✨│ai-help

[pastel oak]

or u go to MMVCServerSIO -> model_dir folder and delete folder number 11 i think it is, just open all folders until you see Venom-TITAN.pth

[unkempt ether]

How can I make an ai voice model trained on a YouTube channel that posted ASMR videos?

I want to make an ai voice model trained on an inactive ASMR youtuber so I can make new ASMR videos and song covers with their voice. What programs and steps would I need to take to go about doing this? Would I have to download all of their videos and put them through a program that isolates their vocals like Lalal.ai? What program would help me do that and once I have the vocals how would I use those to make an ai model? Any advice or links would be appreciated.

[rotund prawn]

for making voice model
you can use rvc disconnected

[unkempt ether]

@brittle wing Thank you! I'll check that out.

@rotund prawn Thank you! I looked at the page and it says I need to load a data set. How do I take the Youtube channel's videos and turn it into a data set?

[rare goblet]

Ayo? @unkempt ether level 1 !!!

[rotund prawn]

np

download it as mp3

make sure there is no other noise or anomaly in background

also remove the empty parts of the vocal

[rotund prawn]

that pause when starting a new sentence

[pastel oak]

adding to what Selimhan said, if theres music there are tools to remove that too. Go to mvsep.com theres plenty of options, or if you have a good gpu you can also download UVR and try it out locally, however if its about background music, then mvsep has the best model currently to remove those flawlessly (its bs-roformer ver 2024.4)

[grizzled forum]

i'm training on applio. should i use the contentvec embedder or japanese-hubert-base if my dataset is in japanese?

[brave swallow]

I don't know I always use hubert

[junior halo]

I know how to remove the reverb out of an audio sample, but is it possible to remove echo? I'm looking into making an actually good sounding voice model of Gabriel from Ultrakill. I'm pretty sure it's the echo in his voice that is screwing up the output quality. The echo sounds like it is coming from a different stereo channel, but I'm not sure how to isolate those.

[dull gazelle]

guys a quick question embender model which one i should use huvert or contentvec?

[junior halo]

If that's not possible, I might just resort to pulling from Gabriel Gaming's streams.

[vale osprey]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• EasyGUI, by rejects Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

[dull gazelle]

guys who can help me for a reason when i use creepe it takes hours and hours to load but rmvpe loads fast, i had to cancel creepe.
i9, 64Ram, 4070

[viral mason]

@rigid mango I sent a dm a bit ago and was just wondering about something for rvc disconncect, again sorry for bothering you but it's been a couple hours

[thin stump]

What was the question? I finally wrote a guide on the mainline colab btw. It did a number to my head trying to process everything https://rentry.co/RVC-Mainline-Colab

[viral mason]

So I was having an issue with the Rvc Disconnected collab not allowing me to train datasets that were really long like 2-3 hours of data ish

it wouldn't process the audio and finished training almost instantly without creating any files

if you want the datasets I have links on hugging face due to the size

[thin stump]

mangio fork problems probably

[viral mason]

how so?

[thin stump]

just wait until he responds

[viral mason]

alright 👍

[cold cave]

How can I make my datasets on my phone?

[violet heron]

MVSEP

[cold cave]

Ik but how can I find voice clips

Like Labeling the Audio

On mobile

[frank latch]

how do i ai covers

1 sec and already got lost

[violet heron]

Files app

[violet heron]

-spaces

[azure marsh]

Suggestions for @frank latch

<:huggingface:1179800228946268270> Hugginface Spaces

• 🆕 Ilaria RVC Zero, by thestingerx Huggingface Spaces
• 🆕 RVC⚡ZERO, by r3gm Huggingface Spaces
• Applio, by IA Hispano Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

More commands: -audio, -colabs, -kaggle, -local, -overtraining, -realtime, -rvc, -help

[violet heron]

Try Ilaria RVC zero

[tough fiber]

tried and doesnt work

[torpid void]

can someone help me

i wanna do voice trolling in valorant but i dont know how to get up the jett voice

[split jasper]

yo

how do i do this

[grand lichen]

how do you set the ai voice up?

[split jasper]

i was wondering if i could get help so i can get the correct version for my computer i am new

[lethal finch]

how do i find the right version for my pc?

[violet heron]

-realtime

[azure marsh]

This interaction has expired, use the command /guides realtime if you wish to see it again.

[brave garnet]

Voicechanger Download (Okada)

⠀
Download for Nvidia GPUs
Version 18a cuda

Download for AMD GPUs
Version 18a directml

Download for Intel GPUs
Version 18a directml

Download for Mac
Version 17b Mac
⠀

Made with love by Ilaria 💖

[violet heron]

@split jasper @lethal finch

[grand lichen]

could I go into a vc for a troubleshoot?

[split jasper]

yeah same

Vc?

[violet heron]

can’t rn

[split jasper]

alr

just

do that?

[violet heron]

There should also be guides and faq for bugs and issue

[split jasper]

Alr and

[lethal finch]

ty

[split jasper]

should I do the one

Under

Or the one you pinged me to?

[violet heron]

Any

[split jasper]

Alr thx

[rare goblet]

Ayo? @split jasper level 1 !!!

[split jasper]

yo

I downloaded it

But it wont open

[violet heron]

Did you unzip everything?

[split jasper]

uhhh

[violet heron]

And which file are you opening

[split jasper]

I am opening the one I downloaded?

[violet heron]

Yeah but which file

[split jasper]

i dont get what you mean

[violet heron]

The file is a zip file, it comes with other files inside of it

[split jasper]

yes

it is

[violet heron]

In order to use the files you need to extract or uncompress

Then you need to run a different file to start the voice changer

Which file are you opening

[split jasper]

alr

so do I click

On the zip

And open the file with everything else

Then press extract all?

[split jasper]

correct

[violet heron]

Click on the zip, click extract all

[split jasper]

Alr

[violet heron]

Select the directory, make sure there is no spaces

Then go to that directory

And open start_http

Rest is in the guide

[azure marsh]

Documentation

📚 All-In-One English documentation

https://docs.aihub.wtf/

Full AI Voice Model Training Guide (Local)

Link: YouTube
Credits: Christopher Villanueva

Model training with Mainline RVC

Link: Rentry
credits: Raven (ravencutie21)

AICoverGen Colab Guide

Link: Google Docs
Credits: Eddy (Spanish Helper)

Create a model with RVC disconnected (colab)

Link: Google Docs
Credits: Angetyde

[grand lichen]

yeah Im lost with this guide, I dont see the RVC folder nor the go-realtime-gui.bat file

[rare goblet]

Ayo? @grand lichen level 1 !!!

[grand lichen]

all I see is this

[violet heron]

That is w-Okada not RVC realtime

-realtime

[azure marsh]

This interaction has expired, use the command /guides realtime if you wish to see it again.

[violet heron]

Read the first guide here

RVC models are completely compatible with w-Okada

[grand lichen]

yeah im still lost

tomorrow could I get some help in vc?

cause im just completely lost with what to even look for

[pastel oak]

Look theres 2 voice changers and while you thought you downlaoded 1 you downloaded 2

https://rentry.co/VoiceChangerGuide

you downloaded this so feel free to follow that

else if you want the RVC one the one you mentioned, follow this

https://rentry.co/rvcrealtimeguide

[olive cliff]

What's the recommended amount of save frequency epochs to use when training on the RVC Mainline Colab notebook? I previously trained on the RVC Disconnected Colab notebook and I wasn't sure what amount to use when training.

[thin stump]

Litsa said he uses 1. Theres a whole discussion about syncing graphs that I don't want to unbag right now. Having to learn what goes on in mainline was hard enough already imo

[olive cliff]

Alright then. I'll go ahead and use "1" I would've used "5" or "10" but I'll go with what Litsa uses.

Also, should I cache all training sets to GPU because I have a dataset that's under 10 minutes?

[thin stump]

you should ask that in model-maker chat. Maybe he'll answer that

[thin stump]

and yeah again, he knows about this "syncing". It's a huge headache to unpack topic that even though thats already explained by codename

so he uses 1

[olive cliff]

I already did that so I'll wait for a response there. Thanks.

