✨│ai-help

[teal rain]

@viral mason sorry to interrupt do you know alot about video ai too?

[viral mason]

If you're going to use the og one yes

[viral mason]

I'm sorry I don't

[teal rain]

np thanks do you know anyone here who is good?

[viral mason]

I'd ask around but maybe some of the mods or helpers might?

Maybe lyery

[teal rain]

TYSM!!!

[viral mason]

No problem! Good luck with your video ai journey ^^

[mortal crag]

I want to model my own voice, but when I try it with Google Colab, it gives me version errors. What should I do? Is there a working Collab?

[rancid pine]

How large of a batch size for one hour of a dataset?

[simple ore]

I have not seen the dataset information about this model, I imagine it needs a set of images (or a video) to learn the motion of objects and stuff

so perhaps every 10th frame of a video

I dont think there are training script to finetune the model though

[viral mason]

If you're using kaggle I'd say use batch 12 (put 6 because it multiplies the batch size)

[rancid pine]

I see. The OG pretrain is simply toggling the "pretrain" on? Or do I uncheck that as well...

[viral mason]

The og pretrain is the one used if you have turned the custom pretrain off

[viral mason]

Oh! Yes keep that option on

The "pretrained" one

Keep that checked

[rancid pine]

Thanks

[viral mason]

Np!

[crystal whale]

@viral mason heyy

[viral mason]

Hello!

[crystal whale]

could u help me w setting up my rvc

[viral mason]

What kind of rvc? The voice changer or are you trying to train a model?

[crystal whale]

voice changer yes

this one right?

i have rtx 4070 super

[wraith jacinth]

<@&1159293204038955078> can you help me

[teal rain]

i see if i figure it out ill share it here if anyone is interested thank you so much!!!

[viral mason]

You can use either that one or Vonovox

[lean linden]

how I make my own voice model?

[short torrent]

on start up in w okada theres no error messages but there's nothing coming from my mic whenever it's on

what could be the issue

[simple ore]

you need to at least show the screenshot of the app screen

[knotty moth]

no one cares

!howtoask

[vernal bone]

❓ 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 lega, 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.

[outer isle]

in Google Colab, when I finished the model for Ariana Grande (Blackiana) model, I got the pretrained error:

Pretrained model sample rate (40000 Hz) does not match dataset audio sample rate (32000 hz)

[knotty moth]

that sounds self-explanatory, please check the model sample rate setting you're using, or use another pretrain that match the sample rate

[rancid pine]

Is it normal for one epoch to take one hour to train?

[prisma silo]

what do i extract the files with?

[craggy bough]

what you normally use to extract files

[prisma silo]

i use winRAR but it doesnt pop up as an option

[rancid pine]

Is it?

[crystal whale]

which one would be better?

[crystal whale]

@craggy boughhey

[crystal whale]

@helper

<@&1159293204038955078> i think my virtual audio cable is bugged

[jaunty shale]

yeah I've been using it after the Google colab thing that happened, I had no issues till yesterday

[viral mango]

When merging in wOkada, what does it mean to merge less than 100% of a model?

[pastel oak]

Percentage ratio. It's a little unnecessarily complicated the way wokada does it
With an example: youd normally say 75 / 25 between two models. In this case model A influence is 3x more than B. You can do that in wokada or you can do 100 / 33 and have the same result

[viral mango]

Got it. Thank you!

[pastel oak]

Can you please be more descriptive I can't do anything with this information cause you're not sure yourself

[pastel oak]

No

You're not using your GPU or you have 500 hours of training data

Probably first

[pastel oak]

Of what program

[celest rover]

Hi everyone,

I’m trying to train a Retrieval-based Voice Conversion (RVC) model, but my PC is CPU-only and too low-spec to handle it locally.

I’ve searched around, but most of the Colab notebooks I’ve found are outdated (from 2023), disabled, or require payment.

I’d really appreciate:

Any working, free Colab notebooks for RVC training

Or if someone’s willing to train the model for me if I provide the dataset

Thanks a ton for any leads! 🙏

[simple ore]

depends on gpu and the dataset size

[brittle bobcat]

kraggle still doesn't work anymore. who wrote the original code?

[simple ore]

what kaggle

what does not work

[brittle bobcat]

https://docs.aihub.gg/realtime-voice-changer/cloud/deiteris-w-okada-fork-kaggle/

that no longer works

i think someone else here also was having issues

but i dont think a lot of people use kraggle

or just us are experiencing the issue

[PYI-133:ERROR] Failed to start embedded python interpreter!
Fatal Python error: init_fs_encoding: failed to get the Python codec of the filesystem encoding
Python runtime state: core initialized
ModuleNotFoundError: No module named 'encodings'

Current thread 0x000079c6cf079340 (most recent call first):
<no Python frame>

it had been working for weeks

[low shard]

Cloud shouldn't be ur first option, you should check if your PC GPU is good enough for locally first

[low shard]

Train (make) RVC Models on cloud:

1. Prepare the Dataset
2. Setup RVC:
   Choose a cloud way to use RVC,

• Google Colabs (max 4 hours of daily T4 16gb gpu not granted for free, but easy to use, there's a paid tier):
  • Applio (UI)
  • Mainline (U, END OF LIFE)
• Kaggles (a bit harder to use and needs phone number but gives 30 hours weekly of better gpus, either T4x2 16gb each or P100 16gb, only free):
  • Applio (UI)
  • Mainline (UI, END OF LIFE)
• Lightning.ai (Kinda hard, needs login, no issue with web uis or anything, but only free 15 credits monthly, Free Studios run 24/7 but require restart every 4 hours. There's a paid tier):
  • Applio (UI)

3. Be sure to know about the tensorboard

Google Colab = Easier but risk of getting disconnected
Kaggle = Harder but more gpu time
If you want the easiest way and for free, is using https://weights.com/ which uses RVC

RVC Inference (use models) on pre-recorded audio on Cloud

You can use either:

• Weights.com: Easiest Possible Ever Automatic
• Ilaria RVC Zero: Fast and free on cloud
• Applio UI Colab: RVC Fork with some extra features like TTS
• RVC AI Cover Maker UI: Automatically Separates the Vocals and Instrumentals, converts the voice and mixes them back

[low shard]

Hello, we can't know your issue with little info

Please Elaborate:

• your PC GPU (Such as Nvidia RTX 3060, AMD RX 9070 XT..)
• your operating system (Like Windows 10, Windows 11, Linux, MacOS 26..)
• what you want to do? There isn't a program that does everything, there's a program for each thing:
  • AI Covers
  • Train RVC Models
  • E girl trolling
  • TTS
  • Roleplay in VC
  • Roleplay in Games
  • etc
• what tutorial link are you using
• a screenshot of the program

[low shard]

Please elaborate more

[sand forge]

yo

[sand forge]

@low shard i need to use okada

idk how

[viral mason]

Vonovox is better in general but deiteris has a bunch of model slots

[rancid pine]

Makes sense

[simple ore]

unless you run it on cpu 🙂

[low shard]

Hello :)
Please Elaborate:

• your PC GPU (Such as Nvidia RTX 3060, AMD RX 9070 XT..)
• your operating system (Like Windows 10, Windows 11, Linux, MacOS 26..)
• what tutorial link are you using (if any)
• a screenshot of the program (if any)

[sand forge]

my gpu rtx 3090

[sand forge]

Windows 11

[sand forge]

im not using tuturiol link

[sand forge]

im using o okada.

