✨│ai-help

[sour sun]

AH I was, in fact, inferring with contentvec

Yup, the embedder was correct as well

All right thanks! Looks like I was just inferring with the wrong thing. I'll cross my fingers and pray this time

Can v2 be downloaded from the og LegacyCore thread on here, or is it somewhere else?

[simple ore]

v2 is in the Applio main branch, we gonna make a release today

(for realtime)

it will include spin-v2 as well

[nova willow]

what is the command for ways to train my own data for voice cloning

[olive escarp]

Does anyone have any advice using the AI voice changer in like FiveM? I’m facing a few issues

[serene swift]

Any app that supports models and can do text to speech?

[vague latch]

i have the cable thing installed and its still not detecting my mic

how do i fix?

[dry maple]

why my okada voice changer only lag on when i use voice models?

[viral mason]

What's FiveM

[olive escarp]

GTA RP

Like gaming

[viral mason]

Depends on which one ur using btw, Vonovox or Deiteris okada

[olive escarp]

Vc client ??

[viral mason]

Might have to lower game settings and possibly the settings of the vc

[viral mason]

Could u send a screenshot of it if possible, that isn't the most helpful info

[olive escarp]

Ah I can’t right now sorry.

https://youtu.be/SxdnGxicJOg?si=TryoachufgYdQQ3j

This one ^

[viral mason]

If you downloaded it from a yt tutorial it's definitely over a year old

Use one of these

-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⛔ Outdated/Discouraged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]

Btw if u have Nvidia I'd use Vonovox but if you don't u can use Deiteris

[olive escarp]

I don’t have Nvidia

I have the other one

Forgot what it’s called

I definitely have more than 6GB tho I have 8GB

[dry maple]

amd?

[olive escarp]

Yes that’s It!

Which one should I use?

[dry maple]

i dont know

my okada lag when i use models

[olive escarp]

Wokanda or Vonovox

[viral mason]

U have to use wokada deiteris since you have amd

It should look similar to what you've got but it just runs better in general

[dry maple]

okada deiteris is other version?

[viral mason]

Yea it's the first guide

If u don't have an Nvidia GPU download that

[dry maple]

ok

i will try

[olive escarp]

OKOK

so I should

get rid of the one I have currently and install this one?

@viral mason I dont see where I can dowload it from that link

[viral mason]

Should be there if you scroll down some

It's in the guide

[viral mason]

Make sure to take a picture of the models u have in it if you have a bunch you want to keep

So u remember what ones to put but in

[olive escarp]

Ive got them haha

Thank you!

Ummm

Where do I

Find the link I dont see it

[viral mason]

I can find it for u lemme look

https://github.com/deiteris/voice-changer/releases/download/b2332/voice-changer-windows-amd64-dml.zip

this is the amd version

[olive escarp]

Thank you

[viral mason]

Np man

If u need any more help like for settings lemme know

[olive escarp]

Ahhh I may do actually

Its currently just opening up

this one feels alot smaller in size download wise?

Is that normal?

[viral mason]

Wdym smaller in size?

[simple ore]

dml version is comparably small

[dry maple]

b ro

i love yu

with just 1 link you solve my problem

[viral mason]

Thanks?

[olive escarp]

The download took way less time

less taxing on my pc already

like this is great

[viral mason]

Yea it's much better and more optimized than the one u had

It should run better too

[olive escarp]

I LIED

I have Nvidia

It asked for my graphics card on this one

its a 3070ti

is that good idk?

[viral mason]

u should be good to download Vonovox then

[viral mason]

-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]

@fierce stream

[paper flare]

bro it dont work

deitris fork

im nvidia

[undone belfry]

why is my ai talking like 10 years after i spoke can somebody help me.

[fierce stream]

does anyone know why voice doesen't transmit in CS2 ??

[nova willow]

what are some good render farms that I can pay with paypal that I can use with ai

[covert ruin]

can someone help me fix my microphone???

when i select it as an option on any platform it doesnt send out audio, but when its set as an input on the voicechanger and i select passthrough i can hear myself

VAC messed everything up help pls

[covert ruin]

anyone?

its really annoying that i cant use my regular mic

[weary socket]

I trained about 10 models yesterday, and all stopped at 40epochs because the discriminator reached overtraining threshold. Is this normal?

I feel the threshold should not be enforced on the discriminator

[viral mason]

I think you shouldn't use the auto overtraining detector

[weary socket]

ok

so you monitor it manually？

[viral mason]

Yes, the spikes that go down you should check those, if it's going up it's bad

[bold aurora]

why does it say no wave file found?

I already copy and pasted the wav file path on Applio

[serene dune]

How can i use realtime voice changer on stream?

[copper ferry]

I downloaded the first one and tried a model from #1175430844685484042 but perf is too high (3400 for instance)

[viral mason]

Perf?

[viral mason]

Look at the guides, all u do is set your mic settings correctly and it should work perfectly with ofc game graphics on the lowest settings

[copper ferry]

Yes

Oh nvm it's okay now

Btw how am I supposed to smoothen the robotic cracks?

[willow lintel]

Anyone know what app would be suitable for a voicechanger in mac i5? Ive been trying so hard but no results. It would be great if someone also gave me a tutor

[hollow wren]

when i use free models it works fine but when i use an imported one it lags and cuts my mic

[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]

This specific W-Okada version may fit with your Apple Mac Intel, but the program will only use CPU which is slow. https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#download-mac-intel

[hallow thistle]

Which NVIDIA GeForce GPU?

[hallow thistle]

What is your PC GPU? It seemed like the other user didn't ask you about this one.

[hallow thistle]

Are you trying to record your W-Okada with OBS Studio or something?

[split pelican]

i need help

[hallow thistle]

What are you looking for?

[split pelican]

Like when i open the start_ thingy and let it loads it says done but it dose not open anything like it just says done

[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]

"start_" thing doesn't mean anything. You either looking for "W-Okada the realtime voice changer", which one specific version has two batch files named "start_http" and "start_https".

[tidal urchin]

Hello. I'm new to this server. I have interest in ai field. So I'm learning machine learning(main language python) and also pytorch deep learning concepts.But I'm confused how can i strengthen my ability more.I want to make cool stuff like other developers do. I am self learner so if u guys have any suggestion on how i can improve my skill, i will be grateful for it.

[hallow thistle]

The thing is there is no absolute way to make your coding skill to be great like other programmers or developers. But when you learn along with coding, you'll get better with it. Especially, the print("text") is one of the first Python commands I learned.

[dusk oyster]

Hi, not sure if this is the correct channel to ask for help with W-Okada but for some reason the voice sounds robotic and staticy. I followed the "Deiteris' W Okada Fork" from the Guides channel in this discord
The thing is, when I use the "Mainline, by RVC-Project" from the Guides channel with an audio file the same voice model works very well so I believe it must be something wrong in my W-Okada settings

[willow lintel]

<PUT IN THE PATH TO YOUR MMVCServerSIO FOLDER HERE> what does this mean

[hallow thistle]

Remember the folder path where "MMVCServerSIO" extracted. Now on Terminal, replace the <PUT IN THE PATH TO YOUR MMVCServerSIO FOLDER HERE> with your folder path, so the command should look something like xattr -dr com.apple.quarantine ~/Desktop/MMVCServerSIO.

[willow lintel]

it says no such files

[hallow thistle]

[willow lintel]

wait hold on lemme copy

[hallow thistle]

What? Did you extract the zip file? So where did you extract it?

[willow lintel]

in the downloads

OOH

i get it

[hallow thistle]

This is the correct channel for asking W-Okada. The thing is, "mainline by RVC-Project" is not W-Okada but rather a different RVC program for different purposes.

[willow lintel]

wait lemme try

[dusk oyster]

When I take a recording of my voice and use the audio file in the main rvc project it sounds good. Is there a way I can make W-Okada work that good too?

[hallow thistle]

What is your PC GPU?

[dusk oyster]

rtx 4060 ti

14th gen i5 cpu and 32gb ram

[hallow thistle]

Now try these settings in your "Deiteris" W-Okada:
Chunk: around 50 - 90 ms
Extra: 2.7 s
F0 Det: rmvpe

[dusk oyster]

still very robotic, and not picking up well/staticy

my perf number is yellow in the 80s when not talking, then goes higher and turns red when i start talking

[hallow thistle]

Increase chunk number up until your perf number green and stable. Additionally, you can try go in Advanced Settings, enable Force FP32 and set Crossfade length to 0.15.

