sonusai 0.14.1__tar.gz → 0.14.2__tar.gz

{sonusai-0.14.1 → sonusai-0.14.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sonusai
-Version: 0.14.1
+Version: 0.14.2
 Summary: Framework for building deep neural network models for sound, speech, and voice AI
 Home-page: https://aaware.com
 License: GPL-3.0-only
@@ -34,6 +34,7 @@ Requires-Dist: pystoi (>=0.3.3,<0.4.0)
 Requires-Dist: requests (>=2.31.0,<3.0.0)
 Requires-Dist: scikit-learn (>=1.3.1,<2.0.0)
 Requires-Dist: sh (>=2.0.6,<3.0.0)
+Requires-Dist: soundfile (>=0.12.1,<0.13.0)
 Requires-Dist: sox (>=1.4.1,<2.0.0)
 Requires-Dist: speechrecognition (>=3.10.0,<4.0.0)
 Requires-Dist: sqlalchemy[mypy] (>=2.0.22,<3.0.0)

{sonusai-0.14.1 → sonusai-0.14.2}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "sonusai"
-version = "0.14.1"
+version = "0.14.2"
 description = "Framework for building deep neural network models for sound, speech, and voice AI"
 authors = ["Chris Eddington <chris@aaware.com>", "Jason Calderwood <jason@aaware.com>"]
 maintainers = ["Chris Eddington <chris@aaware.com>", "Jason Calderwood <jason@aaware.com>"]
@@ -38,6 +38,7 @@ python = ">=3.9,<3.12"
 requests = "^2.31.0"
 scikit-learn = "^1.3.1"
 sh = "^2.0.6"
+soundfile = "^0.12.1"
 sox = "^1.4.1"
 speechrecognition = "^3.10.0"
 sqlalchemy = {extras = ["mypy"], version = "^2.0.22"}

{sonusai-0.14.1 → sonusai-0.14.2}/sonusai/mixture/audio.py

@@ -59,6 +59,7 @@ def get_duration(audio: AudioT) -> float:
     :return: Duration of audio in seconds
     """
     from .constants import SAMPLE_RATE
+
     return len(audio) / SAMPLE_RATE
 
 
@@ -66,14 +67,15 @@ def validate_input_file(input_filepath: str) -> None:
     from os.path import exists
     from os.path import splitext
 
-    from torchaudio.utils.sox_utils import list_read_formats
+    from soundfile import available_formats
 
     from sonusai import SonusAIError
+
     if not exists(input_filepath):
         raise SonusAIError(f'input_filepath {input_filepath} does not exist.')
 
     ext = splitext(input_filepath)[1][1:].lower()
-    read_formats = list_read_formats()
+    read_formats = [item.lower() for item in available_formats().keys()]
     if ext not in read_formats:
         raise SonusAIError(f'This installation of SoX cannot process .{ext} files')
 
@@ -86,6 +88,7 @@ def read_audio(name: Location) -> AudioT:
     :return: Array of time domain audio data
     """
     from .torchaudio_audio import read_torchaudio_audio
+
     return read_torchaudio_audio(name)
 
 
@@ -97,4 +100,5 @@ def read_ir(name: Location) -> ImpulseResponseData:
     :return: ImpulseResponseData object
     """
     from .torchaudio_audio import read_torchaudio_ir
+
     return read_torchaudio_ir(name)

{sonusai-0.14.1 → sonusai-0.14.2}/sonusai/mixture/torchaudio_audio.py

@@ -21,7 +21,7 @@ def read_torchaudio_ir(name: Location) -> ImpulseResponseData:
 
     # Read impulse response data from audio file
     try:
-        raw, sample_rate = torchaudio.load(expanded_name)
+        raw, sample_rate = torchaudio.load(expanded_name, backend='soundfile')
     except Exception as e:
         if name != expanded_name:
             raise SonusAIError(f'Error reading {name} (expanded: {expanded_name}): {e}')
@@ -58,7 +58,7 @@ def read_torchaudio_audio(name: Location) -> AudioT:
     expanded_name, _ = tokenized_expand(name)
 
     try:
-        out, samplerate = torchaudio.load(expanded_name)
+        out, samplerate = torchaudio.load(expanded_name, backend='soundfile')
         out = torch.reshape(out[0, :], (1, out.size()[1]))
 
         if not samplerate == SAMPLE_RATE: