Activity for Speech Signal Processing Toolkit (SPTK)
-
Takenori Yoshimura posted a comment on discussion Open Discussion
Thank you for your report. The bug seems to be fixed in the ticket below. https://sourceforge.net/p/sp-tk/bugs/94/
-
Compile error without including stdio.h : Unknown types 'FILE' and 'size_t'
-
The error was fixed in the git repository. We are very sorry for the delay.
-
Takato Fujimoto committed [3379f4] on Git
minor fix
-
wrapper for SPTK
-
pysptk is distributed
-
Voting in pitch command
-
medfilt is implemented in SPTK4
-
minor fix
-
Fix issue that da command cannot access /dev/dsp
-
The manual has been updated to mention padsp.
-
a little fix of "vc" command to reduce computation time
-
already applied
-
bug fix
-
additonal features of mlsacheck
-
Implemented
-
postfilter
-
postfilter for mel-cepstrum -> SPTK4 postfilter for LSP -> x
-
spectrum input of several commands for spectral analysis
-
SPTK4 solved the problem.
-
range of t in cdist
-
Fixed
-
rawtowav does not have -h option
-
The next major version got away from this kind of implementation. Let me close.
-
The next major version got away from this kind of implementation. Let me closed.
-
installation problem of shell scripts
-
Mac Compiler Segmentation Fault 11
-
Add new option in mfcc for ZMEANSOURCE
-
SPTK4 provides frame -z
-
Segfaults for invalid arguments like "-name=value" instead of "-name value"
-
SPTK4 solved the problem.
-
Add options of MAP estimation to gmm command
-
Implemented
-
error of glogsp/grlogsp
-
Solved
-
losing final frames
-
Thank you for your kind report and sorry for our late reply. In the new major version, SPTK4, we designed synthesis filters so that the length of excitation and that of the filtered signal are same. The above mentioned problem has been solved.
-
-o option in mgc2sp
-
SPTK4 solved the problem.
-
Add the option specifying input order to decimate and interpolate
-
SPTK4 supports input order
-
Conform output signal length to input one in mlsadf
-
SPTK4 solved the problem.
-
GMM-based voice conversion
-
Hosting on Github
-
Migrated.
-
SPTK 4.0
-
Released.
-
lsp2csm command
-
A command which converts autocorrelation to CSM was implemented instead.
-
Add command to convert PARCOR coefficients to log area ratio (LAR)
-
SPTK4 implements the function.
-
Huffman coding
-
SPTK4 implements the function.
-
Entropy calculation
-
SPTK4 implements the function.
-
minor fix
-
Thank you for your suggestion. The requests are successfully merged.
-
Fix compilation with GCC 10 and declare automake foreign
-
Giulio Paci created merge request #3
Fix compilation with GCC 10 and declare automake foreign
-
Thank you for your kind response. Your request has been merged. Sorry for our late reply. Regards, Takenori
-
Several improvements and bug fixes.
-
I also agree to transfer the copyright of my changes to the SPTK team.
-
Hi Mizuki, I have updated the PR according to your request (I assumed your comment "except gmm.c and lmadf.c" only applied to commit ed9d54. I added a new commit since I noticed that one of the error messages did not sound well, even after correction of the spelling error. Best regards, Giulio
-
Mizuki Irabu posted a comment on merge request #1
I am one of the SPTK development members. Thank you for sending us your request. We apologize for the long delay in replying. As a result of discussion among the SPTK development members, we have decided to adopt the following two commits, except for gmm.c and lmadf.c. If you send us a request for these, we would like to merge them. Then, please agree to transfer the copyright to us. https://sourceforge.net/u/giuliopaci/sp-tk/ci/9bc2e42fbdbb3a768737f442ad44c209044cb107/ https://sourceforge.net/u/giuliopaci/sp-tk/ci/ed9d54dc81fde8b79c14d02cf3e2b5d7e841e72f/...
-
remove -s 32 and -s 74 options in ds command
-
minor fix
-
Thank you for your kind report. We reviewed the filter coefficients used in ds and us commands. It was found that some of them are wrong. However, the design of new reasonable filter is not trivial problem due to various factors such as ripple, stopband attenuation, signal energy, and anti-aliasing. We decided to temporarily remove -s 32 and -s 74 options from ds command. We will investigate the suitable design for down/up-sampling.
-
Lgeu created merge request #2
fix filter coefficients in ds
-
bug fix
-
bug fix
-
bug fix
-
bug fix
-
minor fix
-
bug fix
-
bug fix
-
Several improvements and bug fixes.
-
bug fix
-
minor fix
-
modify fwrite to fwritef
-
bug fix
-
bug fix
-
bug fix
-
Revert "bug fix"
-
bug fix
-
Andrew Lee posted a comment on discussion Help
Hello maintainers, I am trying to install the executables on my macOS machine, but the autoconf script (./configure) is telling me “configure: error: cannot run C compiled programs.” I have also tried setting the CC and CXX environment variables in the configure script "i.e. adding the lines export CC=/usr/local/bin/gcc-8 and export CXX=/usr/local/bin/g++-8 at the beginning of the script (which is the GNU GCC compiler installed via homebrew). Please advise as to how I can build this for macOS. -Andrew...
-
Keiichi Tokuda created ticket #114
Entropy calculation
-
Huffman coding
-
Keiichiro Oura modified ticket #95
Possible bug: lpc2lsp
-
Thank you very much for your report! It was fixed. https://sourceforge.net/p/sp-tk/SPTK/ci/b7d3b2a5f27496fe4a88ba9c09f74df98fdc0ec3/
-
Output an error message
-
Ryuichi Yamamoto created ticket #95
Possible bug: lpc2lsp
-
bug fix
-
bug fix
-
Blaine Ayotte posted a comment on discussion Help
I was able to install and run the program but I cannot view any of the outputs. For example I am trying to calculate fundamental frequency. When using commands i can create a .pitch file but what can I do with that? Thanks!
-
fujishita taichi committed [bd2353]
remove the id used in cvs
-
fix bugs of pitch_mark command
-
add the 6th- and 7th-order of modified Pade approximation calculated by Cassia Valentini (The University of Edinburgh)
-
Hi, Thank you for using SPTK. I investigated this issue but as far as I confirmed, there was no problems. I think that the reason for the issue is the designation error of a gamma value. Please check whether the gamma value of mgcep command and that of mgc2sp command are the same. Regards, Taichi Fujishita
-
Nisha G posted a comment on discussion Help
Hi, When I tried to compute the spectrum from mgc with gamma in the range -0.9 or -0.8, the spectrum gets screwed up, unlike that for any other value for gamma. I was wondering if this is an issue with the SPTK implementation. Any help in this regard would be much appreciated. Thanks, Nisha G
1 >
