From bc2cda3dd8b51a2c3cbc8ec043eabd634d88f4e3 Mon Sep 17 00:00:00 2001
From: bshall <benji.l.shall@gmail.com>
Date: Sun, 12 Jun 2022 18:46:08 +0200
Subject: [PATCH 01/14] initial commit and tests

---
 .../loudness_compliance_cpu_test.py           |  8 +++
 .../loudness_compliance_cuda_test.py          |  9 +++
 .../loudness_compliance_test_impl.py          | 60 ++++++++++++++++++
 .../torchscript_consistency_impl.py           |  8 +++
 torchaudio/functional/__init__.py             |  6 ++
 torchaudio/functional/functional.py           | 63 +++++++++++++++++++
 6 files changed, 154 insertions(+)
 create mode 100644 test/torchaudio_unittest/functional/loudness_compliance_cpu_test.py
 create mode 100644 test/torchaudio_unittest/functional/loudness_compliance_cuda_test.py
 create mode 100644 test/torchaudio_unittest/functional/loudness_compliance_test_impl.py

diff --git a/test/torchaudio_unittest/functional/loudness_compliance_cpu_test.py b/test/torchaudio_unittest/functional/loudness_compliance_cpu_test.py
new file mode 100644
index 0000000000..f8dbaeef28
--- /dev/null
+++ b/test/torchaudio_unittest/functional/loudness_compliance_cpu_test.py
@@ -0,0 +1,8 @@
+import torch
+from torchaudio_unittest import common_utils
+
+from .loudness_compliance_test_impl import Loudness
+
+
+class TestLoudnessCPU(Loudness, common_utils.PytorchTestCase):
+    device = torch.device("cpu")
diff --git a/test/torchaudio_unittest/functional/loudness_compliance_cuda_test.py b/test/torchaudio_unittest/functional/loudness_compliance_cuda_test.py
new file mode 100644
index 0000000000..cd310e215f
--- /dev/null
+++ b/test/torchaudio_unittest/functional/loudness_compliance_cuda_test.py
@@ -0,0 +1,9 @@
+import torch
+from torchaudio_unittest import common_utils
+
+from .loudness_compliance_test_impl import Loudness
+
+
+@common_utils.skipIfNoCuda
+class TestLoudnessCUDA(Loudness, common_utils.PytorchTestCase):
+    device = torch.device("cuda")
diff --git a/test/torchaudio_unittest/functional/loudness_compliance_test_impl.py b/test/torchaudio_unittest/functional/loudness_compliance_test_impl.py
new file mode 100644
index 0000000000..4adddb32f2
--- /dev/null
+++ b/test/torchaudio_unittest/functional/loudness_compliance_test_impl.py
@@ -0,0 +1,60 @@
+"""Test suite for compliance with the ITU-R BS.1770-4 recommendation"""
+import os.path
+import zipfile
+
+import torch
+import torchaudio.functional as F
+from torchaudio_unittest.common_utils import load_wav, TempDirMixin, TestBaseMixin
+
+# Test files linked in https://www.itu.int/dms_pub/itu-r/opb/rep/R-REP-BS.2217-2-2016-PDF-E.pdf
+_COMPLIANCE_FILE_URLS = {
+    "1770-2_Comp_RelGateTest": "http://www.itu.int/dms_pub/itu-r/oth/11/02/R11020000010030ZIPM.zip",
+    "1770-2_Comp_AbsGateTest": "http://www.itu.int/dms_pub/itu-r/oth/11/02/R11020000010029ZIPM.zip",
+    "1770-2_Comp_24LKFS_500Hz_2ch": "http://www.itu.int/dms_pub/itu-r/oth/11/02/R11020000010018ZIPM.zip",
+    "1770-2 Conf Mono Voice+Music-24LKFS": "http://www.itu.int/dms_pub/itu-r/oth/11/02/R11020000010038ZIPM.zip",
+}
+
+
+class Loudness(TempDirMixin, TestBaseMixin):
+    def download_and_extract_file(self, filename):
+        zippath = self.get_temp_path(filename + ".zip")
+        torch.hub.download_url_to_file(_COMPLIANCE_FILE_URLS[filename], zippath, progress=False)
+        with zipfile.ZipFile(zippath) as file:
+            file.extractall(os.path.dirname(zippath))
+        return self.get_temp_path(filename + ".wav")
+
+    def test_measure_loudness_relative_gate(self):
+        filepath = self.download_and_extract_file("1770-2_Comp_RelGateTest")
+        waveform, sample_rate = load_wav(filepath)
+        waveform = waveform.to(self.device)
+
+        loudness = F.measure_loudness(waveform, sample_rate)
+        expected = torch.tensor(-10.0, dtype=loudness.dtype, device=self.device)
+        self.assertEqual(loudness, expected, rtol=0.01, atol=0.1)
+
+    def test_measure_loudness_absolute_gate(self):
+        filepath = self.download_and_extract_file("1770-2_Comp_AbsGateTest")
+        waveform, sample_rate = load_wav(filepath)
+        waveform = waveform.to(self.device)
+
+        loudness = F.measure_loudness(waveform, sample_rate)
+        expected = torch.tensor(-69.5, dtype=loudness.dtype, device=self.device)
+        self.assertEqual(loudness, expected, rtol=0.01, atol=0.1)
+
+    def test_measure_loudness_two_channels(self):
+        filepath = filepath = self.download_and_extract_file("1770-2_Comp_24LKFS_500Hz_2ch")
+        waveform, sample_rate = load_wav(filepath)
+        waveform = waveform.to(self.device)
+
+        loudness = F.measure_loudness(waveform, sample_rate)
+        expected = torch.tensor(-24.0, dtype=loudness.dtype, device=self.device)
+        self.assertEqual(loudness, expected, rtol=0.01, atol=0.1)
+
+    def test_measure_loudness_mono_voice_music(self):
+        filepath = self.download_and_extract_file("1770-2 Conf Mono Voice+Music-24LKFS")
+        waveform, sample_rate = load_wav(filepath)
+        waveform = waveform.to(self.device)
+
+        loudness = F.measure_loudness(waveform, sample_rate)
+        expected = torch.tensor(-24.0, dtype=loudness.dtype, device=self.device)
+        self.assertEqual(loudness, expected, rtol=0.01, atol=0.1)
diff --git a/test/torchaudio_unittest/functional/torchscript_consistency_impl.py b/test/torchaudio_unittest/functional/torchscript_consistency_impl.py
index 99c1894725..5ac2decd3c 100644
--- a/test/torchaudio_unittest/functional/torchscript_consistency_impl.py
+++ b/test/torchaudio_unittest/functional/torchscript_consistency_impl.py
@@ -120,6 +120,14 @@ def func(tensor):
 
