Skip to content

How to use it! (viewer)

matt edited this page Apr 30, 2022 · 17 revisions

How to use the AI TTS

From a donator's prospective.

AI TTS uses the uberduck.ai API, which basically means there are a TON of voices you can use with this bot.

Syntax:

<voice>: <message>

(Replace with the voice, and with your message.)

Example:

drake: Hello the Twitch Chat on Twitch.tv

Voices:

Since we are using uberduck, there are thousands upon thousands of voices to choose, while some of them do work (and work very well), some of them don't, so you can test them out on the uberdank website before you donate :)

  1. Head on over to uberduck.ai
  2. Choose a voice! (There is a drop down for a category and the voice itself)
  3. After testing the voice in the browser, the URL should say something along the lines of https://app.uberduck.ai/#voice=<voice here>
  4. Use that voice code (for example kanye-west-rap or drake) in your TTS message.

Note: Include () if a voice ID includes that.

New Features!

Multiple Voices

As of 2.0.0, multiple voices are now available! All you have to do is separate each voice by doing ||. For example:

ninja: hello everyone || spongebob: me next, me next! || kanye-west-rap: ratio

Playsounds!

As of version 2.1.0+, playsounds are now available in the AI TTS Donations program! You separate voices and playsounds still by using ||! Right now, all of the playsounds are from the first 99 playsounds (if your streamer is running version 2.3.0+ it supports ALL of the playsounds!) here.

Example:

ninja: hey guys! i'm ninja real ninja. || (1) || ninja: that's no good! || spongebob: sonic!

Voice Effects!

As of version 2.4.0+, voice effects are now available for the AI TTS! Here is how they work:

The current voice effects are:

reverb: Adds reverb to the TTS
pitchup: Pitches the TTS up
pitchdown: Pitches the TTS down
loud: Makes the TTS loud
android: Makes the TTS sound compressed
autotune: Makes the TTS have a pseudo-autotune effect (not very good right now to be honest)
phone: Makes the TTS sound like its on the phone
muffled: Makes the TTS sound like its muffled. (v2.4.1)

You can use these by adding a . to the end of the voice, then adding the effect, like this: drake.pitchup or kanye-west-rap.reverb

鈿狅笍 Twitch (and some bots) do think that these are URLs being inputted into chat, make sure you at least disable the Twitch automatic censoring of URLs. If a bot times out links, the TTS will still go through.

Example:

kanye-west-rap.pitchup: why do i wound like this??? || drake.pitchdown: i don't really know kanye? || (1) || spongebob.loud: AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA

You can also chain together these effects! Like this: drake.pitchup.reverb or spongebob.pitchup.loud

Example:

kanye-west-rap.pitchup.reverb: woah we are in a big room and my voice sounds so high || drake.pitchdown.reverb: yeah? well we are in the same room and my voice sounds lower || spongebob.reverb.loud: i am in the same room but i am really loud. i am going to scream now. AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA

Fakeyou.com support.

In addition to Uberduck, we now support Fakeyou! (As of v2.5.0+) This means that you can use an extensive library of new voices!

Right now, the voice IDs are a little weird, but they are pretty easy to get once you learn how to do it once.

How to get voice IDs on Fakeyou

  1. Go to fakeyou.com and find a voice you'd like!
  2. Once you chose a voice, click the "See more details about 'Voice' model by 'Username'". Fakeyou.com
  3. In the URL, copy the ID after the / **INCLUDE THE TM: image
  4. Use that as the Voice ID!

Example:

TM:rtyasywg9zb7: This is Android 18, I have no clue what this character is, but don't flame me for it

It, of course, also supports Uberduck voices at the same time

TM:rtyasywg9zb7: Hi spongebob from Uberduck || spongebob: Hi weird ID from FakeYou

And, works with playsounds + voice effects

TM:rtyasywg9zb7.loud: AAAAAAAAAAAAAAAAAAAAAAAAAA || spongebob.loud: dude that's my thing || (1)