feat: Add support for Sox's compression(bitrate) argument

Adds support for bitrate as parameter to output arguments. It is useful converting files to MP3 format. Requires a new version of soundfile(0.11.0) that supports MP3 formats alongside upgraded libsndfile version to 1.1.0. This is for the `test_bitrate_valid` test.

Author: adinhodovic

.github/workflows/run-tests.yml

@@ -31,7 +31,7 @@ jobs:
         python -m pip install --upgrade pip
         python -m pip install flake8 pytest pytest-cov
         python -m pip install coveralls
-        python -m pip install pysoundfile
+        python -m pip install soundfile
         if [ -f requirements.txt ]; then pip install -r requirements.txt; fi
         python -m pip install -e .
         mkdir ${{ github.workspace }}/coverage
@@ -46,7 +46,7 @@ jobs:
     - name: Test with pytest
       run: |
         pytest tests/
-      
+
     - name: Run Coveralls
       run: |
         coverage run -m pytest tests/ > ${{ github.workspace }}/coverage/lcov.info

setup.py

@@ -32,7 +32,7 @@
                 'pytest',
                 'pytest-cov',
                 'pytest-pep8',
-                'pysoundfile >= 0.9.0',
+                'soundfile >= 0.11.0',
             ],
             'docs': [
                 'sphinx==1.2.3',  # autodoc was broken in 1.3.1

sox/transform.py

@@ -317,6 +317,7 @@ def _validate_output_format(self, output_format):
         bits = output_format.get('bits')
         channels = output_format.get('channels')
         encoding = output_format.get('encoding')
+        bitrate = output_format.get('bitrate')
         comments = output_format.get('comments')
         append_comments = output_format.get('append_comments', True)
 
@@ -343,6 +344,9 @@ def _validate_output_format(self, output_format):
         if channels is not None and channels <= 0:
             raise ValueError('channels must be a positive number')
 
+        if not isinstance(bitrate, float) and bitrate is not None:
+            raise ValueError('bitrate must be an float or None')
+
         if encoding not in ENCODING_VALS + [None]:
             raise ValueError(
                 'Invalid encoding. Must be one of {}'.format(ENCODING_VALS)
@@ -364,6 +368,7 @@ def _output_format_args(self, output_format):
         bits = output_format.get('bits')
         channels = output_format.get('channels')
         encoding = output_format.get('encoding')
+        bitrate = output_format.get('bitrate')
         comments = output_format.get('comments')
         append_comments = output_format.get('append_comments', True)
 
@@ -384,6 +389,9 @@ def _output_format_args(self, output_format):
         if encoding is not None:
             output_format_args.extend(['-e', '{}'.format(encoding)])
 
+        if bitrate is not None:
+            output_format_args.extend(['-C', '{}'.format(bitrate)])
+
         if comments is not None:
             if append_comments:
                 output_format_args.extend(['--add-comment', comments])
@@ -398,6 +406,7 @@ def set_output_format(self,
                           bits: Optional[int] = None,
                           channels: Optional[int] = None,
                           encoding: Optional[EncodingValue] = None,
+                          bitrate: Optional[float] = None,
                           comments: Optional[str] = None,
                           append_comments: bool = True):
         '''Sets output file format arguments. These arguments will overwrite
@@ -456,6 +465,8 @@ def set_output_format(self,
                     associated speech quality. SoX has support for GSM’s
                     original 13kbps ‘Full Rate’ audio format. It is usually
                     CPU-intensive to work with GSM audio.
+        bitrate : float, default=None
+                    Desired bitrate. Uses Sox's -C (compression) argument.
         comments : str or None, default=None
             If not None, the string is added as a comment in the header of the
             output audio file. If None, no comments are added.
@@ -469,6 +480,7 @@ def set_output_format(self,
             'bits': bits,
             'channels': channels,
             'encoding': encoding,
+            'bitrate': bitrate,
             'comments': comments,
             'append_comments': append_comments
         }
@@ -1479,7 +1491,8 @@ def contrast(self, amount=75):
     def convert(self,
                 samplerate: Optional[float] = None,
                 n_channels: Optional[int] = None,
-                bitdepth: Optional[int] = None):
+                bitdepth: Optional[int] = None,
+                bitrate: Optional[float] = None):
         '''Converts output audio to the specified format.
 
         Parameters
@@ -1490,6 +1503,8 @@ def convert(self,
             Desired number of channels. If None, defaults to the same as input.
         bitdepth : int, default=None
             Desired bitdepth. If None, defaults to the same as input.
+        bitrate : float, default=None
+            Desired bitrate. Uses Sox's -C (compression) argument.
 