         self._assert_consistency(func, (waveform,))
 
+    def test_measure_loudness(self):
+        if self.dtype == torch.float64:
+            raise unittest.SkipTest("This test is known to fail for float64")
+
+        sample_rate = 44100
+        waveform = common_utils.get_sinusoid(sample_rate=sample_rate, device=self.device)
+        self._assert_consistency(F.measure_loudness, (waveform, sample_rate))
+
     def test_melscale_fbanks(self):
         if self.device != torch.device("cpu"):
             raise unittest.SkipTest("No need to perform test on device other than CPU")
diff --git a/torchaudio/functional/__init__.py b/torchaudio/functional/__init__.py
index 06325da3fe..6d9617162d 100644
--- a/torchaudio/functional/__init__.py
+++ b/torchaudio/functional/__init__.py
@@ -30,11 +30,16 @@
     compute_kaldi_pitch,
     create_dct,
     DB_to_amplitude,
+<<<<<<< HEAD
+    measure_loudness,
+=======
+>>>>>>> 42509325... Fixed linting issues
     detect_pitch_frequency,
     edit_distance,
     griffinlim,
     inverse_spectrogram,
     linear_fbanks,
+    loudness,
     mask_along_axis,
     mask_along_axis_iid,
     melscale_fbanks,
@@ -62,6 +67,7 @@
     "melscale_fbanks",
     "linear_fbanks",
     "DB_to_amplitude",
+    "measure_loudness",
     "detect_pitch_frequency",
     "griffinlim",
     "mask_along_axis",
diff --git a/torchaudio/functional/functional.py b/torchaudio/functional/functional.py
index 067cbc9330..211056a34e 100644
--- a/torchaudio/functional/functional.py
+++ b/torchaudio/functional/functional.py
@@ -11,6 +11,8 @@
 from torch import Tensor
 from torchaudio._internal import module_utils as _mod_utils
 
+from .filtering import highpass_biquad, treble_biquad
+
 __all__ = [
     "spectrogram",
     "inverse_spectrogram",
@@ -35,6 +37,7 @@
     "apply_codec",
     "resample",
     "edit_distance",
+    "measure_loudness",
     "pitch_shift",
     "rnnt_loss",
     "psd",
@@ -1640,6 +1643,66 @@ def edit_distance(seq1: Sequence, seq2: Sequence) -> int:
     return int(dold[-1])
 
 
+def measure_loudness(waveform: Tensor, sample_rate: int):
+    r"""Measure audio loudness according to the ITU-R BS.1770-4 recommendation.
+
+    .. devices:: CPU CUDA
+
+    .. properties:: TorchScript
+
+    Args:
+        waveform(torch.Tensor): audio waveform of dimension of `(..., channels, time)`
+        sample_rate (int): sampling rate of the waveform, e.g. 44100 (Hz)
+
+    Returns:
+        Tensor: loudness estimates (LKFS)
+
+    Reference:
+        - https://www.itu.int/rec/R-REC-BS.1770-4-201510-I/en
+    """
+
+    if waveform.size(-2) > 5:
+        raise ValueError("Only up to 5 channels are supported.")
+
+    gate_duration: float = 0.4
+    overlap: float = 0.75
+    gamma_abs: float = -70.0
+    gate_samples = int(round(gate_duration * sample_rate))
+    step = int(round(gate_samples * (1 - overlap)))
+
+    # Apply K-weighting
+    waveform = treble_biquad(waveform, sample_rate, 4.0, 1500.0, 1 / math.sqrt(2))
+    waveform = highpass_biquad(waveform, sample_rate, 38.0, 0.5)
+
+    # Compute the energy for each block
+    energy = torch.square(waveform).unfold(-1, gate_samples, step)
+    energy = torch.mean(energy, dim=-1)
+
+    # Compute channel-weighted summation
+    g = torch.tensor([1.0, 1.0, 1.0, 1.41, 1.41], dtype=waveform.dtype, device=waveform.device)
+    g = g[: energy.size(-2)]
+
+    energy_weighted = torch.sum(g.unsqueeze(-1) * energy, dim=-2)
+    loudness = -0.691 + 10 * torch.log10(energy_weighted)
+
+    # Apply absolute gating of the blocks
+    gated_blocks = loudness > gamma_abs
+    gated_blocks = gated_blocks.unsqueeze(-2)
+
+    energy_filtered = torch.sum(gated_blocks * energy, dim=-1) / torch.count_nonzero(gated_blocks, dim=-1)
+    energy_weighted = torch.sum(g * energy_filtered, dim=-1)
+    gamma_rel = -0.691 + 10 * torch.log10(energy_weighted) - 10
+
+    # Apply relative gating of the blocks
+    gated_blocks = torch.logical_and(gated_blocks.squeeze(-2), loudness > gamma_rel.unsqueeze(-1))
+    gated_blocks = gated_blocks.unsqueeze(-2)
+
+    energy_filtered = torch.sum(gated_blocks * energy, dim=-1) / torch.count_nonzero(gated_blocks, dim=-1)
+    energy_weighted = torch.sum(g * energy_filtered, dim=-1)
+    LKFS = -0.691 + 10 * torch.log10(energy_weighted)
+    return LKFS
+
+
 def pitch_shift(
     waveform: Tensor,
     sample_rate: int,

