LMAOOOOOO

THE FIRST SAMPLE

I noticed the quality is a bit lower than my Samantha TTS RVC model

https://discord.com/channels/1159260121998827560/1176382189710082109

also its known to be used in Tomodachi Life as a voice template for the Miis

either the site you used produced low quality audio of the model or the dataset itself was low quality because it was probably generated or sourced from a different place. I've actually create my dataset of the RVC model through the Nuance Text-to-Speech website.

I used 8 minutes of random generated text from some Samantha text generator API

ah

Also that voice is from Tomodachi life isnt it?

yes

Yeah Samantha is not a Tomodachi life voice I think

its obviously lower quality due to 3DS' limitations

https://gbatemp.net/threads/tomodachi-text-to-speech-analysis-mod.397923/

The Samantha voice I made is the old Siri/Voice control voice

yeah

I didnt think it was used in Tomodachi life

I researched what was the Tomodachi life tts along time ago

So a TTS that sounds like this is in tomodatchi life?

Oh

https://www.youtube.com/watch?v=tlzrutH6cbo

Tomodachi life has a option in editing the TTS voice for your Mii leading it to make weirder voices like this

that's right you can edit the Samantha tts voice in whatever pitch, "accent", speed, tone, shift, and intonation

Ik you can

But the Samantha voice isnt directly in there

So I always associated it with older iPhones and voice control or GPSs

we have different perspectives towards the TTS voice 🤷‍♂️

Alr

it does mean its possible to make enhanced versions of anyone Mii's voices using a Samantha voice model to create datasets of the modified Mii though it might be a bit difficult for accent and intonation or maybe if I can be used by doing TTS (from Tomodachi Life) to RVC (Samantha) and adjust the pitch to change shift to make new RVC models.

Yes you could do TTS from Tomodachi life and put it through RVC

You could try putting the voices through adobe enhance if you want to enhance the quality

a few people actually made a few RVC models of their Miis

Yeah Ik I saw some a while ago like months ago

Ok I think you are saying that your try to match the Samantha voice as close as you can on todecent-sounding Tomodachi life, record it and then put it through RVC and you will get a decent sounding voice sounding like the Samantha voice but the intonation and accent would be off slightly?

and adjust the pitch in a audio or video editor to create the mii voice

Yeah

and yes there are are easier options, though Adobe Enhance isn't really the best option as that works better for older voice recordings imo

Adobe enhance sucks

IMO

it served its purpose in enhancing audio from old movies but otherwise will muffle the audio. but i think its still funny because it can create really funny vocal sounds from music or sound effects.

It can do really dumb stuff

I love making modified voices through audio/video editing and multiple Ai cloning software (ElevenLabs and PlayHT). I know most people only care about the quality of voice models but at the same time I think we can have more modified voice models more often since there's some recent voice models that fused Vtuber voices together to make a anime girl voice model and last year where someone fused SpongeBob and Squidward's voice to create some fusion models.

You like making Ai voices have strokes?

that too

those are really funny to watch in those ai streams

I mean its sometimes funny yeah

Like Ai Sponge or Ai Peter?

i made a few rvc models using Brian's "*" stroke sounds and chris griffin's loud stroke-ish voice

Nice

i am planning to make more of those ai stream rvc models which i did made datasets for them

Oh nice

It’s OG Siri!!!