✨│ai-help

[kindred kernel]

Output the music and then input it into the rvc

[viral mason]

ohh ok

[kindred kernel]

thats how i do it

Virtual cables

ike i said the new forks are doing something weird

i think its because they dont care about pitch but more on realism

let me try vonovox

[viral mason]

alrighty

for vonovox just run the start file

[hardy yew]

as long as you keep the pitch setting a multiple of 12 (including 0) then it will stay in key

[kindred kernel]

exactly

[hardy yew]

but then, depending on the voice it won't necessarily fit your speech tone range

[kindred kernel]

the forks are just doing something different im telling you

it doesnt match the pitch

[hardy yew]

weeeird, i don't think i've ever encountered that

[kindred kernel]

somethings fucked up

which one do you use

[hardy yew]

Both Tg develop w-okada and vonovox

But, did you say you're just using those apps just for passing singing through them ? Do you mean it's just for recording or do you actually need it to be realtime?

[kindred kernel]

MMVCServerSIO has been my goto for 4 years or so

realtime

[hardy yew]

Alright

[kindred kernel]

ive been doing this for a while trust trust

[viral mason]

he uses this

[kindred kernel]

yes

[viral mason]

there's a lot of dust on it from how old it is lol

[kindred kernel]

not necesseraly the same version

[hardy yew]

Yeah, I wonder how it's possible for it to change the pitch while it's at a multiple of 12

Are you using RMVPE?

[kindred kernel]

IT WORKED PERFECTLY TILL yESTERDAY OR SO

yes there is nothing else to select here

i remember trying vonovox

and i didnt like it so i always come back to mine

the visuals and such

Dude i had like 200 voices + all the pictures for them

[viral mason]

just looks different, works exactly the same but is better

the slot space limit keeps me at tg fork tho

[hardy yew]

slot space limit is gone in beta already

😉

[viral mason]

current (non beta) vonovox has only 8 slots

[kindred kernel]

how do i find out what version i have

[viral mason]

true but is unstable

[kindred kernel]

where do i see that in the files

[hardy yew]

Oh, i just scrolled up to get some more details

You said you changed your audio interface and then it stopped working

[kindred kernel]

Like i said earlier i bought a new audio interface and then it stopped working

lol yes

sry i wrote that before reading

[hardy yew]

Perhaps the config file remembers the old interface which is now gone and the app doesn't handle it correctly and crashes?

[kindred kernel]

i wouldnt know

[hardy yew]

You can check some file that keeps the app's settings and remove the interface name from there, or even rename the file to check if the app starts

[kindred kernel]

where to check

i see

[hardy yew]

(don't know where that file is or how it's named, i don't have that particular app)

((and obviously, i still recommend tg w-okada or vonovox just like Your_Local_Worm, for improved quality and functionality, but I understand why you'd want to stay with your current one as it served you well so far))

About the new apps altering the pitch, what's your new audio interface sample rate?

[kindred kernel]

48k

[hardy yew]

No issue there then i suppose

[kindred kernel]

also no error log or anything

smh

i also thought abt the files being corrupt so i downloaded the same version + newer ones also didnt work

but vcclient worked

from huggingface

thats the best one so far

and last update was like 7 months ago

wtf

first of all

i cant change the sample rate in the tg developer fork

but.

i changed the audio driver from MME to Windows DirectSound and that fixed the weird pitch issue????

the sample rate is at 44k cant change it

[hardy yew]

MME is quite outdated, i think vonovox doesn't even have it as an option anymore

WASAPI i usually the go-to choice

[kindred kernel]

it worked fine on the other one

weird

why cant i change the sample rate?

or how can i change it

[hardy yew]

don't quite remember but perhaps the apps operate on a fixed sample rate for simplicity, there's no use changing it really

which app actually?

[kindred kernel]

tg developer fork

wait

[hardy yew]

in server mode i can switch sample rates with no issue

[kindred kernel]

i think its because its the models sample rate

yeah but start server afterwards

doesnt it change back

?

[hardy yew]

oh, model's sample rate is certainly constant

top left is the model's sample rate

in audio settings is your output sample rate

[kindred kernel]

oh nah im trying to change the one bottom right

yes

i cant change that

i press 48k

then start

it changes back

to 44k

not for you?

[hardy yew]

indeed, same behavior here

[kindred kernel]

yeah

[hardy yew]

i didn't ever bother, don't see a reason to change it

[kindred kernel]

i was only changing it to test if that was the issue

cuz it was 48k

but it was the Monitor driver MME

i also hate the fact that it opens in your browser and not just a standalone window

[hardy yew]

well, another point for vonovox here

[kindred kernel]

dont i have to always select the pth file if i want to change models?

if i iremeber correctly

like select it from my folders?

[hardy yew]

in both apps the model is copied over to app's files on selection and is stored there to be used as preset

you choose the pth file once

[kindred kernel]

isee isee

[hardy yew]

but tg develop is better if you have a huge amount of voices (which AFAIR you do)

latest vonovox has very limited slot amount

beta has much longer list, up to ~100 or something

[kindred kernel]

Nitpicking rn

im just upset that my goto isnt working

[hardy yew]

if you open it in a terminal, isn't there any log that'd help pinpoint the issue?

(I guess if you open it directly by doubleclicking there might be some console that disappears immediately due to the error)

[kindred kernel]

im not sure how i do that

how do i open bat files in the terminal

[hardy yew]

if it's just a bat file

then open it with a text editor and add a line "pause" at the end

[kindred kernel]

alr

[hardy yew]

this will keep it open even if the program fails

so that you can see the output

(and hopefully, some crash log)

[kindred kernel]

nothing

just says "pause" now lol

at the end

dmed u

[gusty moth]

-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

[hardy yew]

well, if there's no error log after the command that starts the app then at the moment i can't really help much

[kindred kernel]

alr thanks anyways

[vagrant marsh]

#✨│ai-help message

I've tried lots of collabs from the docs and none of them works

I've tried lots of them

The collab rvc one doesn't work properly

When I try to test the TTS for example, it says "format not recognised"

I think there's no updated one

[celest violet]

After this, what should i do? @hallow thistle

[hallow thistle]

Has your issue been solved? It seems like a very long conversation that neither one was able to identify actual issue and I'm not sure which to summarize it.

[hallow thistle]

Why did you extract the zip inside a Videos folder? Try extract it to somewhere else like "D:\MMVCServerSIO".

[safe echo]

hey im unable to download VAC Lite (Virtual-Audio-Cable by Muzychenko) as it gives me this error, what should i do from here?

[hallow thistle]

Did you try to run "setup64.exe"? Because "setup64a.exe" is for AArch64 system (e.g. Qualcomm Snapdragon).

[plucky crown]

Is there any delay if I buy 4090, this is not cheap and I'll use all my savings to buy a 4090 PC with Ryzen 7 or Ryzen 9 (which is best here) is 4090 enough or I should save more for a 5090

It took me years to save for a 4090 PC so aaaaaa

Helpp

[hallow thistle]

For W-Okada voice changer** fork** or Vonovox (older W-Okada versions like b2332 and v.1.5.3.18a won't work with RTX 50 series), the performance between NVIDIA GeForce RTX 4090 and RTX 5090 is often negligible, considering if you set chunk and extra settings properly. RTX 5090 is newer, provides more VRAM size more room for gaming and voice changer at the time, while RTX 4090 might be found cheaper.

[celest violet]

i dont have D

[hallow thistle]

[celest violet]

or, i must using external?

[hallow thistle]

You could've buy another hard drive/SSD for more storage. If your PC only has single drive single partition, try extract to somewhere like "C:/Users/username/Downloads/MMVCServerSIO".

[celest violet]

alright i extraxt in that place

[plucky crown]

What does that mean, would 4090 be enough since this is the only PCs I can see in the marketplace which they sell cheaper 2nd hand , for 5090 I have to get it built new

[hallow thistle]

What does this question mean? Weren't you looking for the voice changer or something?