From 05fe1623e8671dddf70db42fb622dc4eb7985932 Mon Sep 17 00:00:00 2001
From: bshall <benji.l.shall@gmail.com>
Date: Fri, 22 Jul 2022 15:26:02 +0200
Subject: [PATCH 02/14] Fixed doc string and renamed function to

---
 .../functional/loudness_compliance_test_impl.py           | 8 ++++----
 .../functional/torchscript_consistency_impl.py            | 2 +-
 torchaudio/functional/__init__.py                         | 6 +-----
 torchaudio/functional/functional.py                       | 8 ++++----
 4 files changed, 10 insertions(+), 14 deletions(-)

diff --git a/test/torchaudio_unittest/functional/loudness_compliance_test_impl.py b/test/torchaudio_unittest/functional/loudness_compliance_test_impl.py
index 4adddb32f2..0fd27128cf 100644
--- a/test/torchaudio_unittest/functional/loudness_compliance_test_impl.py
+++ b/test/torchaudio_unittest/functional/loudness_compliance_test_impl.py
@@ -28,7 +28,7 @@ def test_measure_loudness_relative_gate(self):
         waveform, sample_rate = load_wav(filepath)
         waveform = waveform.to(self.device)
 
-        loudness = F.measure_loudness(waveform, sample_rate)
+        loudness = F.loudness(waveform, sample_rate)
         expected = torch.tensor(-10.0, dtype=loudness.dtype, device=self.device)
         self.assertEqual(loudness, expected, rtol=0.01, atol=0.1)
 
@@ -37,7 +37,7 @@ def test_measure_loudness_absolute_gate(self):
         waveform, sample_rate = load_wav(filepath)
         waveform = waveform.to(self.device)
 
-        loudness = F.measure_loudness(waveform, sample_rate)
+        loudness = F.loudness(waveform, sample_rate)
         expected = torch.tensor(-69.5, dtype=loudness.dtype, device=self.device)
         self.assertEqual(loudness, expected, rtol=0.01, atol=0.1)
 
@@ -46,7 +46,7 @@ def test_measure_loudness_two_channels(self):
         waveform, sample_rate = load_wav(filepath)
         waveform = waveform.to(self.device)
 
-        loudness = F.measure_loudness(waveform, sample_rate)
+        loudness = F.loudness(waveform, sample_rate)
         expected = torch.tensor(-24.0, dtype=loudness.dtype, device=self.device)
         self.assertEqual(loudness, expected, rtol=0.01, atol=0.1)
 
@@ -55,6 +55,6 @@ def test_measure_loudness_mono_voice_music(self):
         waveform, sample_rate = load_wav(filepath)
         waveform = waveform.to(self.device)
 
-        loudness = F.measure_loudness(waveform, sample_rate)
+        loudness = F.loudness(waveform, sample_rate)
         expected = torch.tensor(-24.0, dtype=loudness.dtype, device=self.device)
         self.assertEqual(loudness, expected, rtol=0.01, atol=0.1)
diff --git a/test/torchaudio_unittest/functional/torchscript_consistency_impl.py b/test/torchaudio_unittest/functional/torchscript_consistency_impl.py
index 5ac2decd3c..83d77ec01d 100644
--- a/test/torchaudio_unittest/functional/torchscript_consistency_impl.py
+++ b/test/torchaudio_unittest/functional/torchscript_consistency_impl.py
@@ -126,7 +126,7 @@ def test_measure_loudness(self):
 
         sample_rate = 44100
         waveform = common_utils.get_sinusoid(sample_rate=sample_rate, device=self.device)
-        self._assert_consistency(F.measure_loudness, (waveform, sample_rate))
+        self._assert_consistency(F.loudness, (waveform, sample_rate))
 
     def test_melscale_fbanks(self):
         if self.device != torch.device("cpu"):
diff --git a/torchaudio/functional/__init__.py b/torchaudio/functional/__init__.py
index 6d9617162d..49f8c686e9 100644
--- a/torchaudio/functional/__init__.py
+++ b/torchaudio/functional/__init__.py
@@ -30,10 +30,6 @@
     compute_kaldi_pitch,
     create_dct,
     DB_to_amplitude,
-<<<<<<< HEAD
-    measure_loudness,
-=======
->>>>>>> 42509325... Fixed linting issues
     detect_pitch_frequency,
     edit_distance,
     griffinlim,
@@ -67,7 +63,7 @@
     "melscale_fbanks",
     "linear_fbanks",
     "DB_to_amplitude",
-    "measure_loudness",
+    "loudness",
     "detect_pitch_frequency",
     "griffinlim",
     "mask_along_axis",
diff --git a/torchaudio/functional/functional.py b/torchaudio/functional/functional.py
index 211056a34e..7e7c93a581 100644
--- a/torchaudio/functional/functional.py
+++ b/torchaudio/functional/functional.py
@@ -37,7 +37,7 @@
     "apply_codec",
     "resample",
     "edit_distance",
-    "measure_loudness",
+    "loudness",
     "pitch_shift",
     "rnnt_loss",
     "psd",
@@ -1643,7 +1643,7 @@ def edit_distance(seq1: Sequence, seq2: Sequence) -> int:
     return int(dold[-1])
 
 
-def measure_loudness(waveform: Tensor, sample_rate: int):
+def loudness(waveform: Tensor, sample_rate: int):
     r"""Measure audio loudness according to the ITU-R BS.1770-4 recommendation.
 