[willow lintel]

@hallow thistle srry to ping you but thansk so much, it worked

[hallow thistle]

You're welcome.

[solar bone]

is it possible to export tensorrt stable dfussion from google collab

[short torrent]

There's no sound whenever I use the realtime voice changer inside of games or discord, I set everything up and I can test my mic inside of discord settings and hear the voice changer just fine but whenever i'm in a vc they can't hear any voice being heard i've seen a few threads of this issue but none have had an answer

[dusk oyster]

when im not talking into my mic the perf is all over the place between 60-120 and it seems like the more i raise the chunk that number just keeps getting higher

for some reason when i enabled passthrough i talked and my mic seems extremely loud

[hallow thistle]

Raising your chunk number while perf number always being "red" is not quite possible, unless you run the program alongside a game or there's something wrong with your GPU.

[hallow thistle]

What are you looking for? What is your PC GPU? And did you follow any tutorial video from YouTube before?

[dusk oyster]

it only turns red when i start talking

my gpu is at 1% and cpu at 11% while talking according to task manager

[hallow thistle]

Did you set "GPU" to your NVIDIA GPU and not "CPU" on your W-Okada?

[dusk oyster]

yes, but for some reason my gpu stays between 0-1%

my cpu is only going up to 29

no game running currently btw

[hallow thistle]

That means W-Okada doesn't use your GPU but instead CPU for whatever reason. You either restart W-Okada or restart your PC as solution.

[dusk oyster]

ok i will restart both (sorry i dont know this stuff very well😅)

[knotty moth]

please show the version you're using in a screenshot (first ask me for image perms)

[knotty moth]

for the robotic voice issue, try another model that may sound better

[dusk oyster]

can i have image perms please

[knotty moth]

!give-media-perms 30m @dusk oyster

[dusk oyster]

and how where do i find the version to screenshot?

[hallow thistle]

Just simply screenshot your whole W-Okada.

[dusk oyster]

it was set to 2.7 not 5

for some reason after restarting my pc it reset

[hallow thistle]

[knotty moth]

to recap, what issue are you having now?

[dusk oyster]

here is my advanced settings also i use steelseries for my microphone but when i restarted i launched wokada before steelseries so the screenshot showed my regular mic

[dusk oyster]

i havent tested since i restarted i didnt want to mess with anything lol

ok my perf seems to be more stable now 50-60 but still sounds robotic

[knotty moth]

have you tried few other models?

[dusk oyster]

i will try

[knotty moth]

I think it's mostly the model quality itself

[dusk oyster]

i just tried a different voice and works perfectly fine

[patent lance]

need help
can't download pretrain folder, can't from MMVCServerSIO.exe and nor from method 2's zipped version
It always gets stuck between 410-470 mb and stops
I tried it on a different device, different browsers and different internet connections
still can't download it somehow

[dusk oyster]

the thing is that this one worked good when using voice recording files on the "Mainline, by RVC-Project"

[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]

If the program can't download pretrain models within the terminal, it can either mean that Hugging Face website is blocked in your internet or by firewall.

[patent lance]

sure Ill do that

[patent lance]

yeah but the thing is, I can't download the zipped version of it from website too

[broken urchin]

hello i started training my new voice model on older version of applio like 9 months old, i just saw there's brand new version of applio 3.5.0, can i resume my model training on the new version and how can i do that and what will be the difference between the old and new version

[knotty moth]

it's more of the model quality, maybe not good for voice changer

[hollow wren]

when i use free models it works fine but when i use an imported one it lags and cuts my mic

[hallow thistle]

What are you looking for? W-Okada or other RVC? What is your PC GPU?

[knotty moth]

delete pretrained folder and try again, or try manual download here

https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#how-to-fix-failed-to-download-or-verify

[broken urchin]

can someone help me on this

[hollow wren]

W-okada,every model works fine but once i use a custom one i import aka downloaded thats when it starts lagging and cutting my voice, radeon rx 580 series

[patent lance]

Full GPU Name: Rx 580 2048sp
Operating System: Windows 11
Detailed Description: Trying to download Deiteris' W Okada Fork and whenever I run MMVCServerSIO.exe it gives me the "Failed to download or verify" error, I checked on the website for help and there's a to fix it by installing the zipped version of the pretrain folder and I tried that but it always gets stuck on 410-470 MBs, I tried a different wifi connection, a different device and different browsers and nothing seems to work.
Tutorial Used: "How to fix "Failed to download or verify"

[hallow thistle]

AMD Radeon RX 580 can work but also struggles to run along with a game.

[hollow wren]

just pure discord calls, no game in the background

[dusk oyster]

the guy said he had chunk set to max which is why i had it at 5 in the beginning

[patent lance]

that's what I'm trying to do but it doesn't seem to work

[hallow thistle]

Did you follow any tutorial before this?

[hallow thistle]

Which guy and who told you that?

[hollow wren]

not necessarily i just downloaded the latest AMD version, and opened "start_http", and every model works perfectly fine, but once i tried some of the ones in here thats when my mic started to cut out/lag.

[dusk oyster]

random i met on apex he specifically said make sure the chunk is max

[hallow thistle]

What I asked you still matters. Download and use the better W-Okada DirectML from this guide link instead. https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#download-amd-intel-and-cpu-on-windows The W-Okada version you're trying to use, especially the one that has two "start_http" batch files and many files in the same folder, is outdated.

[hollow wren]

is this the browser version? if so it works fine but it just does not sound as good, it sounds robotic and you can tell its ai.

[hallow thistle]

Setting extra to 5 s somehow does improve audio quality more, but the issue is the audio always cutting off on most GPUs. So, 2.7 s is the balanced number.

[hallow thistle]

It's the browser version, and nah don't judge it bad like that. This always happens when you didn't set anything on W-Okada.

[knotty moth]

only if 5 does really improve the quality over 2.7

[dusk oyster]

lowering it to 2.7 did make it better so maybe should i lower it more?

[patent lance]

^

[hallow thistle]

Any number less than 2.7 s doesn't mean it will give you the best quality.

Instead, lower extra number will make audio to sound in lower quality.

[dusk oyster]

ok so like you said its probably just the model

[hollow wren]

how do i make it sound less robotic and realistic?

[hallow thistle]

The recommended settings for AMD Radeon RX 560, 570 and 580 are rarely documented, even though they do work, these are older than newer Radeon RX GPUs and may likely to struggle with games.
On your Deiteris' W-Okada, try these settings:
Chunk: 200 ms
Extra: 2.0 s
GPU: AMD Radeon RX 580
F0 Det: rmvpe_onnx

[hollow wren]

will that make it sound more realistic

[hallow thistle]

Likely more realistic, but not always the most realistic, as it can also happen to be the voice model you're using.

[hollow wren]

the voice comes out super delayed and still laggy

i changed the chunk to 1157.3ms and now its no longer lagging nor delayed

but it still sounds ai

[hallow thistle]

You can set extra to 2.7 s, but this can cause more delay to your W-Okada.

W-Okada usually works best with NVIDIA GeForce RTX 20xx and greater, but not as this old AMD GPU. GeForce GTX 10/16 GPUs can also work, but can also struggle with games at the same time.

Also, if you have set extra up to 2.7 s but the audio still sound robotic, it can be the model itself, not much you can do about it but you can switch to another voice model.

[knotty moth]

as ever said above, it's mostly the model quality

it may not be perfect nevertheless

esp when handling laughing and some hard parts

[short torrent]

could my gpu be the issue? it's an nvidia gtx 1070 ti but it seems to work perfectly fine other than within discord, and games of course

i didn't follow any old tutorials either

[hallow thistle]

Are you using Deiteris' fork W-Okada?

[short torrent]

yes

[hallow thistle]

Also, you can send your screenshot to here.

[short torrent]

screenshot of my settings?

[hallow thistle]

Your entire W-Okada GUI.

[short torrent]

my mic isn't plugged in right now hold on

[spare veldt]

hey im having trouble with the voice changer. i cant get it to put out any voice as soon as i put in on the virtual audio cable

[short torrent]

this?

[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.