[hardy oar]

Hey

Why when im trying to download a model discord says its a dangerous file and i shouldnt download it

[low shard]

its a default popup to every single download link on discord dw

[hardy oar]

hm

alright ty

Also, whenever i try it the ai keep ignoring my voice and say things like "hhh..ahhhh....hhhhh..."

[digital marsh]

Is llaria, mainline, applio, or mangio rvc better for local?

[low shard]

uhh are you using the real time voice changer? i don't really know but could depend on the model, or your gpu

[low shard]

mainline, the mangio one was also only online (like the og ilaria colab)

[hardy oar]

Yeah got a pretty bad pc maybe

[low shard]

what gpu?

[hardy oar]

idk

let me check

[next geyser]

is there a new version for amd?

[next geyser]

is this ver good?

cclient_win_std_torch_dml_2.0.21-alpha.zip

[low shard]

uhhh i cant use local things, maybe its better u ask in #1192011222023950368 sorry

[next geyser]

wait

is rvc better for amd?

[brave garnet]

RVC Colabs and Spaces

⠀

HuggingFace Spaces 🤗

⠀
Ilaria RVC
EasyGUI port with some improvements, by Ilaria.

RVC-HFv2
Applio port, by r3gm.

AICoverGen
AICoverGen port, by r3gm.

Advanced RVC Inference
Extended version of the GUI with advanced settings, r3gm.
⠀

Made with love by Ilaria 💖

Voicechanger Settings (Okada)

⠀

Settings for AMD GPUs

Don't forget that your models needs to be converted in ONNX!

F0 Det.: rmvpe_onnx (suggested for all series)

7xxx XT cards: 112-128 chunk | +16384 extra
6xxx XT cards: 128-192 chunk | +16384 extra
5xxx XT cards: 192-256 chunk | +8192 extra

RX 580: 192-256 chunk | +8192 extra
RX 570: 192-256 chunk | +8192 extra
RX 560: 256-384 chunk | +8192 extra

Advanced Settings

Protocol : Sio or Rest
Crossfade: 4096 start 0.2 end 0.8
Trancate: 300
Silencefront: Off
Protect: 0.5
RVC Quality: Low

⠀

Made with love by Ilaria 💖

RVC Colabs and Spaces

⠀

HuggingFace Spaces 🤗

⠀
Ilaria RVC
EasyGUI port with some improvements, by Ilaria.

RVC-HFv2
Applio port, by r3gm.

AICoverGen
AICoverGen port, by r3gm.

Advanced RVC Inference
Extended version of the GUI with advanced settings, r3gm.
⠀

Made with love by Ilaria 💖

RVC Colabs and Spaces

⠀

Local Forks 🖥️

⠀
Mainline RVC
Original project, suggested for advanced users,
by the RVC-Project team.

Applio
Simplified, suggested for all, by the Applio team.

RVC Studio
Simplified, suggested for all, by SayanoAI.

Mangio-RVC
Simplified, may not be supported anymore, by Mangio621.

AICoverGen
Simple yet great way to make covers, by SociallyIneptWeeb.

Replay
From the greators of weights.gg, excellent product for everyone.
⠀

Made with love by Ilaria 💖

[next geyser]

how can i fix stuttering

[brave garnet]

RVC Colabs and Spaces

⠀

Local Forks 🖥️

⠀
Mainline RVC
Original project, suggested for advanced users,
by the RVC-Project team.

Applio
Simplified, suggested for all, by the Applio team.

RVC Studio
Simplified, suggested for all, by SayanoAI.

Mangio-RVC
Simplified, may not be supported anymore, by Mangio621.

AICoverGen
Simple yet great way to make covers, by SociallyIneptWeeb.

Replay
From the greators of weights.gg, excellent product for everyone.
⠀

Made with love by Ilaria 💖

Voicechanger Settings (Okada)

⠀

Settings for AMD GPUs

Don't forget that your models needs to be converted in ONNX!

F0 Det.: rmvpe_onnx (suggested for all series)

7xxx XT cards: 112-128 chunk | +16384 extra
6xxx XT cards: 128-192 chunk | +16384 extra
5xxx XT cards: 192-256 chunk | +8192 extra

RX 580: 192-256 chunk | +8192 extra
RX 570: 192-256 chunk | +8192 extra
RX 560: 256-384 chunk | +8192 extra

Advanced Settings

Protocol : Sio or Rest
Crossfade: 4096 start 0.2 end 0.8
Trancate: 300
Silencefront: Off
Protect: 0.5
RVC Quality: Low

⠀

Made with love by Ilaria 💖

RVC Colabs and Spaces

⠀

Local Forks 🖥️

⠀
Mainline RVC
Original project, suggested for advanced users,
by the RVC-Project team.

Applio
Simplified, suggested for all, by the Applio team.

RVC Studio
Simplified, suggested for all, by SayanoAI.

Mangio-RVC
Simplified, may not be supported anymore, by Mangio621.

AICoverGen
Simple yet great way to make covers, by SociallyIneptWeeb.

Replay
From the greators of weights.gg, excellent product for everyone.
⠀

Made with love by Ilaria 💖

Voicechanger Download (Okada)

⠀
Download for Nvidia GPUs
Version 18a cuda

Download for AMD GPUs
Version 18a directml

Download for Intel GPUs
Version 18a directml

Download for Mac
Version 17b Mac
⠀

Made with love by Ilaria 💖

[muted dirge]

Im trying to use a model i found online and continue training it more, where would i be able to get its Gpth and Dpth

All i have currently is the index, pth

[wooden ermine]

guys, do you have link for titan pretrain download?

[jolly thorn]

hello bro i am using pth and index , and tryiing to convert the voice , ut it remains same

[proper shale]

You'd have to ask the original creator for that

[proper shale]

Hm?

What are you using?

[jolly thorn]

https://github.com/JarodMica/rvc-tts-pipeline

[rare goblet]

Ayo? @jolly thorn level 1 !!!

[brave garnet]

RVC Colabs and Spaces

⠀

Google Colabs

⠀
AICoverGen-WebUI
Useful for making quick covers, by Hina.

AICoverGen-NoWebUI
Useful for making covers, doesn't inclued a UI, by Ardha, by Eddy, Hina and Gdr.

RVC Disconnected
To train new voice models, by Kit Lemonfoot.

EasyGUI
The OG interface, by Rejects.
⠀

Made with love by Ilaria 💖

[proper shale]

Seems like it's kinda outdated (last updated 9 or 10 months ago?)

[brittle wing]

does colab still work

[jolly thorn]

do u have/know any good docs, about rvc convert type stuff

[jolly thorn]

apart from that readme is kinda confusing in that repo

[proper shale]

Yeah, for AI covers, just use Applio, or RVC Disconnected if you wanna train

Guides for both: https://docs.aihub.wtf/rvc/cloud/applio-colab/

https://docs.aihub.wtf/rvc/cloud/rvc-disconnected/

[proper shale]

https://docs.aihub.wtf/ has everything basically

[brittle wing]

what are the differences between these two?

[jolly thorn]

is appilo very large? cause ill have to host it somewhere ultimately

thanks for ur help btw

[proper shale]

Applio can do both AI covers and training, and RVC Disconnected has no UI but has a pretty stable training part

https://huggingface.co/spaces/TheStinger/Ilaria_RVC I'd also use this for converting as well cause it's quick ig

[proper shale]

Not sure how large it is - probably like, 3-5GB precompiled?

[jolly thorn]

is appilo gui only?

[proper shale]

yep

i mean there's the No UI colab

and also RVC CLI

https://github.com/blaisewf/RVC_CLI

the ones I listed above, are basically using it online

if you have a capable pc might as well do it locally

https://tenor.com/view/cat-colon-three-happy-kitty-meow-meow-gif-12492309344215380572

[brittle wing]

@proper shale does these work in realtime, like if i play games / talk to my friends on discord

[rare goblet]

Ayo? @brittle wing level 1 !!!

