✨│ai-help

[fossil sage]

I didn't really cate

Care

[viral mason]

@low shard is local tunnel dead for good bc it's been down all day :(

[tender spade]

guys

what happens when i make the

delay

1

sorry

chunk

[viral mason]

what

[tender spade]

to

1

[knotty moth]

the one from #1159289738314919936 ?

im not interested to take it tho

[fossil sage]

No it was muzan from demon slayer dubbed it was a year ago

Assuming you watched demon slayer

[knotty moth]

isn't it dead space?

[knotty moth]

knowing it doesnt mean to have to watch it

[fossil sage]

I removed it

.5 in audacity

[fossil sage]

Anyways I think u get my point

[knotty moth]

except I dont have enough spare time to be willing

[narrow portal]

[oak plank]

Nice emoji

[sharp rampart]

i tried deiteris fork im not getting any audio output same with the most recent alpha version of okada

i set to my vac line and headset monitor

no output

input is my mic

[viral mason]

what alpha version

[sharp rampart]

i liked vonovox but it gets choppy when im on a game like cs

[sharp rampart]

one sec lemme find which one i got

vcclient_win_cuda_2.1.4-alpha

[viral mason]

have you tried lowering the settings of the game?

[viral mason]

don't use that it's outdated

[sharp rampart]

ye still pretty choppy

vcclient_win_cuda_2.0.78-beta this one worked idk how to optimize it tho xd

[viral mason]

have you tried the wokada tg fork?

[sharp rampart]

no i havent can u link me?

[viral mason]

-rt

[patent trellis]

💻 Local Realtime RVC

Guides for Programs that use RVC Models in Realtime for Calls/Games

• Vonovox

A Realtime Voice Changer with similar performance to Wokada Deiteris Fork, with extra features, but supported only for Nvidia GPUs on Windows. and without cloud options GUIDE

• Wokada Tg-Develop Fork

A personal fork (modified version) of Wokada Deiteris Fork, it just adds some Quality of Life improvements to it like supporting Spin Embedder and Audio Effects. Don't expect too much about it since the creator made it originally as a personal project. GUIDE

• Wokada Deiteris Fork

Most suggested WebUI with the best general support for many platforms. GUIDE

⚔️ Wokada Deiteris Fork vs Vonovox

For Windows Nvidia, Both Wokada Deiteris fork and Vonovox have similar performance & quality. Users should read the pros and cons for both and choose based on their differences, such as UI and Vonovox's paid effects.
Read Wokada Deiteris Fork Pros&Cons & Vonovox Pros&Cons

⛔ Outdated/Discouraged

These options are not recommended for use.

• Original Wokada

Not suggested, older versions in youtube tuts are even way worse. GUIDE

• RVC GUI Mainline Realtime

The program is worse compared to the ones above, and much less updated. GUIDE

[viral mason]

it's the second one ^^

[sharp rampart]

mhm see it

ill give it a go let u kno

i have no issues usually but im a uni on my gaming laptop so specs downgraded 💔

[viral mason]

oof, if u can ever afford a desktop get urself a 50 series nvidia gpu

works like a charm

[sharp rampart]

i got a 3090 at home on my desktop i just dont have it at uni okada works like a charm on it

4050 notebook gpu 💔 rn

[viral mason]

ah ok

[viral mason]

never heard of a notebook gpu lmaoo

[sharp rampart]

laptop

its not horrid but when u got multiple things running it just cant keep up

[viral mason]

oof, yea realtime really isn't meant to be on laptop at all

[sharp rampart]

ye i can see that lol

[sharp rampart]

Ok i tried it much cleaner ui i like this one. i got a bunch of errors with torch.cuda.FloatTensor in my console when switching to my gpu from cpu cpu also had no audio

iim just getting no input output audio its all set tho to what it should be

[hoary island]

i installed vonovox

i have the right gpu and pc for it

but when i press start on the voice

it just sits like this

[viral mason]

try closing the app and restarting

[hoary island]

i got it working but it wont go to discord for some reason

[sharp rampart]

im not getting any input or output on this second one

like when i record its blank but my mic works fine normally

[shut jolt]

any of you know how to hear yourself on vonovox? i cant tell if i sound right because i cant hear my self

wokada has a monitoring feature but i cant find vonovox's despite it being a more in date program

[gilded fulcrum]

I bought premium but for some reason it’s not giving me any perks as if I haven’t bought it?

Help:,)

[shut jolt]

Anyone

I’m using it with voicemod as well but that won’t let me hear the voice from vonovox

[viral mason]

Ask in the official vonovox server for that

[knotty moth]

if not vonovox or anything we're affiliated to, we're not responsible for whatever service you've spent on

[gilded fulcrum]

Huh?

I’m just wondering why it suddenly stopped working

[stray copper]

what batch size should I use for 22 minutes and 31 seconds of dataset audio

[knotty moth]

huh? is it just called "premium"?

[gilded fulcrum]

I’ll take a ss

Oh nvm I can’t even send pictures

[stray copper]

so.. batch size?

[gilded fulcrum]

Ok so on dubbing AI I bought the premium so you can have all the voices and the sound board, it’s worked fine since I bought it but today I went to use it and I wasn’t allowed to use any of the premium voices. I had to use my coins to buy premium again

[shut jolt]

Figured out how to hear my voice on vonovox putting this here for any future people with the problem

[knotty moth]

we're not affiliated with that service, which means to not be responsible nor provide support for it

[low shard]

there’s no shutdown official news, might be some maintenence or something, it can happen for servers to be down, it happens for discord too

try other tunnels

Ngrok seems to work for Applio now and its the new default tunnel currently

[low shard]

that’s nice, it was already in the docs btw https://docs.aihub.gg/realtime-voice-changer/local/vonovox/#how-can-i-hear-myself

[low grotto]

can anyone recommend me a good real time voice changer? my pc is kinda choppy so i cant find one..

[austere stream]

something i've been curious about actually, i have two models trained on separate pretrains and if i try to fuse/merge them, despite having identical voices the result is bad
both trained on mainline rvc v2, only difference is the pretrain
is there a reason why this happens?

edit: the inference audio is clean so its not an issue there, sample rate is also the same

[austere siren]

best deitris fork settings for amd rx6600 gpu and intel 10400f cpu

[low shard]

you sure you don’t want to use wokada tg-develop instead for some quality of life features?

theoretically, even tho the mast vajority of helpers and users use windows

[austere siren]

do you reccommend that one? i dont know much about these apps tbh

ill download it now then could you help me out with the settings? much appreciated

[low shard]

yeah

-rt

[patent trellis]

💻 Local Realtime RVC

Guides for Programs that use RVC Models in Realtime for Calls/Games

• Vonovox

A Realtime Voice Changer with similar performance to Wokada Deiteris Fork, with extra features, but supported only for Nvidia GPUs on Windows. and without cloud options GUIDE

• Wokada Tg-Develop Fork

A personal fork (modified version) of Wokada Deiteris Fork, it just adds some Quality of Life improvements to it like supporting Spin Embedder and Audio Effects. Don't expect too much about it since the creator made it originally as a personal project. GUIDE

• Wokada Deiteris Fork

Most suggested WebUI with the best general support for many platforms. GUIDE

⚔️ Wokada Deiteris Fork vs Vonovox

For Windows Nvidia, Both Wokada Deiteris fork and Vonovox have similar performance & quality. Users should read the pros and cons for both and choose based on their differences, such as UI and Vonovox's paid effects.
Read Wokada Deiteris Fork Pros&Cons & Vonovox Pros&Cons

⛔ Outdated/Discouraged

These options are not recommended for use.

• Original Wokada

Not suggested, older versions in youtube tuts are even way worse. GUIDE

• RVC GUI Mainline Realtime

The program is worse compared to the ones above, and much less updated. GUIDE

[fossil sage]

I mean helping other without getting anything in return is spending your time in effort at least you guys aren't gatekeeping though like how they do in the money making discord servers

[knotty moth]

again you don't quite get my point

spoiler: ||do you even think the demon slayer anime creators and voice actors wouldn't mind their voice taken and open sourced?||

[dapper surge]

guys how do u actually use the voices

like i downloaded a voice from #1175430844685484042

and idk what to do wit it

[oak edge]

gang where is my bro @low shard

[tender spade]

how do i fix the voice echoing

when i'm in a vc with someone and they speak, their voice echos back to them

[hybrid fractal]

Hello, I'm running into an issue while training a model on Applio v3.5.

I'm using the 'Legacy core V2' and my sample rate is correctly set to 32k, but I keep getting this error. Could anyone help me figure out what's causing it?

Process Process-1:
Traceback (most recent call last):
File "C:\work\ApplioV3.5.0\env\Lib\multiprocessing\process.py", line 314, in _bootstrap
self.run()
File "C:\work\ApplioV3.5.0\env\Lib\multiprocessing\process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "C:\work\ApplioV3.5.0\rvc\train\train.py", line 579, in run
train_and_evaluate(
File "C:\work\ApplioV3.5.0\rvc\train\train.py", line 1007, in train_and_evaluate
save_checkpoint(
TypeError: save_checkpoint() takes 5 positional arguments but 6 were given

[hybrid fractal]

Also, I set multi_scale_mel_loss to True

[simple ore]

i guess Lyery gave you train.py with an extra change

roll back the change

and simply edit the file manually

[hybrid fractal]

Thank you I'll give it a try!

