#! /usr/bin/env python from numpy.testing import TestCase, assert_equal, assert_array_less from aubio import fvec, cvec, pvoc, float_type from numpy import array, shape from numpy.random import random from nose2.tools import params import numpy as np if float_type == 'float32': max_sq_error = 1.e-12 else: max_sq_error = 1.e-29 def create_sine(hop_s, freq, samplerate): t = np.arange(hop_s).astype(float_type) return np.sin( 2. * np.pi * freq * t / float(samplerate)) def create_noise(hop_s): return np.random.rand(hop_s).astype(float_type) * 2. - 1. class aubio_pvoc_test_case(TestCase): """ pvoc object test case """ def test_members_automatic_sizes_default(self): """ check object creation with default parameters """ f = pvoc() assert_equal ([f.win_s, f.hop_s], [1024, 512]) def test_members_unnamed_params(self): """ check object creation with unnamed parameters """ f = pvoc(2048, 128) assert_equal ([f.win_s, f.hop_s], [2048, 128]) def test_members_named_params(self): """ check object creation with named parameters """ f = pvoc(hop_s = 128, win_s = 2048) assert_equal ([f.win_s, f.hop_s], [2048, 128]) def test_zeros(self): """ check the resynthesis of zeros gives zeros """ win_s, hop_s = 1024, 256 f = pvoc (win_s, hop_s) t = fvec (hop_s) for time in range( int ( 4 * win_s / hop_s ) ): s = f(t) r = f.rdo(s) assert_equal ( t, 0.) assert_equal ( s.norm, 0.) assert_equal ( s.phas, 0.) assert_equal ( r, 0.) def test_resynth_8_steps_sine(self): """ check the resynthesis of is correct with 87.5% overlap """ hop_s = 1024 ratio = 8 freq = 445; samplerate = 22050 sigin = create_sine(hop_s, freq, samplerate) self.reconstruction( sigin, hop_s, ratio) def test_resynth_8_steps(self): """ check the resynthesis of is correct with 87.5% overlap """ hop_s = 1024 ratio = 8 sigin = create_noise(hop_s) self.reconstruction(sigin, hop_s, ratio) def test_resynth_4_steps_sine(self): """ check the resynthesis of is correct with 87.5% overlap """ hop_s = 1024 ratio = 4 freq = 445; samplerate = 22050 sigin = create_sine(hop_s, freq, samplerate) self.reconstruction(sigin, hop_s, ratio) def test_resynth_4_steps(self): """ check the resynthesis of is correct with 75% overlap """ hop_s = 1024 ratio = 4 sigin = create_noise(hop_s) self.reconstruction(sigin, hop_s, ratio) def test_resynth_2_steps_sine(self): """ check the resynthesis of is correct with 50% overlap """ hop_s = 1024 ratio = 2 freq = 445; samplerate = 22050 sigin = create_sine(hop_s, freq, samplerate) self.reconstruction(sigin, hop_s, ratio) def test_resynth_2_steps(self): """ check the resynthesis of is correct with 50% overlap """ hop_s = 1024 ratio = 2 sigin = create_noise(hop_s) self.reconstruction(sigin, hop_s, ratio) @params( ( 256, 8), ( 256, 4), ( 256, 2), ( 512, 8), ( 512, 4), ( 512, 2), (1024, 8), (1024, 4), (1024, 2), (2048, 8), (2048, 4), (2048, 2), (4096, 8), (4096, 4), (4096, 2), (8192, 8), (8192, 4), (8192, 2), ) def test_resynth_steps_noise(self, hop_s, ratio): """ check the resynthesis of a random signal is correct """ sigin = create_noise(hop_s) self.reconstruction(sigin, hop_s, ratio) @params( (44100, 256, 8, 441), (44100, 256, 4, 1203), (44100, 256, 2, 3045), (44100, 512, 8, 445), (44100, 512, 4, 445), (44100, 512, 2, 445), (44100, 1024, 8, 445), (44100, 1024, 4, 445), (44100, 1024, 2, 445), ( 8000, 1024, 2, 445), (22050, 1024, 2, 445), (22050, 256, 8, 445), (96000, 1024, 8, 47000), (96000, 1024, 8, 20), ) def test_resynth_steps_sine(self, samplerate, hop_s, ratio, freq): """ check the resynthesis of a sine is correct """ sigin = create_sine(hop_s, freq, samplerate) self.reconstruction(sigin, hop_s, ratio) def reconstruction(self, sigin, hop_s, ratio): buf_s = hop_s * ratio f = pvoc(buf_s, hop_s) zeros = fvec(hop_s) r2 = f.rdo( f(sigin) ) for i in range(1, ratio): r2 = f.rdo( f(zeros) ) # compute square errors sq_error = (r2 - sigin)**2 # make sure all square errors are less than desired precision assert_array_less(sq_error, max_sq_error) if __name__ == '__main__': from nose2 import main main()