[jolly thorn]

cries in gtx 1650

[proper shale]

that works for inferencing

[jolly thorn]

will it work?

[proper shale]

training might be rough, you'd have to use batch 2 😭

[jolly thorn]

no i have .pth, i just wann convert the audio

[proper shale]

no, those are for audio file inferencing

realtime on colab is kinda

[proper shale]

yeah use RVC CLI then, if you want no UI and stuff

applio for the convenience

[jolly thorn]

ok ok , thanks

[brittle wing]

i have a really bad gpu radeon 5500 xt, the delay is really bad at most times, and its kinda choppy. is it a way to make it like a bit more compatible like few tweaks. or do i need to get a better gpu

really hard to find anything these days cause there arent lots of amd users nowadays

[wispy lodge]

rvc realtime or wokada?

[languid night]

most realistic sounding english voicechanger for every gender in your guys opinion?

[rare goblet]

Ayo? @languid night level 1 !!!

[brittle wing]

rvc realtime

[wispy lodge]

Then you have to adjust sample length so it's more than inference time at least

And possibly reduce extra inference time to 2-3

[pastel oak]

have you read through the rvc guide?

but yea need to raise samplke length, maybe try 0.40 and if it still stutters go up or go lower if not

and the extra part ^

[sick swift]

can someone tell what is the difference in harvest, crepe and rmvpe?

[proper shale]

harvest = uses cpu, slowest out of the bunch, ok quality
crepe = good quality, however slow and susceptible to noise
rmvpe = good quality, fastest

[sick swift]

so rvmpe is just better without any drawbacks?

[proper shale]

yeah, it also helps mitigating some noise as well

it all depends on what u wanna do

some people would even say crepe is better quality, actually, but it's not worth it for realtime stuff tho

[sick swift]

i don't want realtime, i just want to change my audio clips to someone else voice, does that change anything?

or rmpve still best?

[versed herald]

Do you know how to use the program if my computer is not so good?

or if there is a simpler version

[proper shale]

yeah rmvpe best

[proper shale]

basically just use this if u wanna get ai covers done quickly

https://huggingface.co/spaces/TheStinger/Ilaria_RVC

[sick swift]

can you suggest how much epochs are enough for 3 minute audio? is 3 minute audio workable?

[rare goblet]

Ayo? @sick swift level 1 !!!

[proper shale]

3 minutes can work

epochs really vary, i'd say at 200 epochs you get a decent result, but ofc it might need a lot more training

hope you've got tensorboard and the patience to watch graphs go up/down 😭

[sick swift]

what are those 😭 (i am very new to these stuff)

any tutorial you can suggest?

[proper shale]

https://docs.aihub.wtf/rvc/resources/epochs--tensorboard/

though there are some innaccuracies i should sort out on that

just don't mind the "select lowest point in g/total" thing

[sick swift]

oh ok

[sick swift]

how can i determine how much more it needs? is it determined by that tensorboard and graph thing?

[proper shale]

yep, if like, g/mel and g/kl (basically, your best quality indicators) are still going down, keep going

[sick swift]

ok,
another question does batch size have to do anything with quality?

[proper shale]

yeah it kinda interferes, i'd use batch 4 for safety, and since its a small dataset, it should work

[sick swift]

damn i've been using batch_size: 8 and 300 epochs, and it sounds ass

[proper shale]

it could also be your dataset though

[sick swift]

my dataset is also ass 😭, can you do me a favor of listening to it?

[proper shale]

yeah that's probably it then 😭

send in dms i wanna ee

see

[modest trout]

how many epochs do i use for a 5 minute dataset with TITAN?

i used 200 epochs and it sucks 😭

[proper shale]

don't train with titan, use original pretrains instead, low batch size (4), 500 epochs, see if that's any good

if it doesn't go right, it's either the audio you're converting or the dataset

[modest trout]

tysm for help i was desperate

[proper shale]

lmk how it goes

[languid night]

My output and input are switched up, how do I fix it?

[sick swift]

isn't that ok? because i use voicemeter and in output device it always says 'input voicemeter' something. and all of that works fine

[languid night]

nope not for me

[sick swift]

if i were you, i would check sound settings in control panel

[rare goblet]

Ayo? @sick swift level 2 !!!

[languid night]

Alright imma try it

it's switched there too

[sick swift]

is it selected as default?

[languid night]

as my headset

so yea

[sick swift]

is it working?

[languid night]

but I need virtual cable to use in games and discord

[languid night]

yea my headset works with the vc

[sick swift]

so whats not working?

[languid night]

Cable input is in Output

and I need that for using the vc in games and discord n stuff

[sick swift]

its just the names which are exchanged, try selecting cable as default and also select it in discord or any game

[languid night]

alright

nope

[rare goblet]

Ayo? @languid night level 2 !!!

[languid night]

is default and still doesn't work

[sick swift]

probably the 'cable' software problem

[split jasper]

Can someone help me a little?

[languid night]

I can hear things now, but I need the mic and that doesn't work yet

[split jasper]

@sick swift Are you able to help me?

[sick swift]

i am not actually pro or something, i am new here, i guess i can try?

[split jasper]

Wait I am have got it.

[rare goblet]

Ayo? @split jasper level 2 !!!

[sick swift]

did you do the same things you did for speaker?

[languid night]

Nanu you know if there is any other device except virtual cable to use vc for discord and games?

[languid night]

yea

[sick swift]

i would recommend voicemeeter

[languid night]

ight imma try that

is switched up too

[sick swift]

in good case, it should be

just make voicemeeter input and output default in sound settings

and select your mic and headset in voicemeeter application

[languid night]

which of all of these

[sick swift]

huh?

[languid night]

does it even matter?

[sick swift]

oh

[languid night]

thats by playback btw

[sick swift]

probably just select the in 1

[languid night]

ight

doesn't work

[sick swift]

open the voicemeeter software and select your headset in the 'in 1' thing

could be also 'A1'

[languid night]

it works in the app

but I still don't hear anything and the mic doesn't work either

NVM

fixed it

[sick swift]

this is the mic setting, look for the out

[sick swift]

nice

[languid night]

thank you

[sick swift]

welcome

[split jasper]

Hey

So

I downloaded the cable

I removed the zip

And I tried to download it again

And it wont work

[sick swift]

what does it shows?

[split jasper]

Does it have to be a Admin to run?

[sick swift]

i guess no

does it just doesn't run?

doesn't shows anything or something?

[rare goblet]

Ayo? @sick swift level 3 !!!

[split jasper]

Do It click Install

Or is this the wrong thing

[sick swift]

do it

[split jasper]

Alr

[sick swift]

tell what happens next

[split jasper]

I got the Cable

Now how do I get it to work with like discord?

And also it sounds choppy

[sick swift]

what i know is just get a software like voicemeeter

[languid night]

Yo Nanu what are the best settings for RVC?

[sick swift]

i am still noob in RVC 😭, i joined this server today for help

[languid night]

Ah

[split jasper]

no no

[languid night]

oops

[split jasper]

Just

It sounds choppy

When i talk

[sick swift]

oh

[split jasper]

Is my res a big part?

[sick swift]

check your sample rate is the same in cable and your mic/speaker

[split jasper]

Alr

[sick swift]

you can check those in sound settings of control panel

[split jasper]

And

[split jasper]

Is it bad if my Res it to high?

[sick swift]

its not gonna damage something, the probable cases would be it works better or it doesn't work. try experimenting

[split jasper]

Experimenting with what?

[sick swift]

i've not done something like this, i just go for safer side, making sure every number is same in both

[split jasper]

Is that bad?

[sick swift]

oh

i thought you were talking about sample rate

[split jasper]

oh no

Like

[sick swift]

wait

[split jasper]

my voice

When I talk is choppy

Like

Cuts in and out

[languid night]

I should hear the voice changer voice on discord if I got the right settings right?

because I hear my normal voice even though I use the same settings as in RVC

[sick swift]

you tried changing the sample rate right?

[split jasper]

I tried changing the chunk

[rare goblet]

Ayo? @split jasper level 3 !!!