[viral mason]

I think he got snatched by the boogie man idk where he went

[viral mason]

I need the new download to import

[low shard]

could you please elaborate?

!howtoask

[patent trellis]

❓ How to Ask for Help

✅ Before You Ask!

1. Check Docs & Guides: Your answer may already be in the AI Hub Docs or the https://discord.com/channels/1159260121998827560/1159513888199540817 channel.
2. Search the https://discord.com/channels/1159260121998827560/1192011222023950368 : Look for existing posts that solve your issue. Do not invade someone else's post.

📝 How to ask?

Tell your:

• Full GPU Name: (e.g., NVIDIA RTX 3060)
• Operating System: (e.g., Windows 11)
• Detailed Description: What were you trying to do and what went wrong?
• Tutorial Used: Link to the guide you were following.
• Screenshot: A picture of the full error message is very helpful.

🚫 Prohibited Topics (We Will NOT Help With These)

To maintain a legal, safe & ethical community, we will NOT provide help for:

• (E girl, as an example) catfishing/trolling, scamming, impersonation.
• NSFW/Porn.
• Any illegal activities.
  Requests for these topics will be ignored and may result in moderation action.

<:matsuripray:1159685390156967936> Community Expectations

• Be Polite & Patient: Our helpers are volunteers. You may ping the Helpers role once.
• English Only: Please keep all conversations in English.

[low shard]

it would be best you elaborate more for us to help you :)

!howtoask

[patent trellis]

❓ How to Ask for Help

✅ Before You Ask!

1. Check Docs & Guides: Your answer may already be in the AI Hub Docs or the https://discord.com/channels/1159260121998827560/1159513888199540817 channel.
2. Search the https://discord.com/channels/1159260121998827560/1192011222023950368 : Look for existing posts that solve your issue. Do not invade someone else's post.

📝 How to ask?

Tell your:

• Full GPU Name: (e.g., NVIDIA RTX 3060)
• Operating System: (e.g., Windows 11)
• Detailed Description: What were you trying to do and what went wrong?
• Tutorial Used: Link to the guide you were following.
• Screenshot: A picture of the full error message is very helpful.

🚫 Prohibited Topics (We Will NOT Help With These)

To maintain a legal, safe & ethical community, we will NOT provide help for:

• (E girl, as an example) catfishing/trolling, scamming, impersonation.
• NSFW/Porn.
• Any illegal activities.
  Requests for these topics will be ignored and may result in moderation action.

<:matsuripray:1159685390156967936> Community Expectations

• Be Polite & Patient: Our helpers are volunteers. You may ping the Helpers role once.
• English Only: Please keep all conversations in English.

[low shard]

https://docs.aihub.gg/rvc/cloud/applio-kaggle/ got updated, if you just want the direct link, it's the usual IAHispano GitHub Applio link: https://github.com/IAHispano/Applio/blob/main/assets/Applio_Kaggle.ipynb

[tender spade]

okay

[low shard]

huh?

[oak edge]

gangggg how much would it take to pretrain a tamil voice model with tamil speakers

who manages embedders and pretrains

[hybrid fractal]

now it works! thank you

[tender spade]

• Gpu 0 AMD Radeon (TM) graphics Gpu 1 NVIDIA GeForceRTX4050 Laptop Gpu
• Win 11
• when i'm in a discord vc with RVC in bg setup by VB Audio Virtual Cable my output audio is their echo and my voice combined
• For ex, if they say "hello", "hello" will echo from my mic with the voicechanger
• Tutorial : https://youtu.be/SxdnGxicJOg?si=fX-17jrD81EIMLa6
  i'm not sure HOW to provide it in a ss but my settings might be messed up

[viral mason]

oh ty!

[meager sonnet]

Hai I am back, need helpu

[viral mason]

don't use yt tutorials what you have is super old and outdated

[viral mason]

heloo

[tender spade]

4 months old

🤯

[meager sonnet]

Use Vonovox. Worm will help ya, they helped me :3

[tender spade]

wth is Vonovox

[viral mason]

but with a year old voice changer

[viral mason]

-rt

[patent trellis]

💻 Local Realtime RVC

Guides for Programs that use RVC Models in Realtime for Calls/Games

• Vonovox

A Realtime Voice Changer with similar performance to Wokada Deiteris Fork, with extra features, but supported only for Nvidia GPUs on Windows. and without cloud options GUIDE

• Wokada Tg-Develop Fork

A personal fork (modified version) of Wokada Deiteris Fork, it just adds some Quality of Life improvements to it like supporting Spin Embedder and Audio Effects. Don't expect too much about it since the creator made it originally as a personal project. GUIDE

• Wokada Deiteris Fork

Most suggested WebUI with the best general support for many platforms. GUIDE

⚔️ Wokada Deiteris Fork vs Vonovox

For Windows Nvidia, Both Wokada Deiteris fork and Vonovox have similar performance & quality. Users should read the pros and cons for both and choose based on their differences, such as UI and Vonovox's paid effects.
Read Wokada Deiteris Fork Pros&Cons & Vonovox Pros&Cons

⛔ Outdated/Discouraged

These options are not recommended for use.

• Original Wokada

Not suggested, older versions in youtube tuts are even way worse. GUIDE

• RVC GUI Mainline Realtime

The program is worse compared to the ones above, and much less updated. GUIDE

[viral mason]

the first guide

best voice changer around so far and really good

[tender spade]

https://tenor.com/view/jgmm-monkey-think-monkey-meme-ponder-monkey-idea-gif-6401133862108294696

[meager sonnet]

@viral mason My issue is that when I speak. The sentence beginning and ends cut out. Can't quite put it into words.

But imagine beginning and ends being blurred.

[viral mason]

https://tenor.com/view/garn-47-garn47-car-huh-confused-gif-14701624453881581373

[viral mason]

what do u have your settings at?

[meager sonnet]

I guess the voice sounds choppy, is the correct way to put it

[meager sonnet]

Also car 🚗

[viral mason]

try messing with the block size maybe?

[meager sonnet]

As in? Decrease, increase?

[viral mason]

since u have it at 0.10 try increasing it a little each time and testing

[meager sonnet]

What does block size even do?

[viral mason]

controls the delay pretty sure

[meager sonnet]

The higher the delay, the better?

[viral mason]

not really

[meager sonnet]

So why increase it then?

[viral mason]

you're just trying to find a spot to where the delay isn't too bad but not making the voice choppy or broken

[meager sonnet]

I... don't get it -w-

[viral mason]

https://cdn.discordapp.com/attachments/1401477083586101359/1419449901762478243/image0.gif

[meager sonnet]

https://tenor.com/view/car-death-garn47-gif-11470553784351804754

[viral mason]

so like you're trying to go for not too choppy but not so delayed u can't have a conversation with someone

[jaunty shale]

Using Applio Kaggle (Gradio + LocalTunnel)

I've put the ip password into the slots to get access, but then I got this error.

[viral mason]

that happens sometimes

just re run the cell again

[jaunty shale]

I did it like 3rd time already, it worked-

huh.

[low shard]

this might have happened because the localtunnel server were down, be sure to run the latest applio kaggle version and try other tunnels, they recently changed the default tunnel back to Ngrok

[viral mason]

It's up to pure luck, but now u won't have to use that version of applio since the usual applio works again

[jaunty shale]

I created the notebook and imported the newest applio already, tried ngrok but it just shows an error.

it does work now, i refreshed the filebrowser page and it showed the same error basically, it worked before though.
Although, it doesn't really bother me, since I can download the pth and index through applio like I did before.

[viral mason]

if I go back to the normal applio and it gives me a single ngrok error I'm finding the maker and making him get an ngrok error

[jaunty shale]

oh yeah, this too. Tried to refresh the tensorboard.

[jaunty shale]

I used this guide btw
https://docs.aihub.gg/rvc/cloud/applio-kaggle

[low shard]

tried ngrok but it just shows an error.
what error? Ngrok worked fine for both my and @nocturne mural

[viral mason]

oh I don't even need guides for applio as I know what buttons to press

[low shard]

and Vidal updated the warning of the applio kaggle tunnels, saying that it can happen for tunnels to be temporarely broken, that's why there are multiples you should play around with

there can't be fully 24/7 always, reliable, fast and especially free servers for tunnels

[jaunty shale]

I didn't get the screenshot, since I tried doing it like 20 minutes ago, it had a long error

probably because of this

[low shard]

You will need guides one day when things inevitably change

[low shard]

could you please retry with ngrok and check what's the error?

[jaunty shale]

yeah sure give me sec lol

[low shard]

what tunnel are you currently using then?

and btw you could also try Applio UI Lightning.AI notebook

[jaunty shale]

I used gradio and localtunnel.

It worked just fine, its just the filebrowser and tensorboard links weren't accessible anymore, since it showed up the error I showed above.

trying ngrok rn

oh-- wait ngrok works?

huh. okay.

ig it's solved

[low shard]

yeah Ngrok had a temporary issue for Applio, but yesterday me and vidal checked that it worked just like it used to some months ago all fine

[jaunty shale]

alright, also one more thing.

I have uh.. 27 minutes long dataset (truncated/deleted all silence in audacity)
Is there anything I should consider checking/unchecking?

[viral mason]

don't uncheck anything and u don't really need fresh training, just delete the notebook and remake it to have infinite space after each training session, once u finish the model of course

[jaunty shale]

alright, got it, thank you for your help

[fossil sage]

Yes I do

Because you can make them say anything

[tender spade]

can we ALL just clone EVERY SINGLE ANIME VOICE SO PERFECT IT COMES TO A POINT WHERE THE RVC OWNERS GET SUED AND COMEPLELETY BANISHED FROM THE FACE OF THE EARTH

[viral mason]

no

[tender spade]

only

[viral mason]

rvc is super epic and cool so we should not get them sued

[tender spade]

that much

i mean

EVERY SINGLE

fictional

character

https://tenor.com/view/jgmm-monkey-think-monkey-meme-ponder-monkey-idea-gif-6401133862108294696

side characters

back ground characters

even characters that have 10 sec of screentime

[viral mason]

but why

[tender spade]

it would be funny

[viral mason]

https://tenor.com/view/lil-yachty-drake-laptop-duh-gif-8033838636860661307

[fallen yoke]

hi guys what would be the best epochs for training a model that has like 3/2 min audio dataset (?)

[viral mason]

there is no specific epoch count good for any dataset length just train and watch the tensorboard until it sounds good on each one

the line going up = bad
the line going down= good

[elfin nebula]

Is it possible to create an entire language model with a personality, without depending on a cloud service like Chatgpt?

Or do I just cop it and cough up the money?

[tame osprey]

How can i fix the Server Connection problem when i use the RVC VCClient?

where do i get the server URL?

[viral mason]

lemem guess, u got the voice changer from a youtube tutorial

it's old and outdated

[tame osprey]

i got newest one

[viral mason]

what's the link?

or can u send a screenshot?

[tame osprey]

https://huggingface.co/wok000/vcclient000/blob/main/vcclient_win_cuda_2.1.4-alpha.zip

[viral mason]

yup

old

outdated

what's your gpu

[tame osprey]

if u wanna ask me that, i think i only have CPU

its the Intel 3...

[viral mason]

hmm, what computer are you running on currently is it laptop?

[tame osprey]

desktop

[viral mason]

intel can't really run any of the voice changers sadly u might be cooked my friend :(

u can check tho using task manager

[tame osprey]

how?

[tame osprey]

it runs the Voicemod V3 tho

[viral mason]

use V2 it's better, and voice mod is very different compared to Okada

[tame osprey]

:/

[viral mason]

aw dang u really do only have a cpu

[tame osprey]

yea ik

its a office desktop after all

[viral mason]

yea u might need to get your wallet out and find a 30 series nvidia gpu maybe

[tame osprey]

does cheap gpu work?

[viral mason]

that would be the best option if u really want to use this and not go overboard like for a 50 series

[tame osprey]

Im bit eepy, dont mind

[viral mason]

I was able to run wokada detieris on a 1660 but it wasn't the best

[viral mason]

that's ok lol

[tame osprey]

in that case what about voicemod v3, everytime the girl rvc doesnt works well on me

choppy and sounds weird

[viral mason]

bc voicemod is bad lol

the ai voices on there suck tbh

[viral mason]

how much do ya got?

[tame osprey]

no idea, my e-wallet got some issue with the server

[tame osprey]

the russian earrape mic sounds good tho

[viral mason]

dang

[tame osprey]

gimme the gpu than i'll see if i can affort it

[oak wedge]

Hey, does anyone know any good voice ai? Something that is text to speech but not as crappy sounding

[viral mason]

why tts? rvc is much better and more versatile

actually lemme just screenshot it

@tame osprey

[tame osprey]

ya?

honestly

[viral mason]

that's the gpu I used to own

[tame osprey]

what about PNY Geforce RTX3060 TI (8GB)

[viral mason]

lemme check

not bad

[tame osprey]

or GIGABYTE GTX 1660 6G

[tame osprey]

this one?

[tame osprey]

ya that one

[viral mason]

it's not toooo expensive

I would say if u don't have that much money to spend on it u should use that for now

[tame osprey]

yes

aslong it works

[viral mason]

should work decently in vrchat but not clue about normal games, might have to turn down your graphics

[tame osprey]

roblox?

[viral mason]

probably yea

[tame osprey]

Ok, Thx for the info

rn i need to head to sleep cus later i need to wake up in 3 or 2 hours

[viral mason]

oki doki sleep well ^^

[coral wolf]

what do i do when i am at this part

Manual Setup NVIDIA on Windows
Go to Vonovox's Github Repository and download the Latest Stable Release Source Code.

You could also optionally get access to Beta / Early Access versions via Becoming a Vonovox Supporter (and also gaining Premium Effects) or checking for any Free Versions in the Vonovox Official Discord Server. They may have bugs, fixes or settings/options not explained in the guide yet.

After downloading, extract the zip file. Open the folder and run setup.bat.

Vonovox will start downloading everything it needs to run. Be patient as it can take up to some minutes to download everything it needs depending on your internet speed.

Once it's done downloading everything it will display Setup complete! in the command line.

@viral mason

[viral mason]

ok so did u run setup.bat?

[coral wolf]

no

how do i do that

[viral mason]

just open it lol

[coral wolf]

but where is it

[viral mason]

like double click

[coral wolf]

i did setup64

downloaded it but what is nextr

[viral mason]

that's for vb cable or vac lite

open vonovox folder and run setup.bat

[coral wolf]

i am in the github for vonovox and see where it says setup.bat but im unsure how to

like how do i run it

[viral mason]

u just download vonovox, extract the zip file, and then open the folder and run teh setup.bat

by clicking it

with your mouse

[coral wolf]

is it vonovox 169.zip that i download?

then open and extract bthen run setup.bat

sorry i am slow and never done stuff like this

[viral mason]

it's ok

and yes that's what u do

[coral wolf]

ok

its downloading

[viral mason]

after setup has finished and says "Setup complete" run start.bat

[shut jolt]

What do I do to download a weights voice model to vonovox, wokada etc when the weights model doesn’t have a huggingface link

Or can I only download ones with a link like that in the desc

[viral mason]

download it using the 3 dots on teh site, but just a warning because weights is dumb and doesn't name the files it just calls them "model"

you'll have to rename them on ur pc after downloading it

[coral wolf]

can i just share my screen?

[viral mason]

I don't really wanna call rn

[coral wolf]

ok

[viral mason]

it's very simple to just extract it, open one file let it run then close it

just run setup then run start

[coral wolf]

why cnat i send pictures

oh i can

ok so im here

after downloading vonovox169

i exrtact the whole thing?

[viral mason]

run setup, then start after setup is done

[coral wolf]

the setup is taking a long time to even just open is that normal?

[viral mason]

it'll take a while because it's downloading stuff

[coral wolf]

oh alright

ok i did it but

nothing happened

[viral mason]

wdym

[coral wolf]

do i run this

[viral mason]

yup

[coral wolf]

damn it takes so l;ong

[shut jolt]

Rename it to what

The name of the model?

[viral mason]

yea just rename it to whatever the model is

that's why I usually avoid downlaoding off weights to avoid that dum issue

[grand belfry]

how can i fix the res? it goes up to like 40k ms

the whole app itself is really laggy also

[viral mason]

what app?

[grand belfry]

the voice changer

[viral mason]

which one there's 3

did u get yours from a youtube tutorial btw

if u did just uninstall it bc it's outdated and over a year old

[grand belfry]

yeah no ive had it for a while and updated it a couple of days ago, just wanted to try it again

which one do you recommend then?

[viral mason]

could u send as screenshot of it maybe? or send the link u got it from?

[grand belfry]

i cant attach pics

[viral mason]

links should work tho

btw what's ur gpu that is important to what one u download

[grand belfry]

https://github.com/w-okada/voice-changer/blob/v.2/docs_i18n/README_en.md this is the one ive hadd

[grand belfry]

its a low end gpu, havent managed to upgrade yet

r7 370 think it was

[viral mason]

I can't even tell which one this is so it's def gotta be old

[viral mason]

how are you running it at all?

uhm

try the tg fork w-okada, second guide

-rt

[patent trellis]

💻 Local Realtime RVC

Guides for Programs that use RVC Models in Realtime for Calls/Games

• Vonovox

A Realtime Voice Changer with similar performance to Wokada Deiteris Fork, with extra features, but supported only for Nvidia GPUs on Windows. and without cloud options GUIDE

• Wokada Tg-Develop Fork

A personal fork (modified version) of Wokada Deiteris Fork, it just adds some Quality of Life improvements to it like supporting Spin Embedder and Audio Effects. Don't expect too much about it since the creator made it originally as a personal project. GUIDE

• Wokada Deiteris Fork

Most suggested WebUI with the best general support for many platforms. GUIDE

⚔️ Wokada Deiteris Fork vs Vonovox

For Windows Nvidia, Both Wokada Deiteris fork and Vonovox have similar performance & quality. Users should read the pros and cons for both and choose based on their differences, such as UI and Vonovox's paid effects.
Read Wokada Deiteris Fork Pros&Cons & Vonovox Pros&Cons

⛔ Outdated/Discouraged

These options are not recommended for use.

• Original Wokada

Not suggested, older versions in youtube tuts are even way worse. GUIDE

• RVC GUI Mainline Realtime

The program is worse compared to the ones above, and much less updated. GUIDE

[grand belfry]

alright thank you

[viral mason]

np! I have no clue how the tg fork one works but if u need help I can try

[coral wolf]

bro how long does this setup TAKE

[viral mason]

is it doing anything?

the command promt

[coral wolf]

it said press any key to continue

i did

then afterwards nothing happened

it just closed

should i open the start one now?

[viral mason]

it's done then

[viral mason]

yup

[coral wolf]

ok bet

it says extrract all or run

which one should i do

[viral mason]

what

you opened the zip file don't do that

[viral mason]

did you even extract the zip file before running the setup

[coral wolf]

yes i did

[viral mason]

then delete the zip file

so u don't get confused

keep the folder

[coral wolf]

well what should i do here first

[viral mason]

don't do either

delete

the

zip

[coral wolf]

the compressed one or the one before compresseed

[viral mason]

here's teh difference

delete the one that has a zipper

[coral wolf]

ok i deleted that

[viral mason]

idk if u even ran the setup in the actual folder now so I'm confused

[coral wolf]

uhhh

[viral mason]

redo the setup rq in the vonovox folder

[coral wolf]

ok

[viral mason]

and then run the start

srry

[coral wolf]

no IM sorry

im so lost

[viral mason]

it's ok

[slate tide]

how to fix [19156:1007/023640.041:ERROR:gpu_init.cc(523)] Passthrough is not supported, GL is disabled, ANGLE is

[viral mason]

did you get the voicechanger off a youtube tutorial

[slate tide]

yes

[viral mason]

it's outdated and old, delete it

what's your gpu?

[slate tide]

rtx 4060

[slate tide]

where to get the new one?

[viral mason]

-rt

[patent trellis]

💻 Local Realtime RVC

Guides for Programs that use RVC Models in Realtime for Calls/Games

• Vonovox

A Realtime Voice Changer with similar performance to Wokada Deiteris Fork, with extra features, but supported only for Nvidia GPUs on Windows. and without cloud options GUIDE

• Wokada Tg-Develop Fork

A personal fork (modified version) of Wokada Deiteris Fork, it just adds some Quality of Life improvements to it like supporting Spin Embedder and Audio Effects. Don't expect too much about it since the creator made it originally as a personal project. GUIDE

• Wokada Deiteris Fork

Most suggested WebUI with the best general support for many platforms. GUIDE

⚔️ Wokada Deiteris Fork vs Vonovox

For Windows Nvidia, Both Wokada Deiteris fork and Vonovox have similar performance & quality. Users should read the pros and cons for both and choose based on their differences, such as UI and Vonovox's paid effects.
Read Wokada Deiteris Fork Pros&Cons & Vonovox Pros&Cons

⛔ Outdated/Discouraged

These options are not recommended for use.

• Original Wokada

Not suggested, older versions in youtube tuts are even way worse. GUIDE

• RVC GUI Mainline Realtime

The program is worse compared to the ones above, and much less updated. GUIDE

[viral mason]

download vonovox, it's the best one, only works for nvidia

first guide

[coral wolf]

uhh its just on this

this is after i fnished setup

and then did the start

YO

SOMETHING OPENED

[viral mason]

peak

lemme see

[viral mason]

that's it right there

[coral wolf]

lets gooooooooo

so im done??

[viral mason]

go get yourself a nice model and import it

[coral wolf]

lets gooo

[viral mason]

and it better not be a egirl or I'll remove your ball privileges

[coral wolf]

idk what ill get but just for you i ownt

thanks bro

you are the best

[viral mason]

ty, if u need anymore help just lemme know or join the vonovox dc server

[coral wolf]

why cannnot i not hear naything

[viral mason]

I don't think u set your voice settings up correct, change your output to line 1

[coral wolf]

OK

i did

still cannot hear

[viral mason]

hmm

is your input your regular headphone/headset mic?

[coral wolf]

ye

do i have to add a voice model thing first?

[viral mason]

yea lol

[coral wolf]

ok

how do i download

i found sonic the hedgehog one

[viral mason]

u just click the blue linkhttps://huggingface.co/YourLocalWorm/MovieVenom/resolve/main/VenomTomHardLCV1_435e_26100s.zip?download=true

like this one, a mode I made of Venom

[coral wolf]

ooooh i see

and it is downloading

[viral mason]

mk

[coral wolf]

ok i did it

but i still cannot hear

and my mic isnt picking up my talking

[viral mason]

make sure u put the pitch at 0, and have you pressed the start button?

[coral wolf]

no still dosent

ill show you this

[viral mason]

that's odd

maybe try vb cable instead of vac lite

[coral wolf]

uhhh

where is that

[viral mason]

https://download.vb-audio.com/Download_CABLE/VBCABLE_Driver_Pack45.zip

[coral wolf]

goodness there is alot more stuff

which one do i click

and should i delete vaclite?

[viral mason]

nah

[viral mason]

when u click that link it should just download it

[coral wolf]

it did

i have the file open

its just a bunch of stuff

[viral mason]

unzip it, then run setupx64 as administrator by right clicking it

[coral wolf]

ph boy

how do i unzip it

[viral mason]

and remember, delete zip after extraction

[viral mason]

just click something and hit extract all

[coral wolf]

i did it

i dont see where it says run as administrator tho

oh waait i see

[quasi condor]

why are u making em install vb cable

it makes people have alot of issues tbh

[viral mason]

bc vac lite just doesn't work

[quasi condor]

vac lite works for me :p

[viral mason]

only time I get it working is with a whole combination of vb cable, vac lite, fl studio, voicemod, and voice meeter

[quasi condor]

literally there's a whole guide on w okada including vac lite

[viral mason]

send the guide to @coral wolf as Idk how to use it

I only can get it working with a complex setup not everyone would wanna use

[coral wolf]

uhhhh

[quasi condor]

its on -realtime and click GUIDE on Deiteris W Okada Fork

[coral wolf]

well i finished downloading vb cable

so what should i do

[coral wolf]

where is this?

[quasi condor]

u type -realtime on bot commands and click the 3rd one that says wokada deiteris fork

its a whole guide how to use the voice changer

[coral wolf]

#🤖│bots

[fossil sage]

no

[viral mason]

he's using vonovox, is there a reason the vac lite setup isn't in there as well?

[quasi condor]

u could've said that AT THE FIRST PLACE o -o

u didn't even say what type of voice changer he's using

u help him

i need sleep

[coral wolf]

how if they asked you why

[viral mason]

my bad lol

[viral mason]

follow this whole thing rq
https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork

[serene berry]

can someone help with okada voice changer on collab? the server closes after a few seconds of starting

[viral mason]

do u have a good gpu to run it locally?

[serene berry]

Nope I have a GTX 960 4gb, I think its too old for this

[viral mason]

oof

I'm no help at all on non local voice changer stuff sorry :<

[serene berry]

It's okay, I just check the doc you sent above and it also have an alternative for online using kaggle, I will try that Thanks 😄

[viral mason]

good luck ^^

[coral wolf]

i think i will die before i figure this out

[viral mason]

:(

[modern fulcrum]

Question if any good lad/lass can answer. Still learning how to use Applio to train a voice model. When it finishes training, do people post their model with voice examples from the Inference Tab or from the TTS tab? When i try to use it in the Inference tab it doesn't sound exactly like the voice i wanted to train. But when i use the TTS tab and make some changes in the settings i do get it to sound the same from the samples i trained it from. Is that correct or is my training off? 🙏🏼

[viral mason]

inference, tts cannot sing

[modern fulcrum]

Got it. I'm not training one for singing. Just basically have it respond as if it were an alexa but with a custom voice. Then that means the training is off if it doesn't sound the same as the samples i trained it with

[viral mason]

yea if you're doing something like alexa or whatever use tts for outputs

[viral mason]

https://tenor.com/view/oh-moderators-gif-25571015

go away scammer

kill him mods

[fossil sage]

yooooo

got any tips for a dataseet

and yes i read both appilo and ai hub guides

[viral mason]

what specifically are you wanting to know?

[fossil sage]

my rvc model had artifcats

im assuming it was because i accidently put 2 voices

in there

now i gotta retrain

[viral mason]

yea if u do that it won't turn out well

[fossil sage]

it was litterally one second

of someone elses voice

8😭

[viral mason]

I accidentally left a whole section of a episode in a dataset and didn't notice untilmy model came out weird

had to re edit my dataset lmao

[fossil sage]

wait should the model having any breathing

[viral mason]

yes

or it'll sound werid when breathing

[fossil sage]

is it bad if the audio cuts out in the middle of nowhere

all man i also forgot to dereverb

[viral mason]

like the character is talking and it just stops

[fossil sage]

yes

[viral mason]

I think that's ok

[fossil sage]

oh ok well thats bad for tts

can u hop on vc

[teal crystal]

anyone down to vc and help me setup microsoft vibevoice

[fossil sage]

microsoft vibevoice?

whats wrong with confyui vibevoice

[raw socket]

Anyone have sora invitation Code Please?

[teal crystal]

idk what this is

i just saw a video on it

whats the difference?

[viral mason]

leave

[viral mason]

can't rn

[fossil sage]

wnat if you dont talk

what if

[teal crystal]

i've been really trying to get a good platform for TTS

everythings been a lil complicated

[raw socket]

?

[fossil sage]

WHich vibevoice are you downloading

[teal crystal]

which do you recommend

[viral mason]

still can't rn gonna be forced out of the house soon

[viral mason]

what is vibevoice??

[simple ore]

tts

[fossil sage]

im not the best at this and i dont know much about tts but im 99 percent sure https://github.com/vibevoice-community/VibeVoice/tree/main

cuz it has finetuning

@teal crystal hop on vc

[teal crystal]

alright bro

[formal steeple]

anyone know why it says dml?

[soft tiger]

well if anyone can help me with getting a unrestricted sora 2 lmk

[flint gale]

Any place I can post furry AI images at?

Before I get comfortable with this place, and does anyone know where to get prompts for NovelAI?

[tame oracle]

#🏙│ai-images is for AI images

[knotty moth]

if you're not kidding, share the public repository of the dataset in huggingface, etc. and if necessary, proof of consent with the producers/VAs owning the voices

[viral mason]

hopefully he's just jokin around

[fossil sage]

i misread your question

i thought you said " do you even think the demon slayer anime creators and voice actors would mind their voice taken and open sourced"

[brittle wing]

is there any reason for tg okada fork to have chime.aws connections on kaggle browser environment

[dense dagger]

can someone please recommend a good rt voice changer?

[knotty moth]

-rt

[patent trellis]

💻 Local Realtime RVC

Guides for Programs that use RVC Models in Realtime for Calls/Games

• Vonovox

A Realtime Voice Changer with similar performance to Wokada Deiteris Fork, with extra features, but supported only for Nvidia GPUs on Windows. and without cloud options GUIDE

• Wokada Tg-Develop Fork

A personal fork (modified version) of Wokada Deiteris Fork, it just adds some Quality of Life improvements to it like supporting Spin Embedder and Audio Effects. Don't expect too much about it since the creator made it originally as a personal project. GUIDE

• Wokada Deiteris Fork

Most suggested WebUI with the best general support for many platforms. GUIDE

⚔️ Wokada Deiteris Fork vs Vonovox

For Windows Nvidia, Both Wokada Deiteris fork and Vonovox have similar performance & quality. Users should read the pros and cons for both and choose based on their differences, such as UI and Vonovox's paid effects.
Read Wokada Deiteris Fork Pros&Cons & Vonovox Pros&Cons

⛔ Outdated/Discouraged

These options are not recommended for use.

• Original Wokada

Not suggested, older versions in youtube tuts are even way worse. GUIDE

• RVC GUI Mainline Realtime

The program is worse compared to the ones above, and much less updated. GUIDE

[knotty moth]

is there any reason to have concern about that? did you get errors on it?

[brittle wing]

No I just have it detected on anti-script software and needed to know it was essential to allow permissions to run.

[dense dagger]

any good rt vc for intel (r) uhd??.. my laptop is crazy choppy

[brittle wing]

Found out it was essential and required so gave it permissions, just cant be too careful on the internet these days yk

[knotty moth]

what "anti-script"? an adblocker?

[knotty moth]

intel UHD is too ancient, please read the spec requirements first

[brittle wing]

Similar but stops independent domains/scripts running and allows permission management, allowing you to see all connected domains/scripts on a page and independently manage permissions.

Useful to prevent javascript attacks, ads, certain types of site-walls ect

[dense dagger]

ik..😭 but like.. is there nothing that i can use?

[brittle wing]

You able to run on cloud instead of using local device?

[knotty moth]

no much idea but it's more of the frontend or tunnel side, maybe try the one other than ngrok

[brittle wing]

Dw solved it. Chime is needed to make the models upload correctly so gave it perms

[tight linden]

https://www.tiktok.com/t/ZTMkgtoxG/

How to use rvc like this?

[outer isle]

Lightning

[teal crystal]

anyone know how to fix this error in vibevoice?

[odd tide]

chat does voice.ai still work and what is rvc

[thick latch]

Not sure if it's the right channel, but any type to give the voice (vonovox) more flexibility? Like, I'm satisfied with the voice. But doing different sounds like for example, trying to whisper and stuff just breaks it

[viral mason]

what

what's you're gpu don't use voice.ai it's the worst

[odd tide]

what do u use

i have a pretty bad gpu

its like nvidia 970

@viral mason

[odd tide]

sorry its NVIDIA GTX 970

[viral mason]

oh uh, use wokada deiteris fork read the guide it's the third oen

-rt

[patent trellis]

💻 Local Realtime RVC

Guides for Programs that use RVC Models in Realtime for Calls/Games

• Vonovox

A Realtime Voice Changer with similar performance to Wokada Deiteris Fork, with extra features, but supported only for Nvidia GPUs on Windows. and without cloud options GUIDE

• Wokada Tg-Develop Fork

A personal fork (modified version) of Wokada Deiteris Fork, it just adds some Quality of Life improvements to it like supporting Spin Embedder and Audio Effects. Don't expect too much about it since the creator made it originally as a personal project. GUIDE

• Wokada Deiteris Fork

Most suggested WebUI with the best general support for many platforms. GUIDE

⚔️ Wokada Deiteris Fork vs Vonovox

For Windows Nvidia, Both Wokada Deiteris fork and Vonovox have similar performance & quality. Users should read the pros and cons for both and choose based on their differences, such as UI and Vonovox's paid effects.
Read Wokada Deiteris Fork Pros&Cons & Vonovox Pros&Cons

⛔ Outdated/Discouraged

These options are not recommended for use.

• Original Wokada

Not suggested, older versions in youtube tuts are even way worse. GUIDE

• RVC GUI Mainline Realtime

The program is worse compared to the ones above, and much less updated. GUIDE

[odd tide]

what is these names lmao

[odd tide]

Also whats rvc? and when i go to the weights.com thingy shouldnt it give me a zip folder?

[odd tide]

is this the right one?https://github.com/deiteris/voice-changer

[viral mason]

don't download the models off weights bc u gotta rename the files which is annoying get em from here
https://discord.com/channels/1159260121998827560/1175430844685484042

[odd tide]

ohk

thanks

[viral mason]

this one
https://huggingface.co/Shadicti/deiteris-Fork/blob/main/voice-changer-windows-nvidia-b2332.zip

[odd tide]

ahh ok

that link is in the github guide

[viral mason]

ye

[knotty moth]

ye you're cooked if you can't afford a pc build with capable RTX gpu

also I can't guarantee the cloud version with kaggle would work

[austere siren]

best TG develop fork settings for AMD RX 6600 and Intel i5 10400f CPU

[safe flame]

Wthelly is this

[inland bramble]

trying to train my own model on gpt-sovits, and while its pretty good, all my outputs sound like its partially speaking through a snare drum/crunchy. is there a way to edit out this sound with specific eqing in post, or a way to train it so it doesn't end up doing this?

[tulip cloak]

any idea why is this happening?
when i use applio colabs and try to process dataset it keeps saying
Starting preprocess with 2 processes...
0it [00:00, ?it/s]
Preprocess completed in 0.00 seconds on 00:00:00 seconds of audio.

using applio colab for the first time btw

[near hornet]

need help with the set up for amd

i dont know why but i can install that in my cmd

[simple ore]

show the preprocess cell screenshot

[tulip cloak]

i got it done, thanks for the help tho

just did not had to include .zip in the name

[ashen bane]

guys can u put voice model on easeus voicewave

[simple ore]

what moron pays for voice changer?

[ashen bane]

i havn't lol

[carmine hearth]

Excuse me, everyone. Does anyone happen to know the current status of the retrained rmvpe? I seem to recall hearing about training to improve rmvpe on the AI Hub about three months ago, but I haven't heard any related news recently.
Has it already been implemented in Applio? Or was it halted due to critical issues?
Please excuse any awkward phrasing; I'm using machine translation.

