pool/python-librosa
SHA256
2
0
Fork 1
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
factory
python-librosa/python-librosa.changes

398 lines
21 KiB
Plaintext
Raw Permalink Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

-------------------------------------------------------------------
Wed Oct 2 20:59:35 UTC 2024 - Matej Cepl <mcepl@cepl.eu>
- Add csr_matrix-attr-H.patch making the package compatible with
the most recent version of SciPy (gh#librosa/librosa#1849).
- Add sysinfo.py to provide upstream with the required debugging
information.
- Add mark-network-tests.patch marking tests which require
network connection to run.
-------------------------------------------------------------------
Fri Sep 20 22:12:01 UTC 2024 - Matej Cepl <mcepl@cepl.eu>
- Update to 0.10.2.post1:
- Bump minimum matplotlib to 3.5.0
- Update to 0.10.2.post1:
- Multiple accidentals by @erabinov in #1739
- Avoid deprecated matplotlib functions by @bmcfee in #1755
- Mode support by @erabinov in #1756
- Mode support by @erabinov in #1762
- Switch from pkg_resources to importlib.resources by @mattpitkin in #1765
- Add python 3.11 to build matrix by @bmcfee in #1769
- Natural accidentals for out-of-key notes by @erabinov in #1770
- Fix docstring for the writeable argument of util.frame function by @alvingao in #1780
- Update test_util.py to remove problematic "dummy". by @dpwe in #1783
- Updates matplotlib colormap usage to remove deprecation warnings. by @psmskelton in #1782
- Better handling of warnings in test suite by @bmcfee in #1784
- Fix sample rate type annotations by @bmcfee in #1789
- Explicitly allow scalars in decibel conversion by @bmcfee in #1790
- Adding python 3.12 to build matrix, update tests for recent scipy / numpy by @bmcfee in #1808
- Resolve the performance issue of autocorrelate by @alumkal in #1813
- Correct phase advance for odd frame lengths in phase_vocoder by @bmcfee in #1814
- Multichannel peak, onset, and beat detection by @bmcfee in #1766
- Update actions for 2024 by @bmcfee in #1820
- doc copybutton, CI updates by @bmcfee in #1821
- exposed tempo min and max parameters in specshow by @bmcfee in #1822
- remove kwargs from effects.hpss and friends, passthru stft by @bmcfee in #1826
- Docstring and comment improvements by @bmcfee in #1827
- librosa.cite() by @bmcfee in #1829
- 0.10.2 release prep by @bmcfee in #1832
-------------------------------------------------------------------
Wed Apr 17 03:42:24 UTC 2024 - Steve Kowalik <steven.kowalik@suse.com>
- Skip a misbehaving test that can cause segfaults.
- Drop back to ffmpeg 5.
-------------------------------------------------------------------
Sun Sep 24 17:38:22 UTC 2023 - Ben Greiner <code@bnavigator.de>
- Remove pooch upper pin, upstream removed it already for last
version (gh#librosa/librosa#1727)
-------------------------------------------------------------------
Sun Aug 20 16:24:41 UTC 2023 - Ben Greiner <code@bnavigator.de>
- Update to 0.10.1
* VQT octave boundaries by @bmcfee in #1677
* redoing 0.10.0.post1 release by @bmcfee in #1684
* Pin pooch < 1.7 by @bmcfee in #1686
* Post-release 2 for 0.10.0, pin pooch by @bmcfee in #1687
* Stricter checks on yin/pyin parameters by @bmcfee in #1693
* bumping sphinx config, fixes #1671 by @bmcfee in #1699
* fixed bins_per_octave's docs type float to int by @kyaryunha in
#1703
* add volume adjustment(rms-mean) at pitch_shift by @kyaryunha in
#1704
* Several small documentation updates by @bmcfee in #1713
* pin numpy version to < 1.24 in doc environment by @bmcfee in
#1714
* fixed broken 'latest' link generator by @bmcfee in #1716
* README.md typo fix by @Zenitismus in #1706
* CI for documentation builds by @bmcfee in #1718
* reworked lazy loading of matplotlib by @bmcfee in #1722
* Advanced example for patch generation by @bmcfee in #1719
* support pooch >= 1.7 by @bmcfee in #1727
* Consistent conversion of time duration/offset to sample counts
by @bmcfee in #1726
* Update tests for scipy 1.11 by @bmcfee in #1731
-------------------------------------------------------------------
Mon May 29 15:19:42 UTC 2023 - Ben Greiner <code@bnavigator.de>
- Update to 0.10.0.post2
* This maintenance release updates the dependency specifications
to avoid an incompatibility with pooch >= 1.7.0. This
restriction only affects loading of specific example files, and
does not otherwise affect functionality of the library.