[low shard]

great

[low shard]

can you share an entire screenshot of the issue then?

!give-media-perms 1h @sand forge

[sand forge]

no issues, i dont understand how to use it becuase

i want to use it while talking

but it does not work

also im worried if other people voices echos trough it

[low shard]

I see, but sharing a whole screenshot of the program will help me correct your settings and check if you got the updated version

[sand forge]

okay

one second its installing rvc

[low shard]

You're using an old version of original wokada

Original wokada isn't suggested since a lot

you found it on github or youtube?

because all video tutorials are outdated

[sand forge]

it says updated 2month ago

i found it on hugging face

[sand forge]

github

[low shard]

original wokada only has User Interface changes, it doesnt have the performance of Wokada Deiteris Fork, it's not suggested

[sand forge]

it has new interface

[sand forge]

o

so which is latest wokada?

[low shard]

New Interface doesn't mean it's better, it unfortunately doesn't change any other things

[sand forge]

ohh

how do you find the new wokada?

[low shard]

The program depends on what you want to do:

• roleplay in games
• roleplay in calls
• voice change on pre-recorded audios

[sand forge]

probably to do videos

i dont like trolling

or these type of stuff

[low shard]

like, videos of roleplaying in a game or smt?

[sand forge]

i guess

[low shard]

be sure to play on lowest graphics 1080p 60fps cap then

[sand forge]

i do streams etc.

[viral mason]

finally another normal user

[low shard]

-realtime

[patent trellis]

💻 Local Realtime RVC

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

• Wokada Deiteris Fork

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

• 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 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

[low shard]

you can use either wokada deiteris fork or vonovox

[sand forge]

which one you suggest?

[low shard]

vonovox has more recent updates on performance
but it's just better to read yourself the pros&cons of each

[sand forge]

well i have an nvidia gpu

do you think vonovox would be better

if it support multi platform then i would go with it

[low shard]

yeah, that's why I said you can choose either one of them, they both have different pros&cons, that's why it's best you choose for yourself

non windows/nvidia users would only be able to use wokada deiteris fork infact

[knotty moth]

the latter has non-Nvidia gpu & cpu support and is more beginner friendly

[solar crown]

Yo can anyone help me i got my voice changer and everything i hear also myself with the voice changer but cant use it on discord on discord i only hear my own voice

[sand forge]

@low shard

done i installed it

what now sir.

[low shard]

did u install vac lite?

[low shard]

show an entire screenshot of the program and also link the tutorial / download link u used

[sand forge]

here

[low shard]

you also did the default playback and recording tab step?

[sand forge]

didnt do anything

i just put model

i found in find models

idk what else i should do

[low shard]

follow https://docs.aihub.gg/realtime-voice-changer/local/vonovox/#virtual-audio-cable to get vac lite

A VAC (Virtual Audio Cable) makes a fake audio device, used to re-route the audio of different programs

It's used to get the output of wokada/vonovox as the input in other programs

[sand forge]

done

i installed

its in sound setting now

what next?

[solar crown]

I cant send pictures in here

[sand forge]

@low shard

[viral mason]

why the egirl model

[sand forge]

idk i selected

random

idk how to use that voice changer

its very hard

[low shard]

!give-media-perms 1h @solar crown

now you can

[low shard]

are u trying to do e girl trolling or just roleplay on streams btw?

[low shard]

also please dont unnecessarily ping people, he's not an helper

[solar crown]

https://www.youtube.com/watch?v=SxdnGxicJOg&t=15s / https://github.com/w-okada/voice-changer/blob/v.2/docs_i18n/README_en.md

[low shard]

that video is outdated
you got an over year old version of original wokada, and vb audio cable is not suggested for windows
delete everything

I'm guessing you're on windows 10/11, are you trying to do e girl trolling like in the video, roleplay in games/calls or voice change on pre-recorded audios/songs?

[solar crown]

I want to use #1175430844685484042 in discord calls

[low shard]

oh so just roleplay in discord vc, right?

[solar crown]

yea

[low shard]

well, delete everything you got from youtube

-realtime

[patent trellis]

💻 Local Realtime RVC

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

• Wokada Deiteris Fork

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

• 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 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

[low shard]

read the pros&cons of both vonovox and wokada deiteris fork, to choose which one you want

[solar crown]

👍

[simple ore]

@white aurora what applio version is that?

[white aurora]

3.3.1

[simple ore]

why tf it logs at 2, 4, 6

unless you have a tiny set and selected a huge batch size

btw, update to 3.4.0

[white aurora]

There you go

[white aurora]

Currently, I'm using finetuning the original pretrain that RVC comes with.
The dataset consists of the following:

55 hours of singing and speech from 3 languages -> EN, CN and JP. The dataset itself is sourced from GTSinger.
I highly suspect that the model has started to overtrain after around 10-12 epochs. The loss curves flatten out while the generator adversarial loss (g/adv) begins rising, and both generator (norm_g) and discriminator (norm_d) gradient norms spike sharply. I could highly be wrong here though as this is my first time finetuning a model.

[analog obsidian]

did you denoised gtsinger?

btw i'd recommend m4singer instead

[white aurora]

Yeah, I did beforehand

[analog obsidian]

welp there you go, that explains why grads go to stratosphere

denoising pretrains is never a good idea

but also might that you're using an old version of applio

which quite literally had a chatgpt discriminator

they fixed it in the latest version

gtsinger is really a bad dataset, i dont think it would give good results neither as finetune or from 0

[royal kettle]

How did ChatGPT make it worse?

[analog obsidian]

changed a number and thats why no one was able to train a model from scratch

[royal kettle]

A single number?

[analog obsidian]

it was also affecting finetuning but since they're way smaller, the damage doesn't happen fast

while pretrains always exploded

[analog obsidian]

yup...

a single number

maybe gpt thought it was a good idea to change it

if you finetune models i highly advise to use mainline instead

[royal kettle]

Wow, why was old ChatGPT allowed to make changes to sensitive stuff?

[analog obsidian]

well the gpt thing is a theory

[royal kettle]

Why? Didnt you just say Applio is fixed now?

[analog obsidian]

we don't know if they really used gpt

but most of us here believe its indeed gpt

[royal kettle]

I see.

[analog obsidian]

if turns out gpt 3.5 was used
you think it would be safe to use a trainer that potentially may have gpt code?

🤔

but ye they fixed the discriminators.py file and it seems things got improved

[royal kettle]

If its fixed then it doesnt really matter, no?