     .. devices:: CPU CUDA
@@ -1651,8 +1651,8 @@ def measure_loudness(waveform: Tensor, sample_rate: int):
     .. properties:: TorchScript
 
     Args:
-        waveform(torch.Tensor): audio waveform of dimension of `(..., channels, time)`
-        sample_rate (int): sampling rate of the waveform, e.g. 44100 (Hz)
+        waveform(torch.Tensor): audio waveform of dimension `(..., channels, time)`
+        sample_rate (int): sampling rate of the waveform
 
     Returns:
         Tensor: loudness estimates (LKFS)

From 2d91fa2cf1dafd556ad7fe89fc1240ffba8a9f7c Mon Sep 17 00:00:00 2001
From: bshall <benji.l.shall@gmail.com>
Date: Fri, 22 Jul 2022 15:50:46 +0200
Subject: [PATCH 03/14] Moved loudness test out of unittest directory

---
 .../loudness_compliance_cpu_test.py                               | 0
 .../loudness_compliance_cuda_test.py                              | 0
 .../loudness_compliance_test_impl.py                              | 0
 3 files changed, 0 insertions(+), 0 deletions(-)
 rename test/{torchaudio_unittest/functional => integration_tests}/loudness_compliance_cpu_test.py (100%)
 rename test/{torchaudio_unittest/functional => integration_tests}/loudness_compliance_cuda_test.py (100%)
 rename test/{torchaudio_unittest/functional => integration_tests}/loudness_compliance_test_impl.py (100%)

diff --git a/test/torchaudio_unittest/functional/loudness_compliance_cpu_test.py b/test/integration_tests/loudness_compliance_cpu_test.py
similarity index 100%
rename from test/torchaudio_unittest/functional/loudness_compliance_cpu_test.py
rename to test/integration_tests/loudness_compliance_cpu_test.py
diff --git a/test/torchaudio_unittest/functional/loudness_compliance_cuda_test.py b/test/integration_tests/loudness_compliance_cuda_test.py
similarity index 100%
rename from test/torchaudio_unittest/functional/loudness_compliance_cuda_test.py
rename to test/integration_tests/loudness_compliance_cuda_test.py
diff --git a/test/torchaudio_unittest/functional/loudness_compliance_test_impl.py b/test/integration_tests/loudness_compliance_test_impl.py
similarity index 100%
rename from test/torchaudio_unittest/functional/loudness_compliance_test_impl.py
rename to test/integration_tests/loudness_compliance_test_impl.py

From 4243b5a79f0f813b66126e17e6a5f74804dade2c Mon Sep 17 00:00:00 2001
From: bshall <benji.l.shall@gmail.com>
Date: Fri, 22 Jul 2022 16:00:27 +0200
Subject: [PATCH 04/14] Added Loudness transform

---
 torchaudio/transforms/__init__.py    |  2 ++
 torchaudio/transforms/_transforms.py | 26 ++++++++++++++++++++++++++
 2 files changed, 28 insertions(+)

diff --git a/torchaudio/transforms/__init__.py b/torchaudio/transforms/__init__.py
index 527da5c7d2..2dc5bcf555 100644
--- a/torchaudio/transforms/__init__.py
+++ b/torchaudio/transforms/__init__.py
@@ -8,6 +8,7 @@
     InverseMelScale,
     InverseSpectrogram,
     LFCC,
+    Loudness,
     MelScale,
     MelSpectrogram,
     MFCC,
@@ -35,6 +36,7 @@
     "InverseMelScale",
     "InverseSpectrogram",
     "LFCC",
+    "Loudness",
     "MFCC",
     "MVDR",
     "MelScale",
diff --git a/torchaudio/transforms/_transforms.py b/torchaudio/transforms/_transforms.py
index bb5144d713..f89c47a030 100644
--- a/torchaudio/transforms/_transforms.py
+++ b/torchaudio/transforms/_transforms.py
@@ -1251,6 +1251,32 @@ def __init__(self, time_mask_param: int, iid_masks: bool = False, p: float = 1.0
         super(TimeMasking, self).__init__(time_mask_param, 2, iid_masks, p=p)
 
 
+class Loudness(torch.nn.Module):
+    r"""Measure audio loudness according to the ITU-R BS.1770-4 recommendation.
+
+    .. devices:: CPU CUDA
+
+    .. properties:: TorchScript
+
+    Reference:
+        - https://www.itu.int/rec/R-REC-BS.1770-4-201510-I/en
+    """
+
+    def __init__(self):
+        super(Loudness, self).__init__()
+
+    def forward(self, wavefrom: Tensor, sample_rate: int):
+        """
+        Args:
+            waveform(torch.Tensor): audio waveform of dimension `(..., channels, time)`
+            sample_rate (int): sampling rate of the waveform
+
+        Returns:
+            Tensor: loudness estimates (LKFS)
+        """
+        return F.loudness(wavefrom, sample_rate)
+
+
 class Vol(torch.nn.Module):
     r"""Add a volume to an waveform.
 