- Releae 0.10.0
* This release provides many speedups and enhancements to core
operations, faster importing and loading, and full type
annotations.
## New features
* #1485 Added support for “h” (hours) and “m” (minutes) in
librosa.display.TimeFormatter. Vincent Lostanlen
* #1495 Expanded options for bandwidth calculation in
librosa.segment.recurrence_matrix and
librosa.segment.cross_similarity. Tom Xi
* #1529 Accelerated utility functions for squared magnitude
librosa.util.abs2 and librosa.util.phasor. Brian McFee
* #1514 librosa.stft support for pre-allocated output arrays.
Brian McFee
* #1533 Accelerated librosa.util.localmax and
librosa.util.localmin. Brian McFee
* #1552 EXPERIMENTAL Support for just-intonation based variable-Q
transforms librosa.vqt and functional just system (FJS)
notation. Brian McFee, Vincent Lostanlen
* #1579 Imports are now done lazily, significantly reducing
startup time. It is not longer necessary to explicitly import
librosa.display. Brian McFee
* #1587, #1632, #1637 Type annotations. Matan Gover, Brian McFee,
Fabian Keller
* #1647 Harmonic spectrum interpolation librosa.f0_harmonics and
tempogram ratio features librosa.feature.tempogram_ratio. Brian
McFee
## Bug fixes
* #1560 Corrected an error in bandwidth estimation for window
functions that take on negative sample values. Brian McFee
* #1571 Fixed rounding error in multirate filterbank calculation
of sampling rates. Brian McFee
* #1594 Fixed incorrect use of deprecated numpy numerical types.
Mattia Di Gangi
* #1596 Fixed harmonic interpolation for recent scipy releases.
Brian McFee
## API changes
* #1561 librosa.resample now supports axis= for all sample rate
conversion backends. Brian McFee
* #1572 Notation conversion (e.g., librosa.midi_to_note) now
vectorizes operations, and produces numpy.ndarray outputs
rather than list when given iterable inputs. Brian McFee
* #1625 Deprecations now raise FutureWarning instead of
DeprecationWarning.
* #1625 librosa.get_duration parameter filename has been renamed
to path. Brian McFee
* #1643 Removed support for threshold=None in
librosa.zero_crossings. Brian McFee
* #1665 Fixed overflow bug in librosa.feature.rms on integer
inputs. Brian McFee
## Deprecations
* #1521 Removed backward compatibility support for positional
arguments that are now keyword-only. Brian McFee
* #1577 Removed official support for python 3.6. Brian McFee
* #1625 The audioread backend is now deprecated and will be
removed in a future release. Brian McFee
## Other changes
* #1501 Infer square / equal plot dimensions for compatible x/y
axes in librosa.display.specshow. Kian Eliasi
* #1535 Standardized integer type checks throughout the library.
Alexandre Chabot-Leclerc
* #1539 Refactored librosa.display.AdaptiveWaveplot to allow
callback disconnection. Brian McFee
* #1548 Removed deprecated ‘hanning’ window specification from
tests. Brian McFee
* #1551 Removed dependency on contextlib2. Steve Kowalik
* #1565 Efficiency improvements to librosa.autocorrelate. Brian
McFee
* #1566 Changed the default sample rate conversion method from
kaiser_best (resampy) to soxr_hq (soxr). Brian McFee
* #1581 Updated tests for numpy 1.23 compatibility. Brian McFee
* #1617 Accelerated librosa.yin implementation. Brian McFee
* #1640 Updated display module dependencies to require matplotlib
>= 3.3. Brian McFee
* #1643 Accelerated librosa.zero_crossings implementation. Brian
McFee
* #1646 Added support for transposed (vertical) display of
waveforms. Brian McFee
- Drop patches
* remove-contextlib2.patch
* skip-test-data-missing-tests.patch
* update-tests-for-numpy-123.patch
* remove-hanning-from-tests.patch
* update-for-numpy-124.patch
-------------------------------------------------------------------
Mon May 29 13:45:26 UTC 2023 - Dirk Müller <dmueller@suse.com>
- reenable python311 build (numba now available)
-------------------------------------------------------------------
Tue Apr 4 09:59:00 UTC 2023 - Daniel Garcia <daniel.garcia@suse.com>
- skip python311, python-numba is not available for python 3.11 yet.
