✨│ai-help

[simple ore]

C:\Applio

[prime shell]

C:\Users\pauma\OneDrive\Escritorio\applio

this is actually

[simple ore]

who says is is a good idea to unzip 7GB of stuff into OneDrive/Desktop

[prime shell]

idk how to change

[simple ore]

might as well take a dump in your bedroom

[velvet nebula]

Is there a difference between merging in wokada and merging in rvc? And will there be a loss of quality in the model that will be formed when I merging 1 quality model and 1 poor quality model?

[simple ore]

ctrl-x, ctrl-v

[prime shell]

You're so funny, get out of the house

[simple ore]

what, you need a youtube video that would explain how to move folders to a different location?

[prime shell]

yep

please

[analog obsidian]

yes, there will be a quality loss

merging is tricky tbh

is hard to predict the results

i prefer to use rvc for merging because w-okada can be very buggy sometimes

[hallow thistle]

Don't. Extract Applio into somewhere like "C:\Applio" or "D:\Applio"

[velvet nebula]

rvc hasn't 32k merging but w-okada has

[compact drift]

umm how am I supposed to upload a picture over here , I wanna know how to use that voice changer on discord vc's

[analog obsidian]

applio has 32k merging

[compact drift]

@oak plank

[compact drift]

.

[prime shell]

okay

[hallow thistle]

Talk a bit until your name turns light green so you can send your screenshot here. How can you tell me what is your PC GPU? And which W-Okada version are you using?

[prime shell]

ik

[compact drift]

what is W-okada

[hallow thistle]

Wow, so you pinged an admin for this one.

[compact drift]

it will be easy to do that in DM's ig?

[hallow thistle]

W-Okada the realtime voice changer.

[velvet nebula]

For example, model 1 doesn't make whispering sounds very well, but model 2 is good at whispering sounds but can't make subtle sounds, can the merged model make up for the shortcomings of the 2 models?

[compact drift]

if you dont mind can I send you the picture of the voice change am using rn

[hallow thistle]

No.

!howtoask

[patent trellis]

How To Troubleshoot

__**GIVE CONTEXT.**__ 📝

• Don't simply mention your issue, like "my rvc is not working".
• Describe the step you are on, what you're trying to do, the RVC you're using, a screenshot, etc.
• The more context, the better.

__**BE POLITE.**__ <:matsuripray:1159685390156967936>

• Don't be desperate. You can ping a Helper, but if they ignore, they aren't available/don't know the answer.
• It's okay if you're frustrated, but don't take it into this server.
• Don't DM without prior consent.

__**BE PRODUCTIVE.**__ 🤝

• Don't ask for every little instruction. Put your own effort & test things by yourself.
• Don't ask to ask.
• Check if your answer is a Google search away/on our guides website.

If you've changed your mind about your message, feel free to delete/edit it.

[compact drift]

uh fine

.

alr

[prime shell]

and how i can download one model?

i download one model but applio get error

[analog obsidian]

interesting
i havent never tried that uh but i dont think it would fix both problems
because the model still has the influence of both models

[hallow thistle]

That's one of the bad questions I've ever seen. For voice models, find one in #1175430844685484042.

[prime shell]

oka

[analog obsidian]

so if you try to inference whisper, the model may still use the knowledge of the non-whisper model

but first try merging them tho, it may work good, no one knows

[compact drift]

BTW is that okada voice changer free?

[oak plank]

If you want to help someone, it's best to only help instead of rating how bad/good their question is

[hallow thistle]

You download a voice model zip file, extract it to a folder, and copy or move one or two files (.pth and .index) into ".\Applio\logs".

[oak plank]

Otherwise they can wait for someone else for help

[compact drift]

ok @hallow thistle this is the client am using

[hallow thistle]

W-Okada is free.

[compact drift]

so like I can use a minecraft villagr sound though that

[velvet nebula]

I didn't know Applio supports 32k, will there be a noticeable difference in merging? I'll start downloading now

[hallow thistle]

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

[compact drift]

yes I have it , but now how am I supposed to use it on discord?

[hallow thistle]

Also, helping two people at the same time sounds like a bad idea for me. I can't focus many things at once.

[hallow thistle]

[compact drift]

I mean I tried using villager sound on discord through checking the soundboard

but it did not work as inteded

it was like 19k ms

[hallow thistle]

Slow down, please. The W-Okada in your screenshot is the original version of W-Okada, which is outdated.

[compact drift]

oh I see

[analog obsidian]

iirc there's no native 32k merging but applio managed to do it by doing ... idk
i know there's an explanation on how they made that possible but i cant remember
w-okada dev probably did the same as the applio team

[hallow thistle]

The original one and the one from the link I sent to you are two different W-Okada versions.

[compact drift]

so shall I delete the older one?

[hallow thistle]

Yes.

Also, no need to ask me about every step by step on this one. I can only focus on installation, settings and a program needed to be used with Discord.

[compact drift]

downloading it is a bit confusing

[prime shell]

RuntimeError: An error occurred loading the audio: Error opening 'assets\audios\Grabación_16.m4a': Format not recognised.

What is this?

[hallow thistle]

Use mp3 or wav instead of m4a.

[prime shell]

okay

[latent kettle]

Bro you can read the errors. These are very easy to fix. Read them carefully before sending it

[prime shell]

okay

srry

[hallow thistle]

[latent kettle]

m4a is not supposed format so it says to convert it into any other format

[orchid merlin]

the mac download keeps getting corrupted for me alraagghhtt 😼

[latent kettle]

Same it says to extract applio in other directory as I told you above. So read errors carefully

[hallow thistle]

The what? Sorry, but did you download the Windows one when you have Mac?

[orchid merlin]

no im downloading the mac one cuz im using a mac

wait i think i know why

its not working

Ohhhh i was downloading the wrong file 😭

[hallow thistle]

Replying to my message where that message was replied to another person is kinda unexpected.

[orchid merlin]

yes but im having the same issue i was using the old w-okada and it's fried asf

[hallow thistle]

Which CPU does your Mac have? Intel or Apple Silicon?

[orchid merlin]

intel

[hallow thistle]

Download the Intel Mac one.

[orchid merlin]

yeah im doing that now thankx

[tulip cloak]

🎉🎉

[hallow thistle]

I think using external GPU with Intel Mac is possible, but you'll need a Thunderbolt-compatible PCIe Riser card, an NVIDIA GeForce RTX GPU and a power supply for this one.

I'm not sure if Detris ever made W-Okada for this especially, other than having Windows running inside Intel Mac in order to use CUDA features from such GPU.

[brisk orchid]

can someone show me how to use mainline?

im using it for training a model

@steel forge

Anyone?

[latent kettle]

Run go-web.bat

https://docs.aihub.gg/rvc/local/mainline/#mainline

And follow the guide.

[brisk orchid]

ty but

where do i get my voice models pth n index files?

@latent kettle

[strong mantle]

What is the best voice cloner?
I use Applio.

[odd shale]

lower chunk number means less delay, but be careful

Welp, don't expect W-Okada to work that smoothly while playing AAA games at same time

Since AI stuff is more power-demanding than videogames

Altho there you have some recommended settings. (you got 3080 so use 192 chunk and 2.7 on extra)

[odd shale]

Try with 2.7 or 2.5

As the guide states

extra

Well, you can test with each to see which performs better for you.

[silent stratus]

where can i learn how to train a melroformer model

@odd shale is there a cloud notebook where u can do it

[odd shale]

I don't know anything about separation model making Bruhy, sowwy

[tranquil radish]

Where do I put the voice pack?!

Like a app or smth

[odd shale]

Elaborate what you want to do

[tranquil radish]

I downloaded a boy model but idk where to use it to turn on the voice changer

Cuz i bought voice mod and alot of things and I got scammed

So i need a app or smth to import in it the model to be a voice changer

[orchid flint]

https://rentry.co/forkvoicechangerguide

[odd shale]

I think you shouldn't have bought voicemod in first place.

[odd shale]

In that case you can use deiteris fork.

