ModuleNotFoundError: No module named 'aifc'. ModuleNotFoundError: No module named 'sunau'

[hunterhogan]

In Short

• Python 3.13 does not include aifc or sunau
• Librosa uses imports both
• Immediate solution: pip install standard-aifc standard-sunau
• Mid-term solution: add standard-aifc standard-sunau as conditional dependencies for Python 3.13
• Long-term solution: Librosa addresses the issue

Error

Loading tracks info from provider moisesdb_v0.1: 100%|███████████████████████████████████████████████████████████| 240/240 [00:00<00:00, 5311.57it/s]
Traceback (most recent call last):
  File "c:\apps\.py", line 73, in <module>
    , moisesDB[index].audio, amplitudeThreshold): index
      ^^^^^^^^^^^^^^^^^^^^^
  File "c:\apps\.venv\Lib\site-packages\moisesdb\track.py", line 135, in audio
    stems = self.stems
            ^^^^^^^^^^
  File "c:\apps\.venv\Lib\site-packages\moisesdb\track.py", line 128, in stems
    mix = self.stem_mixture(stem=stem)
  File "c:\apps\.venv\Lib\site-packages\moisesdb\track.py", line 103, in stem_mixture
    sources_mixture = self.stem_sources_mixture(stem=stem)
  File "c:\apps\.venv\Lib\site-packages\moisesdb\track.py", line 90, in stem_sources_mixture
    stem_sources = self.stem_sources(stem=stem)
  File "c:\apps\.venv\Lib\site-packages\moisesdb\track.py", line 80, in stem_sources
    source_audios = [load_audio(p) for p in paths]
                     ~~~~~~~~~~^^^
  File "c:\apps\.venv\Lib\site-packages\moisesdb\utils.py", line 18, in load_audio
    samples = list(librosa.load(f, sr=None, mono=False, **kwargs))
                   ~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "c:\apps\.venv\Lib\site-packages\librosa\core\audio.py", line 170, in load
    if isinstance(path, tuple(audioread.available_backends())):
                              ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
  File "c:\apps\.venv\Lib\site-packages\audioread\__init__.py", line 80, in available_backends
    from . import rawread
  File "c:\apps\.venv\Lib\site-packages\audioread\rawread.py", line 16, in <module>
    import aifc
ModuleNotFoundError: No module named 'aifc'

(.venv) C:\apps>pip install standard-aifc
Collecting standard-aifc
  Downloading standard_aifc-3.13.0-py3-none-any.whl.metadata (969 bytes)
Collecting standard-chunk (from standard-aifc)
  Downloading standard_chunk-3.13.0-py3-none-any.whl.metadata (860 bytes)
Collecting audioop-lts (from standard-aifc)
  Downloading audioop_lts-0.2.1-cp313-abi3-win_amd64.whl.metadata (1.7 kB)
Downloading standard_aifc-3.13.0-py3-none-any.whl (10 kB)
Downloading audioop_lts-0.2.1-cp313-abi3-win_amd64.whl (30 kB)
Downloading standard_chunk-3.13.0-py3-none-any.whl (4.9 kB)
Installing collected packages: standard-chunk, audioop-lts, standard-aifc
Successfully installed audioop-lts-0.2.1 standard-aifc-3.13.0 standard-chunk-3.13.0