[spare veldt]

tho could i be because of old version. the one i have got two start files

[low shard]

This is a General AI Server, AI has many fields, so we can't know your issue with little info, and there's many ai realtime voice changers

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
• a screenshot of the program

[hallow thistle]

That's not how you format your question. What is your PC GPU? Did you follow any tutorial before this?

[hallow thistle]

Reduce extra number down to 2.7 s. Anything between 2.8 s and the maximum 5 s can cause audio cutting off.

Additionally, you can set "monitor" to your main speaker to hear your W-Okada.

[short torrent]

Yeah ik, but there's no sound at all

i'll try it anyway but i don't think that's the issue

[spare veldt]

nvidia geforce rtx 3600.
windows 10.
https://www.youtube.com/watch?v=SxdnGxicJOg&t=796s&ab_channel=novision

[spare veldt]

i cant upload a screenshot doesnt let me type it

[hallow thistle]

There's no RTX 3600. You might mean by GeForce RTX 3060.

[spare veldt]

i said geforce yea

[low shard]

that's outdated asf, it uses original wokada version off 2 years ago and vb audio cable causes issues on windows

I'm guessing you got an rtx 3060

[spare veldt]

yea

[low shard]

btw be aware that windows 10 is ending support next month

anyways

delete everything you currently got

-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]

vonovox would be your suggested one because it still gets updates

there are other options like wokada deiteris fork or tg develop tho

but they aren't in active development

[spare veldt]

i see

[spare veldt]

ik. i dont really play alot anymore anyway. but thanks for the help

[low shard]

it's about security, windows 10 could have exploits in the future that makes hackers job easier to get you hacked, plus nvidia gpu drivers will stop on windows 10 one day which can impact performance

[low shard]

Let me know for any issues

[spare veldt]

this pc is probably getting sold soon. i graduated jeager corpse last month starting deployment soon lol

[spare veldt]

i will thanks

[hallow thistle]

While you can still use Windows 10 after its end date, the thing is you won't be receiving any update after that, which will make the system more vulnerable than current Windows 11.

[low shard]

Oh wow, well goodluck

[spare veldt]

thank you

[spare veldt]

is it safe to use the audio cable in the vonovox link

[hallow thistle]

Virtual Audio Cable will work with any program, which also includes W-Okada, Vonovox and VLC.

[spare veldt]

okay tysm

[keen badger]

hELLO GUYS HOW CAN I save the conversation the bot hass with each customer

[low shard]

if you mean VB Audio Cable, that one isn't suggested and should be uninstalled from windows app settings

Be sure to get VAC Lite

(they both are audio cables, but vb audio cable can give issues)

[hallow thistle]

What is this supposed to be?

[spare veldt]

is it the swift_F0 i gotta download in the github for vonovox?

[keen badger]

So I Created A Chatbot, And I NEED To See the other Conversations That Bot Having With The customers so that ı can improve it

[low shard]

swift isn't another program, it's a setting of the program, vac lite is in: https://docs.aihub.gg/realtime-voice-changer/local/vonovox/#virtual-audio-cable

[spare veldt]

ohh tho this new github interior confuse me. what do i click on to download vonovox

nvm got it

[keen badger]

I Created A Chatbot, And I NEED To See the other Conversations That Bot Having With The customers so that ı can improve it

[gentle cradle]

how to use voice changer

[hallow thistle]

Getting your host chatbot to screenshot one's customer sounds wrong, which I don't think it's possible considering if the chatbot is an LLM model, unless you do it for your inner company/individual purposes privately and not to share them outside publicly.

[hallow thistle]

What is your PC GPU? And what are you looking for?

[gentle cradle]

macbook m3 pro

[hallow thistle]

Not a GPU, but rather an Apple Macbook Pro with M3 chip. Try this W-Okada version, but with additional steps than Windows/Linux one. https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#download-mac-silicon

[spare veldt]

uhh do the setup close itself because its not doing anything anymore

downloaded a bunch now its just still

[covert ruin]

can someone help me fix my microphone???
when i select it as an option on any platform it doesnt send out audio, but when its set as an input on the voicechanger and i select passthrough i can hear myself
VAC messed everything up help pls
anyone?
its really annoying that i cant use my regular mic

[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]

Additionally, you can send your screenshot here for much better context.

[covert ruin]

screenshot of what

[hallow thistle]

[covert ruin]

what???

[hallow thistle]

Your W-Okada GUI, isn't it?

[covert ruin]

im using an older version

[hallow thistle]

W-Okada is realtime voice changer.

[low shard]

Can u show a screenshot?

[covert ruin]

this is the old version im talking about

i tried vonovox

didnt work

[low shard]

This is a General AI Server, AI has many fields, so 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 tutorial link are you using
• a screenshot of the program

[hallow thistle]

Use the better W-Okada from this guide link instead. https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#download-nvidia-on-windows

[low shard]

That's an over 2 years old version of original wokada

Uninstall it along with vb audio cable

-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

[covert ruin]

but im using it

[low shard]

You can use wokada deiteris fork, or Vonovox for more updates and better performance

Your choice

[hallow thistle]

W-Okada is realtime voice changer.

[covert ruin]

the newer versions literally didnt work

[low shard]

You shouldn't use the over 2 year old outdated original wokada, it's bad

[low shard]

The newer versions are much better, please try Vonovox or wokada deiteris fork and let me know for any issues

[covert ruin]

i tried them

[hallow thistle]

Don't be hurry. There are other solutions.

[low shard]

Vonovox is suggested for Nvidia because it gets more updates

[covert ruin]

and they do not work

[low shard]

What doesn't work?

Please show the issue

[hallow thistle]

Listen.

[covert ruin]

i know thats what i was told but vonovox literally had no audio output

thats wht i mean it didnt work

and the newest version of w-okada or idk if it was detieris

[low shard]

Vonovox does have the device input and output audio, I think you forgot to scroll down

[covert ruin]

the one where the client is in your browser

the audio was teribble and inaudible

[spare veldt]

it started downloading again. ill try and wait a few minutes

[covert ruin]

no i did the whole set up with help from one of your staff

[hallow thistle]

At least give me more time to explain.

[low shard]

Are you perhaps talking about the monitor option instead of audio output? There's a workaround to make that work too, Im the one who wrote it in the docs lol

[low shard]

Alright

[covert ruin]

oh hi

right

[covert ruin]

yes i nkow the work around from the settings

[low shard]

If you're talking about the monitor option missing, there's https://docs.aihub.gg/realtime-voice-changer/local/vonovox/#how-can-i-hear-myself :)

[covert ruin]

but last time we talked you saw it was turned on and tehre was still no audio

= it didnt work

[low shard]

I don't remember talking to you previously, you sure?

[covert ruin]

and when i tried a loopback test in game it didnt work

turned on hear-myself in discord

and it didnt work

[covert ruin]

yes iom sure

[hallow thistle]

I don't remember helping you before. Maybe you confused me or Nick with someone else here?

[covert ruin]

not you

Nick

[desert lily]

hey can I have some help please

how do I get around the copyright thing? (or can someone do the image generation for me please?)
I am trying to make a pixel art with them (their roblox character) and snorlax but it keeps saying

I wish I could, [my name]—but I wasn’t able to generate that pixel art due to a copyright policy violation. Since your alt Roblox avatar and Snorlax are both recognizable characters, combining them in a new artwork crosses a line I’m not allowed to step over.

That said, your concept is seriously cool. If you ever want help designing an original character inspired by your avatar’s gothic vibe or Snorlax’s sleepy charm—something remixable and totally yours—I’d be thrilled to help you build it from the ground up.

Let me know how you’d like to pivot!

it let me make one with garydos but not snorlax

[covert ruin]

so after all that i just went back to the old version

and suddenly my mic stopped working

but when i enable passthrough i can hear myself...

when i select my regular mic as input setting it doesnt work

[hallow thistle]

Three different people asking for W-Okada in the same channel gonna be mess, but I can still handle them.

[covert ruin]

only when i turn on the voicechanger and enable passthrough

[desert lily]

trouble with copilot image creation - due to a copyright policy violation

[covert ruin]

Ok i have an update, i restarted the audio service from the VAC control panel and i think my mic might be working, but its REALLY quiet like

really really quiet

i made sure the volume is turned up everywhere

[hallow thistle]

You can use any online Stable Diffusion website like PixAi.art to generate image of character you like. While you can use Google Gemini or "OpenAI's Dall-E within ChatGPT" to generate images, there come the line with copyrights.