[analog obsidian]

noobies havent tried training from scratch in applio so we dont know if applio is truly "fixed"

im training from scratch using mainline

i just tested a small finetune of 15 mins and i did noticed things got improved compared to before tho

back then this model was generating garbage spectograms, but after the fix now it seems to be on pair with mainline results

so answering to the question hmmmmmmmmmmmm, yea i think its safe to finetune small models in applio, no idea about training from 0 tho (without a pretrain)

[analog obsidian]

like, a full run of training a model without a pretrain

instead of just 10e

[simple ore]

no it wasn't chatgpt

it would not have made that mistake

it was a manual optimization and missing a very important parameter

like try to spot the difference here

it was manually "optimized" to this

[simple ore]

personally i got rid of gtsinger and switched to M4

gtsinger was very inconsistent

it only took 5e for og pretrain to retuned for spin v2

[simple ore]

i'm gonna do that

fp16, single mel

[marble onyx]

I have the link for early access for perplexity comet browser with free subscription for students
does sending it in the group is against the rules?
Someone please clarify this.
I think it may helpful for the students who can't afford or don't want to spend for subscription

[red kayak]

@analog obsidian moral of the story trust llms more than humans

[analog obsidian]

[arctic badge]

hello i need some help when ever i use realtime voice changer and i speak my voice keep getting cut off any ideas ?

I HAVE THIS VERSION

the cable im using is this

[low shard]

that seems the latest version of wokada deiteris fork, could u please show ur whole settings too while ingame/discord vc?

be sure to play on lowest graphics 1080p 60 fps cap

[arctic badge]

im not even in game and it keep cutting

[arctic badge]

i remember i deleted junk files this morning from my pc

everything was great yesterday

till now

i deleted the software and download it again

[low shard]

f0: rmvpe without onnx
extra: 2.7

on wokada deiteris fork, you can **optionally **use more advanced settings for benefits:

• Advanced Settings -> Force FP32 mode: on (THIS IS OFF BY DEFAULT! Turning this on improves stability. Increases VRAM usage by 200 MB)
• Advanced Settings -> Disable JIT compilation: off for faster loading speed of the program, on for slightly better performance (10-15 ms) for Nvidia only)
• Advanced Settings -> Crossfade Lenght: Controls how smoothly the AI stitches different processed parts "chunks" of your voice back together. 0.1 for fastest voice, 0.15 for improved quality but increases delay by ~50 ms
• Reduce the delay on Windows via the Wasapi / Asio Guide

[arctic badge]

and still having the same problem

[low shard]

you could also try other models

or try Vonovox

[arctic badge]

all my models having the same problem ig its from something i did

[arctic badge]

its not about this honestly i meant whenever i talk and i hear myself its like the voice cut off

not like it used to

[low shard]

what exactly junk files?

[arctic badge]

like unwated files

[low shard]

did you touch anything related to the voice changer?

[arctic badge]

didnt check honestly i deleted them by app called driver booster

[low shard]

thats weird, could u please try reinstalling or using vonovox?

[arctic badge]

i did reinstall actually but still the same problem from where i can download vonovox ?

[arctic badge]

its like the software not catching my voice sometimes

[viral mason]

https://github.com/dr87/Vonovox/releases/tag/v1.6.6

[low shard]

https://docs.aihub.gg/realtime-voice-changer/local/vonovox/

[low shard]

that's the old version

[arctic badge]

when i say like hello it sound hel

[viral mason]

I saw in the server people saying to stay on that version as the current one has issues

Just playing it safe

[low shard]

where did you hear that? could u share a link of that convo?

I haven't heard of any issues related to 1.6.7

[viral mason]

Go to the general page in the Vonovox server

It's not far

[arctic badge]

okay

[arctic badge]

its like im lagging or the site

[low shard]

#1338004708287053894 message this?

[arctic badge]

what now ,

done

[low shard]

setup.bat

[viral mason]

Yes that and this right here
#1338004708287053894 message

[arctic badge]

IS THIS NORMAL ,

?

[low shard]

doesn't seem to happen to me

[viral mason]

Odd

[low shard]

did u run it as admin? u shouldnt

[simple ore]

setup only if you have python installed

[viral mason]

I'm still waiting for the update where there's more than 8 slots for models before I switch completely to vonovox

[arctic badge]

no

[simple ore]

do you have python installed?

[arctic badge]

i guess

[simple ore]

open cmd

type where python

[arctic badge]

in microsftapp

[simple ore]

install python 3.11

[arctic badge]

already

[simple ore]

that's too new

[simple ore]

https://www.python.org/ftp/python/3.11.9/python-3.11.9-amd64.exe

whyyyy

[arctic badge]

he told me to ;-;

[simple ore]

you gonna install shit under windows/system32/runtime and it not gonna work

[arctic badge]

should i remove it ?

what am i supposed to do now ?

@simple ore

[simple ore]

dont run as admin

obviously

[arctic badge]

but i already did

should i alt f4 ?

[simple ore]

close that window

[arctic badge]

done

[simple ore]

download the source code and unzip it to C:\Vonovox

or any other drive

if you have nvidia 5000 series, run setup 5000

otherwise run regular setup

[arctic badge]

i have 3060

[simple ore]

read what it says... if there are errors show them from the start

[low shard]

setup5000 doesn't exist anymore in vonovox

it got merged with normal setup, so there's only 1 setup.bat

[arctic badge]

https://tenor.com/view/no-brain-loading-slow-gif-16361840

i lost u

[simple ore]

okay

setup.bat then

[arctic badge]

btw this is for amd its alright ?

[simple ore]

amd64 is 64-bit windows

[low shard]

amd64 is intended as 64 bit CPU, it's called AMD because AMD invented that extension, every modern computer uses a 64 bit cpu and operating system

[low shard]

seems to be working

[arctic badge]

whats next

[low shard]

wait

till it finishes

[arctic badge]

too many download

[viral mason]

You got this!

You'll live

[arctic badge]

imagine still having the same problem after

[covert ruin]

the voice changer randomly stopped working???

[low shard]

Hello, could you please elaborate more? There are multiple voice changer programs and versions

[ionic sail]

Good evening! Hoping to confirm ideal settings while gaming for Chunk + Extra in W-Okada using a 4070TI? I seem to be having difficulty coming up with a decent combination.

[ebon tapir]

Someone help

How to get voice changer

[knotty moth]

python versions are NOT backward nor forward compatible, it must be exact as what it requires

so if it says 3.11, you should get 3.11

[split lagoon]

when i click start i still continue hear my own voice

help pls

pipeline deleted it says

[errant wyvern]

Does anyone have the download link for the voice changer and virtual audio cable that they could pass on to me please? I can't find it, and I need to download the virtual audio cable too

[ionic sail]

https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#crackle-fix

[viral mason]

did u check if they use nvidia or not

[ionic sail]

Nope. There are instructions for all GPUs/OSs on that guide however.

[short torrent]

is kaggle/collab needed for the voice changers

[hallow thistle]

[hallow thistle]

What do you use W-Okada for? Trolling or catfishing with E-girl voice model? And what is your PC GPU? The virtual cable I'd preferred to use is Virtual Audio Cable lite.