[orchid flint]

Its good for adding extra effects to W-Okoda but thats about it lmao

[odd shale]

https://rentry.co/ForkVoiceChangerGuide

@tranquil radish check this guide and install deiteris fork

[tranquil radish]

Ok ty

Is there is vedio I can follow for instructions

Or it's just the site

[tranquil radish]

I will ask for a refund I hope they accept

[orchid flint]

Just the site, its relatively straight forward.

[odd shale]

Nope, we really don't recommend watching any yt tuts

These get outdated pretty quick

[tranquil radish]

Mhm ok

[tranquil radish]

Tysm

@odd shale tysm

[odd shale]

Yo're welcome buddy.

[tranquil radish]

Is this supported by this server or it's a local one

[odd shale]

Local.

[tranquil radish]

Oh so I found also a voice changer from a content creator called duckus he showed a tut u think the on EU sent is the best or cuz u maybe. know this content creator

[odd shale]

Duckus? I'm pretty sure he used OG W-Okada on his videos along his crystal voice model

Meh, just stick with deiteris.

[tranquil radish]

Idk tbh but the tut was a year ago

[odd shale]

Then it's pretty outdated.

Don't follow yt tuts

[tranquil radish]

Deiteris is the one u sent

[tranquil radish]

Ye ty

[odd shale]

Yep.

[tranquil radish]

I am not the best at pcs and stuff so sorry if I am asking alot of questions

[tranquil radish]

Do I need to pay for this voice changer u sent

[odd shale]

What? of course not.

it's free

[tranquil radish]

Oh ok

It lags and stuff or

[tranquil radish]

??

And will I need from time to time to change it or

[orchid flint]

Just look at the guide every so often to see if it has updated, thats about it. It hasn't updated in like 6 months though so you likely don't need to worry for a while.

[tranquil radish]

Mm ok

[versed edge]

When i have wokada forked idling on default browser or without anything open It utilizes around 90-100% of the gpu but when i hop on roblox it goes down to a consistent 50-60% utilization... Should i be worried? (I use rest protocol cuz it sound better, not sure if that's the reason)

[signal bison]

hey guys, i wanna train my voice for singing but i want good quality.

how should i record my voice to get best quality for training my voice for singing?

for example: recording my voice max 5 min for 3 times with total of 3 audio with 5 min each?

and how much (total_epoch): should i do for training? 50? or 200?

[cinder wyvern]

the applio gradio is stuck on installing requirements

[edgy tangle]

Hii

Someone knows how good is the ryzen 7 5800XT + RTX3060 12gb for training?

[solar shuttle]

hola

necesito crear una voz con IA de Feid, como puedo hacerlo?

[simple ore]

it is serviceable

[edgy tangle]

it would work great? or just good

because I don't think a RTX 4060 8gb will work well

[lyric geode]

@tranquil radish Are you using weights too? Do I have to pay to download the covers I make? I love using Nanachi but Weights doesn't have a good Neco Arc voice model like TopMediAi does

[craggy bough]

you dont have to pay to download the covers

[tranquil radish]

Nah see i don't understand in this things 😭

[lyric geode]

I'm on the app & I can't figure out how to download them

[craggy bough]

try the website

weights.com

the app is kinda buggy

[lyric geode]

It won't even let me listen to them unless I'm online with some kinda service like wifi

[craggy bough]

yeah try using the website

[lyric geode]

Hm ok I might have to remake some covers on the website

[craggy bough]

nope as long as youre logged in they should be there

[lyric geode]

Does the website have a better Neco Arc voice model than the app?

TopMediAi is the best I've found

[craggy bough]

website should have the same models

[lyric geode]

Awe alright at least I love the Nanachi voice model

Thanks

@craggy bough ah I figured it out I just had to share my cover to make it downloadable

[lyric geode]

Which do I choose?

It's making it confusing to download the cover

@tranquil radish Maybe you know?

[tranquil radish]

Really I don't 😭

[lyric geode]

oof

Hmm if output is the first choice maybe that's it I'll try that

[tranquil radish]

It won't work6on phone bgw

BTW

[unreal meteor]

hey guys. i need a little help here. i download UVR5 UI and im a newbie to this. its running and etc, but im lost in the huge quantity of models. if someone has a "manual" hahaha i appreciate

[viscid moss]

https://github.com/Eddycrack864/UVR5-UI/blob/main/info/docs.md#uvr5-ui-documentation

Here's the list of the best models for

And here's the most huge doc about all models:
https://docs.google.com/document/d/17fjNvJzj8ZGSer7c7OFe_CNfUKbAxEh_OBv94ZdRG5c/edit?tab=t.0

[unreal meteor]

omg haha thanks 130 pages

apollo can run in this system?

the model apollo* for restoration

[viscid moss]

Nope

Not in UVR5 UI

[unreal meteor]

oh okay

last question, why have so many models inside there?

[viscid moss]

The idea was to create an appl that has as many models as possible

[unreal meteor]

thank u for your help ❤️ and answer

and its an awsome app

[viscid moss]

ty and ur welcome :3

[unreal meteor]

oh gosh, hold on, u are the creator of it?

[viscid moss]

yep

[simple ore]

depends on the dataset size, 4060 may be faster

[unreal meteor]

im your fan, like literally. u did an awesome job creating it, and i appreciate your work and effor to put this out

so many corporations use the models and limited the usage, and u manage to make a nice interface, and compilation all that. thanks

[viscid moss]

Thanks I really appreciate it

[hot lagoon]

for UVR5 in kaggle is there a way to change the segment size?

[hot lagoon]

Do I also have to check the "Override segment size"

[viscid moss]

nah

[craggy bough]

output

[viscid moss]

for some roformers results could be weird

[hot lagoon]

Do you recommend what segment size I should use

[viscid moss]

tbh default

[hot lagoon]

okay

[edgy tangle]

oh, ok

thanks!

[lament bobcat]

best cpu for voice moddles while playing games or even obs

ryzen series im looking for

X3D or just the X

more cores or more threads help?

[latent kettle]

Logs and weights

[knotty moth]

anything based on ur budget

[low shard]

hello

[broken gale]

if someone can help me in downloading the app i would be thankful if someone can please add me beacuse i am fasing alot of proplems

[hushed steppe]

Having loads of problems with applio

I did the preprocess as usual and it's telling me I didn't?

I think it's an applio problem cuz I used the NoUI ver and it worked

[toxic geode]

I'm trying to train a model using MRF-HiFi GAN vocoder in Codename-RVC-Fork-V3.1.4, from pre-processing datasets, pitch extraction, and generating index is doing fine. but why when i try to start the training im facing this error. I tried using different sampling rate and result is the same, now im using HiFi-Gan and its doing fine.
I'm using GTX 1060 4GB.
i want to try using MRF-HiFi GAN vocoder because i'm not satisfied with the quality of the HiFi-Gan.

[gentle tide]

My GPU is AMD Readeon 7700XT and i have the Alpha Version

[pastel oak]

Fork wokada recommended for amd gpus

https://rentry.co/ForkVoiceChangerGuide download amd/intel one (dml)

[pastel oak]

What app, theres hundreds of. Explain what you need

[broken gale]

some one gave me this link
https://rentry.co/ForkVoiceChangerGuide

[pastel oak]

Whats your gpu, whats your problem

[broken gale]

nivedia gtx 750 ti

[pastel oak]

Its too old, will not work well
You need online hosted, follow kaggle link

[broken gale]

mmm

can u send me the link

[pastel oak]

Its in the rentry link

In the Online Hosted Alternative step

[broken gale]

btw what gpu should i get to work well on it

[pastel oak]

Thats also suggested in the guide, think it was
RTX 2060 or better
or AMD RX 5000-series or better

[broken gale]

ok ty

[tranquil radish]

That's my pc will it run the voice changer

And if not what should I get

This one

https://rentry.co/ForkVoiceChangerGuide

[late flicker]

Someone plz help me with Mangio RVC fork. I did everything right, but the final model is shi*t

Do i have to restart the model?

