aspen.stimuli.modulation_filtered_speech¶
Modulation-filtered speech
Classes
|
Generate modulation-filtered speech. |
-
class
aspen.stimuli.modulation_filtered_speech.ModulationFilteredSpeech(samp_freq=16000, temporal_stopbands='100_200', spectral_stopbands='100_200', spec_samp_freq=1000, gauss_window_alpha=3, spacing_freq=50, spec_standardize=False, spec_db_range=- 1, griffinlim_iter=20)[source]¶ Bases:
aspen.interfaces.abs_common_interface.AbsCommonInterface,aspen.interfaces.abs_stimulus_interface.AbsStimulusInterfaceGenerate modulation-filtered speech.
- Parameters
temporal_stopbands¶ (
str) – Lower & upper temporal modulation frequency of bandstop. Defaults to “100_200”.spectral_stopbands¶ (
str) – Lower & upper spectral modulation frequency of bandstop. Defaults to “100_200”.spec_samp_freq¶ (
int) – Sampling frequency in spectrogram space in Hz. The value determines the width of Gaussian window shift to calculate the spectrogram. Defaults to 1000.gauss_window_alpha¶ (
float) – Width factor of Gaussian window. The value is inversely proportional to the width of the window. (ref: https://www.mathworks.com/help/signal/ref/gausswin.html) Defaults to 3.spacing_freq¶ (
int) – The time-frequency scale for the spectrogram in Hz. This value determines the width of the Gaussian window to calculate spectrogram. Defaults to 50.spec_standardize¶ (
bool) – Spectrogram normalize. Defaults to False.spec_db_range¶ (
float) – The range of narrowing down the spectrogram amplitude. Defaults to -1 (without narrowing).griffinlim_iter¶ (
int) – Number of iteration for Griffin-Lim algorithm Defaults to 20.