Skip to content
A vocoder framework which had been widely used in research community since 1999.
Branch: master
Clone or download
HidekiKawahara Added Japanese manual
Added a detailed technical document of legacy-STRAIGHT, originally
written in 2006.
Latest commit 9646849 Oct 18, 2018
Type Name Latest commit message Commit time
Failed to load latest commit information.
doc Added Japanese manual Oct 18, 2018
morphing_src fixed tracking error Jul 18, 2018
src fixed minor bug in synth Oct 17, 2018
.gitignore fixed tracking error Jul 18, 2018
LICENSE kick_start Jul 16, 2018 README format update: .txt -> .md Jul 20, 2018


The legacy-STRAIGHT is a collection of speech analysis, modification and resynthesis tools.


Set MATLAB path to "src" directory.

Quick start

Paste the following code to MATLAB command window. It generates copy-synthesized output to the variable "syntheszed_signal".

[x, fs] = audioread('vaiueo2d.wav'); 
f0raw = MulticueF0v14(x,fs); 
ap = exstraightAPind(x,fs,f0raw);
syntheszed_signal = exstraightsynth(f0raw,n3sgram,ap,fs);

For running this using GNU Octave, please load signal package.

pkg load signal

Release note

  • [July 19, 2018; Prerelease] The "Quick start" example also runs properly on GNU Octave 4.4.0 on macOS High Sierra (10.13.6)

  • [July 17, 2018: Prerelease] Added documemts. The first release will be on July 24, 2018.

  • [July 16, 2018: Prerelease] This release is a copy of the latest version which was distributed by the first author (Hideki Kawahara) to academic communities. The version is named STRAIGHTV40_007. The last update was July 17, 2016. Kansai TLO has also licensed the legacy-STRAIGHT for commercial use. The licensees of the legacy-STRAIGHT agreed to make the legacy-STRAIGHT open to the public after July 15, 2018.


The legacy-STRAIGHT was supported by many coauthors, contributors, and funding agencies.

Hideki Kawahara, July 16, 2018 (start date)

You can’t perform that action at this time.