...
Loading tracks info from provider moisesdb_v0.1: 100%|███████████████████████████████████████████████████████████| 240/240 [00:00<00:00, 3626.81it/s]
Traceback (most recent call last):
  File "c:\apps\.py", line 73, in <module>
    get, moisesDB[index].audio, amplitudeThreshold): index
        ^^^^^^^^^^^^^^^^^^^^^
  File "c:\apps\.venv\Lib\site-packages\moisesdb\track.py", line 135, in audio
    stems = self.stems
            ^^^^^^^^^^
  File "c:\apps\.venv\Lib\site-packages\moisesdb\track.py", line 128, in stems
    mix = self.stem_mixture(stem=stem)
  File "c:\apps\.venv\Lib\site-packages\moisesdb\track.py", line 103, in stem_mixture
    sources_mixture = self.stem_sources_mixture(stem=stem)
  File "c:\apps\.venv\Lib\site-packages\moisesdb\track.py", line 90, in stem_sources_mixture
    stem_sources = self.stem_sources(stem=stem)
  File "c:\apps\.venv\Lib\site-packages\moisesdb\track.py", line 80, in stem_sources
    source_audios = [load_audio(p) for p in paths]
                     ~~~~~~~~~~^^^
  File "c:\apps\.venv\Lib\site-packages\moisesdb\utils.py", line 18, in load_audio
    samples = list(librosa.load(f, sr=None, mono=False, **kwargs))
                   ~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "c:\apps\.venv\Lib\site-packages\librosa\core\audio.py", line 170, in load
    if isinstance(path, tuple(audioread.available_backends())):
                              ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
  File "c:\apps\.venv\Lib\site-packages\audioread\__init__.py", line 80, in available_backends
    from . import rawread
  File "c:\apps\.venv\Lib\site-packages\audioread\rawread.py", line 19, in <module>
    import sunau
ModuleNotFoundError: No module named 'sunau'

(.venv) C:\apps>pip install standard-sunau

Environment

>py -V
Python 3.13.2

pip show librosa
Name: librosa
Version: 0.10.2.post1
Summary: Python module for audio and music processing
Home-page: https://librosa.org
Author: Brian McFee, librosa development team
Author-email: brian.mcfee@nyu.edu
License: ISC
Location: C:\apps\bandSplitDataDensity\.venv\Lib\site-packages
Requires: audioread, decorator, joblib, lazy-loader, msgpack, numba, numpy, pooch, scikit-learn, scipy, soundfile, soxr, typing-extensions
Required-by: analyzeAudio, moises-db

pip show moises-db
Name: moises-db
Version: 0.0.3
Summary: moises-db
Home-page: https://github.com/moises-ai/moises-db
Author: Igor Pereira
Author-email: igor@moises.ai
License: MIT
Location: C:\apps\bandSplitDataDensity\.venv\Lib\site-packages
Requires: fsspec, gcsfs, librosa, pandas, scipy, tqdm
Required-by: 

[GusionEz]

First Error: ModuleNotFoundError: No module named 'aifc'
Cause:

The librosa library relies on audioread, which requires the aifc module.
However, Python 3.13 removed aifc from the standard library.
As a result, it is no longer available by default, leading to this error.
Solution:

You correctly tried to fix this by running pip install standard-aifc. That was the right approach.
Second Error: ModuleNotFoundError: No module named 'sunau'
Cause:

Just like aifc, the sunau module was also removed from the standard library in Python 3.13.
audioread, and thus librosa, requires this module when handling certain audio formats.
Solution:

You should install it using:

pip install standard-sunau
(It seems you haven't done this yet.)
About librosa and moises-db
librosa is an audio analysis library that depends on audioread to load sound files.
moises-db is a dataset module that uses librosa for processing.
Since librosa uses audioread, it indirectly depends on aifc and sunau.
Solution Steps
Install missing modules:

pip install standard-aifc standard-sunau
Restart your virtual environment:

deactivate
Then reactivate it:

source .venv/bin/activate # Mac/Linux
..venv\Scripts\activate # Windows
If you're using Python 3.13, consider downgrading to 3.11 or 3.12, as some libraries might not yet fully support 3.13.

If the issue persists, reinstall audioread and librosa:

pip install --force-reinstall audioread librosa
Try these steps and let me know if the issue persists!

[hunterhogan]

Thank you very much for such a clear reply! I had already resolved the issue for myself when I posted, the last line of the error is (.venv) C:\apps>pip install standard-sunau, which as you speculated, resolved the problem for me. I wanted to let you guys know about the cascaded effects deprecated modules. I use librosa in one of my packages, but I hadn't encountered the issue yet, so I figured y'all would also be unaware of it.

On a different note for which I would not open an issue, I typed ~60% of moises-db in a stub file, which you can certainly use (and expand 😄 ). https://github.com/hunterhogan/stubFileNotFound