From a02ccbc054e38be325c2d656f17f39c003290a71 Mon Sep 17 00:00:00 2001
From: bshall <benji.l.shall@gmail.com>
Date: Fri, 22 Jul 2022 16:17:46 +0200
Subject: [PATCH 05/14] Removed float types for contants

---
 torchaudio/functional/functional.py | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/torchaudio/functional/functional.py b/torchaudio/functional/functional.py
index 7e7c93a581..0dded5d2dc 100644
--- a/torchaudio/functional/functional.py
+++ b/torchaudio/functional/functional.py
@@ -1664,9 +1664,9 @@ def loudness(waveform: Tensor, sample_rate: int):
     if waveform.size(-2) > 5:
         raise ValueError("Only up to 5 channels are supported.")
 
-    gate_duration: float = 0.4
-    overlap: float = 0.75
-    gamma_abs: float = -70.0
+    gate_duration = 0.4
+    overlap = 0.75
+    gamma_abs = -70.0
     gate_samples = int(round(gate_duration * sample_rate))
     step = int(round(gate_samples * (1 - overlap)))
 

From f968789308d3220324df80835be4a6880cf4c950 Mon Sep 17 00:00:00 2001
From: bshall <benji.l.shall@gmail.com>
Date: Fri, 22 Jul 2022 17:06:08 +0200
Subject: [PATCH 06/14] Moved sample_rate to __init__ in Loudness transform

---
 torchaudio/transforms/_transforms.py | 14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)

diff --git a/torchaudio/transforms/_transforms.py b/torchaudio/transforms/_transforms.py
index f89c47a030..89927e3b82 100644
--- a/torchaudio/transforms/_transforms.py
+++ b/torchaudio/transforms/_transforms.py
@@ -1258,23 +1258,27 @@ class Loudness(torch.nn.Module):
 
     .. properties:: TorchScript
 
+    Args:
+        sample_rate (int, optional): sampling rate of the waveform
+
     Reference:
         - https://www.itu.int/rec/R-REC-BS.1770-4-201510-I/en
     """
+    __constants__ = ["sample_rate"]
 
-    def __init__(self):
+    def __init__(self, sample_rate: int = 16000):
         super(Loudness, self).__init__()
+        self.sample_rate = sample_rate
 
-    def forward(self, wavefrom: Tensor, sample_rate: int):
-        """
+    def forward(self, wavefrom: Tensor):
+        r"""
         Args:
             waveform(torch.Tensor): audio waveform of dimension `(..., channels, time)`
-            sample_rate (int): sampling rate of the waveform
 
         Returns:
             Tensor: loudness estimates (LKFS)
         """
-        return F.loudness(wavefrom, sample_rate)
+        return F.loudness(wavefrom, self.sample_rate)
 
 
 class Vol(torch.nn.Module):

From d8c24afcf08468f6a2101bb50b516984fb47a2e3 Mon Sep 17 00:00:00 2001
From: bshall <benji.l.shall@gmail.com>
Date: Tue, 26 Jul 2022 10:20:39 +0200
Subject: [PATCH 07/14] Fixed doc string

---
 torchaudio/transforms/_transforms.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/torchaudio/transforms/_transforms.py b/torchaudio/transforms/_transforms.py
index 89927e3b82..f6a7e73b0e 100644
--- a/torchaudio/transforms/_transforms.py
+++ b/torchaudio/transforms/_transforms.py
@@ -1259,14 +1259,14 @@ class Loudness(torch.nn.Module):
     .. properties:: TorchScript
 
     Args:
-        sample_rate (int, optional): sampling rate of the waveform
+        sample_rate (int): Sample rate of audio signal.
 
     Reference:
         - https://www.itu.int/rec/R-REC-BS.1770-4-201510-I/en
     """
     __constants__ = ["sample_rate"]
 
-    def __init__(self, sample_rate: int = 16000):
+    def __init__(self, sample_rate: int):
         super(Loudness, self).__init__()
         self.sample_rate = sample_rate
 

From 7e97e153b84333dfa0236bd4341cb9e71f35f710 Mon Sep 17 00:00:00 2001
From: bshall <benji.l.shall@gmail.com>
Date: Tue, 26 Jul 2022 10:29:51 +0200
Subject: [PATCH 08/14] Refactored loudness tests

---
 .../loudness_compliance_test_impl.py          | 42 ++++++-------------
 1 file changed, 12 insertions(+), 30 deletions(-)

diff --git a/test/integration_tests/loudness_compliance_test_impl.py b/test/integration_tests/loudness_compliance_test_impl.py
index 0fd27128cf..67b5b2f602 100644
--- a/test/integration_tests/loudness_compliance_test_impl.py
+++ b/test/integration_tests/loudness_compliance_test_impl.py
@@ -4,6 +4,7 @@
 
 import torch
 import torchaudio.functional as F
+from parameterized import parameterized
 from torchaudio_unittest.common_utils import load_wav, TempDirMixin, TestBaseMixin
 
 # Test files linked in https://www.itu.int/dms_pub/itu-r/opb/rep/R-REP-BS.2217-2-2016-PDF-E.pdf
@@ -23,38 +24,19 @@ def download_and_extract_file(self, filename):
             file.extractall(os.path.dirname(zippath))
         return self.get_temp_path(filename + ".wav")
 