[hallow thistle]

I'm so damn slow for this, but how did you fix it eventually?

[covert ruin]

restart audio service

yup my mic is working... but i had to scream

to hear something come out of it

so may i fix that?

[hallow thistle]

Don't be hurry. Just better wait for my answer huh. On W-Okada, there are volume sliders for input, output and monitor, try increase them both over 100% and hear again.

[covert ruin]

this isnt related to w-okada tho

im talking about my deafult mic

[hallow thistle]

[covert ruin]

???

are you ai

[hallow thistle]

I'm not an AI. But because you went on so fast, I didn't have time to respond.

[covert ruin]

ok im sorry

[hallow thistle]

If your microphone uses onboard sound card like the Realtek one, there's a way to boost your microphone signal up. Any other sound card like Sound Blaster may has their own settings.

But even that enabled, it can also increase the unwanted noise audio coming from your microphone too.

[covert ruin]

ok its turned up but now it sounds like i got noise suppresion from the vc

and i dont have microphone boost option

if my mic is somehow working should i just reinstall vac

[hallow thistle]

That's unfortunate.

On W-Okada, there are volume sliders for input, output and monitor, you can try increase them all over 100%.

I repeat this in case you forgot what I said before.

[covert ruin]

they are already

[harsh kiln]

I have rx 6600, Should I test tg develop or should I go with deiteris okada?

[hallow thistle]

Any fork W-Okada, but Tg Develops one has more recent W-Okada version than Deiteris one.

[harsh kiln]

I asked this question because I already use Deiteris but i have doubts of quality
It works well with some voices, but sometimes it lags
I wanted to see if the TG quality would be better on my card

I'll try it anyway, thank you very much :D

[broken urchin]

what is the best amount of epochs for rvc voice model training for a 90 minute dataset voice model made for speaking?

[humble vortex]

So I was using colab deiteris w okada with ngrok but there seems to be a problem and it is the fact that it disconnects after 20 minutes or so of use, is this normal or is there a way to fix it?
Next, another problem I have is that when I do the first module of the colab for deiteris instead of getting “done porceed with next steps” if see a bunch of text which says that something is not a symbolic link, is this normal or does it need fixing? Could it be the cause of the problem above?
Then, is there a way to save voices on the actual voice changer? I seem to have to upload vocies all over again each time I reopen it after closing the colab.
Furthermore, what voices tend to be the highest quality, and why does it happen this way. I noticed that some vocies seem to just be far more realistic than others.
Oh and on colab there are some GPU’s I can use with the paid tier which I have, are they worth it? I am currently using T4 but L4 and A100 are also available. Are there any ways to optimise cloud deiteris specifically the delay.
Finally, what does index do?
Thanks if anyone bothers answering.

[wind kernel]

Guys, when I try to make ai covers using applio,it didn't even show the exported audio so I have to go to files and in there, there's no converted audio and only the input audio there

Did anyone has a solution or another method to make an ai cover? I tried so many times but still failed. Pls help me

[low shard]

what's your pc gpu? you sure you are on applio 3.5.0?

[low shard]

So I was using colab deiteris w okada with ngrok but there seems to be a problem and it is the fact that it disconnects after 20 minutes or so of use, is this normal or is there a way to fix it?
Yeah that's normal, wokada deiteris fork colab needs a paid tier of google colab

Then, is there a way to save voices on the actual voice changer? I seem to have to upload vocies all over again each time I reopen it after closing the colab.
it's on cloud so no, unless you maybe use wokada deiteris fork lightning.ai port

Finally, what does index do?
In RVC Voice Model context, it's the trained accent

Did you firstly check if you got a good enough pc to do it locally?

[low shard]

none, use the tensorboard https://docs.aihub.gg/rvc/resources/training/#epochs--overtraining

[low shard]

tg develop is more of a personal fork of the already deiteris fork with some Quality Of Life adjustments, like an adjust UI, spin embedder type models and audio effects
Don't expect it to be active, and no it doesn't change quality nor performance

[broken urchin]

what's a tensorboard

[low shard]

The Tensorboard is a tool that allows you to visualize & measure the training of an AI model, through graphs & metrics, it's better you check the docs for more info :)

[broken urchin]

i will download it

how does it check the model though? during training or after training when a voice model is created or how

[low shard]

It's already inside of Applio, if you're using that

[low shard]

during training with graphs

[broken urchin]

yeah i got applio

that's what im using

[low shard]

great then, you can skip the installation step and just check the usage

[broken urchin]

ok which option?

let me check

[broken urchin]

Nick, what is the option called in Applio to check?

is it the "Overtraining Detector"?

that one makes the most sense

[low shard]

you need to only run applio via run-tensorboard.bat

[broken urchin]

oh okay

wait so

i do first run-applio.bat

and then run-tesnorboard.bat

or?

or just the run-tensorboard

[low shard]

yeah

[broken urchin]

i opened it

how do i check now for the voice im training

im super confused with this tensorboard lol

[low shard]

run-applio.bat, start training as said in https://docs.aihub.gg/rvc/local/applio/#training
run-tensorboard.bat, then check how to use the tensorboard: https://docs.aihub.gg/rvc/resources/training/#usage-guide

[broken urchin]

okay thanks

@low shard can i post a screenshot? i think my model overtrained

[strange oxide]

How do you download a model to use on the client?

[viral mason]

Go to voice models in this server and download one, btw what voice changer are you using Vonovox or Deiteris

[strange oxide]

VCClient-native-client

[viral mason]

Idk what that one is, did you download it from a yt tutorial

What gpu do you have btw

[broken urchin]

its the original w okada from github

[viral mason]

Ah

Yea that's bad

-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]

@strange oxide if you have an Nvidia GPU download Vonovox, otherwise download the first one deiteris wokada

[broken urchin]

@viral mason can you help me with my ai model?

im not sure if i overtrained it or not or undertrained

[viral mason]

Just check the spikes on tensorboard, btw anything over 700 epochs I'd say is overtraining

[broken urchin]

can i send a screen?

screensho0t

[viral mason]

Are you able to?

[broken urchin]

not here

can i send you in dms thern

i dont have media perms lol

[viral mason]

Lemme try something I saw in movie once
-# Isaw it here and a mod did it

!give-media-perms 1h @broken urchin

Yes I don't think it worked

Try tho

[broken urchin]

nope

[viral mason]

If not we'll just go to dms lol

[chilly furnace]

https://tenor.com/view/swtinafurry-stan-twitter-gif-24381631

[vale sequoia]

the application wont start up whenever i try to start up in pops the cmd but doesnt open

<@&1159293204038955078>

[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. "The application" doesn't refer to a specific program. You might either looking for W-Okada realtime voice changer, RVC, Stable Diffusion or other Python-related AI programs.

[low shard]

!give-media-perms 1h @broken urchin

or nvm that other guy is helping u in dms

[low shard]

@viral mason @simple ore the applio realtime audio detection seems to be broken on cloud, it doesn't detect any audio devices even after giving it microphone perms and refreshing the devices

This seems to happen on all colab, kaggle and even my lightning.ai port

[simple ore]

yeah, I guess it uses local (server) devices

[viral mason]

how odd

[simple ore]

no client/server change

[simple ore]

at the moment it is not possible

with just a regular gradio interface

the input can potentically be taken from browser mic, but not output

so please dont advice using applio realtime on the cloud

[low shard]

I see thx, I thought it worked fine on cloud since I didn't see any warnings about that but I will make sure users know it's currently not possible if anyone asks
AI HUB Docs will prob be updated with the applio realtime things, so I will add a red warning for cloud users

[raven ginkgo]

-colabs

!colabs

-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

[raven ginkgo]

-help

[spare veldt]

sometimes when i say something my voice repeats in small volumes abit.

gpu - nvidia geforce rtx 3060
windows 10
program is vonovox

[broken urchin]

im a bit confused on this

on Applio when training voice models

the options "audio cutting" and "process effects" you only do those once on the preprocess right?

the first time you start training right

and next time when you want to resume your training

you turn those options off right?

did i understand that correctly

[gentle orbit]

can you help me please I would appreciate it

[broken urchin]

what do you need help with?