[plucky crown]

Yeah would 4090 be enough that gets no delay in voice changing

And also I don't want any buffers while using the ai

[celest violet]

@hallow thistle this place right?

[hallow thistle]

Actually, I don't have any** absolute** recommendation for PC building, but I do have my own idea of a gaming PC that won't actually match your needs. The PC bulding largely depends on your needs, like gaming and running voice changer at the time, not someone else.

[plucky crown]

Voice and ai local run

[hallow thistle]

Simple as that?

[plucky crown]

Yeah

[celest violet]

whats next? @hallow thistle

[hallow thistle]

Go inside that one folder, go inside a folder named MMVCServerSIO, you'll see a program file named "MMVCServerSIO.exe", that's the actual voice changer program.

No, and don't try to ask me for step-by-step that way, the guide doc exists, I'm only to make you think about best working settings.

[celest violet]

okey

where i can find the guide

for this realtime voice changer

[hallow thistle]

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

Don't you think?

[celest violet]

im done, but i cant find after i click the +

how i get the voice models suits to my device

[hallow thistle]

Look for voice models in #1175430844685484042.

[celest violet]

is there any tips? like how much epoch
-# actually what is epoc

[hallow thistle]

You might not need to know about it, but you asked for it.

[celest violet]

so, the more epochs, the more that model has been trained

[pallid rampart]

Hey guys. I downloaded the Vocala voice changer MMVCServerSIO. It can't hear me with any voices except Beatrice. I have a virtual cable. I don't understand what's wrong.

[hallow thistle]

[hallow thistle]

Try Tg Develop's W-Okada voice changer fork instead of the one you're trying to use. Click download https://github.com/tg-develop/voice-changer/releases/download/b2397/voice-changer-windows-amd64-cuda.zip.001 and https://github.com/tg-develop/voice-changer/releases/download/b2397/voice-changer-windows-amd64-cuda.zip.002.

[viral mason]

holy old

try what Namari suggested

[pallid rampart]

which one I should download?

[viral mason]

both

[hallow thistle]

Both.

[plucky crown]

Is 4090 not enough? I only have enough saved for years so I can buy a 4090 PC

[hallow thistle]

Why are you so cared about it? You generally go for what it works, you can't always expect the voice changer to output under < 1 ms.

[pallid rampart]

Did I understand correctly that I need to create 2 archives of 1 file, and then unzip this archive?

[plucky crown]

Is 4090 that bad I'm about to buy ut

[hallow thistle]

Read my question again.

[hallow thistle]

Open the .zip.001 one using 7-Zip or WinRAR. No need to duplicate these split zip files.

[pallid rampart]

ok i did it and what i need to do next : run the app in the folder MMVCServerSio?

[dapper imp]

It's good. The best rig here would be rocking a double gpu with nvidia for AI stuff and probably amd for gaming

Benefit of both worlds

Or I think it works like so🤷‍♂️

[hot lagoon]

I have a focusrite and I keep haveing this problem with Vonovox where it ether cuts noise while playing a game when I start it or it makes noies when I start it.

[hallow thistle]

You got something new to ask about?

[celest violet]

this is optional, right? @hallow thistle

the merge

[hallow thistle]

I don't even understand what you're trying to convey. Try record the video and Vonovox without VRChat or any other program sound interfering.

[hallow thistle]

You don't need to do this one.

[celest violet]

ok, so this voice setting is done, the rest is just setting the VAC input and output, or did I miss a step?

[hallow thistle]

Yes, you missed some steps.

[celest violet]

oh no

[hallow thistle]

You have been asking me back-to-forth, here's your settings:
Chunk: 130 ms
Extra: 2.7 s
Input: your microphone
Output: Line 1 (Virtual Audio Cable)
Monitor: set this to your speakers to hear the program, optional

[celest violet]

sorrry, im too dumb, so I need to ask repeatedly :''D

[hallow thistle]

Actually, I don't like being asked the same questions that have already been answered, so please understand how I try to convey you and also better try ask something new.

[celest violet]

after this step, and start server, its not working after all, did chrome cant handle that?

[hallow thistle]

Make sure you give "microphone permission" to your current Google Chrome tab.

[celest violet]

i do

[hallow thistle]

Can you send the full screenshot of your voice changer interface?

[celest violet]

this

[hallow thistle]

The program seems to work, look at performance section at top right on voice changer. What do you mean the program doesn't work?

Try set pitch extraction model to "rmvpe" not rmvpe_onnx.

[celest violet]

well before this, my voice doesn't come out at all

[celest violet]

after this, my voice come out, but not clear

like hissing

[hallow thistle]

Try another voice model from #1175430844685484042.

[celest violet]

I've tried these 4 types, but none of them are clear, some are like robots, hissing too, did you have a recommendation voice model for beginner?

https://ptb.discord.com/channels/1159260121998827560/1407456715561242645

https://ptb.discord.com/channels/1159260121998827560/1404856994262810816

https://ptb.discord.com/channels/1159260121998827560/1403422896680079390

https://ptb.discord.com/channels/1159260121998827560/1343266090255384586

[viral mason]

I'm horrified

[celest violet]

maybe my device is not enough to use realtime voice changer

[viral mason]

Your GPU is definitely powerful enough to run the voice changer just fine

Not sure why you're having these issues besides possibly just the models you're using being somewhat old or not trained with enough emotion

[hallow thistle]

Terrifying.

[mighty river]

What is the tool for changing facial expression?

[viral mason]

Your face

[celest violet]

this is not good too, https://ptb.discord.com/channels/1159260121998827560/1466674288475246808

[celest violet]

I can show you the results if you want, on the voice channel....

[viral mason]

I can't get in a voice call at the moment I'm sorry

[celest violet]

owh... okey

I'm going to turn off device, maybe rest the device can get some help

[plucky crown]

Helpp

Does all the voice changers here can be run with m4 max or a m3 ultra mac

[kindred kernel]

no it hasnt sadly i was just told to download a different voice changer

[hallow thistle]

That's crazy. Did you extract the files of Tg Develop's W-Okada on your desktop folder? And did you run the program as admin? If these could be issues why neither program worked and none been able to provide a solution.

[kindred kernel]

im having trouble starting the MMVCServerSIO

not tg dev

i was using it for 3 years but now all of a sudden it stopped working after i got a new audio interface + mic

it just shows me the beggining of the cmd prompt then closes

[hallow thistle]

All W-Okada versions even Tg Develop one use the same program name "MMVCServerSIO", except for Vonovox which uses a different way to run the program. If you're still using the older and outdated W-Okada version like v.1.5.3.18a, it's not recommended. Try something else.

[kindred kernel]

exactly thats what i keep being told

[hallow thistle]

This is the program file for Tg Develop's W-Okada fork. https://cdn.discordapp.com/attachments/1159290139609137264/1467193066304831661/image.png?ex=69802646&is=697ed4c6&hm=7670a70d2fcc64ee65cc20518a8dfc4a299736f75974035e715e1a63147a2324&

[kindred kernel]

the one that i have has the "start_http.bat"

[hallow thistle]

And this screenshot is Tg Develop's W-Okada fork. https://cdn.discordapp.com/attachments/1159290139609137264/1463078807937482817/image.png?ex=698057d2&is=697f0652&hm=892cac9ff76dfe56b5340ea9f793ec0acda69071491f2afcc4e8dc825c58e01c&

[kindred kernel]

yes

[hallow thistle]

What do you mean by "yes"?

[kindred kernel]

i mean i know

[hallow thistle]

Did you even download .zip.001 and .zip.002 from GitHub and use WinRAR/7-Zip to open the .zip.001 one?

[kindred kernel]

the thing is i had like 280 voices in that one that isnt working for me

[hallow thistle]

Answer my question.