-------------------------------------------------------------------
Sat Jan 28 17:24:09 UTC 2023 - Ben Greiner <code@bnavigator.de>
- Add update-for-numpy-124.patch
* Replace types from expired deprecation by NumPy 1.24
* Upstream fixed it within a bigger set of changes in
gh#librosa/librosa#1587 and gh#librosa/librosa#1632
- Move to PEP518 build
- Improve test time by testing parallel with pytest-xdist
-------------------------------------------------------------------
Fri Oct 21 06:29:51 UTC 2022 - Daniel Garcia <daniel.garcia@suse.com>
- Add update-tests-for-numpy-123.patch gh#librosa/librosa#1581
- Add remove-hanning-from-tests.patch gh#librosa/librosa#1548
-------------------------------------------------------------------
Mon Sep 26 02:21:59 UTC 2022 - Steve Kowalik <steven.kowalik@suse.com>
- Update to 0.9.2:
* Fixed librosa.show_versions to match package dependencies.
* Fixed a bug in librosa.effects.split when applied to multichannel data.
* Removed test data from main repository and reduced the git attic.
* Accept pre-constructed audioread objects in librosa.load.
* Accelerate librosa.magphase.
* Nearly full support for multi-channel processing.
* Option to disable unicode characters in display functions.
* Significantly expanded the library of example audio clips.
* The default padding mode for most functions (including STFT) is now
zero-padding.
* librosa.load and librosa.stream can now operate directly on open
soundfile objects.
* librosa.display.specshow now uses centered coordinate grids.
* librosa.iirt now exposes API control over resampling modes.
* Maximum frequency is now correctly inferred as Nyquist in onset strength
calculation.
* librosa.effects.deemphasis no longer modifies the input signal in-place.
* librosa.util.frame now correctly works for arbitrary memory layouts and
numbers of axes.
* Corrected a normalization error in inverse CQT.
* ibrosa.cqt now supports arbitrary hop lengths.
* Added a run-time check for minimally supported matplotlib versions.
* Enhanced continuous integration testing for oldest and newest environments.
* librosa.effects.deemphasis, inverse operation of
librosa.effects.preemphasis.
* librosa.display.waveshow, adaptively visualize waveforms by amplitude
envelope when zoomed out, or raw sample values when zoomed in.
- Add patch remove-contextlib2.patch:
* No longer require contextlib2.
- Add patch skip-test-data-missing-tests.patch:
* Due to removing the test data from the main repo, skip tests that require
it, since it's quite large.
-------------------------------------------------------------------
Thu Jan 28 23:09:45 UTC 2021 - Ben Greiner <code@bnavigator.de>
- Disable python36 build because SciPy 1.6.0 dropped support for
Python 3.6. Other packages from the NumPy family do the same
(NEP29)
- Update requirements
- Enable tests. Use librosa-create-pooch-cache.py to provide the
test data in librosa-pooch-cache.tar.gz. These files are under
CC-BY-3.0 license.
-------------------------------------------------------------------
Fri Jan 8 14:29:45 UTC 2021 - Dilawar Singh <dilawar.s.rajput@gmail.com>
- Update to version 0.8.0
* Added pitch tracking (yin and pyin)
* Variable-Q transform
* Hindustani and Carnatic notation support
* Expanded collection of example tracks
* Numerous speedups and bugfixes
* Removed support for Python 3.5 and earlier.