[analog obsidian]

the project got cancelled, it was found the rvc blips are learned during the training, it's not rmvpe's fault

[carmine hearth]

So I guess I'll just have to recheck my dataset if I end up getting blips... Thanks!🫶

[muted crypt]

Hi, I am for people to resize 3d models to real sizes, I work with a 3d modeling company and we are looking to for relationships with 3d modelers, and other creatives. Send me a DM 🙂

[viral mason]

[void raft]

Explain to me like I am 5 years old, please. I have an elf picture. Her long elver ear is pointing to the side at 60 degree. I want her ear to stand upward pointing into the sky. How can I archive it?

[dry bluff]

hi ai

can u help me

AI

HELP

im tryna find a specific voice model

and cant find

[open gorge]

Hello, is there anyone who can help a friend of mine get unbanned? His account got hacked a couple of months ago, and he recently found out he was banned here. Is there a way for him to rejoin?

[craggy bough]

tell him to dm @hushed perch

[open gorge]

Thank you!

[viral mason]

https://discord.com/channels/1159260121998827560/1175430844685484042

[marble thistle]

does anyone know how to fix "An error occurred during voice conversion. Check command line window for more details."

[simple ore]

well... did you?

[marble thistle]

idk where it is

where command line window

[simple ore]

the window with text

[marble thistle]

idk what window

[simple ore]

the window with black background and white text

[marble thistle]

the command promt thing

[simple ore]

yes

[marble thistle]

i jst restarted it and it works now

ty for helping anyway

[inland bramble]

asking for help again regarding this!!!

[brittle wing]

can someone help me set this up? Im fairly new to this type of thing, like if i shared my screen in a VC or smth

[viral mason]

What gpu do u have

It'll depend on which one u get

[viral mason]

https://tenor.com/view/sonic-sonic-the-hedgehog-waiting-bored-impatient-gif-15251437

[narrow portal]

I don't remember these graphs looking like this

This never happened to me until I moved on from Applio 3.2.9 directly to 3.5.0

Should I be concerned?

[knotty moth]

the recent version uses fp16 precision, while the previous one uses fp32. fp16 is prone to causing exploding gradients while it doesn't happen on fp32 (too large grad values that may be caused by inconsistent/poor quality or improperly cleaned dataset).

[narrow portal]

So, do I keep training or should I stop and downgrade to an older Applio version?

[knotty moth]

I'd need to show you a tweaked training script to switch to fp32 but im not on a laptop rn

[inland bramble]

asking again for any answer!

[austere stream]

so i have a weird bug where if i try inferencing using an rvc model on applio it just flat-out sounds bad even if the OG audio is completely clean
same works vice-versa (inferencing an applio model on rvc sounds bad)
is this a known bug/issue/idk or is there a workaround

model quality is also good so im not sure why its like this, its pretty much dissuaded me from using applio at all

[low shard]

!howtoask

[patent trellis]

❓ How to Ask for Help

✅ Before You Ask!

1. Check Docs & Guides: Your answer may already be in the AI Hub Docs or the https://discord.com/channels/1159260121998827560/1159513888199540817 channel.
2. Search the https://discord.com/channels/1159260121998827560/1192011222023950368 : Look for existing posts that solve your issue. Do not invade someone else's post.

📝 How to ask?

Tell your:

• Full GPU Name: (e.g., NVIDIA RTX 3060)
• Operating System: (e.g., Windows 11)
• Detailed Description: What were you trying to do and what went wrong?
• Tutorial Used: Link to the guide you were following.
• Screenshot: A picture of the full error message is very helpful.

🚫 Prohibited Topics (We Will NOT Help With These)

To maintain a legal, safe & ethical community, we will NOT provide help for:

• (E girl, as an example) catfishing/trolling, scamming, impersonation.
• NSFW/Porn.
• Any illegal activities.
  Requests for these topics will be ignored and may result in moderation action.

<:matsuripray:1159685390156967936> Community Expectations

• Be Polite & Patient: Our helpers are volunteers. You may ping the Helpers role once.
• English Only: Please keep all conversations in English.

[low shard]

please elaborate

[low shard]

I dont really use gpt sovits, and i dont think we have any helpers specialized in it, but be sure to use v2pro+ https://github.com/RVC-Boss/GPT-SoVITS/wiki/GPT‐SoVITS‐features-(各版本特性)

[craggy marsh]

hi, so my current pipeline is i speech through my microphone, then it use stt to give my speech to a llm, then from that llm send a answer through tts, i then call applio api to convert that speech into my preference voice. now i stuck as tts phase, how can i make the speech through tts to have more emotions?

[simple ore]

it can simply be switched in settings

[simple ore]

1. you're using a wrong chart 2) you seem to have a miniscule dataset

[knotty moth]

im sure the avg graph would also be like that, so it'd be the dataset quality issue

[simple ore]

average graph gonna have some gaps, but simply having 10 times logged in just 200e means either a tiny dataset or a wrong batch size

[narrow portal]

So, first, the wrong chart, that's fine, but I must know which one should I look at, since most of them show similar statistics, and second, the small dataset is completely intentional since I was making a test model to see how Spin works (it failed so I tried again with Contentvec this time, the one shown in the screenshot)

[simple ore]

use avg loss charts

[surreal magnet]

Hello, quick questions, i used to have deiteris rvc before vonovox were around, is vonovox better in terms of performance compared to deiteris fork ? (i have rtx 3080 10gb)

currently had 90-100ms GPU delay, is this decent enough ?

[reef totem]

hello i just have a question.. i used beatrice model and its works fine realtime response but when i use rvc voice it's super delayed and lagg please help me... im using amd rx 580 gpu

[knotty moth]

there's nothing wrong to either try vonovox and compare the performance, or keep using the deiteris fork, thinking "if it ain't broken don't fix it"

[knotty moth]

the og wokada version is only good for beatrice model, otherwise you should use one of these recommended versions

-rt

[patent trellis]

💻 Local Realtime RVC

Guides for Programs that use RVC Models in Realtime for Calls/Games

• Vonovox

A Realtime Voice Changer with similar performance to Wokada Deiteris Fork, with extra features, but supported only for Nvidia GPUs on Windows. and without cloud options GUIDE

• Wokada Tg-Develop Fork

A personal fork (modified version) of Wokada Deiteris Fork, it just adds some Quality of Life improvements to it like supporting Spin Embedder and Audio Effects. Don't expect too much about it since the creator made it originally as a personal project. GUIDE

• Wokada Deiteris Fork

Most suggested WebUI with the best general support for many platforms. GUIDE

⚔️ Wokada Deiteris Fork vs Vonovox

For Windows Nvidia, Both Wokada Deiteris fork and Vonovox have similar performance & quality. Users should read the pros and cons for both and choose based on their differences, such as UI and Vonovox's paid effects.
Read Wokada Deiteris Fork Pros&Cons & Vonovox Pros&Cons

⛔ Outdated/Discouraged

These options are not recommended for use.

• Original Wokada

Not suggested, older versions in youtube tuts are even way worse. GUIDE

• RVC GUI Mainline Realtime

The program is worse compared to the ones above, and much less updated. GUIDE

[reef totem]

is there any way or compatible for amd gpu like mine?

[knotty moth]

I'd highlight this for you

you can first try the former since it has some improvement features

[deft reef]

Ai bot cover

[fluid dust]

How to fix the crackle in the voices when i speak with voice changer

[low shard]

!howtoask

[patent trellis]

❓ How to Ask for Help

✅ Before You Ask!

1. Check Docs & Guides: Your answer may already be in the AI Hub Docs or the https://discord.com/channels/1159260121998827560/1159513888199540817 channel.
2. Search the https://discord.com/channels/1159260121998827560/1192011222023950368 : Look for existing posts that solve your issue. Do not invade someone else's post.

📝 How to ask?

Tell your:

• Full GPU Name: (e.g., NVIDIA RTX 3060)
• Operating System: (e.g., Windows 11)
• Detailed Description: What were you trying to do and what went wrong?
• Tutorial Used: Link to the guide you were following.
• Screenshot: A picture of the full error message is very helpful.

🚫 Prohibited Topics (We Will NOT Help With These)

To maintain a legal, safe & ethical community, we will NOT provide help for:

• (E girl, as an example) catfishing/trolling, scamming, impersonation.
• NSFW/Porn.
• Any illegal activities.
  Requests for these topics will be ignored and may result in moderation action.

<:matsuripray:1159685390156967936> Community Expectations

• Be Polite & Patient: Our helpers are volunteers. You may ping the Helpers role once.
• English Only: Please keep all conversations in English.

[fluid dust]

anything but jus saying the fucking answer

[low shard]

Please moderate your tone when asking for help and be more polite, read the help guidelines as we need more info, there are multiple programs, for different platforms and hardware, with multiple different settings, so we need to know exactly which one can work for your setup

[fluid dust]

what is a good program?

[inland seal]

vonovox

[low shard]

there isn't a perfect universally good program

[inland seal]

best one ive seen

[low shard]

nope

[inland seal]

oh go on nick

[low shard]

it's not universally compatible

[inland seal]

whats better

[fluid dust]

finally an actual answer

[low shard]

there isn't a best one

[inland seal]

thank you so much nick

[low shard]

it depends on your hardware

[inland seal]

hes asking for a recommendation

[fluid dust]

not dipshit252: !howtoask

[inland seal]

why dont you help him get one?

[low shard]

I can't know which one works for him if I don't know his hardware

[low shard]

vonovox won't work for you if you don't have a modern nvidia gpu with a modern windows installation

[fluid dust]

is it dr87

[low shard]

that's why we ask for users setups

[wraith tangle]

Nick whats best for modern nvidia gpu on a modern linux installation

[fluid dust]

i got nvidia 4070 and win 11

[inland seal]

yeah vonovox would work fine then

[low shard]

vonovox is windows only, your best bet is to use wokada tg-develop fork which is like deiteris with some Quality Of Life enhancements

[low shard]

yeah that's exactly what I meant with "we need more info on the user setup", I just needed that info rather than making a drama but whatever, it will work on that

[wraith tangle]

ill trust you with this one Nick, have a blessed day

sorry we got off on the wrong foot big dawg

[inland seal]

yeah fr nick

you the best

[low shard]

be sure to check: https://docs.aihub.gg/realtime-voice-changer/local/tg-develops-w-okada-fork/ :)

[low shard]

nah its fine dw

I do understand people do not like reading long paragraphs, but at the same time it will be worse if someone wastes an hour trying to install something not compatible with their hardware, that's why we as helpers we make sure to check that first

[fluid dust]

and they all live happily ever after

[inland seal]

🌈 🌤️

[low shard]

we just care about checking the user setup because it happens daily that someone just wants a quick answer then they waste hours on trying to run a program that wasnt made for their hardware, which wastes both our and your time

I also want to apologize if I sounded rude or anything, I hope yall understand it was just a bit of miscommunication

[wraith tangle]

I'll follow the steps on this site and get back to you if anything goes wrong

I'm using PopOS with Nvidia drivers 570, just so you know

[low shard]

Alright, let me know :)

[low shard]

PopOS is Ubuntu based, so for the Virtual Audio Cable step use that command

[wraith tangle]

Yep will do, don't worry

I also have other programs for virtual audio cables just in case I need a workaround, but I'll try your solutions first

[inland seal]

@low shard do you know how I could have optimal settings for best realistic AI voice on my Vonovox windows 10 12th gen intel R core TM i5-12600k 3700 mhz 10 cores with my 16 gb of 4800 mhz ram and my nvidia RTX 3070 Ti?

[eternal sluice]

hi i keep getting a robotic voice i think my params are wrong but cant found the correct ones

[low shard]

your audio setup should look like:

• voice changer:

  • input: microphone
  • output: line 1

• other programs:

  • input: line 1
  • output: headphones

• pitch: play with it, it can depend on different model and your own voice

• f0 method: rmvpe

• AP BWE 48k Upscaler could help

• Use RNNoise Reduction if you're having noise issues

• crossfade duration: 0.15

• extra time: 4.0

[low shard]

Hello, could you please elaborate your pc gpu, operating system and a screenshot of your current program settings?

There are multiple programs for realtime voice changing, so i hope you understand

[eternal sluice]

yes ofc, can i come in dm?

[wraith tangle]

Nick I love you

[low shard]

Let me know for any issues or workarounds, so that they can be added for specific PopOS future users, since we usually don't have so many Linux users

[wraith tangle]

The Web UI works, I'll play around with it for a while

[inland seal]

Thank you so much Nick I love you, do you accept tips?
On the matter, should I keep smart SiNE on, change the block size at all or change the sample rate? pitch smoothing factor?

[low shard]

If you want we can talk here since I can give you media perms (where any helper can help you), or my dms, which would you prefer?

[simple ore]

you wont be able to use a voice changer and anything else, like a game, at the same time with rx580

[low shard]

If you send a screenshot I can also help you with the settings, or its your choice, I'm glad it's working for you

[eternal sluice]

yes thats fine, so im on windows, with a RTX 3050 and i cant send my config withvoice changer

[wraith tangle]

Only thing that's keeping me occupied right now is figuring out where these ALSA audios are supposed to be directing/redirected from

[low shard]

Thank you so much Nick I love you, do you accept tips?
I don't have a paypal account and could technical accept only giftcards, however I want to remind that no user is obbligated to tip helpers and this is more of a dm talk. I'm glad to hear i'm helping you :)

should I keep smart SiNE on
Yeah, if you want a short explaination, it's a smart algorithm to make sure it doesn't process at all most noises

change the block size at all
that controls the delay, you should increase it if you're having delay issues only

change the sample rate?
You should leave it at the 48k default value

pitch smoothing factor?
You might play with this setting only if you feel like the pitch changes too drastically while using the program

[wraith tangle]

Update, I switched to CLIENT rather than SERVER, however I'll need to play around with some sort of Output

Input works fine, it is just an Output that I have issue finding which will be where the result is piped through

[fluid dust]

whoever recommended ts vonox is so saucy

[low shard]

!give-media-perms 1h @eternal sluice

I'm guessing you're atleast on windows 10 or 11

You should be able to send a screenshot now

[fluid dust]

ty for this sir

[low shard]

Your audio setup should look like:

• voice changer:

  • input: microphone
  • output: line 1

• other programs:

  • input: line 1
  • output: headphones

[inland seal]

np vono is so good 🥵

[low shard]

Vonovox is currently the only in active development RVC AI Realtime Voice Changer

[fluid dust]

BRO

THIS

IS

SO

GOOD

[low shard]

that's why it's technically better than wokada and its forks in terms of performance and newer features, but Wokada and its forks have a more widespread compatibility for Linux, Mac, AMD GPU and Intel GPU users

[wraith tangle]

Yeah alright, so should (Line1) just be a random Output device I don't use, but just to be used as a pipe for this?

[low shard]

sorry, with Line 1 I meant any Virtual Audio Cable you use, Line 1 is suggested for windows but you're on linux lol

[wraith tangle]

Ohh alright let me figure something out

[low shard]

btw, be aware that the only currently in active development rvc ai realtime voice changer is vonovox, which unfortunately doesn't support linux atleast for now

I'm just warning you in case you dual boot

[inland seal]

why arent the rest being developed? no demand?

[wraith tangle]

I don't dual boot, I'm fully linux, but it's fine

If Vonovox is available in the near future for linux, or some sort of plug for linux gets developed, I'll hop on that instead

But this should work fine for now, just need to find some sort of proper virtual audio cable software

[low shard]

original wokada by Wok is currently being developed but only for Beatrice (another speech-to-speech type of model ai) which is lightweight but lower quality compared to RVC, so it doesn't help with RVC

Wokada Deiteris Fork added many performance improvements to the original version, but the developer just doesn't seem to want to continue updating it atleast currently, i tried contacting them but had no answer, ofc thats completely their choice since they aren't forced nor paid by the users as its a free program

Wokada Tg-Develop fork is just a modified version with Spin embedder RVC model support, some quality of life features but it's more of a personal fork than of a "revival official fork"

I just realized it kinda became long but hopefully you understand

TL;DR: the major improvements that can be done are performance and feature wise, and there's not many developers who would do that

[inland seal]

get steven hartlet out of here bro

[fluid dust]

jeez nick

u type alot

[low shard]

They just got banned lol

Smh scammers suck, you don't know the amount of them that usually gets automodded daily here lol

[viral mason]

scammers should just jump off a bridge if they have nothing better to do

[low shard]

I heard about the developer possibly adding AMD GPU Support in the future if they get hands on an AMD GPU, but I'm not sure especially for Linux

The majority of users are on windows (like me)

[viral mason]

one day vonovox will be amd compatible

one day

[low shard]

@inland seal btw is it all fine? or do you need any other help?

[low shard]

Hopefully

[fluid dust]

what is the best vonox settings overall?

[low shard]

you got a pretty similar gpu compared to the previous user, could u try those settings?: #✨│ai-help message

[fluid dust]

uhhh

pl

[low shard]

huh? if you aren't aware, you can click the "button" in my message to get redirected to my other message with the best settings

[fluid dust]

i saw

[low shard]

or i can just send you a screenshot

[fluid dust]

whats extra time and crossfade

[low shard]

oh great

[low shard]

• extra time: Gives the model more or less context to work with. lower values will give less delay but can impact kind of quality, higher values do otherwise
• crossfade: controls how smoothly the AI stitches different processed parts "chunks" of your voice back together, usually 0.10-0.15 are suggested

It's kind of technical, but you should be fine with just setting extra time to 4.0 and crossfade to 0.15 if you just care about suggested values for your setup

[fluid dust]

how do i make it louder

[low shard]

increase the output volume

[fluid dust]

igh

[sonic mauve]

Hey guys, does Vonovox work on an amd cpu and an nvidia gpu?

[low shard]

Are you on windows 10 / 11?