[gentle orbit]

voice changer

[broken urchin]

what exactly

[gentle orbit]

we can contact me, I'll show you the error, maybe someone can help (

I can't send screenshots in private messages

[broken urchin]

?

yes you can

accept my friend request

[gentle orbit]

Can we go telegram?

[viral mason]

bro is scared of discord

[jagged elbow]

telegrammmmm is exttremell y dangerous

[viral mason]

fr

kinda sus they want to move platforms

[low shard]

I already replied to you in #1192011222023950368

[low shard]

you only train

[naive dune]

It looks like RVC is the best bet

[broken urchin]

what?

[low shard]

If you want to do Speech To Speech Conversion at the best quality, yup

[low shard]

sorry i meant to say that you:

1. use the same settings & criteria (u can change the save frequency, increase the Total Epoch amount in case you didn't input enough before tho)
2. don't preprocess nor train index again
3. Begin training again and monitor TB & console like before

[digital skiff]

any one help me for best ai app bellder and some trick for VS code ali free use

[broken urchin]

and what happens if you preprocess again and extract features again?

i've been doing that every time i wanted to resume training

did i do something bad?

it would just resume from the checkpoint i had

[pallid seal]

so im new to this, i followed a tutorial on youtube to get a female voice changer to mess around and it sounds mid and is sometimes glitchy

do i need to download the vbcable it says on youtube?

[broken urchin]

vac lite is best

audio cable

[pallid seal]

could u send me the link of it

just in case i dont go to the wrong website

[broken urchin]

docs.aihub.gg

go there and open Deiteris W Okada Fork under voice changer tab on the left

[viral mason]

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

[open flint]

who got wall bang script

roblox

[pallid seal]

Realtime voice changer >> local or cloud

[viral mason]

you're getting thrown into the deepest hole we got

[pallid seal]

where do i install it

i am on the website docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/

[viral mason]

download it

[pallid seal]

yea but which one

theres 3 versions

[viral mason]

oh

[open flint]

why

[pallid seal]

the Offical Original W-Okada made by Wok, the Deiteris' Fork made by Deiteris, and the Tg-Develop's Fork made by Tg-Develop

[open flint]

what did i do

[viral mason]

what gpu do u have

[pallid seal]

nvidia

[viral mason]

if u know what you typed you know what u did bud

[viral mason]

u should download Vonovox

[open flint]

ive said

who

got a

wall bang

script?

for roblox

😄

so i mean magic b ullet

dont u know?

[viral mason]

do it yourself chat

[open flint]

i cant

im asking vchatgpt

BUT HOW

hes saying no and stuf

[viral mason]

idk what wallbang means but I don't wanna

[open flint]

🤡

plz

give

now

idito

dog

[pallid seal]

A Virtual Audio Cable (VAC) is what you need to use the realtime voice changer on Discord & Games.
A VAC (Virtual Audio Cable) makes a fake audio device, used to re-route the audio of different programs.
In AI Realtime Voice Changing context, it's used to get the output of AI Converted Voice Output as the input in other programs such as Discord.
For Windows
Download this: VAC Lite (Virtual-Audio-Cable by Muzychenko)

[pallid seal]

this one correct?

im on the voxonox

[viral mason]

yes vac is to use the voice changer on things like discord or on games

[pallid seal]

okay so i installed it and everything and it says this

can we go dms so i can send the screenshot

[bitter viper]

what is the best vc changer app to put the rvc in ?

[broken urchin]

W Okada Deiteris Fork

download on docs.aihub.gg

[bitter viper]

can u send me the link please and maybe the best setting for my pc r5 5600x + Rx 580

[viral mason]

for non nvidia users

[broken urchin]

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

[viral mason]

vonovox is better tbh but lacks many model slots

[bitter viper]

i have amd card

[broken urchin]

what's the best for nvidia users? i have RTX 3060 Ti and use the w okada

[viral mason]

vonovox

[broken urchin]

i tried that one it sounded much worse t han w okada

[viral mason]

wokada deiteris is good for u then ^^

[viral mason]

strange

[bitter viper]

i dont have the best gpu i have an rx 580

[bitter viper]

ok can u show me your setting

and link of the app please if u can @viral mason

[low shard]

-realtime

[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

[low shard]

since you're on on AMD, you can only use either wokada deiteris fork or tg develop's one

Vonovox is in active development unlike the previous 2 i mentioned, but it currently supports only nvidia gpus on windows

[viral mason]

The creator is working on getting an AMD card tho

[bitter viper]

can u give me a link for amd

[viral mason]

The two that Nick mentioned are in their respective guides

[low shard]

there's no amd link for vonovox, you have to either choose tg develop or deiteris fork, which guides are above

[viral mason]

They're asking (in a not very smart way) for a link to one of the ones you mentioned

[low shard]

Yeah I heard dr87 said he's working on AMD since April and recently said in july that he's trying to get an AMD card since using volunteers would take too much time, well hopefully it comes soon one day

[bitter viper]

Is there an explanatory video because I am new to these things?

[low shard]

nope, all video tutorials are outdated unfortunately

you can read the guides above and let us know for any issues here :)

[viral mason]

You download whichever one u want in the guides, download vac lite, extract the zip files, and install vac lite. Then after that maybe restart your PC and then open the voice changer via the mmvc client exe (idk how to open the tg fork one it's new)

[bitter viper]

voice-changer-windows-amd64-cuda.zip.001
or voice-changer-windows-amd64-cpu.zip
what should i download

[viral mason]

Uhh

@low shard

[agile lynx]

how much the both are weight

the zip files

[bitter viper]

2 gb

[agile lynx]

both?

so unzip both

if they are 2 gigs

if the .zip its more like 3-4 gigs than the zip i think

[bitter viper]

zip

[low shard]

you got an amd gpu, you need the voice-changer-windows-amd64-dml.zip, cuda is for nvidia, dml is for amd, cpu is for cpu

[bitter viper]

okay

[pallid seal]

how do i make the voice changer sound more believable?

i downlloaded the #1159513888199540817 cable

[frozen ocean]

can someone help weioth this

[cyan copper]

How do I look for generic voice models?

[short torrent]

does anyone know a solution?

[elder garnet]

I would like to create a dataset of my voice for singing, How many minutes and WAV files of singing should I put in for the best possible voice quality? Do I also have to speak?

[viral mason]

30 minutes at most is recommended

[elder garnet]

at most? there is no need for more?

[viral mason]

sorry I meant for the minimum

I'm not the best with wording

but at least 30 minutes, if u can get more as long as it's all the same quality the better usually

[elder garnet]

Oh ok I see, how much would it be perfect ? 1 hour ?

[viral mason]

there's really no way to make it perfect but an hour can sound very good

[simple ore]

minimum is perhaps 10, but it has to be a really good set

more than an hour is pointless

[elder garnet]

Oh I see, and for the wav files, is it better to make 30 .wav files of 1 minute or 60 files of 30 sec ?

(over 30 minutes for example)

[sour sun]

Hi, quick question about dataset curation and audio editing

What should the limit be on "trimming" audio waveforms when cleaning up a set vs keeping things as natural as possible?

(1 as opposed to something more cleaned up/"cut," like 2)

[simple ore]

trim silences in audacity, keep breaths

[naive dune]

I just got RVC working on my pc, what is a index for?

[sour sun]

Right, of course, but if these two examples sound 99% the same, is there one that's better for RVC?