[echo meadow]

I somehow can’t put this model in the download model thingy, it’s like not downloading it

https://huggingface.co/FurnTheFurnace/RVCModels/resolve/main/Jack-Black_150e_15000s.zip?download=true

[distant turtle]

-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

[simple ore]

show the logs for preprocess and extract features and training settings

[hushed steppe]

Give me a moment, I closed everything

I'll try again

[hushed steppe]

In the gradio ?

Or the files?

[simple ore]

something is not right with your model's config.json

[simple ore]

in colab

[hushed steppe]

I'm confused

The settings are in the gradio

[simple ore]

if you're using UI colab, then the logs will be in the colab, but settings are in UI

[golden agate]

could somebody help me get an audiokable i can use for a voice changer

[simple ore]

less unnecessary stuff, more the actual logs

[hushed steppe]

Okay wydm

I'm slow guys

[simple ore]

anyway, you got 12 x 3s files, yet your trying to train with batch 15?

[hushed steppe]

Ah

[simple ore]

stretching too thin, my dude

[hushed steppe]

I just don't change the basic stuff

[hushed steppe]

I didn't know </3

[simple ore]

you dont have enough data, you chose batch size intended for 50+ hours of audio

[hushed steppe]

I didn't choose anything

I just left the default

What should I change it toM

*?

[simple ore]

you did

add mode data, at least 5-10 minutes

not 30 seconds

[hushed steppe]

Those at the default settings when you open gradio

I'm not here to argue, I genuinely didn't know

Thank you for letting me know

I lowered it and now it's working

I'll know for next time

[simple ore]

you cant just lower it to 1 or 2 and expect anything good from 30s audio source

[twilit forge]

how do i use the model im training while it trains. Im training locally on applio

[static gyro]

how do i install rvc live?

[low shard]

So, you want realtime voice changer for calls/games?

[static gyro]

calls yes

[low shard]

RVC = Retrieval-based-Voice-Conversion, the best Speech To Speech AI Models (on v2), Inferences (use models) pre-recorded audio (ai covers) and train (make) models. Technically, Mainline RVC does have a go-realtime.bat (aka RVC-GUI), but it's pretty messy and outdated so it's extremely not suggested for realtime.

Wokada = uses RVC for realtime inference. There's 2 main versions, Original made by Wok, and the most suggested one is Deiteris Fork (modified version)

What's your PC GPU?

[low shard]

I told you the differences between rvc and Wokada, so you'd need Wokada deiteris fork for realtime, what's your PC GPU?

[static gyro]

how do i find out?

[low shard]

You can check your pc gpu via:
ctrl+shift+esc (task manager) -> Performance tab -> GPU

How did you exactly train models without knowing your PC GPU?

[static gyro]

GPU 0

NVIDIA GeForce RTX 4070

Driver version:    32.0.15.6603
Driver date:    10/15/2024
DirectX version:    12 (FL 12.1)
Physical location:    PCI bus 1, device 0, function 0

Utilization    1%
Dedicated GPU memory    1.8/12.0 GB
Shared GPU memory    0.1/15.9 GB
GPU Memory    2.0/27.9 GB

[low shard]

Nice, you got a desktop?

Or laptop?

[static gyro]

desktop

[low shard]

Even better

Wait

So how did you exactly train models till now?

Did you really use colab all this time?

[static gyro]

online yes

didnt have an rtx

[low shard]

Because colab is just one of the cloud computing service (remote good PC) services made for people with a bad pc

[static gyro]

had a laptop

[low shard]

Ohh

Do you want me to tell you also how to train rvc models locally? @static gyro

Or do u just want realtime for now

[static gyro]

real time for now

[low shard]

You should really NOT use cloud when you got a good pc

[static gyro]

understood

[low shard]

-realtime

[patent trellis]

💻 Local Realtime RVC

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

• Wokada Deiteris Fork

Most suggested. GUIDE

• Original Wokada

ONLY the latest alpha comes close to the Deiteris Fork performance, older versions in youtube tuts are way worse. GUIDE

• **RVC GUI Mainline Realtime**

Unavailable, the guide is outdated and the program is worse compared to the ones above, and much less updated

[low shard]

1st link, Wokada deiteris fork

[static gyro]

404

[static gyro]

i dont see download

[low shard]

Oh right we changed the url link just yesterday night

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

Here

-realtime

[patent trellis]

💻 Local Realtime RVC

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

• Wokada Deiteris Fork

Most suggested. GUIDE

• Original Wokada

ONLY the latest alpha comes close to the Deiteris Fork performance, older versions in youtube tuts are way worse. GUIDE

• **RVC GUI Mainline Realtime**

Unavailable, the guide is outdated and the program is worse compared to the ones above, and much less updated

[low shard]

Yup the bot is up to date now

[static gyro]

so now my mic is not working

it is a usb

[upbeat mauve]

Any reason why uvr has ensemble all greyed out? i have the latest update and deleted old presets.

[warm swan]

quick question, Replay keeps saying "Server not running" been sitting here waiting on that for about 10 minutes now am i missing some setting or?
v7.2.3

Edit: closing the server and reopening it had it starting around 15 minutes later

(I did the "remove local server" option already)

[twilit forge]

has this started to overtrain yet?

[simple ore]

not even close

[twilit forge]

thought so

[simple ore]

next time show all avg_50 charts

[twilit forge]

where is that

[somber plover]

whats the software to make my ai model to sing

[twilit forge]

?

[simple ore]

like this

[twilit forge]

bet

[random sand]

i have one question after i donwloadet a voice form weights.com it it dident have the index file it only hat a json file

[hot lagoon]

what sample rate is this?

[brisk orchid]

please help me, it says this when i type the correct format

[latent kettle]

What are you doing

[latent kettle]

32khz

[brisk orchid]

im good now

[latent kettle]

Okay

[distant turtle]

thank you so much really!!!, it worked, yesterday I tried it, but it stopped about halfway as it happened before, today I tried again and I understood that I had to insert and activate this code several times, so every 30 minutes I re-entered it 2 or 3 times to be safe, and the runtime lasted 3 hours and stopped only when the video was finished, thanks

[latent kettle]

Yw XD

It was just a guess work. I was not sure. But it worked 😆😆

[distant turtle]

if anyone has the same problem as me, you know this method works

thx

[wide saddle]

are there any free tools i can use to upscale my photos
to look more realistic
that i cld rn locally also
i have a 3060ti and a ryzen 5 5600

[latent kettle]

Great

You can use comfyui

[low shard]

you want to do locally right? either comfyui or real-esrgan or SwinIR

[latent kettle]

You have to install a custom node named "the ultimate upscaler "

[wide saddle]

okay

[low shard]

elaborate:

• what browser are u using
• did u give it microphone perms
• what's the exact issue

[low shard]

it exactly tells you why, repository not found

either privated or delated

also, what do you mean in the download model thingy? here we offer help for multiple programs, not just one

but i'm guessing the download tab on applio

[echo meadow]

Found out why it’s not working, model is empty

that’s why lol

[low shard]

not that the model is empty, it's that the repository is either private or deleted

and the creator left the server

[echo meadow]

A

It had like 20 bytes

I was like how

[low shard]

yeah, just use another model lol

the model is from 2024

[wide saddle]

where can i find this?

[low shard]

Mangio RVC fork is abandoned since 2023, use applio

i think i already told you that

[wide saddle]

where can i find "the ultimate upscaler"

the node he was talkign abt

[low shard]

4gb of ram is extremely outdated, plus that cpu is weak,does this laptop even have a dedicated gpu?

You can check your pc gpu via:
ctrl+shift+esc (task manager) -> Performance tab -> GPU

[signal bison]

how to train my own voice via collab? i always getting some error with training collab

any latest collab link? for training voice

[low shard]

also, the guide is moved to the docs https://docs.aihub.gg/rvc-voice-changer/local/deiteris-w-okada-fork/

[low shard]

firs, what's your pc gpu?

