Switches to using source directions computed as part of the image source model for source directivities. This enables the use of source directivities with non-shoebox rooms.

Author: fakufaku

examples/directivities/source_and_microphone.py

@@ -2,21 +2,15 @@
 import numpy as np
 
 import pyroomacoustics as pra
-from pyroomacoustics.directivities import (
-    CardioidFamily,
-    DirectionVector,
-    DirectivityPattern,
-)
+from pyroomacoustics.directivities import DirectionVector, FigureEight, HyperCardioid
 
 three_dim = True  # 2D or 3D
 shoebox = True  # source directivity not supported for non-shoebox!
 energy_absorption = 0.4
 source_pos = [2, 1.8]
 # source_dir = None   # to disable
-source_dir = DirectivityPattern.FIGURE_EIGHT
 mic_pos = [3.5, 1.8]
 # mic_dir = None   # to disable
-mic_dir = DirectivityPattern.HYPERCARDIOID
 
 
 # make 2-D room
@@ -42,19 +36,15 @@
     colatitude = None
 
 # add source with directivity
-if source_dir is not None:
-    source_dir = CardioidFamily(
-        orientation=DirectionVector(azimuth=90, colatitude=colatitude, degrees=True),
-        pattern_enum=source_dir,
-    )
+source_dir = FigureEight(
+    orientation=DirectionVector(azimuth=90, colatitude=colatitude, degrees=True),
+)
 room.add_source(position=source_pos, directivity=source_dir)
 
 # add microphone with directivity
-if mic_dir is not None:
-    mic_dir = CardioidFamily(
-        orientation=DirectionVector(azimuth=0, colatitude=colatitude, degrees=True),
-        pattern_enum=mic_dir,
-    )
+mic_dir = HyperCardioid(
+    orientation=DirectionVector(azimuth=0, colatitude=colatitude, degrees=True),
+)
 room.add_microphone(loc=mic_pos, directivity=mic_dir)
 
 # plot room

pyroomacoustics/directivities/tests/test_source_directivities_nonshoebox.py

@@ -0,0 +1,73 @@
+import numpy as np
+import pytest
+
+import pyroomacoustics as pra
+
+"""
++ - - - - - - - - - - - +
+|                       |
+|                       |
+|                       |
+|       + - - - - - - - +
+|       |
+|       |
+|       |
++ - - - +
+"""
+
+corners = np.array([[0, 0], [4, 0], [4, 4], [12, 4], [12, 8], [0, 8]])
+source_location = [6, 6, 1.5]
+mic_location = [1, 6, 1.5]
+
+images_expected = np.array(
+    [
+        source_location,  # direct
+        [6, 10, 1.5],  # north
+        [-6, 6, 1.5],  # west
+        [18, 6, 1.5],  # east
+        [6, 6, 4.5],  # ceilling
+        [6, 6, -1.5],  # floor
+    ]
+)
+directions_expected = np.array(
+    [
+        [-1, 0, 0],
+        np.array([-2.5, 2, 0]) / np.sqrt(2.5**2 + 2**2),
+        [-1, 0, 0],
+        [1, 0, 0],
+        np.array([-2.5, 0, 1.5]) / np.sqrt(2.5**2 + 1.5**2),
+        np.array([-2.5, 0, -1.5]) / np.sqrt(2.5**2 + 1.5**2),
+    ]
+)
+
+
+room = pra.Room.from_corners(corners.T, materials=pra.Material(0.1), max_order=1)
+room.extrude(3.0, materials=pra.Material(0.1))
+room.add_source(
+    source_location,
+    directivity=pra.directivities.Cardioid(
+        orientation=pra.directivities.DirectionVector(
+            azimuth=0, colatitude=np.pi / 2, degrees=False
+        )
+    ),
+).add_microphone(mic_location).add_microphone([3, 6, 1.5])
+# We add one microphone with other images visible to make sure we have some
+# image sources not visible.
+
+
+def test_source_directions_nonshoebox():
+    room.image_source_model()
+    visible = room.visibility[0][0, :]
+    directions_obtained = room.sources[0].directions[0, :, visible]
+    images_obtained = room.sources[0].images[:, visible]
+
+    order_obtained = np.lexsort(images_obtained)
+    images_obtained = images_obtained[:, order_obtained].T
+    directions_obtained = directions_obtained[order_obtained, :]
+
+    order_expected = np.lexsort(images_expected.T)
+    images_expected_reordered = images_expected[order_expected, :]
+    directions_expected_reordered = directions_expected[order_expected, :]
+
+    np.testing.assert_almost_equal(images_obtained, images_expected_reordered)
+    np.testing.assert_almost_equal(directions_obtained, directions_expected_reordered)