[sick swift]

for both mic and speaker?

[split jasper]

Where is the other one

[sick swift]

i am talking about sound settings

[split jasper]

Where is that?

[sick swift]

control panel

[split jasper]

Where do I go to get to there

Oh

[split jasper]

Alr what about it?

[languid night]

@sick swift how do I make it work with voicemeteer on discord, so I have the voice changer voice on discord too, cuz currently I only hear myself

[rare goblet]

Ayo? @languid night level 3 !!!

[split jasper]

@sick swift So I dont get what you mean

[sick swift]

right click on cable and go in properties, go to advanced and check the sample rate

[languid night]

Nanu overworking lmao

[sick swift]

it should be same as in the speaker

[sick swift]

and i am not even a mod 😭

[languid night]

you're helpful tho

and I need your help one more time

[sick swift]

make sure you select voicemeeter as default

[languid night]

:)

on discord?

[split jasper]

Wait

[languid night]

or windoes

[split jasper]

How do I get to

Properties

Wait I see

@sick swift its at 24bit 48000 Hz

[sick swift]

yes

[sick swift]

nice

now check this thing of the speaker

[split jasper]

But my chunk is not at that

[split jasper]

Where do I find the speaker?

[sick swift]

chunks are probably different things i guess i dont know about them

[split jasper]

Alr

[sick swift]

in the same sound settings

where you found cable

[split jasper]

Found it

[sick swift]

it would your headset probably

[split jasper]

Now go to Properties?

[sick swift]

yes

[split jasper]

That is at 16bit 44100

[sick swift]

yea that's the problem

[split jasper]

Alr

I will change the cable

[sick swift]

nice

[split jasper]

Done

Now what should I do?

[sick swift]

apply and it should fix the choppiness

probably

[split jasper]

Its still choppy

But its better

[languid night]

I hear my voice with it

not the voice changer one

[sick swift]

this voice changer is in discord?

[split jasper]

So what should me do?

[sick swift]

you have the vb cable application?

[split jasper]

I should

[languid night]

wdym?

[sick swift]

which voice changer is it?

which software?

[languid night]

RVC

[sick swift]

oh

[sick swift]

in discord select the rvc in the mic

[sick swift]

check the sample rate there and the latency

[languid night]

doesn't work

[split jasper]

Latency 7168

[sick swift]

and the sample rate in the applcation?

[rare goblet]

Ayo? @sick swift level 4 !!!

[languid night]

the only one that works is my mic and the one I use for the voice changer (I hear my normal voice with it)

maybe settings in voicemeeter?

[split jasper]

44100

[sick swift]

is there way to select output device voicemeeter in rvc?

[sick swift]

then i should work fine

[languid night]

yea

but that doesn't work on my pc

smh

[sick swift]

probably try reconnecting your headphone physically

[languid night]

nvm

Ill try it

[sick swift]

@languid night the plan is to get your voice from mic to voicemeeter and then from voicemeeter to rvc then to discord

[languid night]

how do I turn off that I hear myself in voicemeeter?

[sick swift]

discord should probably show a rvc input in microphone

[split jasper]

I dont know

Its still choppy

[sick swift]

in microphone setting (voicemeeter) just click the A

[sick swift]

dont know then, when i first tried vb cable i had that problem i guess but then i just moved on to voicemeeter that thing works fine for me

[split jasper]

Should I use voicemeeter?

[rare goblet]

Ayo? @split jasper level 4 !!!

[sick swift]

i would recommend it, you can try

try the banana version, its interface looks complex but has more features

[split jasper]

Alr thx for all of the help

[sick swift]

welcome

@languid night you should try selecting voicemeeter 'in 2' or voicemeeter 'aux input' in rvc output and then turn on the 'B' button in voicemeeter of 'in 2' or 'aux input' whatever you chose in rvc
then in discord select voicemeter 'in 2' or 'aux input' whatever you chose

https://docs.aihub.wtf/rvc/resources/epochs--tensorboard/

this should answer your last question i guess

[pastel oak]

i mean it does, for average model just stop whenever g/total is at its lowest. to determine that you have to keep training to make sure the lowest point is indeed the lowest

if you get more advanced you can also look into the other relevant graphs but for now g/total is enough for a high level model

[gusty kestrel]

An error occurred

Sorry, I could not find models that match your search "egirl voice mod"

[languid night]

sorry I was gone

[muted dirge]

so theres no way to train it further with just the pth and index

[split jasper]

Can you help me?

I have everything installed just its choppy a lot

[proper shale]

yeah no

[pastel oak]

type g/total in the search bar xd

if you had read the docs youd know :p

[pastel oak]

what is "everything"

what program

[split jasper]

Yours

And

Like a Cable

[pastel oak]

What youre saying makes no sense but im gonna assume you downloaded VB-Audio Cable which is shit so download the other cable

[grand lichen]

could I get some help in vc to download and set everything up, I'm lost on the guides

[pastel oak]

https://software.muzychenko.net/freeware/vac470lite.zip

download this, run setup64

[split jasper]

D0 I have to get rid of the zip?

[split jasper]

Oh alr

[pastel oak]

yes extract the folder

[split jasper]

It is VB I think

[split jasper]

Alr I have it downloaded

Should I delete VB cable?

[pastel oak]

yes

the new one is called Line 1

[split jasper]

Alr

[split jasper]

This should help with the

Stuff

[pastel oak]

it is the number 1 fix for crackles

few more stuff to look into if it didnt fix

[languid night]

if I connect my voicemeeter settings to discord I don't hear the ai voice changer voice

[untold frost]

I have a set GPU in my PC, however it doesn't register in the program itself.

[split jasper]

It still goes in and out

[pastel oak]

that is not a GPU on the first screenshot

[pastel oak]

what GPU do you have

and send screenshot of your voice changer

[split jasper]

Wait

Does this matter?

If the Res is to high?

[pastel oak]

well if the res is too high you either dont have a GPU or its not using your gpu

[split jasper]

Oh

Do I select GPU?

[pastel oak]

tell me the name of your gpu

[split jasper]

NVIDIA GeForce GT 1030

But I am about to get a new one

[pastel oak]

that gpu is too weak

[split jasper]

alr

I am getting a new one soon

So it doesnt matter

[pastel oak]

nicee

[split jasper]

Is this good?

NVIDIA GeForce RTX 4060 Ti

[untold frost]

What need I do to fix it, because I do have a modular GPU in my system

[pastel oak]

Yes

[pastel oak]

What is the name of your gpu

[split jasper]

Alr

[untold frost]

AMD Ryzen 5 7600

[rare goblet]

Ayo? @untold frost level 3 !!!

[pastel oak]

that is a CPU

as i said

if you are referring to an integrated gpu inside the cpu, that is not strong enough for the voice changer either way

[untold frost]

I’ll check it out

[honest delta]

Can someone help me make an ai cover? I have no idea what I'm doing lol

[analog obsidian]

http://docs.aihub.wtf/essentials/how-to-make-ai-cover/

[honest delta]

Thank you!

[snow pasture]

When you make a model with only a pince of sounds such as 1 - 3 samples. What do you do to make it work? Do you make the 1 or 3 samples loop for 1 - 2 mintutes?

[analog obsidian]

how many minutes you have* in total? (with the 3 samples merged together)

but the answer is no, you don't have to loop the audio if you're limited in minutes

[snow pasture]

It's around 1 to 2 seconds.

[sharp zodiac]

anyone have the download link?

[analog obsidian]

ah i see, no, don't loop the audio

just in case, rvc devs recommend 10 minutes of minutes for good results
5-8 minutes works just fine
and anything below you're in your own basically

[snow pasture]

K. I'll try and see if it sounds bad or not.

[analog obsidian]

is it a sound effect?

because u can't get good results with 1 or 2 seconds of dataset

so only do it if its a meme or something similar

[sharp zodiac]

anyone have a link for the Local-Mangio fork of rvc

[analog obsidian]

don't use mangio, its outdated

you can choose between mainline and applio

[sharp zodiac]

well its what im used to, the only reason why i dont have it is because i did a clean up and deleted the file by accident

[analog obsidian]

http://docs.aihub.wtf/rvc/local/mangio/

[sharp zodiac]

thanks

wheres the download?

