Utilisation - VOICEVOX Core for PHP

<?php

require __DIR__ . '/vendor/autoload.php';

use Revolution\Voicevox\Core\Enums\AccelerationMode;
use Revolution\Voicevox\Core\Onnxruntime;
use Revolution\Voicevox\Core\OpenJtalk;
use Revolution\Voicevox\Core\Synthesizer;
use Revolution\Voicevox\Core\VoiceModelFile;

// Modifiez les chemins selon l'emplacement d'installation de voicevox_core
$voicevoxCoreDir = getenv('HOME') . '/.local/voicevox_core';
// La version d'ONNX Runtime recommandée est 1.23.2. Si elle ne correspond pas à la version installée, indiquez directement le nom du fichier
$onnxruntimeFilename = $voicevoxCoreDir . '/onnxruntime/lib/' . Onnxruntime::libRecommendedVersionedFilename();
$dictDir = $voicevoxCoreDir . '/dict/open_jtalk_dic_utf_8-1.11';
$vvmPath  = $voicevoxCoreDir . '/models/vvms/0.vvm';

// Texte à synthétiser et ID de style
$text    = 'この音声は、ボイスボックスを使用して、出力されています。';
$styleId = 0;
$outPath = './output.wav';

// Initialisation
$onnxruntime = Onnxruntime::loadOnce($onnxruntimeFilename);
$openJtalk   = new OpenJtalk($dictDir);
$synthesizer = new Synthesizer($onnxruntime, $openJtalk, AccelerationMode::Auto);

// Charger le modèle vocal
$model = VoiceModelFile::open($vvmPath);
$synthesizer->loadVoiceModel($model);

// Synthèse vocale
$audioQuery = $synthesizer->createAudioQuery($text, $styleId);
$wav        = $synthesizer->synthesis($audioQuery, $styleId);

file_put_contents($outPath, $wav);
echo 'Wrote ' . $outPath . PHP_EOL;

php talk.php

$wav = $synthesizer->tts($text, $styleId);
file_put_contents('./output.wav', $wav);

// Avec ttsFromKana
$wav = $synthesizer->ttsFromKana("コノオンセイワ'、ボイスボックスオ'/シヨーシテ'、シュツリョクサレテイマ'ス。", $styleId);
file_put_contents('./output.wav', $wav);

use Revolution\Voicevox\Core\UserDict;
use Revolution\Voicevox\Core\Enums\UserDictWordType;

$userDict = new UserDict();
$uuid = $userDict->addWord(
    surface: 'ボイボ',
    pronunciation: 'ボイボ',
    accentType: 0,
    wordType: UserDictWordType::ProperNoun,
    priority: 10,
);

// Sauvegarder le dictionnaire
$userDict->save('./mydict.json');

// Appliquer le dictionnaire à OpenJTalk
$openJtalk->useUserDict($userDict);

$audioQueryJson = $synthesizer->createAudioQuery($text, $styleId);

// Décoder le JSON puis ajuster
$audioQuery = json_decode($audioQueryJson, true);
$audioQuery['speedScale'] = 1.2;   // Vitesse 1,2x
$audioQuery['pitchScale'] = 0.05;  // Légère élévation de la hauteur

$wav = $synthesizer->synthesis(json_encode($audioQuery), $styleId);
file_put_contents('./output.wav', $wav);

$accentPhrasesJson = $synthesizer->createAccentPhrases($text, $styleId);
$accentPhrases = json_decode($accentPhrasesJson, true);

// Écraser la hauteur et la durée des mores avec un autre style
$updatedJson = $synthesizer->replaceMoraData($accentPhrasesJson, $styleId);

use Revolution\Voicevox\Core\Enums\AccelerationMode;

$synthesizer = new Synthesizer($onnxruntime, $openJtalk, AccelerationMode::Gpu);

// Vérifier que le mode GPU est actif
if ($synthesizer->isGpuMode()) {
    echo 'En mode GPU' . PHP_EOL;
}

$model0 = VoiceModelFile::open($voicevoxCoreDir . '/models/vvms/0.vvm');
$model1 = VoiceModelFile::open($voicevoxCoreDir . '/models/vvms/1.vvm');

$synthesizer->loadVoiceModel($model0);
$synthesizer->loadVoiceModel($model1);

// Vérifier les informations sur les locuteurs contenus dans le modèle
$metas = json_decode($synthesizer->metas(), true);

// Décharger un modèle (lorsqu'il n'est plus nécessaire)
$synthesizer->unloadVoiceModel($model0->id());

use Revolution\Voicevox\Core\Enums\OnExistingVoiceModelId;

$synthesizer->loadVoiceModel($model0, OnExistingVoiceModelId::Skip);

use Revolution\Voicevox\Core\VoicevoxCore;

$audioQuery   = $synthesizer->createAudioQuery($text, $styleId);
$audioFeature = $synthesizer->createAudioFeature($audioQuery, $styleId);

$frameRate        = (new VoicevoxCore)->getFrameRate();
$chunkFrameLength = max(1, (int) round(0.3 * $frameRate));
$totalFrames      = $audioFeature->length();

$stream = fopen('./talk.pcm', 'wb');

for ($start = 0; $start < $totalFrames; $start += $chunkFrameLength) {
    $end = min($start + $chunkFrameLength, $totalFrames);
    fwrite($stream, $audioFeature->slice($start, $end)->render());
}

fclose($stream);