[hallow thistle]

!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 lega, 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.

[hallow thistle]

Read help guidelines before you start ask anything here.

[hallow thistle]

Any Python version greater than 3.10 or 3.11 is too far new, and some packages that were made for Python 3.10 won't gonna work with 3.13.

[hallow thistle]

Also, running with administrator can cause several issues when trying to install or run the program, so try run the program as normal privilege instead.

[hallow thistle]

You seem to use the original version of W-Okada, which "Deiteris" fork W-Okada doesn't look like this.

[short torrent]

could someone send how my mic and virtual cable should look like on discord & games

I feel like i'm using the wrong line

[hallow thistle]

!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 lega, 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.

[hallow thistle]

Read help guidelines before you start asking here.

[wild bobcat]

how can i make an rvc voice out of pitched down recordings of myself

in the sense of:

i talk in the c3 or c4 range (or c2 to c3 range if youre using the other pitch notation system)

and if i feed rvc these pitched down recordings of myself (c2-c3/ c1-c2)

will it be a robotic mess

or will it be able to somewhat do it?

[wild bobcat]

@simple ore

one guy suggested i should perhaps ping for you this

[crystal whale]

<@&1159293204038955078> can yall help me set this up

[pastel oak]

Settings are fine

[hallow thistle]

These settings look fine. How about your entire screenshot?

[crystal whale]

im thinking of switching to vonovox

@hallow thistle

[crystal whale]

thanks

[hallow thistle]

Vonovox is far less documented than Deiteris' fork W-Okada in the guide, so if you know how to use Vonovox you can go for that.

[crystal whale]

isnt it similar?

which one is better?

[hallow thistle]

This is Vonovox. https://raw.githubusercontent.com/dr87/Vonovox/main/docs/images/ui1.png

[crystal whale]

oh no its so complicated

[brittle wing]

Hello! I have installed vcclient_win_cuda_2.1.4-alpha.zip
however i can't choose my GPU RTX 5090 how can i fix this? I get pytorch error in logs

[hallow thistle]

You seemed to use the outdated W-Okada version. What do you use W-Okada for? Trolling or catfishing with E-girl voice model?

[knotty moth]

wrong version

pls get the correct one here: https://github.com/IllIlIlIllIl/voice-changer/releases/tag/b2335

or you could try vonovox https://docs.aihub.gg/realtime-voice-changer/local/vonovox/

[brittle wing]

i can't extract the zip for some reason?

[hallow thistle]

Do you even download them?

[knotty moth]

also:

• what tutorial link are you using/referring?
• what purpose you want to use the voice changer?

[brittle wing]

yeah i have 360MB/s

do i need the both z01 & z02 too?

[hallow thistle]

That's internet speed.

[brittle wing]

None just found the github page

[knotty moth]

found it out of nowhere?

not like from some youtube tutorial?

[brittle wing]

Nope, heared a guy in my irl job using it

@knotty moth do i need MMVCServerSIO.2.zip only or also
MMVCServerSIO.2.z01 and z02?

[knotty moth]

try ask him, did he actually use that alpha 2.x version?

what gpu is his pc spec?

did he actually get it working already?

[brittle wing]

Yeah i guess i dont need voice support then

Thanks for the help tho

[hallow thistle]

Wait what? Are you gonna give up it here?

That's crazy. So have a good day then.

[crystal whale]

i have 2332

should i also download this version

[hallow thistle]

b2335 is for NVIDIA GeForce RTX 50 GPU only. You have GeForce RTX 4070 SUPER, so you do not need to use that version.

[brittle wing]

Got it to work but doesnt output on virtual audio cable

Cable input is set in output tho

nvm

had 2 audio cables

[rose hemlock]

when I speak into the thing, the delay is so long for the ai voice

[quaint plume]

Which rvc voice changer could i download for nividia rtx 3060

[hallow thistle]

!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 lega, 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.

[hallow thistle]

What do you use W-Okada for?

[quaint plume]

trolling

[hallow thistle]

What trolling? With anime character or e-girl voice model?

[quaint plume]

e girl voice one

i downloaded python it wont load the voice changer

[hallow thistle]

[quaint plume]

Like the thing is my voice changer model wont load up

when i click on the https

[hallow thistle]

This is just straight up bad.

[quaint plume]

any fix

[hallow thistle]

E-girl is not allowed here, sorry about that.

[quaint plume]

what about other anime character

I need help downloading the rvc

[covert ruin]

what do you mean

https://ibb.co/r25hFTT4

[low shard]

you're using an over 2 year old version of original wokada

and vb audio cable creates issues on windows

delete everything

im guessing you want to do realtime voice changing for roleplay right? because that won't do ai covers / use on pre-recorded audios

[covert ruin]

it was working just fine ☹️

[covert ruin]

yeah

what should i do then

[low shard]

your version was used in 2023, using that has way worse performance

[low shard]

delete everything

-realtime

[patent trellis]

💻 Local Realtime RVC

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

• Wokada Deiteris Fork

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

• 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 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

[low shard]

get either wokada deiteris fork or vonovox

reach each pros&cons

[low shard]

@covert ruin let me know later for any issues :)

[brittle wing]

hi!
Could anyone help me with finding the right model? (local hosting)
I'm using a MBP m4 pro with 24 gigs of ram

I'm mainly planning on using it for studying/general stuff (asking questions and stuff)

I'm using LM studio btw.

[covert ruin]

which one would u suggest

can i use the same modules there btw?

as in the old version

[low shard]

Are you talking about LLMs?
You could try Gemma3 12b or deepseek-r1 7b, not sure if they could fit but be sure to use quantized versions like q5 or q4

[low shard]

It's better to read each pros&cons, tho vonovox still gets performance updates, your choice

[covert ruin]

cant find the download link for vonovox

[low shard]

they both use RVC v2 Speech-To-Speech Models

[covert ruin]

alright

[low shard]

it's all in the guides, please read them carefully
did you get vac lite first?

and be sure to have deleted what you previously had

[covert ruin]

i checked the guide

[brittle wing]

I've tried these ones:

• Gpt-Oss 20b: Too censored. It also hallucinates alot. When I try searching with it, it keep denying the search results (claiming they're "not possible")/generating the wrong syntax for tool calling. I've tried the uncensored version but I find it too slow (q4/q5)(Tried the David AU version). I'm currently using the 4-bit mlx model.
• Mistral 24b: It's great but it struggles with physics/chem questions.
• Magistral 24b: Great with those questions but doesn't support VL
• DeepSeek(using Qwen) 14b and Qwen 14b: Thinks too much + no VL

I've tried Gemma12b VL too but it tends to get into a continuos loop and crashes 😭

[low shard]

you got vac lite and did the setup thing first, right?