-    def test_measure_loudness_relative_gate(self):
-        filepath = self.download_and_extract_file("1770-2_Comp_RelGateTest")
+    @parameterized.expand(
+        [
+            ("1770-2_Comp_RelGateTest", -10.0),
+            ("1770-2_Comp_AbsGateTest", -69.5),
+            ("1770-2_Comp_24LKFS_500Hz_2ch", -24.0),
+            ("1770-2 Conf Mono Voice+Music-24LKFS", -24.0),
+        ]
+    )
+    def test_loudness(self, filename, expected):
+        filepath = self.download_and_extract_file(filename)
         waveform, sample_rate = load_wav(filepath)
         waveform = waveform.to(self.device)
 
         loudness = F.loudness(waveform, sample_rate)
-        expected = torch.tensor(-10.0, dtype=loudness.dtype, device=self.device)
-        self.assertEqual(loudness, expected, rtol=0.01, atol=0.1)
-
-    def test_measure_loudness_absolute_gate(self):
-        filepath = self.download_and_extract_file("1770-2_Comp_AbsGateTest")
-        waveform, sample_rate = load_wav(filepath)
-        waveform = waveform.to(self.device)
-
-        loudness = F.loudness(waveform, sample_rate)
-        expected = torch.tensor(-69.5, dtype=loudness.dtype, device=self.device)
-        self.assertEqual(loudness, expected, rtol=0.01, atol=0.1)
-
-    def test_measure_loudness_two_channels(self):
-        filepath = filepath = self.download_and_extract_file("1770-2_Comp_24LKFS_500Hz_2ch")
-        waveform, sample_rate = load_wav(filepath)
-        waveform = waveform.to(self.device)
-
-        loudness = F.loudness(waveform, sample_rate)
-        expected = torch.tensor(-24.0, dtype=loudness.dtype, device=self.device)
-        self.assertEqual(loudness, expected, rtol=0.01, atol=0.1)
-
-    def test_measure_loudness_mono_voice_music(self):
-        filepath = self.download_and_extract_file("1770-2 Conf Mono Voice+Music-24LKFS")
-        waveform, sample_rate = load_wav(filepath)
-        waveform = waveform.to(self.device)
-
-        loudness = F.loudness(waveform, sample_rate)
-        expected = torch.tensor(-24.0, dtype=loudness.dtype, device=self.device)
+        expected = torch.tensor(expected, dtype=loudness.dtype, device=self.device)
         self.assertEqual(loudness, expected, rtol=0.01, atol=0.1)

From 4e72bcab238db4d2b45fd92d03beb19fffd956fc Mon Sep 17 00:00:00 2001
From: bshall <benji.l.shall@gmail.com>
Date: Tue, 26 Jul 2022 10:35:40 +0200
Subject: [PATCH 09/14] Added loudness entries to functional.rst and
 transforms.rst

---
 docs/source/functional.rst | 5 +++++
 docs/source/transforms.rst | 7 +++++++
 2 files changed, 12 insertions(+)

diff --git a/docs/source/functional.rst b/docs/source/functional.rst
index edec942731..2945f5cb2c 100644
--- a/docs/source/functional.rst
+++ b/docs/source/functional.rst
@@ -69,6 +69,11 @@ resample
 