[kindred kernel]

yes

but thats not what im trying to get

im trying to fix mine

[hallow thistle]

Please no.

Please don't use the old and outdated W-Okada version, just forget about it and try the more recent one.

[kindred kernel]

yes that was the point

why isnt it working anymore all of a sudden

it worked 2 days ago

what changed

thats what im trying to figure out

[hallow thistle]

Please.

[kindred kernel]

the only thing that changed was my audio interface

so im thinking thats the culprit

the one that i was using was the most robust one imo

[hallow thistle]

Ahhhhhhh.

The voice models inside folder of the older one could be uploaded to newer W-Okada fork, but you should do this on W-Okada fork. https://cdn.discordapp.com/attachments/1159290139609137264/1463079103040192556/image.png?ex=69805818&is=697f0698&hm=fef5527b6217bbcbfb6f7fdab55986e8895e4658bae7ef7313ba776ec4aacbe7&

[kindred kernel]

hopefully

[hallow thistle]

And if you keep asking for the origianl W-Okada, you better don't expect any help from here.

[kindred kernel]

but it doesnt do images?

i tried uploading images it doesnt show them

[hallow thistle]

[kindred kernel]

fr

🤦‍♂️

[hallow thistle]

Why should you expect more from W-Okada fork? Just use it as the full potential and don't complain more.

[kindred kernel]

?

[hallow thistle]

So was this why neither one could help you earlier? You were trying to run the older version, oh fuck it.

[kindred kernel]

VCClient is the best alternative that i have found so far

[kindred kernel]

because nobody else reported having issues with it at the same time i did

so im assuning i did something to break it

not that the client has just gotten too old

by itself

maybe some corrupt audio drivers or something idk

[hallow thistle]

Just let it over, bro. Everyone since 2024 suffered the same issues when trying to get the older versions to work even in the big 26.

There's no excuse by now.

[kindred kernel]

its not like i havent tried other ones but they always have bugs and cons and so on so i never changed

[hallow thistle]

There's no excuse.

Got anything to dispute me and other people on why "the older W-Okada versions" work better than newer ones? If that could make you feel better.

[kindred kernel]

[hallow thistle]

I don't have to care too much about the voice changer, generally, because I don't actually use it as daily basis, nor I am terminally online enough to have over 67 models to use at the time. Even when I run the voice changer fork from my own PC for so many times, it still launches, there's nothing concern about it.

[kindred kernel]

its not about excuses its about workflow and logic

[hallow thistle]

You're still excusing to yourself.

[kindred kernel]

this works for me

as i said the new fork isnt displaying images correctly for me navigating a library of 280 voices without visual cues/thumbnails is a massive downgrade in usability the old version handled this perfectly for my needs

[hallow thistle]

Watch me getting this voice changer to work, with CPU-only 2rd generation Intel Core i3.

[timid bane]

can you help me with my vonovox? For some reason I dont hear the voice changer in discord and im assuming I wouldnt in games either

[hallow thistle]

[kindred kernel]

not the one im using

[hallow thistle]

Lie.

[kindred kernel]

thats tg develop

[timid bane]

?

[kindred kernel]

im talking about the W okada MMVCServerSIO

https://huggingface.co/wok000/vcclient000/tree/main

from here

[hallow thistle]

-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 Tg-Develop Fork, with extra features, but it supports only Nvidia GPUs on Windows 10/11 unlike other options that have wider support. 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

Deiteris' fork (modified version) of wokada that doesn't get updates anymore. GUIDE

⚔️ Wokada Tg-Develop Fork vs Vonovox

For Windows Nvidia, Both Wokada Tg-Develop 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 Tg-Develop 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

[hallow thistle]

Fucking lie.

[kindred kernel]

lmao

why?

rly curious abt this

[timid bane]

so then what do I do? for the voice changer to be heard by other users in discord calls and games?

[kindred kernel]

Consider backreading namari

[hallow thistle]

You don't understand? Or you're trying to ragebait around? I'm also curious about your strategy.

[hallow thistle]

That's how Vonovox is. W-Okada voice changer has input, output and** monitor** audio devices but that's about it.

[timid bane]

so there is no way for people to hear me speak with the voice changer on discord calls and games?

[kindred kernel]

how is it a lie Look at the bot message you just sent It lists "Original Wokada" that is exactly what I linked and what I am using.

[timid bane]

unless I use w-okada?

[hallow thistle]

You must be a joke to think the original W-Okada works better than anything else, because that one W-Okada was released like 3 years ago. Got anything to dispute huh?

The original W-Okada might give** worse** audio quality than W-Okada forks and Vonovox, where most people here cared much about audio quality and such.

[kindred kernel]

i never claimed the audio processing was better i said the UI is better for my specific workflow
there is a diference between performance and usability
managing 280 voice models without image thumbnails (which the new fork breaks for me) is a nightmare
software code doesnt have an expiration date It doesnt just self destruct because its 3 years old It worked 48 hours ago
the variablle that changed was my audio Interface that means the config file is likely pointing to a device ID that no longer exists causing the crash on launch that is a verifiable technical issue not a "joke"

[hallow thistle]

Most importantly, the older versions don't always mean they work well than more recent versions. When you claim "older versions" work better, it makes you look like you have that mindset that prevents you from trying new things.

[timid bane]

is there a way for people to hear me on games and discord calls or no?

nvm

[kindred kernel]

yeah with a Virtual Audio cable

[timid bane]

figured it out

[hallow thistle]

This might give you some idea.

[timid bane]

I thought it was like the others where I could hear myself talking through my headphones AND it display the changer while I talk through apps like discord etc

[hallow thistle]

Since I rarely use Vonovox, I'm not sure what to explain.

[timid bane]

I use the lite version

[hallow thistle]

[kindred kernel]

Lite version only has one but should be enough for you

[timid bane]

Idk I just used whatever the guide said to download

but its working so im cool

[kindred kernel]

if you also use voicemeter it will be

okay good

[timid bane]

any apps yall can suggest I can use while I play games so I can hear myself just incase

[kindred kernel]

It isnt a mindset or a refusal to try new things i literally told you earlier that i did try the others it is a feature requirerment the new forks do not properly load or display the image assets for my library navigating 280 models without visual thumbnails is a workflow breaker for me newer does not automatically mean better for every user if the new version breaks the ui features i rely on then for my specific use case it is a downgrade

[dreamy willow]

chat what would you suggest, run the stuff from my cpu or gpu? i play a game where my gpu is always on 100% load but i know it does not matter that muich, it can just drop my frames, but my cpu floats in between 25-55% (97600x (r7) and 9060xt 16GB )

[hallow thistle]

Still.

[dreamy willow]

it's the w-okada fork

[hallow thistle]

GPU.

[dreamy willow]

thank you

[kindred kernel]

Validating my workflow preference wasnt the question though

the question is technical: the application is crashing on launch because "I THINK" the config file is pointing to a hardware ID that no longer exists (my old interface) i am asking if you know where the configuration file is stored so i can manually reset the audio device to default if not thats fine

[hallow thistle]

Because you cared too much about "your workflows", there's Deiteris' W-Okada voice changer fork (b2332), which while older than Tg Develop's W-Okada, and because Tg Develop one is based on Deiteris one, the b2332 W-Okada lets you to upload an image for a model image cover.

[kindred kernel]

does that fork automatically recognize the existing image mapping from the original folder structure? because re uploading 280 images manually one by one isnt "caring too much" its basic time management i would rather spend 5 minutes fixing a config file reseting the audio device than 3 hours rebiulding a library on a new client that is why i am prioritizing the repair over a migration

[hallow thistle]

The names "VCClient" and MMVCServerSIO refer to the same W-Okada program, by the way. "Deiteris' W-Okada fork", while opens your browser much like Tg Develop's fork, has more similar UI to the original W-Okada (v.1.5.3.18a) which uses its own separated app interface. Look at this screenshot.