pyroomacoustics/tests/test_source_directivity_flipping.py renamed to pyroomacoustics/directivities/tests/test_source_directivity_flipping.py

@@ -3,7 +3,57 @@
 import numpy as np
 
 import pyroomacoustics as pra
-from pyroomacoustics.simulation.ism import source_angle_shoebox
+
+
+def source_angle_shoebox(image_source_loc, wall_flips, mic_loc):
+    """
+    Determine outgoing angle for each image source for a ShoeBox configuration.
+
+    Implementation of the method described in the paper:
+    https://www2.ak.tu-berlin.de/~akgroup/ak_pub/2018/000458.pdf
+
+    Parameters
+    -----------
+    image_source_loc : array_like
+        Locations of image sources.
+    wall_flips: array_like
+        Number of x, y, z flips for each image source.
+    mic_loc: array_like
+        Microphone location.
+
+    Returns
+    -------
+    azimuth : :py:class:`~numpy.ndarray`
+        Azimith for each image source, in radians
+    colatitude : :py:class:`~numpy.ndarray`
+        Colatitude for each image source, in radians.
+
+    """
+
+    image_source_loc = np.array(image_source_loc)
+    wall_flips = np.array(wall_flips)
+    mic_loc = np.array(mic_loc)
+
+    dim, n_sources = image_source_loc.shape
+    assert wall_flips.shape[0] == dim
+    assert mic_loc.shape[0] == dim
+
+    p_vector_array = image_source_loc - np.array(mic_loc)[:, np.newaxis]
+    d_array = np.linalg.norm(p_vector_array, axis=0)
+
+    # Using (12) from the paper
+    power_array = np.ones_like(image_source_loc) * -1
+    power_array = np.power(power_array, (wall_flips + np.ones_like(image_source_loc)))
+    p_dash_array = p_vector_array * power_array
+
+    # Using (13) from the paper
+    azimuth = np.arctan2(p_dash_array[1], p_dash_array[0])
+    if dim == 2:
+        colatitude = np.ones(n_sources) * np.pi / 2
+    else:
+        colatitude = np.pi / 2 - np.arcsin(p_dash_array[2] / d_array)
+
+    return azimuth, colatitude
 
 
 class TestSourceDirectivityFlipping(TestCase):

pyroomacoustics/room.py

@@ -1409,8 +1409,8 @@ def extrude(self, height, v_vec=None, absorption=None, materials=None):
             if len(self.walls[i].absorption) == 1:
                 # Single band
                 wall_materials[name] = Material(
-                    energy_absorption=float(self.walls[i].absorption),
-                    scattering=float(self.walls[i].scatter),
+                    energy_absorption=float(self.walls[i].absorption[0]),
+                    scattering=float(self.walls[i].scatter[0]),
                 )
             elif len(self.walls[i].absorption) == self.octave_bands.n_bands:
                 # Multi-band
@@ -2139,14 +2139,6 @@ def add_source(self, position, signal=None, delay=0, directivity=None):
         if self.dim != 3 and directivity is not None:
             raise NotImplementedError("Directivity is only supported for 3D rooms.")
 
