex_len = min(ma - start, n_len)
o = o_warp.indexing(numpy.arange(ex_mi, ex_mi + ex_len))
out_feature_list.append(o)
return AcousticFeature.concatenate(out_feature_list)
def post_convert(self, start_time: float, time_length: float):
sizes = AcousticFeature.get_sizes(sampling_rate=self.sampling_rate, order=self.order)
keys = ['f0', 'ap', 'sp', 'voiced']
out_feature = self.fetch(
start_time=start_time,
time_length=time_length,
data_stream=self._out_feature_stream,
rate=1000 / self.frame_period,
pad_function=lambda length: AcousticFeature.silent(length, sizes=sizes, keys=keys),
pick_function=lambda segment, first, last: segment.feature.pick(first, last, keys=keys),
concat_function=lambda buffers: AcousticFeature.concatenate(buffers, keys=keys),
)
out_wave = self.vocoder.decode(
acoustic_feature=out_feature,
)
w = out_wave.wave
w[numpy.isnan(w)] = 0
out_wave = Wave(wave=w, sampling_rate=out_wave.sampling_rate)
return out_wave
def concat(self, datas: Iterable[AcousticFeatureWrapper]):
return AcousticFeature.concatenate(list(datas), keys=self._keys)