- Add All zero synthesis filter.
- Add
levdur
. - Add tranposed synthesis filters (
mlsadft
,poledft
,mglsadft
,glsadft
) - Add missing high level synthesis filter class
GLSADF
.
- #63`_: Fix lpc2lsp bug, add lsp2lpc function. Add regression tests for those.
- #55: Add numpy implementation of cdist
- #54: Changes from SPTK v3.11 release. 6 and 7 pade approximatino in lmadf and mlsadf is now supported,
- BUG fix: example_audio_data is now included in the release tar.gz
- c2acr: Fix segfaults for small fftsize
- Fix mcep eps check and input length #39
- Add developer documentation (#34)
- Separate cython implementation and interface (#35)
- Add RAPT (#32)
- Add excite function (#31) @jfsantos
- Fix inconsistent docs about normalization flag for window functions
- Fix test failure in c2dps / ndps2c (#29)
- Building binary wheels for Windows using Appveyor (#28)
- Add Installation guide on windows (#25)
- Start Windows continuous integration on AppVeyor (#24). As part of the issue, binary dependency was updated so that SPTK library can be compiled on linux, osx and Windows as well.
- Remove unnecesarry array initialization (#23)
- Add
pysptk.synthesis
package that provides high level interfaces for speech waveform synthesis (#14) - Add cross-link to the docs
- Add
pysptk.conversion.mgc2b
- Add speech analysis and re-synthesis demonstration notebook (#13)
- Add
pysptk.util.example_audio_file
- Add
fftcep
(#18) - Add
mfcc
(#21) - Cython is now only required to build development versioni of pysptk. (#8)
- Include *.c to pypi distribution
- Initial release