-        if directivity is not None:
-            from pyroomacoustics import ShoeBox
-
-            if not isinstance(self, ShoeBox):
-                raise NotImplementedError(
-                    "Source directivity only supported for ShoeBox room."
-                )
-
         if isinstance(position, SoundSource):
             if directivity is not None:
                 if isinstance(directivity, CardioidFamily) or isinstance(
@@ -2241,8 +2233,11 @@ def image_source_model(self):
                         self.visibility[-1][m, :] = 0
             else:
                 # if we are here, this means even the direct path is not visible
-                # we set the visibility of the direct path as 0
+                # we set the visibility of the direct path as 0.
                 self.visibility.append(np.zeros((self.mic_array.M, 1), dtype=np.int32))
+                # We also need a fake array of directions as this is expected later in
+                # the code.
+                source.directions = np.zeros((self.mic_array.M, self.dim, 1), dtype=np.float32)
 
         # Update the state
         self.simulator_state["ism_done"] = True
@@ -2305,13 +2300,14 @@ def compute_rir(self):
                         src,
                         mic,
                         self.mic_array.directivity[m],
+                        src.directions[m, :, :],
                         self.visibility[s][m, :],
                         fdl,
                         self.c,
                         self.fs,
                         self.octave_bands,
-                        air_abs_coeffs=self.air_absorption,
                         min_phase=self.min_phase,
+                        air_abs_coeffs=self.air_absorption,
                     )
                     rir_parts.append(ir_ism)
 

pyroomacoustics/simulation/ism.py

@@ -31,7 +31,7 @@
 import numpy as np
 from scipy.signal import fftconvolve, hilbert
 
-from .. import libroom
+from .. import doa, libroom
 from ..parameters import constants
 from ..utilities import angle_function
 
@@ -109,61 +109,11 @@ def interpolate_octave_bands(
     return ir
 
 
-def source_angle_shoebox(image_source_loc, wall_flips, mic_loc):
-    """
-    Determine outgoing angle for each image source for a ShoeBox configuration.
-
-    Implementation of the method described in the paper:
-    https://www2.ak.tu-berlin.de/~akgroup/ak_pub/2018/000458.pdf
-
-    Parameters
-    -----------
-    image_source_loc : array_like
-        Locations of image sources.
-    wall_flips: array_like
-        Number of x, y, z flips for each image source.
-    mic_loc: array_like
-        Microphone location.
-
-    Returns
-    -------
-    azimuth : :py:class:`~numpy.ndarray`
-        Azimith for each image source, in radians
-    colatitude : :py:class:`~numpy.ndarray`
-        Colatitude for each image source, in radians.
-
-    """
-
-    image_source_loc = np.array(image_source_loc)
-    wall_flips = np.array(wall_flips)
-    mic_loc = np.array(mic_loc)
-
-    dim, n_sources = image_source_loc.shape
-    assert wall_flips.shape[0] == dim
-    assert mic_loc.shape[0] == dim
-
-    p_vector_array = image_source_loc - np.array(mic_loc)[:, np.newaxis]
-    d_array = np.linalg.norm(p_vector_array, axis=0)
-
-    # Using (12) from the paper
-    power_array = np.ones_like(image_source_loc) * -1
-    power_array = np.power(power_array, (wall_flips + np.ones_like(image_source_loc)))
-    p_dash_array = p_vector_array * power_array
-
-    # Using (13) from the paper
-    azimuth = np.arctan2(p_dash_array[1], p_dash_array[0])
-    if dim == 2:
-        colatitude = np.ones(n_sources) * np.pi / 2
-    else:
-        colatitude = np.pi / 2 - np.arcsin(p_dash_array[2] / d_array)
-
-    return azimuth, colatitude
-
-
 def compute_ism_rir(
     src,
     mic,
     mic_dir,
+    src_directions,
     is_visible,
     fdl,
     c,
@@ -226,11 +176,7 @@ def compute_ism_rir(
             oct_band_amplitude *= mic_gain
 
     if src.directivity is not None:
-        azimuth_s, colatitude_s = source_angle_shoebox(
-            image_source_loc=images,
-            wall_flips=abs(src.orders_xyz[:, is_visible]),
-            mic_loc=mic,
-        )
+        azimuth_s, colatitude_s, _ = doa.cart2spher(src_directions[:, is_visible])
         src_gain = src.directivity.get_response(
             azimuth=azimuth_s,
             colatitude=colatitude_s,