[signal bison]

how to know

[low shard]

second, never use video tutorials for rvc, they are old

[low shard]

You can check your pc gpu via:
ctrl+shift+esc (task manager) -> Performance tab -> GPU

Colab is a cloud computing service (remote good pc) ONLY for people with weak pcs

it shouldn't be your first option, your first option is to check if your pc is good enough to run it locally

[signal bison]

intel r uhd graphic 770

[tranquil radish]

How many rams should I make it +Nivedia gtx 750 ti

[low shard]

a shit that's integrated graphics, please check if you got any other gpus like gpu 0, gpu 1, etc

[signal bison]

i have amd thats why cant use local

AMD readon rx 6700

[tranquil radish]

It's a pc but I can edit it but idk what to replace or add @low shard

[low shard]

holy shit that's ancient, yeah you're way below the BARE minimum which is a gtx 900, also nowdays the minimum standard is 8gb of ram for everything

your pc gpu wouldn't even be supported by the program

[tranquil radish]

Mmm

[low shard]

@tranquil radish your only option is cloud computing services (remote good pc)

you want a realtime voice changer for calls right?

[signal bison]

i have AMD readon rx 6700 but some one said only nvdia works for local

[tranquil radish]

I just wanna prank my friends so games and stuff

[low shard]

AMD is supported locally

your gpu is supported via Zluda (Cuda emulator) on windows

[signal bison]

how to do it is there link to download the working one?

[tranquil radish]

If i buy his would it be good https://abosalahpc.com/product/تجميعة-15000ج/

(That's a computer website)

[low shard]

your only options are:

• pay for an atleast decent pc
• use the kaggle wokada deiteris fork https://www.kaggle.com/code/suneku/voice-changer-public with 30 hours for free weekly, but you need a phone verification

also what games? i fear your pc can't run any online games

[tranquil radish]

There isn't 1 voice changer for calls and games and everything in 1?

[low shard]

can't read arabic, but 8gb of ram with a gtx 900 is the extreme bare minimum to even get supported, even though an rtx 3060 would be suggested if you're on a budget

[low shard]

https://docs.applio.org/applio/getting-started/installation#amd-gpu-support-windows

[tranquil radish]

Mhm

[low shard]

this is applio, an rvc fork, modified version

[tranquil radish]

What's that?!

[low shard]

wokada deiteris fork works for both calls and games, everything that is realtime

if you want inference (use models) on pre-recorded audios, then you'd need another program

[tranquil radish]

Real time means live right?!

[tranquil radish]

Does all look like this

[low shard]

it's not related to your thing dw, it's a modified version of rvc needed for pre-recorded audio

[tranquil radish]

Mmh OK

[low shard]

that's wokada deiteris fork, is that someone's else picture or you got a desktop?

[tranquil radish]

Someone's else Pic

[low shard]

realtime means in live yeah, like using it in a discord vc

[tranquil radish]

Mm

[low shard]

i gave you the options, there's nothing more you can do than that

your pc could have issues even on lowest graphics for games atp

[tranquil radish]

Do u have a pic of the inside design of this voice changer

[low shard]

what?

[tranquil radish]

I am planning to buy a better one

[tranquil radish]

Like

[tranquil radish]

The design of the voice changer the style or they all the same

[low shard]

wdym the style

the picture you sent is how the wokada deiteris fork User Interface looks like

the local guide is https://docs.aihub.gg/rvc-voice-changer/local/deiteris-w-okada-fork/ for people who got a good pc

[tranquil radish]

Mm

[signal bison]

does applio also work for singing?

[low shard]

yes, it's an rvc fork, rvc is Retrieval-based-Voice-Conversion, so speech to speech

[tranquil radish]

I wanna know how kaggle looks like this voice changer

If u have. A pic

[low shard]

the interface is always the same, it's the same program

[tranquil radish]

Oh

Then what's that diffrence

[low shard]

kaggle isn't a program

its a cloud service provider

[tranquil radish]

Mhm

[low shard]

it has it's own user interface for the code cells, all it does is using a remote good pc

and it will run the program's code

[tranquil radish]

This is complex ughhh

[low shard]

AI is more complex than what Chatgpt makes it look like

this is Open Source AI

[tranquil radish]

Mm

Can u add me or smth cuz i would need ur help when I get my new pc

[low shard]

you can use #1192011222023950368 or #✨│ai-help

no need to dm

[tranquil radish]

Oh ok ty

Is there any Arabic Ai supporter?!

[low shard]

nope, this server is english only mostly

it's the most common language on the internet, and we can't really moderate other languages

[tranquil radish]

It's cuz there is a pc I planned to buy it but idk if it will run the voice changer

[tranquil radish]

Mhm ok

[low shard]

what are the specifications

[tranquil radish]

Well i really don't understand but lemme see it again gimme 5 min

I got this info but wait

GPU rx580 nitro plus rgb

Ik it's not enough gimme secs

Rx 580
Core i5 9400f
8 or 16 rams

Idk how many rams exactly

This guy didn't mention it

But it's 16 ddr4

[tranquil radish]

Not sure tho

[low shard]

that's an 8 year old card

it will be the bare minimum for AMD gpus basically to work

[tranquil radish]

So it's bad or what

[low shard]

it's the extremely bare minimum for amd gpus

it means, it's supported, but you won't get a stable experience especially while using it in games

[tranquil radish]

What about calls

And what should I upgrade to make it good

[low shard]

calls it will be decent, but you will basically have to upgrade in some years anyways, it's like you're buying an already old car and expect it to be good for another 8 years

[low shard]

atleast an nvidia rtx card

also

it really depends on your other use cases for the pc

[tranquil radish]

Mm

[tranquil radish]

What does that meab

Mean *

[low shard]

take in consideration that if you're looking for a good experience in the program especially in games, it's better to have n rtx card

but also depends on what other things you use the pc for, like games or other type of usage cases, and also your budget

if you don't have the money, just use kaggle wokada deiteris fork

it's free even though it's on cloud, i think 30 hours are enough weekly for you

[tranquil radish]

Well the RTX card well be kinda expsinve what version should I buy

[low shard]

as i said, it also depends on what other things you use your pc for

[tranquil radish]

Games only

Mine craft pubg roblox free fire

[tranquil radish]

https://www.amazon.eg/-/en/MSI-RTX-4060-Architecture-2X/dp/B0C8BPW1SP/ref=asc_df_B0C8BPW1SP?mcid=78f25f867281313ea5534396a340e200&tag=egoshpadsp-21&linkCode=df0&hvadid=731105053809&hvpos=&hvnetw=g&hvrand=3746784020435360548&hvpone=&hvptwo=&hvqmt=&hvdev=m&hvdvcmdl=&hvlocint=&hvlocphy=9112399&hvtargid=pla-2186271732846&psc=1&language=en_AE&gad_source=1

This one in Egypt for 340$ so it's expsinve

[low shard]

oh those are pretty low end games, since they are even on phones

[tranquil radish]

Oh

[low shard]

are you planning to run the wokada on those games?

[tranquil radish]

Ye

And discord

[low shard]

you should be fine with atleast a gtx 1660 card if you're really short on money and need a pc that will last more, or an rtx 2060 if you got a bit more money and want less delay, just remember that i can't really say much since i don't play those games so i can't guarantee you i have proof of how it runs like roblox on a gtx 1660 card along with the voice changer

just know that you will play games at the lowest graphics while using the voice changer, it's normal to not have more delay in the program

[tranquil radish]

Mhm

Even if the call is on discord but playing rblx will it lag

[low shard]

are you talking about high graphics?

[tranquil radish]

Noo

The voice changer

Like will it lag

[low shard]

should be fine if you close every other program in background and play on lowest graphics

[tranquil radish]

Mhm

[low shard]

@analog obsidian i think a gtx 1660 should be fine to run those games on lowest graphics + discord + voice changer, what do you think?

[tranquil radish]

The proplem the gtx 1660 is 400 dollars in Egypt

[safe slate]

Hi everyone - I was wondering how I would go about trying to make a video like this (https://youtu.be/ulYG7StLoww?si=S9hOTlxa782rsBIT) but based on my favourite movie?