[covert ruin]

no

where may i find that 🥲

[low shard]

Here https://docs.aihub.gg/realtime-voice-changer/local/vonovox/#virtual-audio-cable :D

be aware that it's crucial for any voice changers

[covert ruin]

ohhhhh

the virtual audio cable

vac

i did that

[low shard]

i see you had VB Audio Cable in the screenshot

it's not the same program

if you are talking about vb audio cable, uninstall it from windows app settings https://support.microsoft.com/en-us/windows/uninstall-or-remove-apps-and-programs-in-windows-4b55f974-2cc6-2d2b-d092-5905080eaf98

many users reported issues with vb audio cable on the long run randomly on windows

that's why VAC Lite is suggested instead

[brittle wing]

Is there an uncensored version of gpt oss with mlx q?

[low shard]

I've tried Gemma12b VL too but it tends to get into a continuos loop and crashes 😭
Have you tried using quantizations like q4_k_m? i use ollama instead on a windows nvidia gpu, so i'm not sure about the mac support

[low shard]

also nop i dont think there's any other gpt oss uncensored models, as of now

[brittle wing]

Yeah. It's still kind of slow though

I get 90t/s on mlx

and 20t/s with q4

[brittle wing]

There's a David Au one? Btw why isn't there a fp16 version of it?

[low shard]

There's a David Au one?
yes there is that one, i said i don't think there's any other one

[brittle wing]

I thought you could only quantize a model (while it being useful) only if you quantized from fp16/32?

[low shard]

i mean, you either have good performance or good quality

[brittle wing]

Oh

[brittle wing]

the q5 one

It doesn't really run on my machine 😅

[low shard]

have u tried llama 3.2 vision 11b?

mac isn't really the best for AI as you see

[brittle wing]

😆 Ik

[brittle wing]

I'll try that

[covert ruin]

no i mean i already did that

where may i find the vonovox link tho

[brittle wing]

Thanks

Also are there any 24b< models with good VL and reasoning?

[low shard]

here https://docs.aihub.gg/realtime-voice-changer/local/vonovox/#download-nvidia-on-windows

[brittle wing]

Something maybe close to 14? maybe 20b?

I've tried using 8b and 12b models but they've not been "good enough"?

Btw what do you use locally for general stuff?

[covert ruin]

Thanks for the help! Do you have any recommended settings or any tips

[low shard]

f0: rmvpe
extra time: 3.0
play with the pitch

is it all solved?

[covert ruin]

is there a loopback test option

i really liked that option on the old programm

that i was hearing myself all the time

?

[low shard]

set monitor to ur headphones to hear urself

[covert ruin]

is there such option on the menu or

[low shard]

oh right nvm you're on vonovox not wokada deiteris fork my bad, there's no current built in option like wokada deiteris fork yet, but there is a workaround: https://docs.aihub.gg/realtime-voice-changer/local/vonovox/#how-can-i-hear-myself

[covert ruin]

i dont think its working

im not hearing anything

always somethnig has to go wrong

now im trying to install the other voicechanger and im getting this error https://ibb.co/1BnHFzM

[low shard]

!give-media-perms 1h @covert ruin

[low shard]

show a screenshot of the line 1 properties

and the vonovox settings

[low shard]

are you sure you want to use wokada deiteris fork? vonovox got more recent performance updates

[covert ruin]

from the sound application?

this?

[low shard]

you shouldn't open that

[covert ruin]

oops

[low shard]

did u do this?

[covert ruin]

this?

[low shard]

so you checked listen to this device, clicked apply and ok, right?

after that, in vonovox, click the model preset, then click start

[covert ruin]

ok it started wokring

but its reallyyy buggy

half of my words dont register

its inaudible

[oak edge]

yOO GANG

@low shard you there bro?

ahhh so the doubt is in applio training

it's about like using the output and re applying voice to it again

let's say I have this sample X initially

so I first apply my voice to it and get the output Y

now if I put the Y as input and get the output Z

and keep continuing like this will the quality and pronounciations improve?

[low shard]

can you show again your settings now? be also sure to play on lowest graphics 1080p 60fps cap and to close useless things in background

[low shard]

are you like trying to use the model on other audios, to make artificial data to retrain the model to get better?

[oak edge]

OOPS SORRY I MEANT INFERENCE

not trainin

[low shard]

so like, inference the same audio over and over for it to get better?

[oak edge]

yess

like X->Y

Y->Z

and goes on

get the output, put that as input convert

[low shard]

i never tested such thing, but I don't think it would improve

[oak edge]

again get that output, put it as input convert

[oak edge]

okiee

[simple ore]

that's one of the craziest ideas I've heard... go for it

give it a try

[oak edge]

💀 ayo what

aaahh i'll try

[covert ruin]

play what on olwest graphics

[oak edge]

lemme give it a try and see how it goes

but judging by nature, have nobody ever tried it though?

[low shard]

if you're playing any games, lower the graphics

[low shard]

fr i thought he was talking about artificial datasets before

i mean, i dont see why it would improve tho

[analog obsidian]

no

[oak edge]

[covert ruin]

why are all my words muffled tho

its like i got something in my mouth

[low shard]

please show a screenshot of ur settings

[simple ore]

probably not

[low shard]

put crossfade duration to 0.15, play with the pitch, set extra time at 2.7, maybe lower the output volume

try other models too, not every model is perfect

[covert ruin]

other models are the same

why

it was working perfectly on the old app

i have to scream in order for the app to work

[frosty lion]

i have a problem while loading voices in okada vocie changer

some of the files have a audio logo on them and cant be put on voice changer

if someone knows any answer to that it would be really helpfull

[covert ruin]

why is this error popping up

[knotty moth]

it is corrupted, try do chkdsk and redownload the application if necessary

[covert ruin]

try to do what??

i redownloaded it

already

[knotty moth]

extract using 7zip instead of built in one

[kindred eagle]

I’m looking to train a singing voice model entirely from scratch (no pre-trained weights) I have approx an hour of clean data set. Most RVC/Applio setups seem to require a pretrain. Does anyone here know of an app, tool, or workflow (preferably a Kaggle notebook) that lets me train a high-quality singing voice model purely from my own data?

[analog obsidian]

to train from scratch you need at least 24 hours of data

[kindred eagle]

aight i can source that but where do i train it

[covert ruin]

???

[analog obsidian]

uh sorry i don't know

[covert ruin]

what does this mean

[kindred eagle]

ah sheet

[covert ruin]

does this mean its extracted or

[covert ruin]

i downloaded the other vc

what settings should i change because im still kinda muffled

[low shard]

#📰│dev-updates 👀

[shrewd jay]

I got a problem, I got followed a online tutorial for MMVCS and when I speak into my mic I don't get any audio back and my specs are
CPU: Intel (R) Core (TM) Ultra 7 265F
GPU: NVIDIA GeForce RTX 5070

[low shard]

tg-develop's fork of deiteris fork just got added to the docs 2 minutes ago lol

[low shard]

show a screenshot of the entire program, and give the tutorial link you used

[covert ruin]

huh

[shrewd jay]

https://www.youtube.com/watch?v=SxdnGxicJOg&t=186s

Also I don't got pic perms

[covert ruin]

whats that mean

new update?