[analog obsidian]

always prefer the fullest spectograms

[knotty moth]

you should try the recommended fork version and choose the installation depending on your gpu

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

with the fork version you only need to directly run the MMVCServerSIO.exe and it will open the UI in your browser

[knotty moth]

you can try the tg that can work on RTX 5000 series

[knotty moth]

please read the docs here

https://docs.aihub.gg/rvc/resources/inference-settings/#also-known-as-index-rate-it-determines-the-level-of-influence-of-models-index-file

[worldly notch]

im having this problem with UVS UI 5 in colab where it always cancels it self and give ^C which im not sure why, it seems to happen when i use long file like 30+ mins it always finishes the steps and then i get the error i noticed the same with hugging face i also get an error at the same step. from what i can see i think colab is running out of memory and it stops the code automatically as i noticed the ram usage is at almost full before it does that so my question anybody faced that and whats the solution and if possible can i limit the ram usage on the colab? i cant attach screenshots for some reason

[wind egret]

Can someone please help me with this? evenn when running BLIP 2 on google collab it refuses to work #1417078192136523816 message

[knotty moth]

splitting it into at most 10 mins each should work

colab could handle at most 20 mins iirc

[worldly notch]

its kinda hard since the audio files im trying to work with are 1+ hours long at least and i do multiple iterations per file at least 3, however using it locally seems to work fine 2 GB of memory only and no errors so if there's a way to limit the usage i think that should fix it im. not sure though

[knotty moth]

you could manually split the audio in audacity with labels and export, also try not to cut inbetween voice sentences

may not sound convenient but it's recommended

[wind egret]

I feel like there's absolutely nothing online regarding what i'm looking for, feel stupid using chatgpt for this goal because it's imagining things.

[jovial flame]

i downloaded MMVCServerSIO and when i open it a blank cmd shows up what do i do

[low shard]

Please elaborate your PC GPU, Operating System, tutorial link and show a screenshot of the issue

!give-media-perms 1h @jovial flame

[jovial flame]

Nvidia gtx 1050 TI windows 11 , i just opened it as admin and now its downloading ty for replying

[wind egret]

Ok, clearly what i wanted doesn't exist or is unusable for me, i do wanna know tho, is there a way to setup Stable Diffusion to generate images at ease under 4gb vram gpu? Like maybe it needs lower resolution or something?

Because when i generate at 512x512 it takes like 13 min or so for 1 gen

[low shard]

You shouldn't need to open it as an admin, can u share the tutorial link you're using? Because there's different programs

[severe violet]

I'm currently using W-Okada in conjunction with a third-party software, and I'm experiencing some latency issues. Specifically, when I speak into my microphone, there is about a one- to two-second delay before my voice actually comes through. I'm looking for any possible solutions to reduce this latency or if there's alternative software that might perform better.

[steady python]

yo can ayone help me setup wokada i cant hear myself

[short torrent]

I set deteris w-okada fork for cuda and everything and the voice changer seems to work fine in all of the tests but whenever I talk over discord it shows that there's no sound? I have a GTX 1070 Ti I know it's not the best but when I test my mic within discord settings it sounds perfectly fine and everything runs smoothly, it's only when I try talking in a voice call or within a game where there's zero sound being heard or indicated
There were a few threads about this issue but none had an answer so I was wondering if anyone experienced knows why this is happening? (still haven't found a fix)

Should I try upgrading to windows 11 really quick?

[viscid topaz]

Go to User Settings → Voice & Video.

Make sure the Input Device is set to your virtual microphone (the one created by Detrs/W-Okada, usually something like “VB-Audio Virtual Cable” or “VoiceMeeter”).

Sometimes Discord “remembers” your physical mic from when you tested in settings, but in a call it reverts to “Default.” That’s why it works in test but not in calls

[short torrent]

Is there a way to set the virtual cable as default?

They're already set up in settings but how could I stop discord from going back to the default in calls?

[viscid topaz]

You've already set up virtual cable in discord setting

[short torrent]

Maybe it's something with my actual settings tell me if anything looks like the case

[viscid topaz]

Looks alright idk bro wait till a actual helper some and help you 👍

[short torrent]

I've been waiting for a few hours

[viscid topaz]

Yea no one was helping so I thought Id give it a try sorry

[short torrent]

Not just waiting obviously but no one has answered

[short torrent]

it's ok

[viscid topaz]

Do you see green bar move in calls or does it stay dead

[short torrent]

I haven't tested it in awhile but it didn't show any green when i last tried it a few hours ago

It also doesn't work in games so I was thinking it might not be something with discord

[viscid topaz]

Yea, mb jumping on the discord angle too fast

[short torrent]

I might just try vonovox but every time I start it closes after the warm up

[wheat ledge]

why i dont find woman voice

[severe violet]

I'm currently using W-Okada in conjunction with a third-party software, and I'm experiencing some latency issues. Specifically, when I speak into my microphone, there is about a one- to two-second delay before my voice actually comes through. I'm looking for any possible solutions to reduce this latency or if there's alternative software that might perform better.

[severe violet]

can someone help me with w-okada?

[fathom owl]

soemone know my mic is cutting while i speak?

[jovial flame]

does anyone know why my ping keeps rising when im using ht tp://127.0.0. 1:18888/

my ping in real time voice changer keeps rising up and when i turn passthru mode it just becomes 1ms my network isnt bad i just dont know why its doing this

[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 start asking. What is your PC GPU? Did you follow any tutorial before this? And which W-Okada version are you using?

[hallow thistle]

What is your PC GPU? And did you follow any tutorial before this?

[hallow thistle]

[hallow thistle]

You can send your W-Okada screenshot to here.

[low shard]

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 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.

[viscid topaz]

😅

[fathom owl]

mb
Full GPU Name: rx 6750xt
Operating System: Windows 11
Detailed Description: with any voice model the first like 5 sec sound good then it starting to freeze saying 1/2 word
Tutorial Used: https://www.youtube.com/watch?v=IS_SPQVv5iY

[low shard]

that's outdated

[hallow thistle]

Tutorial videos on YouTube about W-Okada especially are outdated. There's a better W-Okada for this.

[low shard]

delete everything

dont use video tutorials

-realtime

[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

[low shard]

use either wokada deiteris fork or tg develop

[fathom owl]

oh

ok ty guys

[hallow thistle]

Let me know if you're looking for settings or any issue trying to install/run the program.

[fathom owl]

ok ill say if something goes wrong

[fathom owl]

i can use both woakda but not vonovox if i have amd gpu right

[hallow thistle]

Vonovox is NVIDIA-only as of now.

[viscid topaz]

@low shard

[low shard]

The scammer is executed

[low shard]

vonovox is nvidia windows only
you can use either wokada tg develop or deiteris fork

[steady python]

uhh i fixed it its a 2050

bugt idk the optimal settings

rtx 2050

[fathom owl]

ok one last question for installation its the exact same way as the one i had installed?

[hallow thistle]

How do you fix that? Which W-Okada version are you using?

[steady python]

the

one posted

wait

lemem take a ss

the cuda one

[low shard]

not exact same, please read the guide

[steady python]

the problem is w okada registeres noice from my speaker and reregisters it in call

[low shard]

that's a previous version of the v2 beta of original wokada, it's not suggested

[steady python]

oh

can u send a new one

[low shard]

you want a realtime voice changer for calls/games right?

[steady python]

yeh

trolling

/ catfishing my friends

[hallow thistle]

This version is the true latest version of the original W-Okada, where its performance is almost that closer of Deiteris fork W-Okada, but not really preferred to try anyway.

[hallow thistle]

What?

[steady python]

yeh i mcatifishing my friends

for funsies

for proof lmao

cus i got banned

[hallow thistle]

"Catfishing" means to pretend to be someone else to do something bad to others. It's not the same as "trolling" but still questionable.

[steady python]

warned*

[low shard]

Catfishing isn't allowed in this server, you have been warned in sapphire and verbally and won't get help, be aware that you will risk further actions if you keep asking for it

[steady python]

oh

so what shld i use

tht u recommend

[hallow thistle]

Use your real voice instead.

[steady python]

sadly im not a femboy :

[steady python]

ITS FOR FUNSIES

WITH MY FRIENDSSS

uhh

so do i hget helt 😭

[broken urchin]

can i post my applio screenshot here?

can give me media perms

im wondering if im overtraining

my voice model

[fathom owl]

@hallow thistle all work fine with this new one ty :)

and its even faster

[marble spruce]

is there anyway to fix the delay and cutting out?

[fathom owl]

@marble spruce that one helped me

[sinful adder]

anyone know why my voicechanger is not working like its no sound is coming out

i did download everything needed

[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]

Did you follow any tutorial before this?

Make sure to read help guidelines before start asking here.

[sinful adder]

yes i did but i cant seem to hear myself

[hallow thistle]

That sounds bad. What is your PC GPU? Which W-Okada version are you using? And did you use VB-Cable?