[low shard]

as i said, if you don't have money, just use kaggle wokada deiteris fork

[tranquil radish]

This will work on games and discord?!

[low shard]

it's the same program just running on a remote good pc gpu provided by kaggle, ofcourse it will give you only 30 hours weekly for free and you HAVE to verify your phone number

[hot lagoon]

is it overtraining

https://tenor.com/view/erm-aksuali-veli-vel-nerd-nerd-emoji-gif-11908435316515627035

[tranquil radish]

There is none for free forever?

[low shard]

ofcourse not

gpus are expensive

you're lucky there even are services that provide free time

[tranquil radish]

I am talking about the voice changer

[tranquil radish]

Mm

[low shard]

the voice changer alone is open source and free, but kaggle, lets you use their remote good pcs, so ofcourse there's a limit

you are expecting to use pcs that cost thousands to be for free forever to your usage

[low shard]

do you really need more than 30 hours weekly

[tranquil radish]

I really don't know

How much does it cost?!

[low shard]

it's free, 30 hours weekly

[tranquil radish]

What if forever

[low shard]

there's no such service that give you a lifetime access to their gpu

there's other services like google colab which have paid plans, which are depending on their monthly subscription or pay as you go

you can't expect to run ai on cheap hardware or for free

[tranquil radish]

Mmm

[tranquil radish]

Well the pc I am getting is for 15000 egy means 300$ so it's expensive in Egypt

So it's kind of expensive but the prices in Egypt isn't the best

[low shard]

what pc are u getting

[tranquil radish]

It's a case collection

Not a official or smth

[low shard]

what matters are the specifications

[tranquil radish]

Well ik

[tranquil radish]

Well i told u the specifcations

[low shard]

oh that one with rx 580? well i wouldn't really know if it will be good for gaming + voice changer + discord

i can't assure you everything will go fine with low delay

[tranquil radish]

Mmm

[hot lagoon]

ot?

[tranquil radish]

Is low delay means the voice would progress fast or it will take time

[low shard]

delay = the time it takes to run the program

more delay = will take more time to respond, will be slower

less delay = faster time to respond

[tranquil radish]

Mmm good

[low shard]

no what i'm saying is your pc surely won't be fast at all for it, so even if it works i would say it will be slower to reply

just use kaggle dude

[tranquil radish]

That's kaggle dude?

[low shard]

that's not kaggle

[tranquil radish]

your only options are:

• pay for an atleast decent pc
• use the kaggle wokada deiteris fork https://www.kaggle.com/code/suneku/voice-changer-public with 30 hours for free weekly, but you need a phone verification

also what games? i fear your pc can't run any online games

[tranquil radish]

This thing

[low shard]

kaggle runs on cloud

it's on a remote good pc

it doesn't depend on how good is your pc

[tranquil radish]

Dude this is so complex 🙂

[low shard]

local = runs on your pc
your pc right now is bad for local because it's weak

[tranquil radish]

Is kaggle the program or what

[low shard]

cloud = runs on a remote good pc
kaggle will use their own good pc to RUN the code

[tranquil radish]

Ohh

[low shard]

Wokada deiteris fork is the program

[tranquil radish]

Now I got it

[low shard]

Kaggle is just a site that allows you to use their pcs

[tranquil radish]

Ohhh

Ye now I got it

Wokada works on games and discord right?

[low shard]

yes.. it's the realtime voice changer that uses RVC for realtime inference (usage of models)

[tranquil radish]

Ohh ye

And kaggle just makes it run faster

Right?

[low shard]

kaggle is just a cloud computing service, all it does is give you the permission to run code on their own good pc

[tranquil radish]

Ye

If i runned wokada alone it will lag so hard right

Without using kaggle

[low shard]

it wouldn't even run

[tranquil radish]

Dam🙂

[low shard]

your pc right now is below the minimum standards

[tranquil radish]

U mean the old one I told u about or the new one I was gonna buy

Like

THis one

[low shard]

the old one, the new one will work but not really much good and wouldn't suggest it for long time usage since it's already old

[tranquil radish]

Rx 580
Core i5 9400f
8 or 16 rams ( didn't mention)

[low shard]

just use kaggle

no need to buy a new pc if you don't have enough money

[tranquil radish]

I do have but

[tranquil radish]

Well it's 30 hours only that's the issue

Cuz idk for how much time I will use jt

[low shard]

https://docs.aihub.gg/rvc-voice-changer/cloud/w-okada-kaggle/ here's the kaggle guide if you want

[tranquil radish]

Is there a paid version for kaggke

[low shard]

30 hours are already alot for free

[tranquil radish]

Ik

[low shard]

for kaggle no, for colab yes, another cloud service, but there's no such thing as lifetime plan, you have to pay monthly subscriptions

[tranquil radish]

Oh sad

[low shard]

why would they make a lifetime plan

no matter the price, it costs them alot to power those gpus

so they would lose money

[tranquil radish]

Mmm

[low shard]

seriously, try to make 30 hours enough rather than paying money for an already old pc

[tranquil radish]

Well see i would already get this pc for games and more my question is just will there be diffrence in the voice changer when I buy this one
Rx 580
Core i5 9400f
8 or 16 rams

[low shard]

the difference is, it won't be fast running games + roblox + discord

[tranquil radish]

Mhmm

[low shard]

pcs are expensive

just know that pc won't last long

[tranquil radish]

Mine?

[low shard]

the one you're going to buy

it already uses a 8 year old gpu

[hot lagoon]

Do you recommend any graphic card that I should get for a? 2000$ build

[tranquil radish]

Mm

[mortal peak]

So bro she has an i5 4th gen with 4 rams ddr3 and u want her to get a "rtx 2060"

Like bro

Wth?

[low shard]

the user would want to buy a new whole pc, not just replacing a single component

so i told what gpu would be the good for the games they played + discord + voice changer

also, i suggested a gtx 1660 too

[mortal peak]

It's going to be good

[low shard]

not really much for games + voice changer + discord

[mortal peak]

@tranquil radish That's gamerfighter's gpu

[mortal peak]

Ye ye

[tranquil radish]

Oh

[low shard]

wtf? spy?

[tranquil radish]

Nah that's our friend

[mortal peak]

I meant gpu

[low shard]

you know right AI takes ALOT of computing? especially realtime

[mortal peak]

Ye ye

Ima programmer

Like I'm not that good

Btw

Do u recommend the 9400f for the cpu or ehat

Alr

[tranquil radish]

Ty

[low shard]

not really needed to do an argument here for this, everything is explained in the guide https://docs.aihub.gg/rvc-voice-changer/local/deiteris-w-okada-fork/ along with suggest settings for each gu, and the guide tells it clear that you can't expect much from old gpus especially for games

@tranquil radish speak only english here

[tranquil radish]

Oh sry I didn't say smth bad anyways

[mossy gull]

Hello guys. I need a TTS system that can run through the command line (not a GUI-based executable). It should support RVC models so i can use those models. Are there any good options available? I would really appreciate any help 🙏

[low shard]

none

RVC is meant for Speech to Speech

not Text To Speech

[mossy gull]

I thought maybe there are already a work around of it.

[low shard]

The only way is to use Applio (with ui) that basically uses microsoft edge tts api to make text then uses it as an input in RVC, but it won't be emotional at all

[mossy gull]

So is there are TTS that you can recommend?

[low shard]

and it depends on the internet, unless you use another TTS of your own as a "base"

but no, there's no single way to have a native way to 'transform' sts to tts

[low shard]

this is the only way you could use TTS with RVC (ofcourse, you can use any tts-generated text as input)

but yeah it won't be as good as a tts ai like elevenlabs

[mossy gull]

Thanks for the help , I am currently gonna try "TTS" library , Let's see if that works then maybe Ill use "TTS_RVC" to do the conversion

THis also has tts but its bad , lol
Let see what happens

[slender osprey]

In order for me to have a higher speed in the MMVC do I need a better internet?