[sacred temple]

What can I use instead of llaria rvc

[hallow thistle]

What makes you think some other W-Okada versions could use your "older" W-Okada version's folder that way?

[hallow thistle]

Applio RVC.

[sacred temple]

thx

[kindred kernel]

yeah i know i tried deiteris yesterday aswell it was buggy sadly

[hallow thistle]

What's even wrong with your PC system? Why won't yours get that lucky like other people?

[kindred kernel]

fr

now youre asking the right questions

lets figure it out

[hallow thistle]

The recommended settings for both W-Okada fork versions are:
Chunk: 130 ms
Extra: 2.7 s
Pitch extraction/F0: rmvpe (not rmvpe_onnx)

I don't use the older and outdated W-Okada versions, so I'm not sure what's the setting for them.

[kindred kernel]

just started deiteris btw

it was working yesterday lmao

[hallow thistle]

Try extract either W-Okada fork to somewhere like "D:\MMVCServerSIO" or "E:\MMVCServerSIO", not in "C:\MMVCServerSIO" or your desktop folder. Make sure you don't run the program as administrator.

[kindred kernel]

let me try

still the same

[hallow thistle]

Try another web browser app.

[kindred kernel]

wait now it works i think

deiteris

why no rmvpe_Onnx?

[hallow thistle]

For NVIDIA GPU, the regular "rmvpe" F0 model works better for it, while rmvpe_onnx works for any GPU (Intel Arc/AMD Radeon RX) but sometimes slightly worse on NVIDIA GPU.

[kindred kernel]

alralr

[rigid dove]

.

[tough bronze]

Hello guys, I wanted to ask you something. I am using a virtual cable, downloaded its driver, restarted my computer, started testing voices, and can hear myself perfectly altered in the launcher, but how do I enable the altered voice in Discord, for example? I enabled Virtual Cable by default for Windows and set it as the input in Discord, but I simply cannot be heard. Could you possibly advise? Maybe I can find a solution somewhere here in the various sections? I don't want to waste my time or yours.

[hardy yew]

Did you set the virtual cable as output in the voice changer? Which app are you using?

[kindred kernel]

Just uninstalled my Focusrite Audio interface drivers and that fixed my original W-Okada

it starts now

i fucking knew it 💔

[spark epoch]

hi

hii

[viral mason]

Helo

[spark epoch]

Hello, how can I create a cover?

I'm a new girl on the server

mmmmm

[viral mason]

You can use Applio, the docs are right here

-Rvc

[patent trellis]

📚 RVC Documentations

AI HUB Docs

https://docs.aihub.gg

🍏 Applio Docs

https://docs.applio.org

[viral mason]

All you need is applio which can be used locally on your computer or online on a website, and a site to seperate vocals and music from a song

[spark epoch]

I don't know anything. Is there an explanatory video?

I don't have a computer

[viral mason]

No pretty much all videos are outdated

[viral mason]

Oh

Idk how to help you then I'm sorry, you could ask Namari

They should be around here somewhere

[spark epoch]

Okay, no problem! Thank you for your trouble with me.

[viral mason]

You're welcome, sorry I couldn't really get much done

[spark epoch]

It's okay, I was pleased to meet you! ❤❤❤

[viral mason]

Likewise! I hope you have better luck today with someone who knows more ^^

[kindred kernel]

i plug in my Focusrite Scarlett 2i2 4th Gen Audio Interface and the voice changer doesnt work

i plug it out

it works

💀

[spark epoch]

Hello, may I ask you a question?

[kindred kernel]

me?

yes

[spark epoch]

Yes

[kindred kernel]

sure but im no expert

[spark epoch]

Is the company "ai hub" The company's brother "weights"؟

[kindred kernel]

sry no clue lol

[spark epoch]

Hahaha, it's okay

[kindred kernel]

but im pretty sure weights is a company and ai hub is just a discord

Weights IS SHUTTING DOWN??

[spark epoch]

Okay, but I accessed the AI Hub platform through Google, and it says there that AI Hub creates audio overlays.

[kindred kernel]

maybe different ai hub?

like i said im no expert idk :(

[spark epoch]

No, I entered through an invitation link.

[kindred kernel]

this discord?

[spark epoch]

But what is the AI Hub platform? And does it really create AI-powered covers?

[kindred kernel]

i dont know man im sorry

wish i could help

you can check this out

https://docs.aihub.gg/essentials/how-to-make-ai-cover/

https://docs.aihub.gg/

[spark epoch]

Okay, let me try.

[tawny viper]

hey can u help

me

pls

[spark epoch]

What

Are you fine

[tawny viper]

im fine

[viral mason]

Nope, pretty much not related at all anymore

[tawny viper]

how can i fix cannot start voice conversion no gpu avaible?

[spark epoch]

Oh... is ai hub still making covers?

[tawny viper]

help

[viral mason]

I'm unsure what you mean

[tawny viper]

i cant send picture

[viral mason]

You can use applio for covers but weights was the only free website that let you do that

[tawny viper]

it says cannot start voice conversion - no GPU available. You can still access settings and configuration.

[spark epoch]

Okay, thank you

[tawny viper]

help

@viral mason can u help

[viral mason]

Did you get the voice changer off a yt tutorial?

[tawny viper]

what

[tough bronze]

yes. realtime voice changer client ,last version, i guess

[viral mason]

Do you have the link you used to download it? You may be using something old

Voice changer client is outdated, the only two I'd recommend anyone is Wokada tg fork and Vonovox

[tough bronze]

https://huggingface.co/wok000/vcclient000/tree/main
downloaded this file : 'vcclient_win_cuda_2.1.4-alpha'

[viral mason]

Yep that's super old

What is ur gpu on your pc

[tough bronze]

ryzen 7800x 3d

super old? lol. where i can find new newest version?

[tough bronze]

can u send me a link,ty

[viral mason]

Sure yea

-rt

[patent trellis]

💻 Local Realtime RVC

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

• Vonovox

A Realtime Voice Changer with similar performance to Wokada Tg-Develop Fork, with extra features, but it supports only Nvidia GPUs on Windows 10/11 unlike other options that have wider support. 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

Deiteris' fork (modified version) of wokada that doesn't get updates anymore. GUIDE

⚔️ Wokada Tg-Develop Fork vs Vonovox

For Windows Nvidia, Both Wokada Tg-Develop 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 Tg-Develop 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]

https://github.com/tg-develop/voice-changer/releases/download/b2397/voice-changer-windows-amd64-dml.zip

Here ya go

The UI is a little different from the one you have rn but this one runs way better

[tough bronze]

Why did you ask about my gpu if I can set my video card in the settings?

[viral mason]

Because there are Nvidia versions of the voice changers and AMD versions

I don't wanna send the wrong thing

[tough bronze]

My fault, you meant my graphics card, I have a 4070 ti) can u resend

[viral mason]

Ohh ok

Ye one sec

https://github.com/tg-develop/voice-changer/releases/download/b2397/voice-changer-windows-amd64-cuda.zip.001

https://github.com/tg-develop/voice-changer/releases/download/b2397/voice-changer-windows-amd64-cuda.zip.002

You'll need these two

Extract the first one then drag and drop the second one into the first one

The second one cannot be extracted but is still needed idk why

[tough bronze]

ok love u

[tough bronze]

! C:\Users\kojda\Downloads\voice-changer-windows-amd64-cuda.zip.001: Unexpected end of archive
This error occurs when I simply open the ZIP file

[viral mason]

Hmmm

Edit the file name on ur PC and delete the part where it says .001 after the word .zip

[tough bronze]

! C:\Users\kojda\Downloads\voiceChanger.zip Unexpected end of archive
the same thing f
This error even occurs when I try to unzip the file.

