daga2026_dode/dodecahedron_sphere.py at main · spatialaudio/daga2026_dode · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
# see https://github.com/spatialaudio/daga2026_dode

# paper figure 4(a)

import matplotlib.colors as colors
import matplotlib.pyplot as plt
import numpy as np

from matplotlib import cm
from pyfar.classes.coordinates import cart2sph
from util_cap import (figure_path,
                      pressure_impulse_response_with_filterbank,
                      surface_velocity,
                      vnm_dodecahedron,
                      )


plt.rcParams.update({'font.size': 10})
figw, figh, dpi = (37/2)/2.54 / 2, (37/2)/2.54 / 2 * 20/16, 600
# 37 cm is the column width for Uni Rostock DIN A0 poster
# we go for 37/2 cm with 600 dpi and font size 10
# and use \includegraphics[width=37cm] in tex
# this handling looks nicer than png renderings
# with original size 37cm and font sizes >= 16

fig, axs = plt.subplots(nrows=2, ncols=1,
                        figsize=(figw, figh),
                        subplot_kw=dict(projection='3d'),
                        constrained_layout=True)

# sphere sampling such that plot_surface has
# nice triangulation patches
M = 2**8
u = np.linspace(0, 2*np.pi, 2*M)
v = np.linspace(0, np.pi, M)
# get unit sphere with polar/azimuth convention:
x = np.outer(np.cos(u), np.sin(v))
y = np.outer(np.sin(u), np.sin(v))
z = np.outer(np.ones(np.size(u)), np.cos(v))
phi, theta, _ = cart2sph(x, y, z)
Theta = np.array([np.reshape(theta, (-1)), np.reshape(phi, (-1))]).T
print(Theta)

# dode parameters
L = 2**12  # time samples
N = 24  # SHT max order
fs = 32000  # Hz
rho0 = 1.2041  # kg/m^3
c = 343  # m/s
R = 0.4 / 2  # m
ac_dly_smp = 200
r = ac_dly_smp / fs * c + R  # m
dBmax = 0  # we normalise max value to 0 dB
dBspan = 36
dBmin = dBmax - dBspan
# Hz -> for N=24, the spatial aliasing is properly modelled
# for this third-octave:
fm = 4000
bw_oct = 1/3

# gl = np.ones(12)
dither_dB_span = 2
np.random.seed(3)  # for reproducible/consistent output
gl = 10**(np.random.uniform(-dither_dB_span,
                            +dither_dB_span,
                            12)/20)
beta_kaiser = 5
alpha_cap = 20 * np.pi/180
Lfir = 21
n0 = 10

# velocity on surface
vnm = vnm_dodecahedron(N, beta_kaiser, gl, alpha_cap)
velocity = surface_velocity(vnm, Theta)
velocity = np.reshape(velocity, theta.shape)
velocity_dB = 20*np.log10(np.abs(np.copy(velocity)))
velocity_dB -= np.max(velocity_dB)
velocity_dB += dBmax
velocity_dB[velocity_dB < dBmin] = dBmin
C_velocity_dB = np.copy(velocity_dB)

# pressure on surface
f = np.logspace(np.log10(2), np.log10(fs//2), num=2**10)  # Hz
pressure_ir = pressure_impulse_response_with_filterbank(
        vnm, f, Theta, r, R, L,
        c=c, rho0=rho0, fs=fs,
        farfield_flag=False, flag_pfe_only=False,
        Lfir=Lfir, n0=n0, num_fc=27)
P = np.fft.fft(pressure_ir, axis=1)
df = fs / P.shape[1]
f_fft = np.arange(P.shape[1]) * df
idxl = np.where(f_fft <= fm*2**(-bw_oct/2))[0][-1]
idxh = np.where(f_fft > fm*2**(+bw_oct/2))[0][0]
print(f_fft[idxl], f_fft[idxh])
P_dB = 10*np.log10(np.sum(np.abs(P[:, idxl:idxh])**2, axis=1))
P_dB = np.reshape(P_dB, theta.shape)
P_dB -= np.max(P_dB)
P_dB += dBmax
P_dB[P_dB < dBmin] = dBmin
C_P_dB = np.copy(P_dB)

# velocity sphere plot
min, max = C_velocity_dB.min(), C_velocity_dB.max()
bounds = np.arange(dBmin, dBmax+3, 3)
norm = colors.BoundaryNorm(boundaries=bounds, ncolors=256)
color_map = cm.viridis
scalarMap = cm.ScalarMappable(norm=norm, cmap=color_map)
C_colored = scalarMap.to_rgba(C_velocity_dB)
surf = axs[0].plot_surface(
    x, y, z, rstride=1, cstride=1,
    facecolors=C_colored,
    cmap=color_map,
    antialiased=True)
m = cm.ScalarMappable(
    cmap=surf.cmap, norm=norm)
m.set_array(C_velocity_dB)
clb = fig.colorbar(m, location='left', pad=0.1, shrink=0.8, ax=axs[0])
clb.set_label(r'velocity $v(\mathbf{\theta})$, dB rel max',
              loc='center', labelpad=0)

# pressure sphere plot
min, max = C_P_dB.min(), C_P_dB.max()
bounds = np.arange(dBmin, dBmax+3, 3)
norm = colors.BoundaryNorm(boundaries=bounds, ncolors=256)
color_map = cm.magma
scalarMap = cm.ScalarMappable(norm=norm, cmap=color_map)
C_colored = scalarMap.to_rgba(C_P_dB)
surf = axs[1].plot_surface(
    x, y, z, rstride=1, cstride=1,
    facecolors=C_colored,
    cmap=color_map,
    antialiased=True)
m = cm.ScalarMappable(
    cmap=surf.cmap, norm=norm)
m.set_array(C_P_dB)
clb = fig.colorbar(m, location='left', pad=0.1, shrink=0.8, ax=axs[1])
clb.set_label(r'pressure $p(\mathbf{\theta})$, dB rel max',
              loc='center', labelpad=0)

for i in range(2):
    axs[i].set_xticks([0], labels=[''])
    axs[i].set_yticks([0], labels=[''])
    axs[i].set_zticks([0], labels=[''])
    axs[i].set_xlim([-1, 1])
    axs[i].set_ylim([-1, 1])
    axs[i].set_zlim([-1, 1])
    axs[i].set_aspect('equal')
    axs[i].text(1.1, 0, -1, 'x')
    axs[i].text(0, 1, 1, 'z')

plt.savefig(figure_path()+'dodecahedron_sphere.png', dpi=dpi)