[analog obsidian]

please go to the link and find this

its literally the first thing in the guide xD

[sharp zodiac]

wait why is it a 7-zip file?

[rare goblet]

Ayo? @sharp zodiac level 3 !!!

[analog obsidian]

it was always a 7zip file

u extract it on your desktop

or whatever

[sharp zodiac]

okay, in all honesty i dont know which version i used before, but the gui has a familliar layout so i just assumed.

[analog obsidian]

maybe u used mainline version

http://docs.aihub.wtf/rvc/local/mainline/

which is this

[sharp zodiac]

if its a 7-zip file i have no clue for how to get it to work

[analog obsidian]

https://www.7-zip.org/download.html

download 7zip

[sharp zodiac]

i have it

[analog obsidian]

extract the 7zip file in any place you want, then run go-web.bat

[snow pasture]

It's a sound that's similar to banjo kazooie speech.

[sharp zodiac]

i got it working, thanks

[untold frost]

Radeon RX 7600

[analog obsidian]

amd gpu trying okada?
https://github.com/deiteris/voice-changer/releases/download/b2020/voice-changer-windows-amd64-dml.zip download this instead, its emojikage's fork which boosts amd perfomance

also we have a channel dedicated for okada help #🔍│help-w-okada
this channel is for rvc help

[untold frost]

My issue is that the GPU isn't showing up in the slot, only the CPU

Despite there being a modular GPU installed

[analog obsidian]

yes because u downloaded cuda version which only works in nvidia cards

instead download what i sent

your gpu should appear there

[untold frost]

So should I delete the other file?

[analog obsidian]

yes delete your installation and use the new one instead

your amd gpu will appear in the version i sent

[untold frost]

I see,

I'll try it out

Thank you :)

Last question, srry, so just run this application here?

[analog obsidian]

yes

[untold frost]

And it's supposed to be browser-based?

[analog obsidian]

yep its going to open your browser and u can configure your stuff there

[untold frost]

Okok

Tankuuuu!!! <33

[analog obsidian]

[pastel oak]

hope the issue has been resolved thanks to lyery stepping in

lmk if anything else is needed

[untold frost]

I believe so :D

Tyvm :)

[grand lichen]

can anyone vc right now to help me set up the rvc and stuff? the guides messed me up, lmk if you can vc tho

just ping me 🙏

[dusky loom]

i'm trying to upload my mp3 for the weights bot but it says "1 files uploading failed" because whenever it opens the file explorer it lags and i have to restart discord and i couldn't do anything

[sharp zodiac]

@proper shale

[proper shale]

Is Command prompt still open

[sharp zodiac]

yes

[proper shale]

oookay that's kinda confusing

[sharp zodiac]

okay lte me close and reopen

[proper shale]

If you can send what command prompt says after you try to process data, that would also help 🙏

[sharp zodiac]

is this good?

[proper shale]

doesn't seem good - says directory name is invalid, for some reason

[sharp zodiac]

what does that mean?

[glad zealot]

Not sure if it will work but try replacing all \ with / in you path

[dull sand]

So i downloaded a voice model but it dosent accept zip files

what do i need to do?

[dull gazelle]

how can i know if this is overtrained?

[jolly thorn]

it has very large files, hosting it somewhere will be costly , do u know about anything lightweight?

[rare goblet]

Ayo? @jolly thorn level 2 !!!

[violet heron]

All ai will be large in file sizes

[shrewd nebula]

ay yall is a 1:00 of speaking enough for a dataset?

[violet heron]

I would get more if possible

[oak jewel]

i downloaded the app from the link you sent but it doesnt see any output devices, do you know how to fix the problem?

[analog obsidian]

because that version is for amd only

[oak jewel]

yes i have 6700 xt

[analog obsidian]

can u show me a screenshot of the interface?

it just says cpu?

[oak jewel]

guess i cant upload pics here, shall i just send it on dm?

though it does show both cpu and rx 6700 xt

[rare goblet]

Ayo? @oak jewel level 1 !!!

[oak jewel]

oh now i can send pics

[analog obsidian]

its using your gpu already

[oak jewel]

yes im asking i dont see any output devices not the gpu

or is there any way to run it on windowed mod rather than browser?

[analog obsidian]

oh

do you have virtual audio cable installed?

[oak jewel]

no, but i have tried different versions of this and they actually showed me some output devices but they were only working with cpu

thats why i downloaded the one you mentioned but now i cant see any output devices :(

[analog obsidian]

u have to install virtual audio cable first

[oak jewel]

any suggestions?

as in which app

would be nice to convert it to a virtual input device to use the voice in discord etc

[analog obsidian]

download virtual audio cable

also idk if this might work but open the configuration window in a chromium based browser (chrome, edge, opera)

[oak jewel]

tysm, yeah i searched it on internet and it works when i enable listening function from windows settings, now i need to work on how to make it my input sound

[misty stream]

how do i set total_epoch higher than 1000 in mainline

https://cdn.discordapp.com/attachments/710940317683023912/1254258232583655514/2024-06-22_22-11-51.mp4?ex=6678d681&is=66778501&hm=2d1677111a06904a845343104df53383036d725fc4f57542ea0863eea6c8e3fa&

[proud elbow]

quality and variation matter

[zinc sky]

im using applio rvc, it doesnt seem to convert the ooooooooh sound in the background from my vocal audio file. how do i fix it?

[dull gazelle]

im training a model idk if the graph looks good it still training tho.

[summer flame]

doing rvc thru google colab (shitty pc lol) and im having trouble cuz it keeps cancelling out due to the colab runtime

and now its not working overall

cant even launch it

[sharp field]

Hey does any1 here know how to use mangio?

[proud elbow]

not bad, just keep going

[rare goblet]

Ayo? @dull gazelle level 4 !!!

[dull gazelle]

300 epoch

how can i know when i should stop? when that thing starts to rise. i only know how to read that g/total

[proud elbow]

look at g/mel & kl as well

[proud elbow]

prob around 8-8.5k

[summer flame]

need help with this 🙏

[sharp field]

can someone help me please? I'm dying

[dull gazelle]

i use the applio 3.2.0 for me its better than the colab

[dull gazelle]

? how do you know that

[dull gazelle]

with what?

[sharp field]

it's hard to explain

[rare goblet]

Ayo? @sharp field level 1 !!!

[sharp field]

like it won't show anything when i click process

me no understand

[summer flame]

not exactly sure what that is

[rare goblet]

Ayo? @summer flame level 1 !!!

[proud elbow]

u can test some interesting checkpoint around that, or try train further to see if it still keeps improving

[dull gazelle]

ok look if you have that disc C, create a new folder i create the folder in the desk or download part, then try to put your dataset (song, voice or whatever you will use to train in that folder) after that, copy the direction of that example C:\Users\jack\Downloads\Voice Song example and paste that in the Enter the path of the training folder

[dull gazelle]

and if you cant tell me what the cmd is showing you

[proud elbow]

try not to have spaces in the model name

[sharp field]

ok so create folder outside of RVC folder

[dull gazelle]

i do that usually

and put in that folder the mp3 or wav

[sharp field]

:0 i got this @dull gazelle

dis gud?

[dull gazelle]

yes

continue now

[sharp field]

ok thank you

[dull gazelle]

[sharp field]

it work

ok i go let it train

[summer flame]

so i managed to get it to work with my own pc but its being super goofy and repeats over and over

[dull gazelle]

ohhh

i thought you were talking about applio

https://applio.org/

but for that i recommend you using CABLE

rmvpe only

[summer flame]

wdym cable?

sorry im new to this

[dull gazelle]

imma send you a video explaining that

[summer flame]

oh okay

yeah i was confused cuz i saw a friends ss of his and he has cable input but i dont

[dull gazelle]

https://www.youtube.com/watch?v=IS_SPQVv5iY&ab_channel=JarodsJourney

watch that video JarodsJourney explain how to use that with CABLE driver. if you have more questions ask me i use it to play vr chat xd

[summer flame]

i dont have audio cable

no frickin wodner

[dull gazelle]

install it

[sonic agate]

KLM 4.1 or TITAN?

