Realtime AI Voice Changer Installation - Beatrice v2

[@mik3lang3lo]

I always appreciate the installation guides, thank you

[@Jarods_Journey]

Your welcome :)

[@BluecoreG]

for those who try to run start http bat and the window immediately closes, the fix is opening cmd, navigating to the folder with that bat and running it that way

[@lilredchingofficial]

i opened the start file and its not showing any numbers and stuff help me please wait nevermind it started

[@unreleasedmuzik2]

@@lilredchingofficial it did?

[@stickmanland]

Yet another killer video. Keep it up!

[@BichotaValentina]

it keeps saying "failed to fetch" when I try changing the model

[@josecastellanosargueta7370]

I can't seem to get any voices to work... After the installation, the menu would pop up just fine. But after the configurations, no sound would ever be played in the headphones. After closing it and opening a second time, the terminal would simply close by itself, not loading the menu fully, and giving me "error to fetch" messages. What could I do?

[@Jarods_Journey]

You may have to delete and re-unzip the package. However, for no sound being played I'm not too certain as there are a variety of factors that can be checked. I'd first check that it does work outside of the voice changer first, then move back into the voice changer. If you have other devices you can try, I'd recommend that as well

[@CinemaTgrinya]

When i click on the models it is saying "Type error : failed to fetch" is there anything more to download i need your help please?

[@NokhchiTankist]

Cant wait for the next video

[@pragmata7997]

where we can download beatrice models?

[@ganagallachetan213]

github

[@Kumatanium]

So, a CPU....what would you recommend from low to high tier??

[@Jarods_Journey]

Well, the faster the CPU, the better. There are many hardware channels out there to check out

[@TaylorLaurieAndTheAutumnLeaves]

Lol

[@sh1kimori]

When I go to start http he opens the window for me but then disappears immediately

[@Vanubis-]

but how does it work on discord how can they hear me

[@SkyeBoredom]

Answer this one please

[@Headhunternn]

hi. how to export the A.I. voice as an audio file..?

[@TonalDesigns]

When you're doing the training one, please include how you overcome the lisp-like thing that it gives on english voices if you've managed it! Haha

[@Jarods_Journey]

Thanks! I'll talk a little about it, but that seems to be due to the base model being trained on Japanese (as the author is Japanese)

[@TonalDesigns]

@@Jarods_Journey I suspected as much, I've tried playing with learning rates, gradients and sample sizes so far. Starting to worry getting a really good English speaking model would require a new checkpoint, which I sure as sugar don't have the GPU horsepower for! 😅

[@abdollah008m]

helppp when i open start_http rapidly reject and not working when i want open it

[@Semi-Sam-t9h]

That happen to me too

[@aTsuKo_9]

Hello Im wondering if we can use the w okada while playing game on mobile ? Like together while the pc running the w okada ? Thank you and sorry for my English (:

[@aTsuKo_9]

And let me know how to connect to mobile..

[@dirtyrainbowsaltwater8490]

the word english and korean is not there

[@ItsAaronPlayz]

When I had a 4gb ram laptop the voice delays alot

[@FadlyOmagat]

I have a problem when I start recording, my voice is not there at all😢

[@Cody2564]

when i open for the second time the Start button gone miss? how do i fix this??

[@SinisterSpatula]

Quality is a bit lacking in this one. RVC is way better.

[@Jarods_Journey]

Definetely agree. It's a compromise, but it's much faster

[@aryanhooshi]

​@@Jarods_Journey it is defo very fast, but whats the point when the its a pain to listen to the voice it generates?, its just not worth training a beatrice model to get such low quality model

[@Jarods_Journey]

Could be quite useful for gaming where you wanna use a voice changer, don't necessarily need the best mic for that. Plenty of examples but kitboga comes to mind where quality isn't stellar. It's "good enough" to be relatively convincing

[@gafra_]

@@Jarods_Journey I’m just curious and I don’t really understand, can this voice changer also use RVC? Is Beatrice v2 the whole point of it or is it just a type of model

[@helloofthebeach]

@@gafra_ RVC and Beatrice are different model types, and this program handles both of them. This voice changer been around for a couple of years but only handled RVC until a few weeks ago because Beatrice is brand spanking new.

[@cdlamagiela]

Hello Jarods how to use Beatrice in audacity or other daw ? thank a lot if ur awnser

[@Jarods_Journey]

Not too sure, as of rn, the inference code is only in wokadas voice changer, it would require a little digging to figure out how to get inference code from the training scripts

[@cdlamagiela]

@@Jarods_Journey i see wish soon , thx for ur awnser

[@v2zel960]

can you use this for discord and gaming like valorant?

[@basharalassad1073]

the real time voice change models dont seem to work

[@Ucantguardzay]

Can we get a updated tutorial

[@rajajawa-t9p]

how to upload my recording file to change the vocal

[@aimanOsu]

what the different between mmvc and vc?

[@glenfoxh]

Been trying to set this up for VRchat, but I don't know how to create the proper overrides that let you see this as an input or an output. All I can do with it is, hear my own voice change, via the mic and speakers. I can't even use it to record my voice with Window's own voice recorder, for noting sees it as an input. Is there some setting that can be used to make this actually useful as a real time voice changer?

[@emeriktremblay3881]

any solution to this ?