[low shard]

i just checked the tutorial you used, it's outdated asf
delete everything

are you trying to do roleplay or e girl trolling/catfishing?

[low shard]

new version , check #📰│dev-updates

[shrewd jay]

Im trying to troll my friends in vc

and other voice changers are paid and doesn't sound good so I heard that this one was good

[covert ruin]

which of the 3 versions tho

great after i downloaded my whole audio is completely messed up, the vc doesnt work at all

???

and the whole installation is so complicated

for what

great and the voicechanger doesnt even reopen at all

every single thing that couldve went wrong went wrong wtf

can i get help

[shrewd jay]

Nevermind got it working

[solar laurel]

how do i use the gvoices in voice model
the links go to the website but how do i know the best settings
im using voice changer client

[viral mason]

which voice changer there's multiple

[solar laurel]

like which voice or client?

[viral mason]

which client

[solar laurel]

realtime voice changer

[viral mason]

there's the original, deiteris, and vonovox

I'd ask u to send a screenshot but I can't give you perms

[solar laurel]

onnxgpu-cuda is all i see

i could dm u

or link me the bestr one to download

[viral mason]

which gpu do u use, nvidia, intel, amd?

that will depend on which one u can use

[solar laurel]

nvidia

[viral mason]

u can download either of these then

-rt

[patent trellis]

💻 Local Realtime RVC

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

• Wokada Deiteris Fork

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

• 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 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]

if u don't need a bunch of voices at once loaded and want better quality u should use vonovox

[solar laurel]

okk thank you

let me try ovnox

[solar laurel]

is there a video i can watch to set it up easier

[viral mason]

no, all you do is download it, unzip the file, click on setup, then after that finishes press start

[solar laurel]

yeah thats what im doing does set up close itself

or when it stops downloading i close it

[steep hemlock]

quick question, can vonovox generate text-to-speech?

[viral mason]

when it's done just close it, then open the start

[viral mason]

I don't think it can, it's used for speech to speech

but u can ask in the official discord server

[solar laurel]

ohh this looks tuff

could you help me get my first voice running so i understand how to use it

do i just download the file form the website extract it and it does all the settings for it automatically?

[viral mason]

lemme see if I have the power to give u image perms
-# I probably don't

[solar laurel]

xddd

[viral mason]

wdym?

ye download it, extract then click the file that says startup

[solar laurel]

no no i meant the voices

i have vonovox open rn

[viral mason]

ohh

just download the voice from here
1175430844685484042

[solar laurel]

?

[viral mason]

oops

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

here lol

[solar laurel]

ok ok let me try to do it myself and if i get lost if u can help me i appreciate it

cause this is different to the one i was using

this 1 looks better

[viral mason]

btw make a folder for all the voices u download so u don't lose them

like just a new folder

[solar laurel]

yep dw i alr did

does vonovox do the settings for me alr or do i have to tweak it

cause the one i downloaded sounds so poo comapred to the sample

[solar laurel]

on the weights website do i click the 3 dots and download

all the ones im trying just sound demonic

[viral mason]

ye

in case u want any of my models
https://www.weights.com/user/your_local_worm

[solar laurel]

when downloading models am i meant to change pitch formant and stuff

or no

[viral mason]

it depends, if it's a female model and you're a guy with a deep voice you may have to up the pitch

formant usually is just for fun

[solar laurel]

ok thank youu

im understanding this now

the omni man one is so valid xdd

[steep hemlock]

do u know any other TTS software i can download

becuase i cna remember i used one a year ago but i cnat remember the name

[solar laurel]

when i use my virtual cable the mic is not working anymore idk what happened

it only picks up if i switch it to my main mic but only in app

[ember cave]

whenever i run my start,https and the command prompts runs it instantly disappears within 2 seconds and doesn't even download the stuff

@low shard

[solar laurel]

can anyone help me its not picking up the voice

[ember cave]

run help for help

[solar laurel]

wdym

[viral mason]

Your input should be your regular mic and your output is cable

[ember cave]

can u help me too

ill give u 5$

[viral mason]

What

[solar laurel]

ohh tyyy

[viral mason]

Lemme know if that works ^^

[ember cave]

@viral mason can u help me fix too

whenever i run start.https BAT FILE it runs and then disappears

and nothing happens

[solar laurel]

it worked

[ember cave]

nth downloads

hello

?

.

[solar laurel]

did you do setup first

[ember cave]

which set up

i downloaded file

[solar laurel]

above start

[ember cave]

and made the shortcut

[solar laurel]

theres setup

[ember cave]

no

ima do that

[solar laurel]

ohh ur doing that video

i think

[ember cave]

ye

[fickle bough]

Is w okadas voice changer is what everyone uses ? Or most people

[hot lagoon]

Dose uvr work with a 5090 and also what fork support the 500 series?

[shrewd jay]

Does anyone got a good rvc model?

[viral mason]

Go look here
https://discord.com/channels/1159260121998827560/1175430844685484042

[fickle bough]

Ive tried downloading it ran the http file and it just opens and closes lol

[ember cave]

very bad graphic card

you need smth custom

to use this

@solar laurel

help me

man

[viral mason]

Patience

[ember cave]

alright

[viral mason]

Which voice changer are you trying to use? And what gpu do u have

[ember cave]

[fickle bough]

Which gpu does it need to work

[ember cave]

will this run on a 4070

[fickle bough]

I’m guessing an 1060 gtx 6gb will not get it to work huh

[viral mason]

Well depending on which one u have depends on which one u can use

[ember cave]

why am i getting ignored

[viral mason]

Is that Nvidia, Intel, or amd

[ember cave]

someone help me

ill give a nitro

[fickle bough]

Nvidia my bad

And Intel cpu

[crystal whale]

i can help

[ember cave]

FR

?

[crystal whale]

yeah

[ember cave]

whenever i run my start,https and the command prompts runs it instantly disappears within 2 seconds and doesn't even download the stuff

[crystal whale]

which voice changer are u using

[viral mason]

Use this one, there are two voice changers u could use, Vonovox and wokada deiteris

-rt

[patent trellis]

💻 Local Realtime RVC

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

• Wokada Deiteris Fork

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

• 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 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

[fickle bough]

goat thank youu

[viral mason]

Vonovox has better quality and other stuff, but only has 8 model slots

And deiteris is not asssssss good but it's got over 100 model alots

[fickle bough]

hmm what would you rec

[crystal whale]

thats the only disadvantage?

8 slots?

[crystal whale]

whats your gpu?

[ember cave]

4070

[viral mason]

If you don't need many models loaded at once I'd say Vonovox since it's the best quality wise and won't be as delayed

[crystal whale]

you downloaded the wrong version

[ember cave]

oh did i

[shrewd jay]

Yo guys what do you use for ur voice changer

[crystal whale]

check dm

[crystal whale]

wait wdym better quality? like better quality audio?

or is it faster

[viral mason]

It makes models sound more realistic when speaking and such,

It also has eq and other effects built in

[crystal whale]

whats eqq

[fickle bough]

do i have to download the vc cable? im tryna use it for fl studio to record vocals lol

if its possible to do that w the ai voice thing idk

[viral mason]

Basically like microphone suppression

To keep out background noise from messing with the voice changer

[fickle bough]

not sure how do i get models?

