This plugin uses the Kaldi-based Montreal Forced Aligner to phonetically segment audio data based on corresponding text files. It uses MaryTTS to predict the text pronunciation.

In your project directory, place the audio and text files under your build directory like this:

build
├── text
│   ├── utt0001.txt
│   ├── utt0002.txt
│   ├── utt0003.txt
│   ├── utt0004.txt
│   └── utt0005.txt
└── wav
    ├── utt0001.wav
    ├── utt0002.wav
    ├── utt0003.wav
    ├── utt0004.wav
    └── utt0005.wav

The wav files have to be downsampled to 16 kHz (we recommend using SoX for this).

If you are on a linux system, ensure that the libcblas.so.3 file is on the library search path, since it is not bundled with the Linux release of Montreal Forced Aligner. For details, see the installation notes.

We hope that this issue is resolved in a future release; until then you can install the missing library by running

sudo apt-get install libatlas3-base

on Debian-based distributions (such as Ubuntu).

Please see the instructions at https://plugins.gradle.org/plugin/de.dfki.mary.voicebuilding.marytts-kaldi-mfa

To customize the directories configured as input for the forced alignment, you can override them like this in the project's build.gradle:

convertTextToMaryXml {
    srcDir = file("some/other/text/directory")
}
prepareForcedAlignment {
    wavDir =  file("$buildDir/wav")
}

Run

./gradlew runForcedAlignment

after which the resulting TextGrid files will be under build/TextGrid.

In order to use the TextGrids for voicebuilding with MaryTTS you have to replace the sil and sp labels with _ and convert them to lab files in Xwaves format.

This can be done simply by running

./gradlew convertTextGridToXLab

Note that you can also override the label mapping:

convertTextGridToXLab {
    labelMapping = [sil: 'pau', a: 'ah'] // etc.
}

If you want to use convertTextGridToXLab alone you may want to override the default TextGrid-directory which is build/TextGrid/forcedAlignment:

convertTextGridToXLab.tgDir = file("$buildDir/TextGrid")