[@glenfoxh]

@@emeriktremblay3881 Yes and no. My setup is to make Dubbing AI work for VRchat. I had given up on trying to use this voice changer, for one, I could not get it to work without a lot of lag. But the setup I made works with Dubbing AI. And I feel Dubbin AI sounds a lot better anyway. I downloaded some vb-audio virtual cables for free. For VRchat use, I have my headset mic as the Dubbing mic INPUT. Set the virtual CABLE input as the HEAR MYSELF output. Then in VRchat, instead of setting Dubbing as my VRchat mic, I set the virtual CABLE output as my mic. I turned on HEAR MYSELF in Dubbing. And for some reason, the lag in VRcaht I was getting from using Dubbing directly (was about a one second delay) as my mic, is near gone. I don't know how or why this works better, I only care that it does. So my suggestion is, get Dubbing AI, and the vb-audio virtual cables, all for free. You can also buy one standard voice in Dubbing AI for $0.99, without paying for a subscription or the expensive perpetual for pro.

[@MolkCartone]

@@emeriktremblay3881 Downloading a virtual Audio Cable

[@NuchiAsaki]

Install VB-CABLE (it's free), set the output device in the voice changer to CABLE input and set CABLE output as your default microphone in windows and vr chat.

[@DBZGirlUwU]

Mines be so laggy

[@troyamato8617]

So you just gave up on Tortoise TTS? I’m confused that used to be #1 focus on the channel?

[@Jarods_Journey]

It's on the backburner - there are newer architectures out that I am trying out that serve a different purpose than tortoise. Tortoise is great, but I've pretty much done everything I've wanted to with it for the time and skills I have. As open-source goes, everything is out there, so people with the means and the skills can fork off of it and continue their ways with it. And even though its great, there are imperfections in the architecture; repeating, skipping, artifacts at the ends, okay-ish zero shot cloning, imperfect speaker tone matching (which RVC takes care of), etc. My knowledge of ML isn't yet deep enough to modify the architecture and make any significant changes to fix it.

[@strawbunnylol]

hey can u make a tutorial on how to get this detected into streamlabs ?

[@Cliker_cookie_1999]

Is a ryzen 5600G an alright cpu for this

[@Liloon-jk2lk]

why isnt it have a gpu selection?

[@cryptosean-x9k]

how do i restart after exiting everything?

[@Sultankhalid127]

really great wanted to start a manga read along where characters talk thanks so much !~!!!

[@cryptosean-x9k]

how do you restart after exiting everything?

[@Sultankhalid127]

@ start the program like you did the first time

[@MyersVC]

great video but how do i get custom voices like kai cenat or someone

[@maighe_tv2848]

Damn this video should explode

[@tenC4D_]

Does it work with intel or amd cpu ?

[@obezcan74]

Yo bro can you also explain the advanced settings part since it has also changed with v2

[@Jarods_Journey]

I'll do a little explaining in the follow-up

[@hikarukun5126]

I would like to use it when roleplay Dungeons & Dragons as I am playing for a young female wizard but I am so laaaaaaazy that I just look videos how it works and don't event want to try

[@Kasuuchszz6969]

windows10 on work?

[@BarishaillaMUNNA]

Can i upload audio file from my gallery and download the converted audio file???

[@Jarods_Journey]

Not sure what you mean here, but its probably not the best workflow to upload a file, convert it, and record it on wokada

[@BarishaillaMUNNA]

@@Jarods_Journey ok i understand. 😇

[@GEEKDW]

are u able to use this on discord?

[@ClearsightYT]

How would you connect this to Discord etc?

[@MolkCartone]

Downloading a virtual audio cabel

[@SRGAPEX]

where to get models for it for Beatrice v2

[@almeshan]

did you find ?

[@rons96]

in vst3 plugin it runs even faster

[@EliteVid69]

how to use it on discord

[@MrHeadshot23]

Does this work with discord?

[@La.aa0]

Is it possible on AMD?

[@Sonickuforever]

Can't seem to do anything on my Mac, any help?

[@sjoerdglasbergen]

download mac version not windows cuda

[@Kawims]

@@sjoerdglasbergen but the whole thing is different on mac

[@lanxsvk]

how do i fix delay

[@Kasuuchszz6969]

working windows 10?

[@ساکنسنسیرو]

It doesn't work for me

[@memomii2475]

This doesn't sound good. It obviously sounds like a VC, unlike the other one if you have a good GPU.

[@babalasagna0]

virus?

[@GrimGearheart]

For an RTX 4090, Core i9 13900k, do you think I could run this while gaming, vtubing, and livestreaming?

[@EroonYT]

Yo, i have AMD 7800XT, Ryzen 7 5700X3d and i do voice change, streaming, vtubing and gaming, so u are just fine.

[@himura-lee]

Can I train in paid google colab?

[@Jarods_Journey]

Collab is possible, but I havent set it up there so it's a different process

[@rwshopee5688]

how to add on game or discord

[@emix1400]

IT WORKS

[@fumieshi3071]

does it work with rvc models?

[@LavanyaKanagarla]

this voice changer sounds like shit, w-okada is the only best voice changer as of now

[@Ryploy]

Is it free ?

[@Jarods_Journey]

Yes, requires you to install it though

[@Insidestoryland]

please help to train model on my voice using anaconda. and save the .pth file in my pc.. create a python script to crate text to voice in my audio. Please.

[@orobocigano]

Looking forward for a real time accent canceling!

[@evolutionaer2161]

[@AyzStudio-c3h]

Awesome

[@kryttee]

Shame that you can only use japanesee pitched voices with this. Like this whole JVC corpus model is based on japanase voices and even though you can speak English with them they all sound asian.

[@Jarods_Journey]

Yup, that is a huge weakness, I'm thinking of ways to try and mitigate this though for training purposes so it should be possible to train English better

[@Mosen_xd]

perfect

[@xo_yonsi]

FIRST

[@Jarods_Journey]

Yes indeed :)

[@GWGS_YT]

gg

[@xeno3676]

hii i keep hearing myself can i stop it ?

[@MrHeadshot23]

Does this work with discord?