 .. autofunction:: resample
 
+loudness
+--------
+
+.. autofunction:: loudness
+
 
 :hidden:`Filtering`
 ~~~~~~~~~~~~~~~~~~~
diff --git a/docs/source/transforms.rst b/docs/source/transforms.rst
index 4fe5acce26..e2d046fc39 100644
--- a/docs/source/transforms.rst
+++ b/docs/source/transforms.rst
@@ -90,6 +90,13 @@ Transforms are common audio transforms. They can be chained together using :clas
 
   .. automethod:: forward
 
+:hidden:`Loudness`
+-------------
+
+.. autoclass:: Loudness
+
+  .. automethod:: forward
+
 :hidden:`Feature Extractions`
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 

From 93f350fd07b0697a32d54099d3b2c9bd782fadc4 Mon Sep 17 00:00:00 2001
From: bshall <benji.l.shall@gmail.com>
Date: Tue, 26 Jul 2022 10:38:15 +0200
Subject: [PATCH 10/14] Replaced constant bias with variable

---
 torchaudio/functional/functional.py | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/torchaudio/functional/functional.py b/torchaudio/functional/functional.py
index 0dded5d2dc..b179979185 100644
--- a/torchaudio/functional/functional.py
+++ b/torchaudio/functional/functional.py
@@ -1667,6 +1667,7 @@ def loudness(waveform: Tensor, sample_rate: int):
     gate_duration = 0.4
     overlap = 0.75
     gamma_abs = -70.0
+    kweight_bias = -0.691
     gate_samples = int(round(gate_duration * sample_rate))
     step = int(round(gate_samples * (1 - overlap)))
 
@@ -1691,7 +1692,7 @@ def loudness(waveform: Tensor, sample_rate: int):
 
     energy_filtered = torch.sum(gated_blocks * energy, dim=-1) / torch.count_nonzero(gated_blocks, dim=-1)
     energy_weighted = torch.sum(g * energy_filtered, dim=-1)
-    gamma_rel = -0.691 + 10 * torch.log10(energy_weighted) - 10
+    gamma_rel = kweight_bias + 10 * torch.log10(energy_weighted) - 10
 
     # Apply relative gating of the blocks
     gated_blocks = torch.logical_and(gated_blocks.squeeze(-2), loudness > gamma_rel.unsqueeze(-1))
@@ -1699,7 +1700,7 @@ def loudness(waveform: Tensor, sample_rate: int):
 
     energy_filtered = torch.sum(gated_blocks * energy, dim=-1) / torch.count_nonzero(gated_blocks, dim=-1)
     energy_weighted = torch.sum(g * energy_filtered, dim=-1)
-    LKFS = -0.691 + 10 * torch.log10(energy_weighted)
+    LKFS = kweight_bias + 10 * torch.log10(energy_weighted)
     return LKFS
 
 

From 01f979f1d8651677ff43b17fe86b4fd851e841da Mon Sep 17 00:00:00 2001
From: bshall <benji.l.shall@gmail.com>
Date: Thu, 28 Jul 2022 11:13:17 +0200
Subject: [PATCH 11/14] Added expecttest and parameterized to integration test
 workflow

---
 .github/workflows/integration-test.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/integration-test.yml b/.github/workflows/integration-test.yml
index 4eef95dddb..13e82ab73d 100644
--- a/.github/workflows/integration-test.yml
+++ b/.github/workflows/integration-test.yml
@@ -29,7 +29,7 @@ jobs:
       run: |
         python -m pip install --quiet --upgrade pip
         python -m pip install --quiet --pre torch -f https://download.pytorch.org/whl/nightly/cpu/torch_nightly.html
-        python -m pip install --quiet pytest requests cmake ninja deep-phonemizer sentencepiece
+        python -m pip install --quiet pytest requests expecttest parameterized cmake ninja deep-phonemizer sentencepiece
         python setup.py install
       env:
         USE_FFMPEG: true

From b173b58883f2874a6fb4498e2cfb4a38256cae02 Mon Sep 17 00:00:00 2001
From: bshall <benji.l.shall@gmail.com>
Date: Thu, 28 Jul 2022 17:20:30 +0200
Subject: [PATCH 12/14] Added scipy to integration test workflow

---
 .github/workflows/integration-test.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/integration-test.yml b/.github/workflows/integration-test.yml
index 13e82ab73d..cb777cb4eb 100644
--- a/.github/workflows/integration-test.yml
+++ b/.github/workflows/integration-test.yml
@@ -29,7 +29,7 @@ jobs:
       run: |
         python -m pip install --quiet --upgrade pip
         python -m pip install --quiet --pre torch -f https://download.pytorch.org/whl/nightly/cpu/torch_nightly.html
-        python -m pip install --quiet pytest requests expecttest parameterized cmake ninja deep-phonemizer sentencepiece
+        python -m pip install --quiet pytest requests expecttest parameterized scipy cmake ninja deep-phonemizer sentencepiece
         python setup.py install
       env:
         USE_FFMPEG: true

From 9c54cd7a67053457475b8fdbba66919e449e536e Mon Sep 17 00:00:00 2001
From: bshall <benji.l.shall@gmail.com>
Date: Tue, 2 Aug 2022 15:40:10 +0200
Subject: [PATCH 13/14] Removed torchaudio_unittest module from loudness test

---
 .github/workflows/integration-test.yml        |  2 +-
 .../loudness_compliance_cpu_test.py           |  8 ----
 .../loudness_compliance_cuda_test.py          |  9 ----
 .../loudness_compliance_test.py               | 45 +++++++++++++++++++
 .../loudness_compliance_test_impl.py          | 42 -----------------
 5 files changed, 46 insertions(+), 60 deletions(-)
 delete mode 100644 test/integration_tests/loudness_compliance_cpu_test.py
 delete mode 100644 test/integration_tests/loudness_compliance_cuda_test.py
 create mode 100644 test/integration_tests/loudness_compliance_test.py
 delete mode 100644 test/integration_tests/loudness_compliance_test_impl.py

diff --git a/.github/workflows/integration-test.yml b/.github/workflows/integration-test.yml
index cb777cb4eb..4eef95dddb 100644
--- a/.github/workflows/integration-test.yml
+++ b/.github/workflows/integration-test.yml
@@ -29,7 +29,7 @@ jobs:
       run: |
         python -m pip install --quiet --upgrade pip
         python -m pip install --quiet --pre torch -f https://download.pytorch.org/whl/nightly/cpu/torch_nightly.html
-        python -m pip install --quiet pytest requests expecttest parameterized scipy cmake ninja deep-phonemizer sentencepiece
+        python -m pip install --quiet pytest requests cmake ninja deep-phonemizer sentencepiece
         python setup.py install
       env:
         USE_FFMPEG: true
diff --git a/test/integration_tests/loudness_compliance_cpu_test.py b/test/integration_tests/loudness_compliance_cpu_test.py
deleted file mode 100644
index f8dbaeef28..0000000000