[cold kindle]

Hello I’m new here

I build an AI agent that automates almost everything and has a very powerful memory that becomes your personalized agent, and it’s very proactive doesn’t always wait for you to tell it what to do, you can automate tasks using simple and natural language. It’s called Jotium, I’d be glad to have 5 early tester

[tough bronze]

Sorry for tagging you, but could you send me a link to the GitHub file? Unless, of course, I can solve the problem myself.

[hardy yew]

you need to have both .001 and .002 files downloaded in the same place

then just extract the .001 one

don't rename them

[tough bronze]

ill try now

[quick kite]

can someone help me out
im having issues with the voice changer not going to harvest

@crude bolt @primal valley @serene totem @unique blade Do you know how I can harvest/crepe in my voice change software?

[unique blade]

no

[primal valley]

No, imagine using ai to change your voice.... smh my head

[quick kite]

i forgot you're morally better than the rest of the internet

[primal valley]

I'm morally the greatest on the internet
thank you

[quick kite]

you got it

[tough bronze]

only .001?

[hardy yew]

it should use both files, it's a single zip split into two files

"unexpected end of archive" means it read first file but didn't find the second

so if it extracts without error that'll mean it used both files and you're good to go

[tough bronze]

yea thank u it works for me i already unziping them

witout preblems

[void flume]

Which voice change software do you use? Did you make your own or are you using vonovox or w-okada?
It helps if you share more details about what you set up basically.

[quick kite]

i just uninstalled

the one i was using from this youtube tutorial

which one should i do?

[viral mason]

any from youtube are old

[quick kite]

thats what im thinking

and they had the nerver to play ads

[viral mason]

what is ur pc graphics card? (Nvidia or AMD)

[quick kite]

4070SUper

[viral mason]

nicee

u should use vonovox

I'll get u the download

[quick kite]

i just downloaded it

[viral mason]

oh

this one right?
https://github.com/dr87/Vonovox/archive/refs/tags/v1.6.9.zip

[quick kite]

nope haha

ill download it

[void flume]

the beta has problems, the 1.6.9 version is probably what you want to use

[quick kite]

its setting up

[viral mason]

https://tenor.com/view/kiss-boys-kisser-mega-kot-brick-gif-11900063593491011307

[viral mason]

that's the one I sent them :3

[void flume]

I know ;p

[void flume]

https://software.muzychenko.net/freeware/vac470lite.zip
also use VAClite instead of VB-Auidio

[viral mason]

oh yea

[void flume]

all those youtube videos make you install vb-audio for some reason...

[viral mason]

bc they don't know the good stuff

[quick kite]

its installing rn

but

if this works

YES I HAD TO USE VB AUDIO

I was like wtf...

but if this work I got something for you @viral mason

[viral mason]

what is it? :o

[quick kite]

well lets see if ts works

then i got you

[viral mason]

I thank u in advance

[quick kite]

just a thank from me to you

cause i hate youtube tuts

especially with the annoying music just talk to me like a human

[viral mason]

ikr?

[void flume]

I had to do shopping earlier so I couldn't immediately get on the topic, but I am not sure why the people you pinged said 'no'.

might be the misunderstanding on 'my voicechanger'.

[quick kite]

dont know but i keep it pushing

its discord so

anyone can be however they want

[viral mason]

yea some ppl on discord are really annoying

some not

[quick kite]

okay now its running

[viral mason]

nice

[quick kite]

Okay

we lit what to do

[viral mason]

ok so first scroll till u find the input and output

[void flume]

So, in Input you put your microphone
in output you put line 1 (virtual audio cable)
and monitor you use if you want to hear a preview of the changed voice.

[viral mason]

and do what Rumi said

[quick kite]

its in there

[void flume]

You also need to load in a voice model in one of the profiles and use that.

[viral mason]

should test one of mine trust

mine are peak

[quick kite]

pass it

[viral mason]

uhhh

[quick kite]

my goat

[void flume]

I agree, worm is both nice and makes peak voice models

[viral mason]

idk what u really want

[quick kite]

trying to sound like a girl

[viral mason]

I am starting to get sea sick

[quick kite]

dont do that lol

[viral mason]

any reason whyyyy?

[quick kite]

building a music track and i want to hum out the tunes in a proper girl voice to see what it would sound like before finding a artist

who can recreate it back to what i want it too

[viral mason]

oh ok

thank god

uhhhh

[quick kite]

what other reason would it be fore

[viral mason]

just check hall of shame

[void flume]

A lot of people uhm... break one of the server rules basically

[quick kite]

lmao thats funny

[viral mason]

fishing for cats you could say

[quick kite]

nah i just need it because the ai i was using wasnt doing it like i envisioned in my head

and my voice to deep for pitching to do the sole job

[viral mason]

anyway I literally made a good singing voice last night
https://huggingface.co/YourLocalWorm/Synthvmodels/resolve/main/KasaneTetotalkcvlc1.5_215e_7740s.zip?download=true

it can go really high pitched without breaking

[quick kite]

once i export

and add the preset

i cant hear it from my speakers

[viral mason]

Rumi what was that thing u can do to hear vac lite?

[void flume]

"Listen to this Device" in Sound Settings (Settings App > Sound > Advanced)

but I suspect that their default audio device just changed to line 1

[viral mason]

ohh yea

it does that

[quick kite]

i changed that back

[void flume]

vaclite is only needed for real time voice changing. If you just want to change the voice of an existing file, you shouldn't even need that.

[quick kite]

i need to see what it sound slike real time but hmm

dont know where to hear it

even on discord my mic says it doesnt pick it up but

[void flume]

Something is wrong I guess.
I should note that the voice changer will convert your voice or whatever voice to a lower volume by default, so you do need to either bump up the output volume or use gain.

[quick kite]

gotcha yeah im just trying to hmm

hear it from the software

[void flume]

but if discord isn't even picking up audio from the mic, then something is off

[quick kite]

but mabe im doing something wrong

ill restart pc

@viral mason how to hear back the audio in real time

wait

[viral mason]

not sure I use a complicated setup that's different than normal

[quick kite]

hold on i think im cooking

can you jump in a call with me?

im trying to test and see

[viral mason]

nah my partner has kidnapped me into a call

[quick kite]

no prob

i have a bit of a echo

[viral mason]

hmm

[quick kite]

is it because my speakers are picking up the audio back into the mic

so i should use headphones

[viral mason]

I don't understand how ppl get that issue I have never had that

[void flume]

Likely yeah

[quick kite]

hmm maybe not a way around it...

[void flume]

There is, but you need to keep the volume of your speakers low

and then use a Noise Gate setting to filter out softer sound

it's not ideal though

[quick kite]

yeah even chatgpt says use headphones...

because it wont interfere.

[void flume]

Closed Ear headphone, or open ear with sound proofing panels may work better

[quick kite]

and i need the speakers for the music aspect

let me grab my airpods

okay i think i got it all to work

i think the last part is when i do the discord test because its the only test that i can hear my voice back in real time my voice sounds the same im worried in the music software im using it will be like that as well any pointers

[viral mason]

u can check on discord by using this

make sure in discord u mic is line 1 and output is ur headphones

[quick kite]

but once its set on line 1 for voice my mic doesnt work anymore

[viral mason]

weird

[void flume]

The mic should be in use by vonovox

The browser page that loads up when vonovox is launched needs to remain open, and it needs permissions to use the microphone and line 1 for input and output

[quick kite]

Input MIC In Elgato Wave 3
Output Headphones Apple Airpods
On Vonovox

[viral mason]

there's the issue

ouput has to be line 1 in Vonovox

[leaden veldt]

could i get some help here? https://discord.com/channels/1159260121998827560/1467643299748647073 i don’t know how to activate this

[quick kite]

okay but once the output is line 1 i cant hear sound anymore from my headphones/speakers

[viral mason]

idk how to fix ur issue I am out of ideas

help rumi

[void flume]

[Vonovox] | [Discord]
Input: Elgato Wave 3 | Input: Line 1 (virtual audio cable)
Output: Line 1 (virtual audio cable) | Output: Airpods

Then try the "Lets check" test in discord

[leaden veldt]

i dont know what any of this is or means

im new to it

[viral mason]

I saw ur trying to use an egirl model

why?