[sinful adder]

RTX 5060 TI im not sure for the version and yes i am using vb cable

i think its 1.5.3.18a

[spiral elm]

Dumb question 😅 .. i use my laptop ( 4070 ) for w-okada.. if i upgrade to a 5080 laptop, will i see a mayor difference on the delay? Atm my laptop can do 32 chunk .. which is like.. 1 second and a few millie delay-ish

[hallow thistle]

There are two voice changer programs that can run on GeForce RTX 50 GPU: fork W-Okada b2335 and Vonovox.

[willow flicker]

anyone got good voice changers that are free?

[hallow thistle]

I gotta sleep.

[willow flicker]

???

[sinful adder]

ok i will try it thx

[willow flicker]

do you know any free voice changer?

[chilly furnace]

-wokada

[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

[thick latch]

I just tried okada, works kinda well. Is there any program which is better? As in maybe less heavy, less delay or something like that

[knotty moth]

@willow flicker this^

[viscid topaz]

⚔️ 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

[thick latch]

Yeah I was going through the docs posted there rn

Is Vovonox an easy install like okada? As in download a zip and click a .exe

[low shard]

Catfishers aren't allowed to get help

[viscid topaz]

Yes, Vonovox is designed for easy installation, similar to Okada

[low shard]

that's old

dont use video tutorials

delete everything

-realtime

[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

[low shard]

vonovox would be suggested for you

[thick latch]

I tired this other one which requires a lot of complciates stuff

Python, Pytorch download this, that Jesus christ

[low shard]

Original Wokada off youtube tuts could be harder to download

[low shard]

what's your pc gpu and operating system?

vonovox is nvidia windows only

[thick latch]

ryzen 5 4650, rtx 3060 (12gb normal) and 32gb ram

[low shard]

I'm guessing windows 10 or 11, right?

[naive wadi]

why okada voice changer sounds robotic?

how can i fix that

[thick latch]

10 yeah

[low shard]

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 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.

[low shard]

great, btw be aware that windows 10 support is ending next month

-realtime

[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

[low shard]

vonovox is best for you in terms of performance

[thick latch]

It depends on the voice. But I tried to follow some tips as in: if the original dub is japanese, speaking in other language makes it sound extra robotic

[bleak smelt]

which do you guys suggest for making a good voice model? what out of the collections do I use and what would be the best settings?

[naive wadi]

i did talk in japanese (jotaro's voice) and it sounds robotic but it didnt happened in the past

but i had some settings in the past

but i forgot what setting i did

[low shard]

I'm guessing you want an RVC STS AI Voice Model, can you tell me your pc gpu and operating system?

[low shard]

can you please elaborate what I asked? you could be using some old version off youtube

[low shard]

^

[bleak smelt]

Pref something not locally lol

[naive wadi]

NVIDIA GeForce GTX 1650 SUPER, Windows 11, the voice sounds robotic and echoes, No Tutorial.

[low shard]

You sure? you checked that your pc specs and they aren't good enough?
Locally would be suggested since you don't have any gpu free time limit (if you got a good one)

[low shard]

alright ur specs are good enough, could u send a screenshot?

!give-media-perms 1h @naive wadi

[naive wadi]

oke

i just watched one video how to dowland it, let me send it

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

[low shard]

that's outdated like the rest of the other youtube video tutorials, delete everything

[naive wadi]

damn

ehm

where do iam going to dowland it then?

[low shard]

I'm guessing you want an ai realtime voice changer for roleplay in games/calls? since there's a different program for ai covers or pre-recorded audios

[naive wadi]

yes

[low shard]

-realtime

[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

[low shard]

vonovox would be mostly suggested for performance in your case

[naive wadi]

hm

thanks for the help

[low shard]

yw and lmk

[thick latch]

I'm basically on the same boat, I do some starwars/half-life roleplay stuff in vrchat

[naive wadi]

wut

(i dont know short terms of words)

[viscid topaz]

you're welcome and let me know

[naive wadi]

oh oh oke

[low shard]

yw = short for "you're welcome"
lmk = short for "let me know"

[bleak smelt]

Yeah i dont mind paying for the extra resources, I dont have enough space on my hdd anymore

[naive wadi]

oke

[thick latch]

I download from the same youtube video (lol)

I thought that as its a github link, it was probably updated. Maybe I'm wrong

[low shard]

https://tenor.com/view/half-life-gordon-freeman-no-way-half-life2-gordon-gif-22072856

I'll play half life one day lol

[low shard]

well, the original wokada technically does still get updates, but the creator (wok) is focusing on Beatrice voice models mostly, which are yes lightweight but lower quality than RVC

While other voice changes like the modified version from deiteris/tg-develop, and the other voice changer Vonovox,
actually add performance and RVC-related updates

like original wokada updates are only focused on beatrice as you can see, which barely anyone uses or trains since people usually care more about quality rather than lightweight

[knotty moth]

sounds like an excuse to hide your capable gpu like rtx 2060/higher

[thick latch]

I guess trying Vonovox is my next step then

[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]

I mean, even if they got a capable gpu, it seems like their decision that they want to use cloud to save space and to rent faster gpus :)

[bleak smelt]

My gpu is Geforce RTX 2070 Super lol

[knotty moth]

renting gpu for voice changer sounds wasteful when there's free alternative

would still be okay for more intensive ones like training

[low shard]

that is capable and we can help with local if you want, ofc it can take some GBs (usually is 30gb free for training would be good),
Do you also want me to send you local links, or are you okay with just cloud?

[low shard]

They want to do RVC Training, not AI Realtime Voice Changing

[bleak smelt]

I have like 6gbs of free space lol, so probably the cloud route

[low shard]

I mean it's their choice, it's good we make users aware if they weren't about their hardware, but it's not like they should be forced to do local lol

[spiral elm]

soo with Vonovox i can't use another pc's gpu and use it on my other pc through my local ip network?

[low shard]

holy 😭 , i would suggest cleaning up even for just letting ur pc breath a sec

but yeah all cloud links are in #✨│ai-help message

[knotty moth]

if I were replying in a separate forum thread, I wouldnt be confused like this

[bleak smelt]

Which would you recommend? I'm trying to get the best quality one (and I have a good quality dataset) i haven't trained a model in a good 2 years so I forgot how to do a lot of this haha

[severe violet]

have two versions. The old one works really well in terms of quality, but the newer one, even though I’m using the same voice modle, performs really badly. There’s a lot of cutting, the latency is terrible, and I couldn’t find any solutions on the web—most guides only show the basic setup for the app. For reference, my PC specs are RTX 3060 with an i5-10600K.

[simple ore]

both are wrong

[naive wadi]

NVIDIA GeForce GTX 1650 SUPER, Windows 11, The Voice Echoes (Vonovox)

[low shard]

Fair point, #1192011222023950368 is more organized, anyways it's all good dw :)

[low shard]

Quality related there isn't much difference, like those are just different cloud providers that run the same programs

I'd suggest Applio since it gets more updates about the providers;

• lightning.ai has around free 80 hours of gpu monthly and cheap renting, and saves all files too and needs phone number
• kaggle gives 30 hours weekly and needs phone number, it can go up to 60 hours if you link your paid google colab pro subscription
• google colab free tier isn't much tbh, paid tier could be better

So lightning.ai and kaggle seem the best choices with the most amount of hours

[low shard]

let me know!

[low shard]

both are old versions of original wokada

delete them

please tell your pc gpu and operating system, I'm guessing you want an ai realtime voice changer for calls/games

[severe violet]

i thought i had the latest version cause that where the github download link took me "hugging face"

so where do i get the newer one?

[naive wadi]

Vovonox The Voice is Echocing idk why, NVIDIA GeForce GTX 1650 SUPER, Windows 11

[low shard]

Oirignal Wokada isn't suggested anymore, the 1.5.3.18a is an over 2 years old version
the 2.0.78 beta is a v2 beta version, which changes mostly just Beatrice models which no one makes because they are worse quality compared to RVC

[low shard]

it depends on your PC GPU and Operating System (since I'm guessing you want an ai realtime voice changer for calls/games, and not a pre-recorded audio / ai covers program, since those are 2 different programs)
could you please elaborate that?

[low shard]

can you show a screenshot of your settings please?