         See Also
         --------
@@ -1513,6 +1528,11 @@ def convert(self,
             if not is_number(samplerate) or samplerate <= 0:
                 raise ValueError("samplerate must be a positive number.")
             self.rate(samplerate)
+        if bitrate is not None:
+            if not isinstance(bitrate, float) or bitrate <= 0:
+                raise ValueError("bitrate must be a positive float.")
+            self.output_format["bitrate"] = bitrate
+
         return self
 
     def dcshift(self, shift: float = 0.0):

tests/data/output.mp3

@@ -17,6 +17,7 @@ def relpath(f):
 INPUT_FILE4 = relpath('data/input4.wav')
 OUTPUT_FILE = relpath('data/output.wav')
 OUTPUT_FILE_ALT = relpath('data/output_alt.wav')
+OUTPUT_FILE_MP3 = relpath('data/output.mp3')
 NOISE_PROF_FILE = relpath('data/noise.prof')
 
 
@@ -371,6 +372,35 @@ def test_encoding_invalid(self):
             self.tfm.set_input_format(encoding='16-bit-signed-integer')
         with self.assertRaises(ValueError):
             self.tfm._input_format_args({'encoding': '16-bit-signed-integer'})
+    def test_bitrate(self):
+        self.tfm.set_output_format(bitrate=320.0)
+        actual = self.tfm.output_format
+        expected = {
+            'file_type': None,
+            'rate': None,
+            'bits': None,
+            'channels': None,
+            'encoding': None,
+            'bitrate': 320.0,
+            'comments': None,
+            'append_comments': True
+        }
+        self.assertEqual(expected, actual)
+
+        actual_args = self.tfm._output_format_args(self.tfm.output_format)
+        expected_args = ['-C', '320.0']
+        self.assertEqual(expected_args, actual_args)
+
+        actual_result = self.tfm.build(INPUT_FILE, OUTPUT_FILE)
+        expected_result = True
+        self.assertEqual(expected_result, actual_result)
+
+    def test_bitrate_invalid(self):
+        with self.assertRaises(ValueError):
+             self.tfm.set_output_format(bitrate='320.0')
+        with self.assertRaises(ValueError):
+             self.tfm._output_format_args({'bitrate': 320})
+
 
     def test_ignore_length(self):
         self.tfm.set_input_format(ignore_length=True)
@@ -427,6 +457,7 @@ def test_file_type(self):
             'bits': None,
             'channels': None,
             'encoding': None,
+            'bitrate': None,
             'comments': None,
             'append_comments': True
         }
@@ -449,6 +480,7 @@ def test_file_type_null_output(self):
             'bits': None,
             'channels': None,
             'encoding': None,
+            'bitrate': None,
             'comments': None,
             'append_comments': True
         }
@@ -477,6 +509,7 @@ def test_rate(self):
             'bits': None,
             'channels': None,
             'encoding': None,
+            'bitrate': None,
             'comments': None,
             'append_comments': True
         }
@@ -499,6 +532,7 @@ def test_rate_scinotation(self):
             'bits': None,
             'channels': None,
             'encoding': None,
+            'bitrate': None,
             'comments': None,
             'append_comments': True
         }
@@ -533,6 +567,7 @@ def test_bits(self):
             'bits': 32,
             'channels': None,
             'encoding': None,
+            'bitrate': None,
             'comments': None,
             'append_comments': True
         }
@@ -567,6 +602,7 @@ def test_channels(self):
             'bits': None,
             'channels': 2,
             'encoding': None,
+            'bitrate': None,
             'comments': None,
             'append_comments': True
         }
@@ -601,6 +637,7 @@ def test_encoding(self):
             'bits': None,
             'channels': None,
             'encoding': 'signed-integer',
+            'bitrate': None,
             'comments': None,
             'append_comments': True
         }
@@ -629,6 +666,7 @@ def test_comments(self):
             'bits': None,
             'channels': None,
             'encoding': None,
+            'bitrate': None,
             'comments': 'asdf',
             'append_comments': True
         }
@@ -657,6 +695,7 @@ def test_append_comments(self):
             'bits': None,
             'channels': None,
             'encoding': None,
+            'bitrate': None,
             'comments': 'asdf',
             'append_comments': False
         }
@@ -1748,6 +1787,27 @@ def test_bitdepth_invalid(self):
         with self.assertRaises(ValueError):
             tfm.convert(bitdepth=17)
 
+    def test_bitrate_valid(self):
+        tfm = new_transformer()
+        tfm.convert(bitrate=320.0)
+
+        actual = tfm.output_format
+        expected = {'bitrate': 320.0}
+        self.assertEqual(expected, actual)
+ 
+        actual_res = tfm.build(INPUT_FILE, OUTPUT_FILE_MP3)
+        expected_res = True
+        self.assertEqual(expected_res, actual_res)
+
+        tfm.set_output_format(file_type='mp3', bitrate=320.0)
+        tfm_assert_array_to_file_output(
+            INPUT_FILE, OUTPUT_FILE_MP3, tfm, skip_array_tests=True)
+
+    def test_bitrate_invalid(self):
+        tfm = new_transformer()
+        with self.assertRaises(ValueError):
+            tfm.convert(bitrate=0)
+
 
 class TestTransformerDcshift(unittest.TestCase):