I modified in the speech-dispatcher configuration file this lines:
DefaultVoiceType "FEMALE2"
DefaultLanguage "hu"
DefaultModule espeak

QVector<QVoice> voices;
tts = new QTextToSpeech("speechd", parent);
voices = tts->availableVoices();
foreach(const QVoice &voice, voices)
{
qDebug() << QString("%1 - %2 - %3").arg(voice.name()).arg(QVoice::genderName(voice.gender())).arg(QVoice::ageName(voice.age())); <-- this line result "hungarian - Unknown Gender - Other Age" two time
		
	}

And call tts->say(const QString &text) method.
If I set voice with setVoice method voices[0] or voices[1] is same result, the male gender speaking.

Thank you for answer, I founded a partially solution.
The flite is used native, to using speech dispatcher required libspeechd2, libspeechd2-dev and speechd2. Recompile the qt environment the speechd plugin is available, but only speaking with male voice. The available voices name, gender and age is "hungarian - Unknown Gender - Other Age"
The spd-say is working all espeak gender and age.

You should start your application with the QT_DEBUG_PLUGINS environment variable set to 1. This will give you more clues about what is happening with the modules.