py_responsivevoice icon indicating copy to clipboard operation
py_responsivevoice copied to clipboard

German Male not correct working

Open AnnikenYT opened this issue 5 years ago • 1 comments

Hey, more a question than an issue. I used this packet for a bit now, and I really love it. But today, i ran into an issue. I created a new file with this code:

engine = ResponsiveVoice(lang=ResponsiveVoice.GERMAN)

def talkToMe(audio):
    try:
        os.remove("audio.mp3")
    except:
        print("no file found")
    print(audio)
    engine.say(audio, mp3_file="audio", gender=ResponsiveVoice.MALE)

talk("hallo")

As far as I'm concerned, it should say the content of audio as a German Male. But it says it as a German Female. Does anyone know why?

Thanks in advance!

PS: That's the working script:

engine = ResponsiveVoice(lang=ResponsiveVoice.GERMAN)

def talkToMe(audio):
    try:
        os.remove("audio.mp3")
    except:
        print("no file found")
    print(audio)
    engine.say(audio, mp3_file="audio", gender=ResponsiveVoice.MALE)

I don't see a difference, but maby I'm blind c;

AnnikenYT avatar May 16 '20 10:05 AnnikenYT

voices have been scrapped from the website, see data dump here

Something seems to have changed, i have noticed sometimes i get randomly female or male responses, i believe responsive voice is caching some utterances (which respect selected voice) but deprecated some of them (which reverts back to a lang default if not cached)

I also noticed using en-us vs en for example gives different results, might be worth experimenting with lang code variations

JarbasAl avatar Jun 22 '20 16:06 JarbasAl