[brittle wing]

so question

what does index ratio do?
median filter?
resampling?

[summer flame]

so rn i have it set to that and its just making funny noises

its not picking up my voice properly

[dull gazelle]

tell me where do you want to use the vcr

i use it for videogames

[summer flame]

on discord for a video with some friends

[dull gazelle]

ok

hold on

[coarse hinge]

I only have rmvpe_onnx and crepe_tiny everything else is N/A: on my F0 Det

I need help

[summer flame]

you prob have the wrong version

thats how it was for me too

what gpu do you havbe

[coarse hinge]

4090

[summer flame]

oh

hm

nvm idk then

[dull gazelle]

once you have that ready, download voicemod so you control

and enable push to talk on discord

[dull gazelle]

you downloaded a wrong version pretty sure that.

imma try to find

a version

hold on

[coarse hinge]

Do I need to uninstall or just delete the old one?

[summer flame]

is there any chance the reason im having issues is amd gpu?

[rare goblet]

Ayo? @summer flame level 2 !!!

[dull gazelle]

i use that

[summer flame]

cuz ive noticed most of you have insane nvidia gpus

[coarse hinge]

i downloaded the directML oops LOL. ||I will just delete the old one||

[rare goblet]

Ayo? @coarse hinge level 1 !!!

[dull gazelle]

https://www.youtube.com/watch?v=zHFtwA0cKzc&t=64s&ab_channel=SiliusXix

follow what that guy is doing and try getting the one i have

or if there is a newest try tha but must be onnxgpu cuda

if you use windows

imma sleep :j i helped many people today have a good night dudes

use chatgpt if you have a problem with the cmd

[proud elbow]

just install the gpu-cuda version and follow the guide https://rentry.co/VoiceChangerGuide
※ better use VAC from the link attached in the guide, which is known better

[thin stump]

KLM 4.1 is getting close to being better over Titan because they bring in actual VA's to do recordings. The only problem is that its noise sensitive, which was a problem the Rin e3 had

@dull gazelle #1235952130855010365

[violet heron]

No you don’t need voicemod

@summer flame download VAC Lite and run the setup64

Set that as your output on Okada

And your input on the program you want to use it for

[violet heron]

https://tenor.com/view/omw-to-spread-misinformation-all-over-the-internet-gif-26525511

[violet heron]

Or if you have a different way to extract output and make it a input use that

Think so

[keen coyote]

all these models buns 🙏

[dull gazelle]

voicemod lets you, use soundboards -.-, push to talk, turn off mic, use background music etc. i play vr chat tho

[violet heron]

You don’t need a program to turn off your change your mics volume buh

Most apps also have push to talk too

Most features you can achieve with stuff like voicemeter or similar stuff

(Tho i still recommend downloading vac lite over voicemeters built in one since it’s more reliable)

And plus w-Okada has built in settings for passthru and output

[violet heron]

Index ratio controls accent strength with the .index file
Median Filter is only for harvest iirc
dunno about resampling

[thin stump]

voicemodel is like an off-brand drink. You want the real cola with either RVC or okada

[violet heron]

worse is that this person is mixing the knockoff with the real one

[thin stump]

brainwashing on youtube is a huge problem lol. Largely outdated info

[violet heron]

(Or not knowing about the name brand colas existence is worse can’t really tell)

[violet heron]

“Errm you have to install PyTorch and python for the voice changer” (Total misinfo, installs it for you)

Tutorials on YouTube suck nowadays

I just read random Reddit comments and wikis dedicated towards the topic

[lost bison]

Does anybody know how to make metal screaming vocals work?

[dull sand]

How do i get started with this

[dull gazelle]

is this orvertrained? 340 epoch i will try 600 to see if that thing can go down

[turbid meteor]

what error is this

[proper shale]

don't think so, check g/mel and kl as well

[vestal island]

can i ask why my GPU: appear l ike cpu0, cpu1, cpu2 ? instead of CPU and (graphic card ) ?

[proper shale]

[proper shale]

cause that's how the DirectML version is

aka the AMD version

[vestal island]

oh, thx

[dull gazelle]

guys i have RTX 4070, i9, 64RAM when i open applio and i choose crepe for a reason the cmd stops loading but the processing keeps going i mean I had to cancel everything because I waited for about 2 hours and it didn't load, it just stayed at what it had been at the beginning, the percentage in the cmd didn't go up when I tried to do the part with the Pitch extraction algorithm with crepe, but with the others it loads quickly. a guy told he has a 4060 ti and he can use crepe but my RTX 4070 is not that dif from his 4060 ti isnt it?

[proper shale]

what hop length were you using?

and how many CPU processes were you using as well? normally this should be set to 2, to prevent errors

[dull gazelle]

32 hop lenght and how can i set the cpu processes to 2?

[dull gazelle]

?

[proper shale]

literally one of the first settings in the GUI, if it has a different name on Applio idk

[dull gazelle]

i mean when i am training that crepe option

[rare goblet]

Ayo? @dull gazelle level 5 !!!

[dull gazelle]

but if you are talking about GUI this one loads in 7 seconds each audio

[proper shale]

no not that

[proper shale]

it should be right next to the model's name, somewhere near that

unless ofc they changed the GUI to not include the option to change CPU processes

[dull gazelle]

ohhh you mean this?

[proper shale]

nope, still not that

it should've been up there along the name, architecture and sample rate stuff

[dull gazelle]

what do you use to make models?

:n xd

[proper shale]

thats just wonderful 😭

[dull gazelle]

imma google it but tell me what program do you use

[proper shale]

personally, I'd use Mainline RVC

but eh I'm stuck on Colab ig

[chrome sable]

i'm using the "dominant mommy" rvc model (because it's really good) but there's about 2-3 seconds where the audio is really quiet after i start talking that gets longer when i turn the extra up. is this a problem with the model or with my client?

only happens on that model which is unfortunate because it's the best i've come across, but i just want to know if i can fix it

[limber finch]

guys is there any way to convert my own voice to rvc form without using google colab

since yesterday im tryna find a way

google colab doesnt work

[dull gazelle]

try local, install applio

just google applio

[limber finch]

is it free

[dull gazelle]

yeh

[dull gazelle]

this?

[snow rose]

Hi, so i wonder how people are making ai covers with custom lyrics, at first i thought they maybe using some text to speech then mix it up, but no it always have same rythm and beat as original song. Can someone explain? I can provide some examples of ai covers that i seen with custom lyrics

[cyan parcel]

Is it important that the audio material have a whole pitch range or it doesn't matter? I mean do I need wave files were the singer singing high and low, or can I take one of the song and don't care about?

[finite galleon]

Possibly synth v

[snow rose]

hmm?

[twin bough]

Anyone know what settings to tweak on Applio to make "S" sounds to sound natural and not robotic? can't seem to figure it out

[brave swallow]

train with more wav, also what was the lowest value?

[twin bough]

not training my own models, using ones others made

[brave swallow]

Hmm i see, maybe they didn't train enough

[stark wadi]

is there a way to install mainline rvc on mac?

[rare goblet]

Ayo? @stark wadi level 2 !!!

[pastel oak]

theres mac steps here https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/blob/main/docs/en/README.en.md

[low shard]

send the google colab link

do it locally only if u got a good pc

[stark wadi]

does it work for m1/m2/m3?

[limber finch]

actually i tried it with too much links and any of them didnt work

[rare goblet]

Ayo? @limber finch level 1 !!!

[limber finch]

how can i do? it's a laptop with amd ryzen 7 and 12 gb ram

[low shard]

send me the google colab links, you are prolly using an outdated one

[violet heron]

That’s Mangio I think

Still no

(Also don’t help other people when you’d not know what your doing either)

https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI

[low shard]

im not really sure if u can do it locally tbh, what gpu?