[slender osprey]

Or is there another way?

[low shard]

iirc yeah there's cli python packages, but all they do is use microsoft edge tts as an input

unfortunately you can't directly 'convert' the architecture

the quality and emotion will all depend on how good is the TTS base if you're going to use it as an input in rvc

[low shard]

elaborate:

• what's ur pc gpu
• what tutorial link did u follow
• what is the issue

[mossy gull]

Ahhh... There are better TTS ik , better then eleven labs the problem is my hardware . I will need 14GB v-ram at minimum 😦 . And since they are new there are no quantized model also...

[low shard]

@mossy gull https://docs.aihub.gg/tts/tts-tools/ this could help you too, it's a list of pretty good tts tools

[slender osprey]

My gpu is a Radeon RX 580, I followed the tutorial in the manual of the latest version, the problem is that I would only like to increase the speed at which the voice is generated, with as little delay as possible 🫡

[low shard]

AI requires good hardware yeah, unfortunately the only Open Souce TTS that is better than 11labs at the moment is Nari Dia TTS 1.6b, which is way better at english dialogues, not sure what language you need

[low shard]

close all programs in background, show a screenshot of your wokada and also tell what programs are you using wokada with

i remind you that your pc gpu is pretty old btw

[mossy gull]

Need eng , but V ram is the prob....
btw my main target is like a robotic voice , So i guess mid level tts will work

[low shard]

a robotic voice? then why don't you just use Applio since as it uses ms edge as a base it won't have any emotion

or do you mean low quality too?

[mossy gull]

I need smt that will work on CMD , cause I am making a script take the tts output and match the facial expression of a character in one go , just by giving the text prompt for the tts and a image of the character

[low shard]

put game graphics to the MINIMUM, cap fps to 60
check the triangle
use asio for a more complex but less delay guide: https://rentry.co/lessdelayasio
operagx eats more resources for their fancy effects, might want to try chrome or firefox

or, buy a better pc gpu since you can't expect much from a 8 year old card

it's already running at the lowest possible settings in the voice changer quality 😭

extra 0.1 is the worst quality control you can put

[low shard]

ohh, might want to try smt like https://github.com/Atm4x/tts-with-rvc

[mossy gull]

I tried!

[low shard]

it's basically what Applio does but in a python package

[low shard]

it doesn't work?

[slender osprey]

F

[mossy gull]

Is there a restriction placed? It seems I cant send files here

I want to show the output

[low shard]

did you also tried the other things i said above?

[low shard]

!give-media-perms 1h @mossy gull

well you'd generally need to be level 10 to send images

[viscid nimbus]

Are there any free AI create song sites where I can create my own model?

[slender osprey]

I'll see now

[low shard]

could you elaborate more?

[low shard]

what's your pc gpu? it's better to check local first

[mossy gull]

import sys
from pathlib import Path
from tts_with_rvc import TTS_RVC

script_dir = Path(os.path.dirname(os.path.abspath(sys.argv[0])))

model_path = script_dir / "KittyKI.pth"
index_path = script_dir / "KittyKI.index"
tmp_directory = script_dir / "temp"
output_directory = script_dir

tmp_directory.mkdir(exist_ok=True, parents=True)

tts = TTS_RVC(
    model_path=str(model_path),
    index_path=str(index_path),
    f0_method="rmvpe",
    device="cpu",
    tmp_directory=str(tmp_directory),
    output_directory=str(output_directory)
)

generated = tts(
    text="Hello, world! This is a test.",
    pitch=0,
    index_rate=0.75,
    resample_sr=32000,
    filter_radius=1,
    rms_mix_rate=0.5,
    protect=1.33,
    is_half=False,
    verbose=True
)

print(f"Audio generated successfully and saved to: {generated}")```

This is the script I used

I have tried difffrent f0 methods but its still the same 😭

[low shard]

have you tried with other models too?

[mossy gull]

yeah

I tried with one of thier recommended models too

same thing

[low shard]

rmvpe is the best one

[low shard]

also, what's your pc gpu?

[mossy gull]

Ohh , I just treid just in case

[mossy gull]

1660 super , but as you can see i run that on cpu

is that the prob?

[low shard]

I mean, if you put cuda, it will run faster

just saying for better performance

[low shard]

have you also tried playing with the pitch?

[mossy gull]

I know that much , I tried that way cause this script will be used in a lot of peoples machines thats why (since a lot of people does not have a GPU)

[mossy gull]

Yes , as you can hear that the pitch is not the prob, The prob is the way it talks . I mean why tf that accent coming from 😭

[low shard]

oh wait, you didn't set a ms edge tts vocie

it's better to set a tts voice that has the same sex and language as the RVC model one

[simple ore]

tts_rvc has it, I bet

[low shard]

yeah you can choose the voice

it's in the repo demo script so

[mossy gull]

will check rn

[simple ore]

I use a different script with Applio

[mossy gull]

This isnt half bad!
Using "TTS" library

[low shard]

yeah deffo better than before

@mossy gull remember:

• in most cases rmvpe is the best, crepe might be a bit more soft but it's not really robust at all for noise, while fcpe is kinda like rmvpe but less accurate and more soft
• always use a tts edge voice as much similar in language and sex to the rvc one
• always play with the pitch

[mossy gull]

Thanks bro I will let u know my progress 🙂

[graceful patio]

yo bro

what setting should i use in uvr 5

for de echo

could i ask u that

do yk lmao

the defualt is 0.3

on mvsep

and i could use uvr 5

but im lazy

i got the site open might as well

[fiery rain]

3080 10GB good for okada fork?

[glacial pollen]

more than enough

[knotty moth]

the default settings are fine

but you can try different values though it won't be much different

[silver ravine]

i used the find model thingy right here #🔍│find-models message but when i clicked on the model it’s just only allowing me to like make the model sing and i can’t actually download it or anything is there a way to download these

[silver ravine]

nevermind i figured it out, the download button was just very tiny and i didn’t see it

[austere siren]

i need help in terms of editing the batch size, data set and etc. i dont know where to find them

[green ermine]

Guys I used to use control net fast with xl but right now gen speed is so low how can I fix that ?

[cinder crown]

Hi, I'm having a problem with my voice changer. It sometimes lags. I downloaded the latest version as recommended.

The problem occurs when the ping and MS speeds spike.

Screenshot lag : https://prnt.sc/vJ48pIxmygA2

Screenchost config: https://prnt.sc/rfWk0eUQ1VXD

AMD RYZEN 6800 16gb my GPU.

[strange mountain]

I see 2 versions of voice changer, which one should I download? One is vcc and the other is mcvv.

[latent kettle]

Have you downloaded comfyui

[latent kettle]

The fork is better

[strange mountain]

wdym 1 or 2 mb i cant understand*

[latent kettle]

The one with mcvv

-realtime

[patent trellis]

💻 Local Realtime RVC

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

• Wokada Deiteris Fork

Most suggested. GUIDE

• Original Wokada

ONLY the latest alpha comes close to the Deiteris Fork performance, older versions in youtube tuts are way worse. GUIDE

• **RVC GUI Mainline Realtime**

Unavailable, the guide is outdated and the program is worse compared to the ones above, and much less updated

[latent kettle]

@strange mountain wokada deiteris fork

[strange mountain]

so MCV right?

[latent kettle]

Which 2 versions are you looking?

[knotty moth]

install and use comfyui manager to install required custom nodes in the workflow

[strange mountain]

yesir im bad english 😭

this mmvc?

ok ty

[low shard]

yea

[low shard]

have u checked https://docs.aihub.gg

[low shard]

that's latest wokada deiteris fork
uninstall vb audio cable, get vac lite from the same guide where you got the program
uncheck sup1
set extra to 2.7
set chunk to 192
check the triangle

[low shard]

the "vcc" one is commonly called original wokada, the "mcvv" oone is commonly called wokada deiteris fork, get the wokada deiteris fork

what's ur pc gpu btw?