[sonic mauve]

im on windows 11

[low shard]

Then yeah, it's going to work fine

[sonic mauve]

ohh thank you

[low shard]

Here's the guide: https://docs.aihub.gg/realtime-voice-changer/local/vonovox/

[sonic mauve]

i just set up Vonovox for the first time and I have no clue what I’m doing, been trying to find good settings for realtime voice changer but there’s barely any tutorials out there, do you know where i could look??

oh mb i could've just read the guide thoroughly lol

[sleek edge]

not opening

[sonic mauve]

never mind there are way too much information

[low shard]

do you still need help? if you tell me your full gpu name I can help with optimal settings for your setup

[sonic mauve]

i have an rtx 3060 mobile*

[low shard]

please try the Precompiled Vonovox version where you need to only run start.bat

[low shard]

your audio setup should look like:

• voice changer:

  • input: microphone
  • output: line 1

• other programs:

  • input: line 1
  • output: headphones

• pitch: play with it, it can depend on different model and your own voice

• f0 method: rmvpe (or fcpe for a less accurate pitch but less delay experience)

• Use RNNoise Reduction ONL>Y if you're having noise issues

• crossfade duration: 0.10

• extra time: 3.0 or 2.7

• block size: controls your delay, play with it to make sure to not have choppy audio, you could set lower values for lower delay if your gpu can handle it, else set it higher

[viral mason]

what about block size?

that's also important

[low shard]

the message sent too early, I adjusted that and other settings

anyways, there isn't a "perfect" value known for that user's GPU

[sonic mauve]

which embedder do i use?

[low shard]

that depends on the model you're using, most models are contentvec, use spin only if its specified in the model you want to download

[sonic mauve]

Should I mess with pitch smoothing, harmonic strength, and voice texture? I’ve got no clue what they do. Also, which F0 method gives the best quality, and should I go with ASIO or WASAPI for the audio backend?

[viral mason]

always use rmvpe as pretty much all models use it

for f0

probably shouldn't mess with pitch smoothing, harmonic strength, and voice texture. And for which mic to use no clue but just have it be
input: regular headset/headphone microphone
output: line 1

[sleek edge]

start bat does this

[viral mason]

I think u should press any key to continue

[uneven wraith]

guys why doesnt my voice work?

i cant put in any media

[viral mason]

lemme guess, u got the voice changer from a youtube tutorial and it's not working really well

[uneven wraith]

yeah

its not working at all

[viral mason]

thought so, it's outdated delete it

what's your pc gpu?

[uneven wraith]

whats a good one to use now then?

[uneven wraith]

nvidia

[viral mason]

do u know the numbers for it? like 1660 5070 ect

[uneven wraith]

3060

[viral mason]

nice

[uneven wraith]

rtx

[viral mason]

I'd suggest vonovox as it's the best one and is the only one currently still getting updated

-rt

[patent trellis]

💻 Local Realtime RVC

Guides for Programs that use RVC Models in Realtime for Calls/Games

• Vonovox

A Realtime Voice Changer with similar performance to Wokada Deiteris Fork, with extra features, but supported only for Nvidia GPUs on Windows. and without cloud options GUIDE

• Wokada Tg-Develop Fork

A personal fork (modified version) of Wokada Deiteris Fork, it just adds some Quality of Life improvements to it like supporting Spin Embedder and Audio Effects. Don't expect too much about it since the creator made it originally as a personal project. GUIDE

• Wokada Deiteris Fork

Most suggested WebUI with the best general support for many platforms. GUIDE

⚔️ Wokada Deiteris Fork vs Vonovox

For Windows Nvidia, Both Wokada Deiteris fork and Vonovox have similar performance & quality. Users should read the pros and cons for both and choose based on their differences, such as UI and Vonovox's paid effects.
Read Wokada Deiteris Fork Pros&Cons & Vonovox Pros&Cons

⛔ Outdated/Discouraged

These options are not recommended for use.

• Original Wokada

Not suggested, older versions in youtube tuts are even way worse. GUIDE

• RVC GUI Mainline Realtime

The program is worse compared to the ones above, and much less updated. GUIDE

[viral mason]

just follow that super awsome guide, the first one

[uneven wraith]

ok

tysm

[viral mason]

if u have any questions ask here ^^

and avoid yt tutorials they don't know anything about the good stuff

[uneven wraith]

oh alr

ty

[viral mason]

np

[sonic mauve]

I also had a voice changer from a YouTube tutorial called Realtime Voice Changer by Okada, but it made my PC heat up quite a bit. Is Vonovox a better alternative? It seems to run cooler from what I’ve seen

[viral mason]

that's because the one u had was outdated and had way worse optimization

all 3 new ones run better, and have better code

[sonic mauve]

Makes sense, I didn’t realize mine was that outdated. Glad to know the newer ones are better optimized.

do you know what index files do?

[viral mason]

they are the file with the model's "accent" bacially how they talk/how they say stuff

it's needed to make the model sound more like itself

[sonic mauve]

Ohh I see so if I want it to sound more like my accent, I don’t need the index file then?

[viral mason]

not really no, it's kinda confusing

since models are really just voice filters if u think about it, it'll have your accent when u speak through the voice changer

[uneven wraith]

once i set up my app can u help me with my settings?

🙏

[viral mason]

it's why when a model is singing over a song where the original singer is say british they'll become british

[viral mason]

vonovox right?

[uneven wraith]

yea

[viral mason]

def

[uneven wraith]

im extracting the stuff rn

im not rly into these typa stuff

[viral mason]

in what way

did u read the rules, if you're egirl trolling I cannot help u and you'll be sent to a meat grinder

[uneven wraith]

wait

pause

[viral mason]

no like

you'll die

/j

[uneven wraith]

hm

[sleek edge]

it just leaves the application

[uneven wraith]

😭

[sonic mauve]

what does this do?

[uneven wraith]

i retreat all i ve said

🙏

[viral mason]

basically a pitch shifter

[viral mason]

u should just download vonovox 1.6.9 as it's the most recent one

[sonic mauve]

Should I just leave it as it is?

[sleek edge]

i just downloaded this 2 days ago bruh

[uneven wraith]

btw why do ppl sell stuff like this that u can get for free

literally

😭

[viral mason]

yea you won't need it, it's just for fun

[sonic mauve]

gottcha thank you very much

[viral mason]

because they're greedy and stupid

take advantage of ppl who dunno where the actual good stuff is

[uneven wraith]

i js extracted everything

leme check

what i gotta do

next

[viral mason]

https://github.com/dr87/Vonovox/releases

[viral mason]

did u get vac lite

[uneven wraith]

ye

[sleek edge]

i already have done this

[viral mason]

cool, so open the folder for it and run the setup64, not 64a or the regular setup

for vac lite

[uneven wraith]

okok

i did

that

already

i downloaded

vonovox

169

what do i do now

[viral mason]

alr so now run the setup.bat file in vonovox

then when it's done (it takes a while) run the start.bat

[sleek edge]

start.bat is gonna get smoked

[uneven wraith]

hmm

okok

i trust u

[viral mason]

I gotta take a dump brb

[viral mason]

run

[uneven wraith]

ait

[sleek edge]

guess what

[uneven wraith]

u gon get viruses

ur system 32 will get deleted

[sleek edge]

sybau grey

im sending it so i can fix the problem

[uneven wraith]

btw i js got ur ip

off that

[sleek edge]

ok buddy

[uneven wraith]

lol

im js joking

[sleek edge]

yeah like i didnt realize

foid

[viral mason]

Chicken butt

Sorry

What happened

Lol

[uneven wraith]

sir

tell me

what do i do here

[viral mason]

Hmmm

[uneven wraith]

\

[viral mason]

Ok follow these settings rq, u don't need to mess with the whole program
blocksize: 3.0/2.7 (it depends just mess with it)
crossfade duration: 0.15
extra time: 4.0

Everything else just leave as is

[uneven wraith]

blocksize is max 3.0

1.0*

[viral mason]

Uhh

Wdym

[uneven wraith]

\

[viral mason]

Click any button to continue, then once that closes run the start.bat

[viral mason]

Oh

[uneven wraith]

yh

[viral mason]

Uhhh

See what works for you, lower it some and if it's too choppy raise it

Just gotta mess with it

[uneven wraith]

ima try on max

hope it works

[viral mason]

And since you're using a female voice try leaving it around pitch 3 or 6, if you're a guy

If you're a girl just leave at 0 might work fine

Depends on your voice

[uneven wraith]

mk

tysm

il play around w the settings

[viral mason]

Np, remember to put in the index btw

[uneven wraith]

whats the best girl voice

???

[viral mason]

Shhh do that on your own I can't help with that

[uneven wraith]

wdym put it in the index

???

[viral mason]

See where it says index file and it says none

Go find the index file for that model

[uneven wraith]

ok

[viral mason]

Should be in the same folder

[uneven wraith]

so

i should make voice file

go in the vile for voice changer

?

[viral mason]

I recommend putting all your voices you download in a new folder

Just all in one thing to keep track of them