[pastel oak]

should work yes

[thin tusk]

Where i found run-install.bat?

[rare goblet]

Ayo? @thin tusk level 1 !!!

[upbeat pine]

https://www.google.com/url?q=https://colab.research.google.com/drive/1u1brjK8IZt647UsbZuGYfW29oFM2I4tk?usp%3Dsharing&sa=D&source=editors&ust=1703256679653319&usg=AOvVaw2eSNYAzG5TkFBXENJi6qkk

This colab of mine isn't working, does anyone have another colab they can recommend to me?

[brittle wing]

what the hell? what does this mean

[rare goblet]

Ayo? @brittle wing level 2 !!!

[torn oar]

Is there any other webui you would recommend or is that that the best one?

I'm very new to this

[stark wadi]

I'm blind, and so I tried to install RVC mainline on the mac but its not working. Tried for an hour with no luck. RVC mainline is the most accessible for me with a screen reader, but I can't get it to install. Can someone help me? I tried to install homebrew for an hour and tried to download the missing Aria2 that the RVC script was complaining about but I had no success. Anyone know what I might be doing wrong? Thanks.

[mystic pawn]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• EasyGUI, by rejects Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

[violet heron]

Some stuff is not translated

There is some ui modifications out there for it tho

Forks usually are more bloated or have unnecessary features if you just just want audio to audio inference with models

[torn oar]

Got ya

What about text to audio?

[violet heron]

TTS? Some forks have tts support but its just generating normal tts and inferwncing it over

some forks with tts are applio and ilaira RVC

[violet heron]

https://github.com/Blanc-dot/RVC-Edits there is a gui mod for mainline RVC by blanc that removes unnecessary features and translates everything

[torn oar]

Thanks!
Any recommendations when it comes to TTS?

[rare goblet]

Ayo? @torn oar level 1 !!!

[violet heron]

I’ve seen people just use something like edgetts and then convert that with RVC

Like I said some forks have edgetts built in

[torn oar]

Got ya got ya, I just thought that you might have had a preference

[quartz lake]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• EasyGUI, by rejects Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

[violet heron]

There is also GPT SoVits which is made by the same person who made RVC but RVC models aren’t compatible with it

[torn oar]

I'll check it out, thanks for the quick answers!

[violet heron]

This server mainly focuses on RVC so don’t really expect support here

[snow rose]

.

[violet heron]

Though I would recommend going on osai / sanctuary if you need help with GPT sovits

[snow rose]

@violet heron

[violet heron]

A way you can achieve this is if you sing it yourself then inference it over

[snow rose]

🫥

[violet heron]

There is some ai that would automatically turn your audio to a audio but I don’t really know of some

I’ve seen people use suno to make songs and use RVC to make it their singer of choice

[thin stump]

Yeah those "custom lyrics" are sung by them and their voice is converted through rvc. The rest is clever audio mixing. I'll post an example in #1159290752195633273

[violet heron]

Don’t really know for custom lyrics

[snow rose]

well... then i should probably ask for some singing tutorial

I know whole kick back for example but im horrible singer

[thin stump]

You could achieve a similar effect with synth v but it's definitely not user friendly

like i dont even know how it works

synth v lyrics > convert through rvc ai model

[snow rose]

I did whole stayed gone from hazbin hotel by sukuna and gojo but if i could change some things in lyrics it would be much better

[thin stump]

but do check out Suno too. I think you can make custom lyrics of already copyrighted songs

[snow rose]

Yeah but you cant use your voices, if nothing changed beacuse i wasnt using suno for like 2 months

[violet heron]

You can generate it then use RVC to make it in your voice

[snow rose]

hmm

kinda yes

[patent schooner]

anyone know how to stop the voice changer repeating what i say

?

[woeful crow]

hi! what is the best uvr model when extracting stems like drums, bass and other??

[summer flame]

is that not basically the same thing as voicemeeter?

i font want to download a shitload of audio cables

[violet heron]

It’s basically just voice meters audio cable but it’s actually good

[summer flame]

its not getting it connected to discord that im having problems with

its the rvc its self

[violet heron]

what’s the problem with it

[summer flame]

it doesnt pick up my voice properly and makes random noises

[odd shale]

-rvc

[azure marsh]

Documentation

AI HUB Docs

https://docs.aihub.wtf/

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• Full AI Voice Model Training Guide (local) by Christopher Villanueva
• Model training with Mainline RVC by Raven (ravencutie21)
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[violet heron]

Input/output section

Might want to configure some of the sup ones aswell

[summer flame]

ive adjusted damn near everything i can and nothings worked

[craggy jewel]

quick qustion here but what does epoch mean

i see it in the title of some models

[violet heron]

Training cycles, more or less doesn’t mean better

[gentle moat]

what the hell is applio, should i trust downloading off it?

[rare goblet]

Ayo? @gentle moat level 1 !!!

[violet heron]

Fork of RVC that tries to be bigger then it is

[gentle moat]

so im good downloading off it?

[violet heron]

Yes

[gentle moat]

thx gng

[tame mica]

a certain cat will see this

[low shard]

https://docs.aihub.wtf/rvc/resources/epochs--tensorboard/

[low shard]

yes, its a fork of rvc, its safe to download

[low shard]

[low shard]

vidalcat

[stark wadi]

Can someone help me install RVC on mac?

[torpid siren]

Hi can someone help me with making ai cover? I have pretrained model but cant succesfully make ai cover bcs i cant succesfully setup a setting. I mean i made few covers with this models but every voice cover seems to be bad

[rose arrow]

what's the major difference between RVC and w-okada?

[proper shale]

I'd say check our docs, to know how to make better covers n stuff:

https://docs.aihub.wtf/

Also, I hope you aren't using the pretrains as a model

[proper shale]

Okada uses RVC to convert in realtime.

RVC just does audio conversion

[quiet ginkgo]

How much epochs should I do for 30 mins of dataset? (rmvpe_gpu)

[rose arrow]

thanks

[proper shale]

go for 500 & check graphs

if it ain't overtrained by then you can keep going

cause like, asking for an accurate epoch prediction is impossible, every model is different

[proper shale]

You're welcome :)

if it gets a better result in the end...

[viscid spire]

i need help installing rvc

plz

[low rapids]

in google colab w okada m not able tp upload the files
wot's the issue

[quiet ginkgo]

Why did It stop

[rose arrow]

does anyone here use steelseries sonar for noise cancellation? can someone teach me how to configure it while using okada or rvc

[wispy pawn]

watch the tutorial

[wispy pawn]

that happened to me just leave that link and run another googe colab see what happens

[snow rose]

@violet heron using suno for custom lyrics is bad idea

[rare goblet]

Ayo? @snow rose level 7 !!!

[snow rose]

XD

[violet heron]

Eh it was worth a try I guess lol

[violet heron]

No new RVC forks have proper Mac support, use a online version like Ilaria RVC zero instead

-spaces

[azure marsh]

<:huggingface:1179800228946268270> Hugginface Spaces

• 🆕 Ilaria RVC Zero, by thestingerx Huggingface Spaces
• 🆕 RVC⚡ZERO, by r3gm Huggingface Spaces
• Applio, by IA Hispano Huggingface Spaces
• AICoverGen, by r3gm Huggingface Spaces
• Advanced RVC Inference, by r3gm Huggingface Spaces
• RVC v2 Huggingface version, by Clebersla Huggingface Spaces

More commands: -audio, -colabs, -kaggle, -local, -overtraining, -realtime, -rvc, -help

[formal vessel]

so is that like the browser version?

[snow rose]

okay so i guess i need to try singing...

Or learn for next month or longer synth V

[violet heron]

It runs off HuggingFaces severs

Yes it’s completely in browser

[formal vessel]

so how do i still get the voices in

[lapis tinsel]

Hi everyone

Is there anyone online?

[sonic pond]

Anyone know why I only have crepe_tiny and regular crepe has (N/A)?

Never mind I might have been dumb and installed the wrong one I shall try to reinstall.