[cinder crown]

Hey bro, I put this in PowerShell: "ForEach($PROCESS in GET-PROCESS audiodg) { $PROCESS.ProcessorAffinity=4; $PROCESS.PriorityClass='High' }"

and it didn't work. It was working fine. Do you know how to fix it?

[low shard]

what's ur pc gpu? what tut link are u using?

[cinder crown]

RX 6800 16 GB Vram

[low shard]

what tutorial link are you using?

what's also the issue exactly?

[cinder crown]

https://docs.aihub.gg/rvc-voice-changer/local/deiteris-w-okada-fork/#download-amd-intel-and-cpu-on-windows

[cinder crown]

#1369899088874704987

[cinder crown]

Finally, I left some audio for you to listen to by entering the following command:

PowerShell "ForEach($PROCESS in GET-PROCESS audiodg) { $PROCESS.ProcessorAffinity=4; $PROCESS.PriorityClass='High' }"

The modulator started to act poorly, the performance increased, and lag was heard.

[low shard]

let's talk there

[brisk robin]

hi

[low shard]

hello, if you need help, elaborate

[brisk robin]

am looking for RVC GUI zip or anything similar to it

[low shard]

RVC GUI is extremely outdated

don't watch youtube tutorials for RVC / Wokada

a 1 year old tutorial is prehistorical in ai field

what's your pc gpu and what you want to do?

[brisk robin]

oh okay

[low shard]

I can suggest you better programs, tell your pc gpu and what you want to do

just never follow video tutorials for those programs since they get old very easily

@brisk robin You can check your pc gpu via:
ctrl+shift+esc (task manager) -> Performance tab -> GPU

[rose hound]

@rare cliff

now

where is link

😭

[low shard]

link for what?

[rose hound]

voice changer

[tulip cloak]

nick

please guide my man

[low shard]

what's ur pc gpu?

realtime for calls/games?

[rose hound]

amd

[rose hound]

discord or insta

[low shard]

which

[rose hound]

amd radeon hd 7000

[brisk robin]

is there like threads, handles, utilization, speed .............

[low shard]

that's an over decade old gpu

i'm more surprised that people still have that

[rose hound]

yea cus my new one is broken after exams ill get new one

[low shard]

your current pc would be extremely too weak

the only option is either using cloud (remote good pc) with limited free time or just wait till u get the new one

[rose hound]

so i cant use voice changer?

[low shard]

which one do u want to do?

AI takes alot of computing, it can't run on bad hardware

[brisk robin]

ohh

[low shard]

what? i asked you a question #✨│ai-help message

[rose hound]

just a girl voice changer

[low shard]

your current pc is too weak, i told you your only options #✨│ai-help message

do u want to use cloud with limited time and more harder to setup or just wait till u get a new pc

[rose hound]

ok ill wait ty

[low shard]

yw

[rose hound]

so there is not voice changer i can use now?

[low shard]

the only way you could right now is via cloud, which has limited time for free and is harder to setup

it would be better u just get a better pc

ai needs a good pc

it can't run on an over decade old pc

[rose hound]

okk

[brisk robin]

do you recommend applio

[low shard]

what's your pc gpu? what you want to do?

there isn't a single program for everything

and ai can't run on any hardware

[brisk robin]

so what's the best option that i can use

[strange mountain]

rtx 3050

[low shard]

what's your pc gpu and what you want to do?

if you don't tell me this

I can't help you

[low shard]

nice, if u want realtime voice changer for calls, then just get wokada deiteris fork

-realtime

[patent trellis]

💻 Local Realtime RVC

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

• Wokada Deiteris Fork

Most suggested. GUIDE

• Original Wokada

ONLY the latest alpha comes close to the Deiteris Fork performance, older versions in youtube tuts are way worse. GUIDE

• **RVC GUI Mainline Realtime**

Unavailable, the guide is outdated and the program is worse compared to the ones above, and much less updated

[low shard]

1st link

[strange mountain]

oh ok ty

[low shard]

yw

[strange mountain]

by the way when i speak sometimes it comes out in another language sometimes it is distorted 💀

[low shard]

show a screenshot of ur wokada

!give-media-perms 1h @strange mountain

[strange mountain]

i go to school now can you remove it and after i come back from school can i explain it to you later?

[brisk robin]

Intel(R) HD Graphics 4000

[low shard]

that's cooked, weak integrated graphics

do you have any other gpus?

[rose hound]

so is there is any video of via cloud so i can setup?

[low shard]

alr

[low shard]

no, all videos are outdated

there's only https://docs.aihub.gg/rvc-voice-changer/cloud/w-okada-kaggle/

but it's just better to get a better pc

[rose hound]

what is this?

[knotty moth]

he meant the Nvidia or AMD radeon gpu

[low shard]

a guide to read

[rose hound]

bro how much time it take to setup?

[low shard]

all you need to do is just read it

no, it's not a 5 seconds program like chatgpt

[rose hound]

i read but ...

there is no video?

😭

[low shard]

no, all videos are outdated

a 1 year old youtube tutorial is gonna just cause you more issues with old programs

just read the guide

[austere siren]

@low shard hey can i please vc with you by chance, i need help

real quick

[rose hound]

ok i need to follow all steps?

[low shard]

yes

[low shard]

i can't vc

explain your issue via text here

[austere siren]

no worries, just want to know where i can edit the batch size and data base

to fit the voice model

[low shard]

are u making ur model or trying to modify the one from soneone's else

[austere siren]

its someone elses

can i send screenshots in here?

[low shard]

you can't simply

and that won't fix the model

[austere siren]

the creator put specific batch size numbers and etc under the voice model link

do i have to apply that

?

or is it just something else

[low shard]

that's just an information of how it was trained

it's not a setting for the voice changer or smt

[austere siren]

ohhh omg im so slow

i thought i was meant to copy the settings

lmao

thank you boss

[low shard]

lol yw

[austere siren]

one last question, does okada run well with AMD?

[low shard]

wokada deiteris fork does, ofcourse it's not as good as nvidia but it's more improved

[low shard]

be sure to NOT use youtube tutorials, those use an old version of original wokada

[knotty moth]

wdym editing batch? it's just some training configuration

like a cooking recipe after you've cooked it

[austere siren]

could you please send the link

[austere siren]

yeah i just figured that out lol thanks though

[austere siren]

im getting failed to fetch error in the client, how do i fix this

[vestal helm]

if i wanna train on 40k should i remove the audio/data that don't 20 khz?

[lilac wasp]

which settings i should use here for 5060Ti ?

[vestal helm]

https://rentry.co/forkvoicechangerguide#download-for-nvidia-gpu-on-linux

[lilac wasp]

thank you

[vestal helm]

im guessing trhis

[low shard]

What's your PC GPU and what you want to do exactly

[low shard]

That's original wokada, don't use it

-realtime

[patent trellis]

💻 Local Realtime RVC

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

• Wokada Deiteris Fork

Most suggested. GUIDE

• Original Wokada

ONLY the latest alpha comes close to the Deiteris Fork performance, older versions in youtube tuts are way worse. GUIDE

• **RVC GUI Mainline Realtime**

Unavailable, the guide is outdated and the program is worse compared to the ones above, and much less updated

[low shard]

1st link

Wokada deiteris fork

[low shard]

The link got changed to the docs one

[hallow thistle]

Are you sure that person uses Linux? Because your fork W-Okada guide Rentry link has #download-for-nvidia-gpu-on-linux instead of #download-for-nvidia-rtx-5000-series-gpu-on-windows at the end.

[lilac wasp]

im getting this error can someone help?

[hallow thistle]

This could indicate that you didn't select a voice model you uploaded on W-Okada.

Also, install W-Okada into somewhere like C:\MMVCServerSIO or D:\MMVCServerSIO. It doesn't have to be on your desktop folder all the time, as it can cause some issues.

[lilac wasp]

thank you

can you look at that too

wat is that mean

[hallow thistle]

This is a warning from one of packages in W-Okada, telling you one of its code is deprecated and will be replaced by another new one. Although you can simply ignore it, as it less likely has to do with the program.