[viral mason]

U download models from the voice model section or off of weights.gg

There's a blue link in the voice model section in this server
https://discord.com/channels/1159260121998827560/1175430844685484042

Look for any character or person u want

It's probably there

[viral mason]

Yes it's needed for it to work properly

[stoic zephyr]

what do i do if mine keeps like cracking and taking SO long

it just keeps cutting in and out

[viral mason]

I'm not sure about that

It's most likely VB cable causing the crackling

Download vac lite it's in the guides

-rt

[patent trellis]

💻 Local Realtime RVC

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

• Wokada Deiteris Fork

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

• 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 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

[fickle bough]

i cant hear nothin lmao

im using a focusrite too btw

[viral mason]

What's that

[crystal whale]

@viral mason hi i downloaded vonovox

could u help me set it up

[fickle bough]

audio interface

[fickle bough]

im tryna hear my voice on fl studio

Pressing start warming up voices idk what I’m doing 💀

[viral mason]

Use your regular mic/headset microphone for input, and output as vac lite or VB cable

[viral mason]

The only way I figure this out is slightly complicated

[fickle bough]

Rlly

[crystal whale]

yea

wasapi or asio?

@viral mason

[fickle bough]

I’m pressing start and it says warming up voice convo plz wait and idk just keeps loading lmao

[azure patio]

guys what rvc are available rn

[viral mason]

wasapi is less confusing so use that

[viral mason]

if it does that uh

try closing the app and reopening

[viral mason]

are you trying to train a voice model or use a realtime voice changer?

rvc doesn't stand for realtime voice changer btw

[fickle bough]

How do I open the app without going into the start batch file everytime

[viral mason]

you have to open it with the start file every time

or it won't work lol

[fickle bough]

Oh true well not sure how I can get this to work with fl 😭💀

[viral mason]

it would be a complicated setup

you'd need what you have now plus voice meeter, and a second vb cable thing

[marble cedar]

hey is there a better way to run / host local models, and predictions, than replicate - i was trying to get a model to work but am kinda stuck

[crystal whale]

@viral mason what abt these settings?

[viral mason]

I would ask dr87, the person who made the voice changer as I don't use vonovox rn and don't know the optimal settings for it

I think the discord server is in the guide to vonovox

if I try sharing here I can't

[crystal whale]

u can dm me

[viral mason]

sure!

[twin estuary]

how can i reduce this ?

wtf

@viral masonu have any idea why ?

@simple orecan u help ?

[viral mason]

what are you trying to do with the voice changer?

[twin estuary]

i wanna make shorts for ytb

[viral mason]

it shouldn't be going that high if you're not playing any video games while using it

what settings are u using and what version is that

[twin estuary]

last version

[viral mason]

could send a screenshot this doesn't help lol

[viral mason]

what settings do u have, chunk size ect

[viral mason]

ah, those are kinda bad I believe those are the ones given when u first download it, us these instead

[twin estuary]

the problem will be fixed ?

[viral mason]

most likely

if not I don't know what would help

[pliant knoll]

Hi, Can someone help me create the most professional way of creating a model of RVC
-I have DataSet (30 Mins)
I want someone with good GPU who can train this for me around (1000 Epoche)
I want it to be highly professional
(I have a lot of gigs like these)
I can pay a bit

Dm me/ Add me

[viral mason]

do it yourself, it's free

[brisk trail]

Hey peoples! I am a newbie model maker (I'm not planning to get the role at the moment tho) and I wanted to know what's the best model trainer out there right now that I can use, I use applio which most the times leaves heavy artifacts when I say the letter "s" or "f" and I wanted to know how to fix that, perhaps another trainer can help me?

[fossil glen]

-colab

[patent trellis]

📒 Google Colab Notebooks

Google Colab is a Cloud (Remote Good PC) Service. While the Free plan provides up to 12 hours of daily usage, the GPU is typically available for only about 4 hours each day on average.

• **Applio**

by IA Hispano
Google Colab

• **RVC Mainline**

by Hina
Google Colab

• **UVR5 NO UI**

by Eddy
Google Colab

• **UVR5 UI**

by Eddy
Google Colab

• **Wokada Deiteris Fork**

by Deiteris & Hina
Google Colab

• **Hina's Modified Original Wokada**

Google Colab

• **RVC-AI-Cover-Maker-WebUI**

by Shiro & Eddy
Google Colab

• **FaceFusion UI**

by Nick088
Google Colab

• **FaceFusion NO UI**

by Nick088
Google Colab

• **Music Source Separation Training (Inference)**

by Jarredou & Makidanye
Google Colab

[ebon crag]

so i just the onnxgpu cuda 18a file, and got it setup, even got a custom voice in. the issue i run into is, i have a 5060ti 16gb gpu, and a ryzen 7 5800x cpu. when i attempt to use my 5060ti in the gpu slot, no voice comes through with all the settings correctly, but when i switch it to cpu, it works okay-ish, is that an issue on my end that i can fix?

[rancid pine]

Are these normal graphs while training or should I restart altogether?

[grizzled quartz]

who got a clarence claymore voice changer

[pastel oak]

You're using an old ass version that came out before the RTx 5060 it won't work

-rt

[patent trellis]

💻 Local Realtime RVC

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

• Wokada Deiteris Fork

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

• 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 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

[pastel oak]

Use vonovox guide

[stray hull]

how to install?

[hazy ice]

i have only one working voice and it beatrice who can help me

[proven aurora]

Any ways to optimize the client with AMD GPUs?

[wild charm]

e

how to stop 5-8 second delay when talking

[stray hull]

how to use the voice changer in valorant? icant hear my team mates

[low shard]

Hey, could you please show a screenshot of your program and valorant settings?

!give-media-perms 50m @stray hull

[stray hull]

wait ill send

[low shard]

I'm guessing you're talking about the realtime voice changer?
It would be glad if you show a screenshot of your current program, you might have an old one or bad settings

[wild charm]

ight lemme ss rq

gimme a sec

[low shard]

!give-media-perms 55m @wild charm

[stray hull]

here

@low shard

[low shard]

show also your program settings please

[low shard]

Yep, you got an over 2 years old program, and vb audio cable gives issues randomly on windows

dont trust youtube video tutorials for realtime voice changers

[stray hull]

i have a question

my mic in game settings will be vb settings?

ang my output settings will be my headphone?

[low shard]

The issue is your whole setup (both programs) is old, it was used back in 2023, but that's really outdated nowdays

It's better you delete everything

[stray hull]

is there any i could do?

and how to fix?

after i deleted ?