[naive wadi]

ok

[severe violet]

I have Windows 11 and RTX 3060 and yes I want to use it for video games / calls

[low shard]

• RNNoise Reduction
• extra time: 3.0
• crossfade duration: 0.15 can help with a bit better quality at the cost of some delay (around 50 miliseconds of delay)
• output device: Line 1
• pitch extraction method:
  • rmvpe: usually the best precision and robustness, most suggested
  • swift: pretty new, theoretically it should be more precise than rmvpe, but it's not as much tested so it's more of experimental
  • fcpe: less precise and robust compared to rmvpe, but gives lower delay and might make it sound more human soft, suggested only for lower end devices (which isn't your case)
• AP-Bwe-48k Upscaler: can help with reaching better frequencies
• pitch: there's no perfect setting, you gotta play around with this on different models, lower values can sound more masculine, higher values sound more feminine

[low shard]

-realtime

[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

[low shard]

Vonovox would be the most suggested in terms of performance in your case

[thick latch]

damn

Vonovox has almost 0 delay compared to the okada I was using

The voice seems less robotic, but its cutting my last word short for some reason

[naive wadi]

theres one problem now

my voice lags in discord calls

[thick latch]

In calls or tester?

Because the discord tester lags

But its not a reflection of how it will actually sound

[naive wadi]

oh

oke

but uh

i want to hear the ai voice too

but when i make output device Line 1

i cant hear the voice but others can

[severe violet]

sorry but i didn't get it

realtime

[round panther]

Hey guys which fork of W-Okada would you recommend for a 3060 Mobile (6GB)? I just don't know which would be better Deiteris or the other? Or is there anything better? I'm really just kind of dipping my toes in.

[low shard]

be sure to finish your sentences, to have a good microphone and to put it nearer to u

[low shard]

https://docs.aihub.gg/realtime-voice-changer/local/vonovox/#how-can-i-hear-myself

[low shard]

Those are guides link, Vonovox is suggested for best performance in your case :)

[naive wadi]

Thanks!

[severe violet]

is Vonovox free?

[naive wadi]

ye

also uh

[low shard]

What's your Operating System? I'm guessing you want a realtime voice changer for calls/games

[naive wadi]

my voice still echoes idk why

[low shard]

You're welcome

[low shard]

Yes, it's closed source (you can't see or edit the code), but it's free, except a patreon monthly subscription for some paid audio effects like low quality mic (which aren't related to like how many models you can use or smt, it's just for 'realism' effect)

[naive wadi]

ohhh

i think i know why my voice echoes

it detects the voices from the pc too

[low shard]

lower your pc volume too

[naive wadi]

ok its fixed now

thanks

[low shard]

you're welcome, is it all solved now?

[naive wadi]

yep

no any problems

[severe violet]

god bless you my friend

[low shard]

great, have a nice day

[naive wadi]

have a nice you too

[low shard]

let me know for any issues ofc, you still have to finish installing lol

[dapper pawn]

Hello, does anyone in here have any experience with XTTS_v2? Ive hit a wall with it and im trying to train a voice to use for my an Ai assistant. I don't have any experience with this stuff as I'm completely new to programming and what not

[low shard]

btw is it all fixed for you?

[round panther]

Win10 Pro however I also run Arch in a VM if it would be any better. Not lack of technical understanding generally, just wondering about the communities opinion and suggestions. I'm unfamiliar with audio processing entirely.

[round panther]

Yes for calls and games

[dapper pawn]

The furthest i could get is following a guide to get the voice models that come with XTTS to work in the inference portion but getting the custom voice files i want to use on there is a different story

[severe violet]

ye just 1 small issue i don't know how to install it i didn't veen find a download link for Vonovox on the web

help!

is that the site? https://www.vonova.ai/

[low shard]

For windows nvidia, it's suggested to use Vonovox instead

because Wokada Deiteris fork and tg develop don't get much updates especially about performance and rvc-related

while vonovox is in active development

-realtime

[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

[low shard]

this isn't related to vonovox at all

[low shard]

Please read the guide https://docs.aihub.gg/realtime-voice-changer/local/vonovox/

[round panther]

I appreciate it. Thank you.

[thick latch]

Seems to be working better than the old Okada I was using for sure.

Almost no delay, which helps with movement in vrchat as before I had to wait 3 seconds to do so

Its still kinda robotic depending on the tone, I need to find a good settings sweetspot and keep my voice stable too

Do you guys prefer with or without index?

[low shard]

play with the pitch and different models

[round panther]

One more question, is there an existing guide or can you point me in the right direction in what you would suggest to do in order to apply VST filters to the output with out paying for a subscription service?

[thick latch]

Assuming how much superior this one is in performance and quality that might be a skill issue on my side yeah

[low shard]

index is the trained accent, it's not a necessity, it could add a kinda auto tune effect on high values, kind of a try and test tbh

[low shard]

be aware that RVC can't do always super realistically non speech sounds, like screams or some certain laughs

[low shard]

• Vonovox does have built in (lower delay than an external one) audio effects like that, but some are paid
• Wokada Tg Develop does have some other audio effects (built in, so lower delay) which are all free and open source (but this isn't in active development so won't get as many performance updates)
• You could try yourself to play around using an external program for VST filters to any voice changer, but be aware that the previous guide for doing this on wokada deiteris fork was removed after people reported issues with it, so there's no current vst filter external guide

[round panther]

Thank you I appreciate it, detailed and exactly what I needed to know. ❤️

[thick latch]

okay DAMN

[low shard]

You're welcome, let me know for any other issues

[thick latch]

Now I found the setup

[low shard]

Soo, it's all solved now?

[severe violet]

thank you so much for the guide

[thick latch]

Well it was never an actual problem, we could say that I'm finally satisfied with the outcome

I appreciate the attention

Its so good to click a tutorial that has steps like
"Okay so download this, click that. Wait install. Just press start" and it actually works

[severe violet]

just 1 last question in w-okada i used to dwnload models and add then to the software is it thz same with Vonovox? can i do that?

[thick latch]

Different from
"Okay so you need to download python, open cmd, but you can't forget to configure the path, also you need to give some really complicated commands that probably won't work because you need the MMH but never forget the NNH that is needed for the JJ-88"

Like dude its 2025 lel

[low shard]

You're welcome! Feel free to ask me for any other issues

[low shard]

lol

Well I'm glad to hear it's all solved and good for you

[thick latch]

Is it normal that if in the game/discord/whatever you change from the audio cable to your original microfone and use your original voice, the original has lost a lot of gain? Like its very low

[low shard]

Vonovox also has a built in updater for future updates btw

[low shard]

Sorry I couldn't understand much your message, but about the audio setup for when you want to use Vonovox:

• vonovox:
  • input: microphone
  • output: line 1

other programs:

• input: line 1
• output: headphones

let me know if I misunderstood what you said

[severe violet]

just 1 last question in w-okada i used to dwnload models and add then to the software is it thz same with Vonovox? can i do that?

[analog obsidian]

disable auto gain in discord

[low shard]

If wioth "models", you mean the same RVC STS AI Voice Models that you used in Wokada, then yeah

[severe violet]

i know i said "last questiron" but just 1 more! so how do i download? so based on the guide i have to download source code and im not that good with github so what do i press?

[low shard]

in Assets -> Source Code .zip

[thick latch]

Nono like. I have that exact setup
Vovonox
Input: my microphone
output: digital audio

Program
input: digital audio

What I meant is that when in the program I change from the digital audio input to my actual microphone (So my voice goes back to normal as its using my microfone as a direct imput now) for somer eason my microphone gain gets very low. But that might be because the program is open

[severe violet]

am i on the right page? cause i'm kinda stuck

nvm! found it

[grizzled crystal]

Rvc V2 is not working

[latent kettle]

!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.

[viral mason]

lmao of course William Afton doesn't know how to use it silly old man

[dense spruce]

hello

https://www.youtube.com/watch?v=H8MJ6CsRsJs

can anybody identify which ai was used here?

it doesnt sound suno ai to me

or is it suno premium?

[stark lion]

hi, what´s the best Voice Changer now?

[viral mason]

What gpu do u have? Nvidia, AMD, intel

[sonic garden]

-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

[viral mason]

Kaggle better 🐠🫧

[stark lion]

AMD graf

[low shard]

Graf?

Could you please give your full GPU name?

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

[stark lion]

AMD Radeon TM Graphics

[low shard]

Oh that's integrated graphics, which isn't good for local ai tasks as it's not really powerful

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)

Your main option seems to be cloud