[leaden veldt]

just to mess around with some friends

[viral mason]

hmm

ok

[void flume]

I can't post any recommended settings without knowing their gpu at least. It's possible they downloaded the wrong build of deiteris's fork.

[quick kite]

@viral mason

it works

you're the goat

@void flume you are too

Can I dm you to get the prize?

[viral mason]

sure

[tawdry linden]

can i still get help here?

orrrr

cause i do need help

with applio

i cannot train anything

it just dosent train

any epochs

how do you reset applio?

or how do you fix this?

here is what is happening now

[celest violet]

is windows wasapi is the best setting compared to other?

[tawdry linden]

i run it locally, not colab

what do i do?

[astral rune]

ANYONE GOT MOMMY RCV I CAN HAVE

[tawdry linden]

has anyone experienced this?

?

[inner elbow]

i have intel gpu which one should i install

[tawdry linden]

i still need help fixing my applio

anyone?

?

[viral mason]

you're cooked

[inner elbow]

😭

[tawdry linden]

?

[tawdry linden]

can you help me?

[viral mason]

what issue are u having?

[tawdry linden]

applio

can't train shit

this is all i get

i can't train jack shit and it's infuriating

plz help me

[viral mason]

eh..

[normal shoal]

How do u train?

[tawdry linden]

i do it like the vids tell me

i preprocess

extract data

set values and options and train

[viral mason]

I have no idea how to use it locally

[normal shoal]

What vid

[tawdry linden]

tutorial videos

[normal shoal]

Where at bro

[viral mason]

any yt tutorals use outdated stuff, applio is still used but you may have an outdated version

[tawdry linden]

like any tutoial video

[normal shoal]

Dude

[tawdry linden]

btw i use v3.6.1 which looks the most recent

[viral mason]

o nvm lol

[tawdry linden]

so what do i do?

[viral mason]

idk what would be causing it as I use applio on kaggle and not locally

[tawdry linden]

what is a kaggle?

[viral mason]

it's like google colab but better

[tawdry linden]

show me

im gonna try it

can i dm u for future help?

[tawdry linden]

show me the link

on how to use it

?

[viral mason]

for the site u have to phone verify btw
https://www.kaggle.com

[tawdry linden]

weird

i didnt have to

can u tell me how to run applio plz

?

[viral mason]

go to settings, u have to phone verify to use the gpu in kaggle

sorry my partner is keeping me from trping consistently

[tawdry linden]

it is fine

i just need to do this rq

does it disconnect from runtime?

[viral mason]

nah, u get 30 hours of free runtime per account every month

[tawdry linden]

wat

actually ill do kaggle later

for the other model

how do u prevent google colab from runtime disconnection

[viral mason]

u can't there's a set timer for that

[tawdry linden]

i thought you could import code?

[viral mason]

probably but I haven't used colab in literal years

[inner elbow]

is there any replacment for intel iris xe?

[viral mason]

what's that?

[inner elbow]

an intel gpu type

[viral mason]

get amd or nvidia

[inner elbow]

i got a laptop+too poor for that shit

[viral mason]

ur cooked on anything good then lol

[inner elbow]

i see

[viral mason]

sure yea

[void flume]

sure, but I may not be online atm

[quick kite]

Oh it was only for worm because she helped first lol

[violet crag]

where can i download all voice models from my collection in weights.gg?

because weights.gg is going shut down in march 31st

[viral mason]

did you make them?

[violet crag]

no, i just adding on my favorites collection

[viral mason]

Just go to the collection and download them I guess, I'm not sure how to find the collections this isn't the weights.gg server

[violet crag]

i see, i just trying to download a collection of voice models from weights,gg but idk how

[viral mason]

I see you're in the official discord server for weights, I'd ask there ^^

Probably more people who would know

[hallow thistle]

You don't need help anymore. Just go for whatever works for you, anyway. End of the story.

[hallow thistle]

No.

[hallow thistle]

Basically another Intel integrated GPU that's the level of "UHD Graphics".

[pallid rampart]

Hi! What settings should I use to make things more convenient? Also, for Input Volume and Output Volume. Sorry to disturb you from your work.

[surreal sierra]

Can I make RVC Covers on my phone?

[hallow thistle]

For mobile phones, only online websites work.

[hallow thistle]

Chunk: 140 ms
Extra: 2.7 s
Pitch extraction: regular "rmvpe"

[brazen gorge]

Sorry, but neither Last.fm or Spotify know the tracks for Best of Romance: Atif Aslam by Atif Aslam.

[trim onyx]

has there been any major changes to ai voice changers?

[chilly furnace]

no

[viral mason]

Vonovox existing kinda

It's the best one currently

Nvidia only tho

[cunning night]

i m new how to create my own RVC v2 voice in .pth?

[hollow salmon]

hi

how can i adjust the delay of the voice?

im using the vcclient

[viral mason]

That's old

What is your pc graphics card

Like AMD or Nvidia

[hollow salmon]

3080

nvidia

[viral mason]

Bet

You should use wokada tg fork I'll get the links

-rt

[patent trellis]

💻 Local Realtime RVC

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

• Vonovox

A Realtime Voice Changer with similar performance to Wokada Tg-Develop Fork, with extra features, but it supports only Nvidia GPUs on Windows 10/11 unlike other options that have wider support. 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

Deiteris' fork (modified version) of wokada that doesn't get updates anymore. GUIDE

⚔️ Wokada Tg-Develop Fork vs Vonovox

For Windows Nvidia, Both Wokada Tg-Develop 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 Tg-Develop 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

[hollow salmon]

so the second guide right

[viral mason]

Yup

https://github.com/tg-develop/voice-changer/releases/download/b2397/voice-changer-windows-amd64-cuda.zip.001

https://github.com/tg-develop/voice-changer/releases/download/b2397/voice-changer-windows-amd64-cuda.zip.002

You'll need to extract the 001 zip then place 002 into the folder for 001

Since the second one cannot be extracted but is needed in the same folder as 001

[hollow salmon]

and i cant use the regular Virtual cable

i have to use VAC Lite

[viral mason]

U can use the regular VB cable but it's recommended to use vac lite over that one

[hollow salmon]

the first one is damaged for me idk why

[viral mason]

Just rename the file by remove the part that says .001

After the .zip

[hollow salmon]

didnt fix it

[viral mason]

Idk what else to do

[hollow salmon]

i will redownload

[viral mason]

I didn't have issues unzipping the zip file

Maybe Namari can explain this

[hollow salmon]

workedn ow

[viral mason]

Nice

Must've been a glitch earlier then

So just unzip, place the 002 zip inside the folder for 001 and run mmvcserversio

It's an exe file

[hollow salmon]

yep running rn

now a website opened with settings

[keen apex]

where can i make rvc Covers of songs

[obsidian swift]

Where can i find the discord server of OpenClaw or alike?

[viral mason]

Applio

[viral mason]

That's the voice changer 👍

[hollow salmon]

thx

how can i go about reducing the delay?

[keen apex]

thx is it the one on github?

[solar mauve]

hi yall, i dont know how use these voice models and need help

[viral mason]

I personally lower the chunk to around 120 something

And extra is always at 2.7

[viral mason]

Depends on what you want to use them for, for covers you would need to use applio but for realtime there's different voice changers that can be used like Vonovox or wokada tg

[solar mauve]

I wanted to try using real-time AI voices, like voice changers, but I don't know what program to use or the best settings, it's been a long time since I tried and I don't remember anything and I lost my programs

[solar mauve]

may i dm you and ask you there?

[viral mason]

What's your PC graphics card? Amd or Nvidia

I can get u one rn

A voice changer

[solar mauve]

Nvidia

[viral mason]

Is it like a 30 series, 50 series

[solar mauve]

40

[solar mauve]

its good or bad?