[brittle wing]

anyway to run the voice changer on a intel cpu? i have the winnonxdirectml 1.5.3.15 version

[strange mountain]

uhm in original woka where 192 ms?

[frosty heart]

What's wrong?

[hallow thistle]

[simple ore]

dont run as admin

[hallow thistle]

You use RVC AI Cover Maker? Never heard of this one, but I use Applio.

[hallow thistle]

!howtoask

[patent trellis]

How To Troubleshoot

__**GIVE CONTEXT.**__ 📝

• Don't simply mention your issue, like "my rvc is not working".
• Describe the step you are on, what you're trying to do, the RVC you're using, a screenshot, etc.
• The more context, the better.

__**BE POLITE.**__ <:matsuripray:1159685390156967936>

• Don't be desperate. You can ping a Helper, but if they ignore, they aren't available/don't know the answer.
• It's okay if you're frustrated, but don't take it into this server.
• Don't DM without prior consent.

__**BE PRODUCTIVE.**__ 🤝

• Don't ask for every little instruction. Put your own effort & test things by yourself.
• Don't ask to ask.
• Check if your answer is a Google search away/on our guides website.

If you've changed your mind about your message, feel free to delete/edit it.

[hallow thistle]

You wanna run W-Okada with only CPU? Does your PC have any GPU?

[brittle wing]

integrated cpu

gpu*

[hallow thistle]

Running W-Okada with only CPU or an integrated GPU is not recommended.

[brittle wing]

alright, but still i wanna try

would it work on a 8th gen i5 by chance?

[hallow thistle]

If you really wanna run W-Okada with only CPU anyway, there's this guide. https://docs.aihub.gg/rvc-voice-changer/local/deiteris-w-okada-fork/

[brittle wing]

alright thanks

[hallow thistle]

For an actual situation like you wanna use W-Okada with a game, use cloud service instead. There's Kaggle, but you'll have to register with your phone number on this one. https://docs.aihub.gg/rvc-voice-changer/cloud/w-okada-kaggle/

[brittle wing]

alright ty

wait but its saying that kaggle will use my cpu's gpu aswell is it like a system where it uses both the cloud processing power and my cpu

[hallow thistle]

You confused? Kaggle doesn't use your PC CPU for it to process. It runs on its own server.

[brittle wing]

ye but on the tutorial its saying to select a gpu

WAIT OH

OHH

nvm

[hallow thistle]

@sacred crescent To find which proper program where you can use RVC voice models, go here. No need to hop into my direct message just for this.

!howtoask

[patent trellis]

How To Troubleshoot

__**GIVE CONTEXT.**__ 📝

• Don't simply mention your issue, like "my rvc is not working".
• Describe the step you are on, what you're trying to do, the RVC you're using, a screenshot, etc.
• The more context, the better.

__**BE POLITE.**__ <:matsuripray:1159685390156967936>

• Don't be desperate. You can ping a Helper, but if they ignore, they aren't available/don't know the answer.
• It's okay if you're frustrated, but don't take it into this server.
• Don't DM without prior consent.

__**BE PRODUCTIVE.**__ 🤝

• Don't ask for every little instruction. Put your own effort & test things by yourself.
• Don't ask to ask.
• Check if your answer is a Google search away/on our guides website.

If you've changed your mind about your message, feel free to delete/edit it.

[sacred crescent]

I am just looking for a good RVC program to use

[hallow thistle]

Which one are you looking for? The AI cover or realtime voice changer? There are two specific programs.

[sacred crescent]

I got a kanye AI model and just want to use it for lyrics

[hallow thistle]

You can't simply ask for a download link like this. What is your PC GPU?

[sacred crescent]

3070

[hallow thistle]

https://docs.aihub.gg/rvc/local/applio/

[sacred crescent]

I extracted and it just came up with a CMD.exe saying AI Models??

[hallow thistle]

Any idea to elaborate?

[sacred crescent]

Like what do I do now?

It said press any key to continue and now nothing is on screen

[hallow thistle]

Are you sure you have downloaded the compiled one? Because the compiled one isn't supposed to crash upon running.

If you wonder what I mean about the compiled one, I mean this link. https://huggingface.co/IAHispano/Applio/tree/main/Compiled

[sacred crescent]

Yeah got the 2.9 one

[hallow thistle]

More like V3.2.9.

Applio is supposed to look like this if successfully launching.

[sacred crescent]

Yeah didn't work like that for me let me try and run it again

It says move the applio folder to spaces without path

so straight into C drive aye?

just moved it and still says the same thing

[hallow thistle]

If you have extracted Applio into your desktop, do not do that. Extract it to somewhere like C:\Applio or D:\Applio.

[sacred crescent]

Okay let me try that

[timber bramble]

Guys how do I get RVC working

[hallow thistle]

!howtoask

[patent trellis]

How To Troubleshoot

__**GIVE CONTEXT.**__ 📝

• Don't simply mention your issue, like "my rvc is not working".
• Describe the step you are on, what you're trying to do, the RVC you're using, a screenshot, etc.
• The more context, the better.

__**BE POLITE.**__ <:matsuripray:1159685390156967936>

• Don't be desperate. You can ping a Helper, but if they ignore, they aren't available/don't know the answer.
• It's okay if you're frustrated, but don't take it into this server.
• Don't DM without prior consent.

__**BE PRODUCTIVE.**__ 🤝

• Don't ask for every little instruction. Put your own effort & test things by yourself.
• Don't ask to ask.
• Check if your answer is a Google search away/on our guides website.

If you've changed your mind about your message, feel free to delete/edit it.

[timber bramble]

So other people can hear it

[low shard]

elaborate more

[timber bramble]

I did

[timber bramble]

.

[hallow thistle]

how can I get my RVC working
so everyone can hear me
I think you mean W-Okada the realtime voice changer.

[timber bramble]

Currently

No one else can hear the voice changer

[low shard]

@timber bramble

RVC = Retrieval-based-Voice-Conversion, the best Speech To Speech AI Models (on v2), Inferences (use models) pre-recorded audio (ai covers) and train (make) models. Technically, Mainline RVC does have a go-realtime.bat (aka RVC-GUI), but it's pretty messy and outdated so it's extremely not suggested for realtime.

Wokada = uses RVC for realtime inference. There's 2 main versions, Original made by Wok, and the most suggested one is Deiteris Fork (modified version)

• what's your pc gpu
• what do you want to do
• what tut link did u use

there's thousands of ai programs

[hallow thistle]

RVC and realtime voice changer (W-Okada) are two different programs.

[low shard]

i can't predict which are u using

[timber bramble]

Im using

[low shard]

i just hope you aren't using youtube tutorials, because those are old asf for wokada and rvc

[timber bramble]

realtime voice changer (W-Okada)

Sorry

[low shard]

what's ur pc gpu? what tutorial link did u use?

if u don't remember the tutorial link, u can share a screenshot of the program

[timber bramble]

I have a 3080Ti

[low shard]

!give-media-perms 1h @timber bramble

[shrewd stag]

Hey gang I'm new to this stuff how do I make an ai cover

Channels are confusing

[hallow thistle]

You can type each information in one message at once.

[low shard]

what's ur pc gpu?

[timber bramble]

This is what im using

[low shard]

that's old original wokada

[timber bramble]

Oh

[low shard]

lemme guess

you used youtube?

-realtime

[patent trellis]

💻 Local Realtime RVC

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

• Wokada Deiteris Fork

Most suggested. GUIDE

• Original Wokada

ONLY the latest alpha comes close to the Deiteris Fork performance, older versions in youtube tuts are way worse. GUIDE

• **RVC GUI Mainline Realtime**

Unavailable, the guide is outdated and the program is worse compared to the ones above, and much less updated

[shrewd stag]

Lemme check hold up

[low shard]

use the 1st link

wokada deiteris fork

[timber bramble]

Yeah

Ok

[low shard]

this is wayyyyy better

in temrs of performance

and less bugs

[hallow thistle]

As I expected. You're using the original W-Okada, which is outdated.