-------------------------------------------------------------------
Thu Jan 30 15:57:15 UTC 2020 - Todd R <toddrme2178@gmail.com>
- Update to version 0.7.2
+ New Features
* added liftering support to MFCC inversion. *Ayoub Malek*
* added `pathlib` support to `core.load`. *Stef van der Struijk*
* added "mu-law" compression and expansion (companding). *Brian McFee*
+ Bug fixes
* fixed errors in dynamic time warping with `NaN` inputs and detection
of invalid warping paths. *Frank Zalkow, Brian McFee*
* fixed frame-length normalization in spectrogram-based RMS energy
calculation. *Shunsuke Hidaka*
* fixed the length of signals produced by `core.tone`. *Brian McFee*
+ Documentation
* removed math formatting from documentation. *Vincent Lostanlen*
* corrected documentation for DTW. *Eric Semeniuc*
* corrected typos in time-frequency reassignment spectrogram
documentation. *Scott Seyfarth*
* removed `dct_type=None` as an option in MFCC documentation. *Brian McFee*
+ Other changes
* expanded normalization modes in Mel basis construction. *Brian McFee*
-------------------------------------------------------------------
Fri Dec 6 18:44:56 UTC 2019 - Todd R <toddrme2178@gmail.com>
- Disable python2 support since dependencies dropped python2
-------------------------------------------------------------------
Thu Nov 21 04:36:07 UTC 2019 - Todd R <toddrme2178@gmail.com>
- Use github source
-------------------------------------------------------------------
Tue Nov 19 20:52:02 UTC 2019 - Todd R <toddrme2178@gmail.com>
- Update to version 0.7.1
+ New Features
* `util.frame` now supports multidimensional data. Includes a new helper function `util.stack` for contiguous concatenation.
* `core.griffinlim_cqt`: Phase retrieval from constant-Q magnitude spectra.
* Enhanced compatibility with HTK's MFCC implementation: `effects.preemphasis` and `lifter=` parameter in MFCC.
* `util.shear` utility for shear-transforming 2D arrays.
* `core.reassigned_spectrogram`: time-frequency reassigned spectrogram.
+ Bug fixes
* Added a missing parameter to `griffinlim_cqt`.
* Correct FFT normalization discrepancy in rms calculation.
* Corrected an error in automatic tuning correction for CQT.
* Fixed seeking behavior in `librosa.stream` when operating on file-handle inputs.
* Fixed a boundary condition check in full-sequence DTW.
+ Documentation
* Fixed a typo in `stream` documentation.
* Visualization of reassigned spectrograms.
* Improved documentation for `core.stft`.
* Expanded documentation of `feature.melspectrogram`.
+ Other changes
* Expose frame parameters in `onset` and `chroma_stft` functions.
* Removed warning filter reset, and changed the Python2 deprecation
warning to class `FutureWarning`.
* Fixed an upstream deprecation warning from `np.asscalar()`.
* Beat and tempo estimation now support prior distributions.
* `util.valid_audio` now enforces memory contiguity.
* Improved x-axis decoration types in `display.waveplot`.
* Ensure memory contiguity of audio buffers after resampling.
* Code-base audit for proper usage of `times_like`.
* Deprecated `core.ifgram` in favor of `reassigned_spectrogram`.
* Allow initial phase estimates for `griffinlim` methods.
* Accelerated `segment.lag_to_recurrence` and `segment.recurrence_to_lag`. Deprecated `util.roll_sparse`.
* `A440=` parameter has been deprecated across the library in favor of a standardized `tuning=` parameter.
-------------------------------------------------------------------
Thu Jul 25 20:34:49 UTC 2019 - Todd R <toddrme2178@gmail.com>
- Update to version 0.7.0
+ New features
* `librosa.core.stream`: Stream generator to process long audio files into smaller pieces. *Brian McFee*
* `librosa.core.load`: Replaced the default audio decoder with `pysoundfile`, and only use `audioread` as backup. *Brian McFee*
* `librosa.core.griffinlim`: Phase retrieval from magnitude spectrograms using the (accelerated) Griffin-Lim method. *Brian McFee*
* `librosa.feature.inverse`: New module for feature inversion, based on the Griffin-Lim phase retrieval algorithm. Includes `mel_to_audio` and `mfcc_to_audio`. *Brian McFee*
* `librosa.core.lpc`: Linear prediction coefficients (LPC). *Adam Weiss*
* `librosa.sequence.rqa`: Recurrence Quantification Analysis (RQA) for sequence alignment. *Brian McFee*
* `librosa.beat.plp`: Predominant local pulse (PLP) for variable-tempo beat tracking. *Brian McFee*