[viral mason]

pretty good

are you looking for realism or wha

just to have fun?

[solar mauve]

mhhh, both tbh, i have some voice in mind i alr saw, but as i said i dont know how i should start and allat

[viral mason]

oh yea! voice settings should be like this
input: regular headphone/headset mic
output: line 1 (vac lite)

and swap around for games and discord

[sullen heron]

Why am I getting this? There are audios in the logs folder of this experiment

[sullen heron]

Fixed it by just turning audio cutting to simple instead of skip, my bad lol

[gusty moth]

-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

[clear oasis]

Can someone help me? I can only create AI covers on my phone, and I usually do it with Huggingface or RVC Zero, but there are others. But I never really understood how to actually configure it.

AICoverGen
rvc_zero
RVC_HFv2

I find it easier to use the first two, so I use them more, but I don't know how much I should correctly adjust in Index Rate and Filter Radius,RMS mix rate,Protect rate (this is AICoverGen), or, Envelope Ratio etc.

If someone could explain it to me better, I admire English, I love the music, but it's not my native language and sometimes I get confused and lost.

[patent trellis]

❓ Help Guidelines

If you need urgent help, please checkout our AI Hub Docs or ask for help here following the [Guidelines](#1402790586028789830 message)

[tough vine]

Keep your index rate 0.5–0.7; going higher makes it sound exactly like the character but adds robotic glitching
Set your Filter Radius to 3 to smooth out jumpy pitch
and keep Protect Rate at 0.33

nudge the RMS Mix Rate toward 0.25

Transpose: use +12 for Male-to-Female and -12 for Female-to-Male

[tawdry linden]

how do you reset applio local?

or fix it?

because i still cant train anything on applio local

[clear oasis]

Thanks! I'll try <33

[tough vine]

Lmk how it goes

[tough vine]

delete the venv folder and the .conda folder then run the run-install.bat (windows)

[tawdry linden]

is that how u reset it?

[tough vine]

yes

install.sh if linux

[tawdry linden]

is this how it should look?

[patent trellis]

❓ Help Guidelines

If you need urgent help, please checkout our AI Hub Docs or ask for help here following the [Guidelines](#1402790586028789830 message)

[tough vine]

missing requirements.txt and run-install.bat

go back to applio on github, download as zip and then move the missing .bat

once u move the files delete env

then run installer

[tawdry linden]

should i download from relases or the main code?

[tough vine]

releases

its more stable

[tawdry linden]

2 links

which one should i download?

[tough vine]

Applio 3.6.1

[tough vine]

zip

[tough vine]

then do this

[tawdry linden]

not compiled?

or do i get compiled?

[tough vine]

just download the zip

trust

[tawdry linden]

fine fine

i got compiled

[tough vine]

yep

lmk if u need anything else

[tawdry linden]

ill dm u for future help

sound good?

[tough vine]

sure

[hallow thistle]

Download Applio RVC from this repository link. https://huggingface.co/IAHispano/Applio/

[heavy dove]

is there an archive of the voices that were hosted on weights, i was going to snag the rosedoodle rvc on huggingface but the download link leads nowehre

[hallow thistle]

There's #1175430844685484042 for voice models, although it less likely to features a model link direct from Weights.com

[heavy dove]

oof ye tried looking through this discord too but those links also didnt work, so was just seeing if there was anywhere else to look, thank you though

[tawdry linden]

for kaggle do u get 30 hours daily or jst 30 hours?

[hallow thistle]

30 hours in** a week**.

[tawdry linden]

how do you fix this?

i'm trying to run applio on kaggle

but i always get this error

how do you fix this?

[hallow thistle]

"ngrok" tunnel isn't always needed for Applio RVC, but try change a variable line Tunnel = "Ngrok" into Tunnel = "Gradio + LocalTunnel" and the Applio RVC code cell will instead use "Gradio and Loca.lt" websites.

[tawdry linden]

how do u download ur model files for future use?

[hallow thistle]

[tawdry linden]

wdym

can u dl model files to ur computer or no?

[hallow thistle]

Still a simple step anyone can do. To download a voice model from #1175430844685484042, go to a post there, spot a link (usually looks like https:// huggingface.co/Author/repository/resolve/thevoicemodel.zip), click on that link to download zip file to your local drive.

[tawdry linden]

no no

the model i am currently training on applio

[hallow thistle]

You could've explain.

[tawdry linden]

but is it still doable?

?

tell me tmrw

i have to go to bed

or dm me the solution

[hallow thistle]

How did you not know how to download your trained voice model from Applio RVC? On Applio RVC, go to "training" tab, scroll way down, and you'll see options to download your trained model.

[tawdry linden]

ohhh

ok thank u

[hallow thistle]

The another way to download your model(s) is to go to your "loca.lt" website as circled in my screenshot, type your LocalTunnel password, you'll see a file manager one. There's a download button somewhere there.

[spring socket]

Hello! I have 4060 for stream & games, I want to buy a 2nd GPU for W-Okada only, is 1050Ti 4GB enough for running w-okada smoothly while gaming with my 4060?

[viral mason]

it wouldn't run too well tbh on that low of a graphics card

[spring socket]

Yes, I just want the 1050ti to run the RVC while I stream for games, because my 4060 is full of work hahaha

2nd gpu is for okada only~

Thankyou for answering Wormmm

[viral mason]

only for wokada? hmm

might actuall work decently

[spring socket]

Yes, the rest is handled by my 4060

Beast

[hallow thistle]

NVIDIA GeForce RTX 4060 would only work well for W-Okada voice changer and Vonovox, meanwhile same programs might struggle with GeForce GTX 10 series if settings that could work with RTX 40xx being used with GTX 10 one.

[spring socket]

Maybe Increasing the Chunk size and reducing the extra works?

With my 4060 I usually still use 192 / 280 chunk & 1.5 Extra too hahaha

To make okada still runs while I use DLSS Ultra performance while gaming if stream 😭😂

[viral mason]

extra should always be at 2.7 for the best results

[queen glen]

anyone here use okada ? mines like broken

[spring socket]

Yup, but in my past experience reducing it to 1 / 1.5 still works but not as good as 2.7, Idk but it's make the work not too hard right

Reducing it reduce the workload of gpu right?*

[viral mason]

yea I use it

[spring socket]

Best app 🫶🏻

[viral mason]

what graphics card do u have? (amd or nvidia)

if it stopped working u prob have a reallllllllly old one

did u get it from a yt tutorial?

[queen glen]

nvidia

[hallow thistle]

Which NVIDIA GeForce RTX?

[spring socket]

@hallow thistle So, is it okay to buy 1050 Ti just for w-okada to runs and I can talk while I game/stream with my 4060?

[queen glen]

5070

[hallow thistle]

Yes, increasing chunk value might help, but with slightly more audio delay. Extra 2.7 s is the recommended value for better audio quality but that's about it.

[spring socket]

Alright! thanks namari and worm

[queen glen]

nvidia geforce rtx 5070

[spring socket]

I just need it so I can talk 😂

[queen glen]

its pretty good on every game at max ive never had and gpu issues

[hallow thistle]

RTX 2060 or newer might be better, or go for GTX 1050 Ti unless when your PC is smaller or power supply provides lower wattage.

[queen glen]

i mightve botched the files but i dunno what im doing

[hallow thistle]

Either try Vonovox or Tg Develop's W-Okada fork, as both programs can work with RTX 50 series but also competitive at audio quality. Older versions like b2332 and v.1.5.3.18a aren't known to compiled for RTX 50.

-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 Tg-Develop Fork, with extra features, but it supports only Nvidia GPUs on Windows 10/11 unlike other options that have wider support. 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

Deiteris' fork (modified version) of wokada that doesn't get updates anymore. GUIDE

⚔️ Wokada Tg-Develop Fork vs Vonovox

For Windows Nvidia, Both Wokada Tg-Develop 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 Tg-Develop 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

[vapid wagon]

Join up!