--- a/test/integration_tests/loudness_compliance_cpu_test.py
+++ /dev/null
@@ -1,8 +0,0 @@
-import torch
-from torchaudio_unittest import common_utils
-
-from .loudness_compliance_test_impl import Loudness
-
-
-class TestLoudnessCPU(Loudness, common_utils.PytorchTestCase):
-    device = torch.device("cpu")
diff --git a/test/integration_tests/loudness_compliance_cuda_test.py b/test/integration_tests/loudness_compliance_cuda_test.py
deleted file mode 100644
index cd310e215f..0000000000
--- a/test/integration_tests/loudness_compliance_cuda_test.py
+++ /dev/null
@@ -1,9 +0,0 @@
-import torch
-from torchaudio_unittest import common_utils
-
-from .loudness_compliance_test_impl import Loudness
-
-
-@common_utils.skipIfNoCuda
-class TestLoudnessCUDA(Loudness, common_utils.PytorchTestCase):
-    device = torch.device("cuda")
diff --git a/test/integration_tests/loudness_compliance_test.py b/test/integration_tests/loudness_compliance_test.py
new file mode 100644
index 0000000000..7ac8dba9e3
--- /dev/null
+++ b/test/integration_tests/loudness_compliance_test.py
@@ -0,0 +1,45 @@
+"""Test suite for compliance with the ITU-R BS.1770-4 recommendation"""
+import zipfile
+
+import pytest
+
+import torch
+import torchaudio
+import torchaudio.functional as F
+
+
+# Test files linked in https://www.itu.int/dms_pub/itu-r/opb/rep/R-REP-BS.2217-2-2016-PDF-E.pdf
+@pytest.mark.parametrize(
+    "filename,url,expected",
+    [
+        (
+            "1770-2_Comp_RelGateTest",
+            "http://www.itu.int/dms_pub/itu-r/oth/11/02/R11020000010030ZIPM.zip",
+            -10.0,
+        ),
+        (
+            "1770-2_Comp_AbsGateTest",
+            "http://www.itu.int/dms_pub/itu-r/oth/11/02/R11020000010029ZIPM.zip",
+            -69.5,
+        ),
+        (
+            "1770-2_Comp_24LKFS_500Hz_2ch",
+            "http://www.itu.int/dms_pub/itu-r/oth/11/02/R11020000010018ZIPM.zip",
+            -24.0,
+        ),
+        (
+            "1770-2 Conf Mono Voice+Music-24LKFS",
+            "http://www.itu.int/dms_pub/itu-r/oth/11/02/R11020000010038ZIPM.zip",
+            -24.0,
+        ),
+    ],
+)
+def test_loudness(tmp_path, filename, url, expected):
+    zippath = tmp_path / filename
+    torch.hub.download_url_to_file(url, zippath, progress=False)
+    with zipfile.ZipFile(zippath) as file:
+        file.extractall(zippath.parent)
+
+    waveform, sample_rate = torchaudio.load(zippath.with_suffix(".wav"))
+    loudness = F.loudness(waveform, sample_rate)
+    assert pytest.approx(loudness.item(), rel=0.01, abs=0.1) == expected
diff --git a/test/integration_tests/loudness_compliance_test_impl.py b/test/integration_tests/loudness_compliance_test_impl.py
deleted file mode 100644
index 67b5b2f602..0000000000
--- a/test/integration_tests/loudness_compliance_test_impl.py
+++ /dev/null
@@ -1,42 +0,0 @@
-"""Test suite for compliance with the ITU-R BS.1770-4 recommendation"""
-import os.path
-import zipfile
-
-import torch
-import torchaudio.functional as F
-from parameterized import parameterized
-from torchaudio_unittest.common_utils import load_wav, TempDirMixin, TestBaseMixin
-
-# Test files linked in https://www.itu.int/dms_pub/itu-r/opb/rep/R-REP-BS.2217-2-2016-PDF-E.pdf
-_COMPLIANCE_FILE_URLS = {
-    "1770-2_Comp_RelGateTest": "http://www.itu.int/dms_pub/itu-r/oth/11/02/R11020000010030ZIPM.zip",
-    "1770-2_Comp_AbsGateTest": "http://www.itu.int/dms_pub/itu-r/oth/11/02/R11020000010029ZIPM.zip",
-    "1770-2_Comp_24LKFS_500Hz_2ch": "http://www.itu.int/dms_pub/itu-r/oth/11/02/R11020000010018ZIPM.zip",
-    "1770-2 Conf Mono Voice+Music-24LKFS": "http://www.itu.int/dms_pub/itu-r/oth/11/02/R11020000010038ZIPM.zip",
-}
-
-
-class Loudness(TempDirMixin, TestBaseMixin):
-    def download_and_extract_file(self, filename):
-        zippath = self.get_temp_path(filename + ".zip")
-        torch.hub.download_url_to_file(_COMPLIANCE_FILE_URLS[filename], zippath, progress=False)
-        with zipfile.ZipFile(zippath) as file:
-            file.extractall(os.path.dirname(zippath))
-        return self.get_temp_path(filename + ".wav")
-
-    @parameterized.expand(
-        [
-            ("1770-2_Comp_RelGateTest", -10.0),
-            ("1770-2_Comp_AbsGateTest", -69.5),
-            ("1770-2_Comp_24LKFS_500Hz_2ch", -24.0),
-            ("1770-2 Conf Mono Voice+Music-24LKFS", -24.0),
-        ]
-    )
-    def test_loudness(self, filename, expected):
-        filepath = self.download_and_extract_file(filename)
-        waveform, sample_rate = load_wav(filepath)
-        waveform = waveform.to(self.device)
-
-        loudness = F.loudness(waveform, sample_rate)
-        expected = torch.tensor(expected, dtype=loudness.dtype, device=self.device)
-        self.assertEqual(loudness, expected, rtol=0.01, atol=0.1)

From 2dbeab0b0eb850d9d4503d2b5df43829d0b82e30 Mon Sep 17 00:00:00 2001
From: Caroline Chen <carolinechen@fb.com>
Date: Wed, 3 Aug 2022 11:44:27 -0400
Subject: [PATCH 14/14] nits

---
 docs/source/transforms.rst                         | 2 +-
 test/integration_tests/loudness_compliance_test.py | 3 ++-
 2 files changed, 3 insertions(+), 2 deletions(-)

diff --git a/docs/source/transforms.rst b/docs/source/transforms.rst
index e2d046fc39..9077616805 100644
--- a/docs/source/transforms.rst
+++ b/docs/source/transforms.rst
@@ -91,7 +91,7 @@ Transforms are common audio transforms. They can be chained together using :clas
   .. automethod:: forward
 
 :hidden:`Loudness`
--------------
+-----------------
 
 .. autoclass:: Loudness
 
diff --git a/test/integration_tests/loudness_compliance_test.py b/test/integration_tests/loudness_compliance_test.py
index 7ac8dba9e3..d9473cfa50 100644
--- a/test/integration_tests/loudness_compliance_test.py
+++ b/test/integration_tests/loudness_compliance_test.py
@@ -42,4 +42,5 @@ def test_loudness(tmp_path, filename, url, expected):
 
     waveform, sample_rate = torchaudio.load(zippath.with_suffix(".wav"))
     loudness = F.loudness(waveform, sample_rate)
-    assert pytest.approx(loudness.item(), rel=0.01, abs=0.1) == expected
+    expected = torch.tensor(expected, dtype=loudness.dtype, device=loudness.device)
+    assert torch.allclose(loudness, expected, rtol=0.01, atol=0.1)