[violet heron]

Rmvpe and FCPE are better then crepe anyways

[sonic pond]

Weird, I've only heard about crepe being the best, I'll try the other ones though! Though I already reinstalled because I think those options also displayed (N/A).

[violet heron]

What’s your GPU?

The DirectML versions are for AMD/Intel GPUs

the Cuda version is for Nvidia

[sonic pond]

NVIDIA GeForce RTX 4070 Ti

[sonic pond]

Yeah, I installed the wrong one by mistake.

[rare goblet]

Ayo? @sonic pond level 2 !!!

[brave garnet]

RVC Colabs and Spaces

⠀

Google Colabs

⠀
AICoverGen-WebUI
Useful for making quick covers, by Hina.

AICoverGen-NoWebUI
Useful for making covers, doesn't include a UI, by Ardha, by Eddy, Hina and Gdr.

RVC Disconnected
To train new voice models, by Kit Lemonfoot.

EasyGUI
The OG interface, by Rejects.
⠀

Made with love by Ilaria 💖

[hollow vapor]

Hey,
Anyone has an idea why I can only use "crepe_tiny" or "rmvpe_onnx" in the setting? I have a AMD GPU and downloaded the DirectML version, or are there only these 2 options available?

[wispy lodge]

DirectML has only these options available

[hollow vapor]

Ahh good to know, thought I did an installing mistake, thanks!

[topaz dome]

hello! does anyone have a very beginner video guide for someone looking to do an ai voice changer? i'm very new to all of this and have no idea where to start. i apologize if this is in the wrong chat

[dreamy meadow]

how are you guys making ai models

[dreamy meadow]

i didnt understand it

[violet heron]

That’s why it tells you how

If you told you directly you still wouldn’t

[obtuse igloo]

can someone help me get the voice changer app? i got a nivida graphics card, please help 😭

[violet heron]

-realtime

[azure marsh]

This interaction has expired, use the command /guides realtime if you wish to see it again.

[violet heron]

First guide should have everything you need

Settings, setup, downloads, fixes, etc

[astral perch]

This is not necessarily a question for RVC but I was wondering if anyone knows if there exists an AI that translates and sings a song wtih the vocals and voice of the original song that original was sang in a different language. I have seen people use it but I don't if they're using RVC or something else

[proud elbow]

skill issue

[wet light]

my RVC GUI Can't run, can someone help me?

Running with the runtime Python, Please wait.
The system cannot find the path specified.
Press any key to continue . . .

It's always showing this message when I try running it as Administrator

[valid sierra]

where to put plugins

[blazing axle]

public url not working

in applio colab

[low shard]

If the public url is not showing, just checkmark the share_tunnel in the cell

[blazing axle]

i did

can u plz help me

[low shard]

Didn't a link show to you?

[blazing axle]

only the local

[low shard]

Like can you show me a screenshot of what you see

[blazing axle]

i sent it in dm

[rare goblet]

Ayo? @blazing axle level 1 !!!

[low shard]

All done

[vale osprey]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• EasyGUI, by rejects Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

[warm iris]

-colab

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• EasyGUI, by rejects Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

[warm iris]

-local

[azure marsh]

🖥️ Local stuff

• 🍏 Applio, by IA Hispano GitHub
• Mangio-RVC-Fork, by Mangio621 Huggingface
• RVC Studio, by SayanoAI Huggingface
• AICoverGen, by SociallyIneptWeeb GitHub
• Replay, by Replay Team Website
• Original RVC, by the RVC-Project team GitHub
• GPT-SoVITS, by RVC-Boss GitHub

Credits to Faze Masta and Antasma for compiling these links.

More commands: -audio, -colabs, -overtrain, -realtime, -rvc, -help

[sonic agate]

what is the best pretrain for talking voices?

[marble forge]

anyone know how to fix?

[silent peak]

I’m trying to make a text to speech but am having trouble figuring out the steps to all these different collars

[rare goblet]

Ayo? @silent peak level 1 !!!

[crude bolt]

you done all previous steps ? if not - do that.

[marble forge]

dude ive had this setup for months almost a year now infact and now it just stopped working of course i did them all

[crude bolt]

i cant tell from 1 picture what you have done in the past.

[marble forge]

ok

well what do i do

im so frustrated

thats the last thing it said

[brittle pivot]

it doesnt work for me

for some reason

[brittle wing]

-colabs

[azure marsh]

☁️ Google Colabs

• Applio, by IA Hispano Google Colab
• RVC Disconnected, by Kit Lemonfoot Google Colab
• RVC Mainline, by Hina Google Colab
• AICoverGen-WebUI, modded by Hina Google Colab
• AICoverGen-NoWebUI [English], by Ardha, fixed by Eddy, Hina and Gdr Google Colab
• AICoverGen-NoWebUI [Spanish], credits to Eddy, Hina and Gdr for translating and fixing Google Colab
• Ilaria RVC, by thestingerx Google Colab
• EasyGUI, by rejects Google Colab
• UVR5 NO UI for Google Colab, by Eddy Google Colab

More commands: -audio, -hf, -local, -kaggle, -overtraining, -realtime, -rvc, -help

[brittle wing]

-rvc

[azure marsh]

Documentation

AI HUB Docs

https://docs.aihub.wtf/

🍏 Applio Docs

https://docs.applio.org/

✨ More guides

• Full AI Voice Model Training Guide (local) by Christopher Villanueva
• Model training with Mainline RVC by Raven (ravencutie21)
• AICoverGen Colab Guide by Eddy (Spanish Helper)
• Create a model with RVC disconnected (colab) by Angetyde

[low spoke]

Hi Guys

I've been inactive on social media and everything, so I got out of the loop. I've noticed that Colab no longer works for making covers. Could you please instruct me on the new method for making AI-generated covers without needing to download programs?

[rose arrow]

anyone know how to configure steelseries sonar for noise cancellation when using w-okada or rvc?

[proper shale]

https://huggingface.co/spaces/TheStinger/Ilaria_RVC - pretty fast space for quick & easy covers.

https://docs.google.com/document/d/1YbXcLFPaGjhOdG5NFkK3QrucCEpHZBwFUxkeMO8aB18/edit - guide for it

For everything else (other Colabs): https://docs.aihub.wtf/

[azure marsh]

Documentation

📚 All-In-One English documentation

https://docs.aihub.wtf/

Full AI Voice Model Training Guide (Local)

Link: YouTube
Credits: Christopher Villanueva

Model training with Mainline RVC

Link: Rentry
credits: Raven (ravencutie21)

AICoverGen Colab Guide

Link: Google Docs
Credits: Eddy (Spanish Helper)

Create a model with RVC disconnected (colab)

Link: Google Docs
Credits: Angetyde

[lilac lily]

I'm having an issue

Traceback (most recent call last):
File "D:\Mangio-RVC-v23.7.0\runtime\lib\site-packages\gradio\routes.py", line 437, in run_predict
output = await app.get_blocks().process_api(
File "D:\Mangio-RVC-v23.7.0\runtime\lib\site-packages\gradio\blocks.py", line 1349, in process_api
data = self.postprocess_data(fn_index, result["prediction"], state)
File "D:\Mangio-RVC-v23.7.0\runtime\lib\site-packages\gradio\blocks.py", line 1283, in postprocess_data
prediction_value = block.postprocess(prediction_value)
File "D:\Mangio-RVC-v23.7.0\runtime\lib\site-packages\gradio\components.py", line 2586, in postprocess
file_path = self.audio_to_temp_file(
File "D:\Mangio-RVC-v23.7.0\runtime\lib\site-packages\gradio\components.py", line 360, in audio_to_temp_file
temp_dir = Path(dir) / self.hash_bytes(data.tobytes())
AttributeError: 'NoneType' object has no attribute 'tobytes'

Anyone knows how to fix?

[thorny juniper]

someone help

[proper shale]

Something may not have been selected properly... double check everything you've selected, and also make sure to use the file paths for audios, and not the dropdowns

[ornate summit]

My RVC was corrupted so I deleted the files to reinstall it, but when I did it still had all the data like I never got rid of it. Does it require an actual uninstall like an application?