[hallow thistle]

[queen glen]

i sound so dumb for saying this i followed the guid to the download page and im a bit confused which file i should download

[hallow thistle]

Download https://github.com/tg-develop/voice-changer/releases/download/b2397/voice-changer-windows-amd64-cuda.zip.001 and https://github.com/tg-develop/voice-changer/releases/download/b2397/voice-changer-windows-amd64-cuda.zip.002.

[queen glen]

just finished up and this one no audio either like okada

so strange

[hallow thistle]

What do you mean "no audio"?

[pallid rampart]

Whats wrong? 🙁

[hallow thistle]

Try increase chunk number up to 320 ms, else show your full screenshot.

If you're running a game at the same time the voice changer running, you might try to reduce your game settings.

[pallid rampart]

What exactly should I reduce in the game so that the voice changer works well and doesn't distort the voice?

[pallid rampart]

I can hear my voice now, but it's a little hard to hear and doesn't work well.

Maybe I should increase the chunk size even more?

[hallow thistle]

Try use "Virtual Audio Cable lite" instead of VB-Cable, as well as change these follow settings:
Sample rate: 48000 Hz
Audio driver: WASAPI
Input: (WASAPI) microphone
Output: (WASAPI) Line 1 (Virtual Audio Cable)
Monitor: (WASAPI) speakers/headphones, this one optional to hear the program

Make sure to increase either input or output audio volume up until audio sounds clear.

The more "chunk" value generally makes audio delay further, but also provides safer specific spot that the voice changer audio wouldn't lagging.

[pallid rampart]

Ok, thanks

[pallid rampart]

Thanks, that helped! I just don't understand the Virtual Audio Cable Lite thing. Also, how can I enable noise cancellation? My microphone picks up the sound of my chair and even the keyboard.

[hallow thistle]

These "noise cancellation" options only available in "client" audio model on voice changer; these same options would only greyed out when set to "server" audio mode.

[pallid rampart]

So, I won't be able to set these settings in client mode? Or will it just not work properly there?

[hallow thistle]

There are trade-offs between these two audio modes: "client" for easier input/output** with** noise suppression but less flexible, "server" for less audio latency and flexible but no noise suppression.

[pallid rampart]

oh.. Ok

[pallid rampart]

What do I need to increase in such cases? :_

[viscid scarab]

help me

[pallid rampart]

I should increase chunk number up?

[hallow thistle]

Are you playing Minecraft with "RTX Ray Tracing" enabled? It's an expected behavior when running a graphic-demanding game alongside W-Okada voice changer at the same time, the perf number spikes up as well. Usually, certain entry-to-mid-tier NVIDIA GeForce RTX (xx50, xx60) GPUs struggle with these, meanwhile higher ones like 4080 or 3090 might be able to run both programs fine. The "12 GB" VRAM size of RTX 3060 variant might help perform a bit better but that's about it.

[pallid rampart]

So I have to turn the RTX 3060 back on? Or do I need to somehow disable RTX Ray Tracing in Minecraft?

Ah... so I have RTX 3060 enabled there.

[hallow thistle]

There are potential solutions if you not yet thinking to switch to newer one:

• Try decrease Extra value down to 2.3 s, this one will decrease GPU usage but with a trade-off by less audio quality as well.
• Try increase chunk number up to 800 ms, as what I said earlier, the more chunk value significantly makes more audio delay.
• Try reduce video settings in your Minecraft, either if the game has RT enabled or running as regular vanila one.
• Close any unrelated programs in background that you might not use.
• Make sure the "processing unit" is not set to "CPU".

[plucky crown]

Hi hopefully you see my message is it working well with that PC specs? I'll be buying the same PC specs because I couldn't afford a 4090 😭😭

[plucky crown]

Helpp is this ok specs and will not lag

[hallow thistle]

Did you check this Facebook page account if it legit or not? Just wanna make sure there won't be any fraudulent there.

[plucky crown]

This is my final pick since is the cheapest 3090 I told the seller to go to me or get a trusted rider and it's cash on delivery but no reply yet, what do you think of the specs for this one, I want no lag in voice changer pls

[plucky crown]

Plss replyy as I want to make sure the PC I buy is not gonna lag on this

[hallow thistle]

Despite having similar 24 GB VRAM size, NVIDIA GeForce RTX 3090 is a slightly the downgrade of RTX 4090 but still capable of running games at 2160p just fine. If your budget still not enough to buy an RTX xx90 GPU,** don't be too exaggerated about your vision over reality**. For a slightly cheaper one with lower specs but still capable of running games at least 1440p, you might check out GeForce RTX 5080 or RTX 4080. All mentioned high-end GPUs can work with Vonovox and W-Okada voice changer, by the way.

[plucky crown]

But isn't the vram too small since 3090 is 24gb

And also I'll use it with heavy video editing and music softwares

[hallow thistle]

Why would you? Neither single program uses over 24GB VRAM like that, but higher Stable Diffusion models (like Flux.1) sure would. W-Okada generally needs at least 8 - 12GB to work well, non-realtime RVC program like "Applio RVC" uses significantly lower VRAM than W-Okada/Vonovox for simple inferencing (AI cover); video editing softwares (like Adobe Premiere Pro) generally use 2 - 4 GB minimum for "hardware" render, whereas DAW mainly uses CPU, sometimes uses minimal or does not use GPU resource at all, unless one VST plugin needs a GPU to display its GUI.

[plucky crown]

But is 3090 slower/doesn't work Is that the issue why I should get latest but smaller vram

[hallow thistle]

Please think more before you reply too quick to catch up the pace.

[plucky crown]

I don't get it

[hallow thistle]

If you care too much about "VRAM" size and specific professional tasks, there's "NVIDIA RTX** Pro**". These GPUs are not for gaming, they cost more expensive than typical GeForce RTX GPUs; some higher-end models generally have larger VRAM sizes than RTX xx90 GPUs, few entry-to-mid-tier models are on par with GeForce RTX xx80 GPU of VRAM size. They are made for rendering certain hardcore scenes from softwares (e.g. Blender, AutoCAD, Adobe Premiere Pro) where the software performance matters than some flashy gaming. Also, almost every RTX Pro GPU can also run "W-Okada voice changer" as much as the consumer GeForce RTX.

Here, I give pretty much information for you to consider for your needs. If you still think "RTX 3090" is not enough for hardcore performance, I'm not sure why would you come here to complain.

[plucky crown]

Was the message removed

I didn't get to read it

So should I get it I'm using all my savings earned for years on this so it's serious, I'm upgrading from dell optiplex 9020

[hallow thistle]

Dell Pro/OptiPlex is a basic business/professional desktop PC for office settings, not a typical gaming PC. For elaborate hardcore gaming ones, there are Alienware and Asus ROG.

[plucky crown]

Yes

So should I get the 3090 and no lag for the voice softwares here

[plucky crown]

I am getting 3090

[hallow thistle]

That's your choice.

[plucky crown]

But does it get no lag with that gpu

[hallow thistle]

[plucky crown]

But is it enough, what specs is your pc

[hallow thistle]

While you might get minimal or no lag when running W-Okada voice changer with GeForce RTX 3060, you might get a bit more delay instead if set Chunk to a certain value spot, while leaving "Extra" value at 2.7 s. There's no absolute predictable result, this is still a subjective debate.

[hallow thistle]

This decade-old laptop has no GPU appearing in Task Manager.

[viscid scarab]

pls help me someone

[tough vine]

What’s up

[arctic badge]

@tough vine hi today i changed my gpu and this happened

[tough vine]

hello

[arctic badge]

hi can u help me please

[arctic badge]

im having this error here

[tough vine]

sec

[tough vine]

ok

try reinstalling pytorch