* `librosa.feature.fourier_tempogram`: Fourier Tempogram for representing rhythm in the frequency domain. *Brian McFee*
* `librosa.core.pcen` Per-channel energy normalization (PCEN) now allows logarithmic range compression at the limit power->0. *Vincent Lostanlen*
* `librosa.effects.pitch_shift` supports custom resampling modes. *Taewoon Kim*
* `librosa.core.cqt` and `librosa.core.icqt` Forward and inverse constant-Q transform now support custom resampling modes. *Brian McFee*
* `librosa.segment.path_enhance`: Near-diagonal path enhancement for recurrence, self- or cross-similarity matrices. *Brian McFee*
* `librosa.segment.recurrence_matrix` now supports a keyword argument, `self=False`. If set to `True`, the recurrence matrix includes self-loops. *Brian McFee*
* `librosa.core.piptrack` now supports a keyword argument, `ref=None`, allowing users to override the reference thresholding behavior for determining which bins correspond to pitches. *Brian McFee*
* `librosa.segment.cross_similarity`: Cross-similarity function for comparing two feature sequences. *Rachel Bittner, Brian McFee*
* `librosa.onset.onset_strength_multi` now supports a user-specified reference spectrum via the `ref` keyword argument. *Brian McFee*
* `librosa.core.resample` now supports `mode='polyphase'`. *Brian McFee*
* `librosa.onset.onset_strength_multi`: Setting `aggregate=False` disables the aggregation of onset strengths across frequency bins. *Brian McFee*
+ Bug fixes
* `librosa.effects.pitch_shift` now preserves length. *Vincent Lostanlen*
* `librosa.core.pcen` Dynamic range compression in PCEN is more numerically stable for small values of the exponent. *Vincent Lostanlen*
* `librosa.core.ifgram` Instantaneous frequency spectrogram now correctly estimates center frequencies when using windows other than `hann`. *Brian McFee*
* `librosa.sequence.dtw` Fixed a bug in dynamic time warping when `subseq=True`. *Viktor Andreevitch Morozov*
* `librosa.core.pcen` now initializes its autoregressive filtering in the steady state, not with silence. *Jan Schlüter, Brian McFee*
* `librosa.segment.recurrence_matrix`: `width` parameter now cannot exceed data length. *Brian McFee*
* Filter bank constructors `mel`, `chroma`, `constant_q`, and `cq_to_chroma` are now type-stable. *Vincent Lostanlen, Brian McFee*
* `librosa.core.icqt` Inverse constant-Q transform has been completely rewritten and is more numerically stable. *Brian McFee*
+ Removed features (deprecated in v0.6)
* Discrete cosine transform. We recommend using `scipy.fftpack.dct`
* The `delta` function no longer support the `trim` keyword argument.
* Root mean square error (`rmse`) has been renamed to `rms`.
* `iirt` now uses `sos` mode by default.
+ Documentation
* Improved the documentation of PCEN. *Vincent Lostanlen*
* Improved installation documentation. *Darío Hereñú*
* Improved code style for plot generation. *Alex Metsai*
* Improved the documentation of spectral features. *Brian McFee*
* Improved the documentation of MFCC. *Brian McFee*
* Removed a redundant link in the `util` documentation. *Keunwoo Choi*
* Improved the docstring of `recurrence_matrix`. *Brian McFee*
* Improved the docstring of `load`. *Andy Sarroff*
+ Other changes
* The `output` module is now deprecated, and will be removed in version 0.8.
* More informative exception handling. *Jack Mason*
* `librosa.core.resample()` now supports `mode='fft'`, equivalent to the previous `scipy` mode. *Brian McFee*
* More efficient length-aware ISTFT and ICQT. *Vincent Lostanlen*
* Nine librosa functions now store jit-compiled, numba-accelerated caches across sessions. *Brian McFee*
* `librosa.core.load` no longer relies on `realpath()`. *Brian McFee*
* All spectral feature extractors now expose all STFT parameters. *Brian McFee*
* Refactored `librosa.cache`. *Brian McFee*
* Thanks to `np.fft.rfft`, functions `stft`, `istft`, `ifgram`, and `fmt` are faster and have a reduced memory footprint. *Brian McFee*
-------------------------------------------------------------------
Wed Feb 20 19:30:09 CET 2019 - Matej Cepl <mcepl@suse.com>
- Upgrade to 0.6.3:
- This release contains a few minor bugfixes and many
improvements to documentation and usability.
- Also, SPEC file is run through spec-cleaner.
-------------------------------------------------------------------
Fri Nov 2 04:48:42 UTC 2018 - Todd R <toddrme2178@gmail.com>
- Initial version
Reference in New Issue View Git Blame Copy Permalink