[low shard]

Yeah, there are newer programs, but you gotta delete what you currently have

uninstall vb audio cable from windows app settings

delete the original wokada folder and zip

[stray hull]

and then?

[low shard]

then after that

-realtime

[patent trellis]

💻 Local Realtime RVC

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

• Wokada Deiteris Fork

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

• 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 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

[low shard]

get either vonovox or wokada deiteirs fork (or even tg-develop's wokada fork which is in the docs)

You can choose which, by reading each pros&cons

[stray hull]

i install vonovox

is this?

[low shard]

You got an over 2 years old outdated program (original wokada), and vb audio cable creates issues on windows

What's your PC GPU? I'm guessing you're on windows 10/11

[wild charm]

let me check rq

[low shard]

this is VAC Lite which is a must, click yes

[low shard]

If you don't know how to check your pc gpu on Windows, do:
ctrl+shift+esc (task manager) -> Performance tab -> GPU

[stray hull]

i already installed now

what to do now in game?

[low shard]

did you do the recording and playback setup?

[low shard]

You got AMD Integrated Graphics, that is pretty too weak to run any AI Locally and it would only use your CPU which isn't good

Your pc is too Weak for Wokada locally, You got 3 options:

• Buy a pc with a better GPU if possible
• Run it locally (on ur pc) using the CPU mode of the wokada fork which has better performance https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/ (but this isn't suggested as it could be unstable)
• Use **cloud **(remote good pc):

About Cloud, there are different services:

• Kaggles (30 hours weekly of better GPUs, T4x2 & P100, harder to use, requires an account and a phone number, doesn't allow a Web User Interface for free it usually doesn't get detected but if it does you could get in trouble):
  • W-Okada's Deiteris' Fork Voice Changer Kaggle (the most free gpu hours)
• Lightning.AI (75 hours monthly of free T4 gpu, harder to use, requires an account and a phone number, allows web user interfaces) :
  • W-Okada's Deiteris' Fork Voice Changer Lightning.AI (the safest for free)
• Google Colabs (4 hours daily of free T4 gpu, easy to use, require only a google account) :
  • W-Okada's Deiteris' Fork Voice Changer Google Colab (currently works only on google colab PAID tier)

[low shard]

you're still using the over 2 years old program, delete original wokada, that's not vonovox

delete everything you previously had please, that is too old

using that program will get you worse performance

[wild charm]

ima be honest i have no idea but thanks frrr

ima find sum gpus

[low shard]

The issue is your PC is too weak to run AI, you can either buy a new one, or use a remote good pc (cloud) for a limited free time (which depends on the platforms),
which do you choose?

[wild charm]

uhhh

ngl gon choose 2nd option

[stray hull]

what should i download here?

[low shard]

Sure,
About Cloud, there are different services that you can choose depending on your needs:

• Kaggles (30 hours weekly of better GPUs, T4x2 & P100, harder to use, requires an account and a phone number, doesn't allow a Web User Interface for free it usually doesn't get detected but if it does you could get in trouble):
  • W-Okada's Deiteris' Fork Voice Changer Kaggle (the most free gpu hours)
• Lightning.AI (75 hours monthly of free T4 gpu, harder to use, requires an account and a phone number, allows web user interfaces) :
  • W-Okada's Deiteris' Fork Voice Changer Lightning.AI (the safest for free)
• Google Colabs (4 hours daily of free T4 gpu, easy to use, require only a google account) :
  • W-Okada's Deiteris' Fork Voice Changer Google Colab (currently works only on google colab PAID tier)

Each link is a guide, please let me know if you don't understand something

[low shard]

download the zip source code

[wild charm]

ight thanks

https://tenor.com/view/tf2dance-tf2meme-medic-tf2-medic-hooh-gif-20895057

[stray hull]

i need to restart after set uping?

[low shard]

Nop, let it keep downloading

[low shard]

well, let me know

[stray hull]

yes still installing and downloading

what should i do now?

[low shard]

delete that one

close it, then delete the original wokada folder

it's not needed

please forget about what they say in video tutorials, all info like crepe is all outdated

[low shard]

click start.bat

[stray hull]

ok

what shoul i do now?

[low shard]

f0: rmvpe
extra time: 3.0
input: microphone
output: line 1

in game:
input: line 1
output: your headphones

then, click start

[stray hull]

im hearing my voice now

how do i know is on in game?

i only hear my voice but the voice changer is good

[stray hull]

wat is wasapi?

in my audio>?

[pastel oak]

Windows Wasapi is an audio system your pc uses

There's nothing relevant you need to know about it in the context of the voice changer

[pastel oak]

Test on discord or any playback with the Line 1 virtual cable as Input

[sand forge]

hello.

@pastel oak would you be able to assist me

[pastel oak]

Just ask someone will help

[sand forge]

voice changer does not seem to be working

i dont know how i can know if its working or how i am able to use the voice changer in voicechat

im using elmo model

its cartoon chracter

[pastel oak]

Outdated version

[sand forge]

is it model fault?

[pastel oak]

-rt

[patent trellis]

💻 Local Realtime RVC

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

• Wokada Deiteris Fork

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

• 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 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

[pastel oak]

Wokada deiteris fork or vonovox both are good

[sand forge]

i installed it from -rt tho

[pastel oak]

Everything's explained in the guidea

[pastel oak]

You didn't

[sand forge]

i think i did the

[pastel oak]

No

[sand forge]

official w okada.

not deitris

[pastel oak]

Official is not in -rt at all

[sand forge]

?????????

u sure

if you click this it will redirect you to the offical github.

[low shard]

you shouldn't use that one,

Note that those 3 links are just for reference to the Source Code Github Repositories of both projects, you should instead follow the guide below

[sand forge]

deiteris fork does not work well for me

i dont know how to install

[low shard]

Did you read the wokada deiteris fork guide? could u tell me what you didn't understand?

[sand forge]

when i get the file

what i do

unzip or 7zip

@low shard

what i do now?

[low shard]

7zip is just a program to unzip the file

[low shard]

do you use a download manager? the file still seems downloading

[boreal magnet]

Guys i have the problem that when my chat is too long it starts to lag around, does anyone know how to fix this without starting a new chat?

[sand forge]

Sadly it wont let me

install

i tried all browsers

[pastel oak]

there is a compiled download link on the guide

stop clicking on github you dont seem to know what youre doing

[vital solstice]

Can you do voice to voice AI model with GOOGLE Colab ?

[sand forge]

oh thanks

[karmic cairn]

anyone who can help in n8n and meta developer call back url config??

[sand forge]

@pastel oak done,

[pastel oak]

rest of the steps are on the guide but to sum it up
f0 det rmvpe
chunk 200
extra 2.7
select your gpu

audio server is better
input [windows wasapi] your mic
output [windows wasapi] line 1 (virtual cable)
monitor is optional to hear urself with the voice changer, your normal headphones

[rancid pine]

Are these normal graphs while training or should I restart altogether?

[sand forge]

output does not have line1

is this correct sir.

[pastel oak]

input your mic

but the output is correct yea

[sand forge]

\