Dev. sneak peek

[0131]

To get some new stuff up here while im working on the programm, some sneak peeks from my development (Will add some more from time to time)

Talk with ctitan | +

introduction.PNG (62.2 KiB) Viewed 3880 times

She will guide you, but don´t make any mistakes she remembers. | +

profile.PNG (30.68 KiB) Viewed 3877 times

[Cian]

Omg this looks amazing can't wait

[0131]

Currently working on the deckmanager.

deckmanager.png (99.1 KiB) Viewed 3791 times

[Simon]

I don't want to bother you 0131, but "Configurate" is not a real word... In case you didn't do it on purpose it should just be 'configure' (probably). Love your work though!

[0131]

Ok fixed it, thx!

[Inside7shadows]

Apparently it was obsolete in 1913, but still a word!
Configurate

[Simon]

I'm sorry for mistakenly correcting your 20th century english, you see English isn't my native language either, so I'm not a pro at the old versions . (Pro at (old) Dutch, though!)

[0131]

Currently trying to work with sphinx4 speech to text, using live recording.
Should be a nice feature to add voice commands to the viewer, because most time hands have better things to do while teasing.
Here is my first result.

sphinxtest1.png (23.26 KiB) Viewed 3660 times

But i have to see how reliable this is...


There are also other possible applications

[Inside7shadows]

Words like Configurate are generally accepted informally, mostly due to a recent president who was known for making up strange words.

Currently trying to work with sphinx4 speech to text, using live recording.

Speech to text is bananas. You are bananas. The length and effort you go for your perversion is astounding. That being said, I'm constantly fucking up with that card every time it comes up.

[0131]

I am also trying to add text to speach (TTS) after i read about it in the todo-list of another viewer.
After messing around with Mary and then implementing FreeTTS with mbrola voices (a slightly better voice) i was very depressed , the quality is fucking horrible and it takes a few seconds to generate the voice-output.

There are many realy good TTS voices/systems out there but they are expensive as fuck and hard to combine with java.

good_example.mp3

(77.65 KiB) Downloaded 131 times

bad_example.wav

(387.88 KiB) Downloaded 108 times

I made some example files if you want to listen, maybe tell me your thoughts after listening.
- Would you like such an feature? (There are many cases this could be handy for example if blindfolded)
- How do you think about the quality does it matter to you?

After many mindstorming these are the options:
- No Text to Speech
- FreeTTS (very bad quality and performance)
- My favorite:
There are some good sites (especially one) out there which offer an (as far as i think) limitless free online demo at their webpage to sell their produkt.
We can use that!
Either generate a voice file for each card, which is much work but also the best way. (Problem a card text can have many different versions with different values)
OR i could try to cheat a little bit and hook up with their free demo send some http requests and snatch the voicefile from their server. (Would require an online connection while teasing and the delay would also be unpredictable)

Anyway if you have suggestions or if you have some knowlege about TTS let me know

//Update: Bad example added.