#✨│ai-help
1 messages · Page 295 of 1
Do i need to delete the stuff i already downloaded?
yea except the voices
@viral mason Uhmm i can't redownload it it keeps telling me "Virtual Audio Cable (Lite) is already marked to delete (In German)
Also, how do you make it look like it's in a cave for the sake of sound reverberation?
?
thing ur supposed to use - @rare parrot
I use voicemod
Sorry what?
rvc gui
like to train a voice?
real time voice changer
Is that the Voice Changer or what is it? I'm sorry i'm kinda new to all of it
rvc gui is a completely diff thing to realtime
what gpu do u have
alr so everything must be good but the only thing changed is crossfade?
how to train a voice btw
try and see if it improves it
So how can i redownload the stuff or better said how do i delete the stuff i downloaded before properly so that it doesnt show me its marked to delete
that should be enough i will figure it out myself
thanks @viral mason @low shard for help
Anybody who has good experience still here?
You delete the files with the voice changer
Yeah i think i did but im not sure can i see somewhere if they're fully gone?
If u deleted them using the recycling can on your PC they're gone forever
I had them in my downloads folder clicked on them once and then hit the trash can on top middle
Should be deleted then
I'm trying to send ya a pic but idk how since it keeps popping up the same window its German though so idk if yiou understand it
You mean this?
Yup!
Mk, now just run setup.exe pretty sure
Should be fine
Ok so i hit the ok button and it all closed what now? Just follow the tutorial further?
Mhm
Now i need what ?
The Nvidia download in the guide, none of what you are looking at rn
If u have Nvidia 30 series and above you can use Vonovox but if u don't have Nvidia at all u can't use it
i do ´have just saw many diffrent opinions but there is surely better one
but ill try it thanks
If you're going for the best one I'd say use Vonovox
But it says here aswell that i need them right? Or am i in a wrong tutorial here
Only file you need is the zip that's what I downloaded and it works fine
So where can i find it then there is no other link that i can open
in the download section, scroll down
I finally got myself a pc. 64gb ram 5070ti gpu. Any tips on how to use ai to edit certain parts of images? Idk much and this is all confusing for me even with guides.
yup
Alr and now just reedit the Voices back? And the "Trial" Voice should be gone?
how can i make w okada voices sound good and do you reccomend ally good rvcs
@nova girder turn your game graphics down
I found the mistake but the voice keeps cracking like idk i didnt have those problems before any clues? It sounds kinda like a voice crack
these are the settings I use and they work fine but I use a 5070ti
I kinda use the same Chunk but lower "Extra" it keeps cracking on some Sentences though not all the time but still often...and before the restart i didnt have those problems idk though what exactly should have brought them
do u have bad bg noise or ppl talking in the bg?
RVC means Retrieval-based-Voice-Conversion, not realtime voice changer
also if you came froma 2 year old youtube tutorial about rvc gui realtime voice changer, that's outdated asf
- Check Docs & Guides: Your answer may already be in the AI Hub Docs or the https://discord.com/channels/1159260121998827560/1159513888199540817 channel.
- Search the https://discord.com/channels/1159260121998827560/1192011222023950368 : Look for existing posts that solve your issue. Do not invade someone else's post.
Tell your:
- Full GPU Name: (e.g.,
NVIDIA RTX 4060 8gb vram desktop) - Operating System: (e.g.,
Windows 11) - Detailed Description: What were you trying to do and what went wrong?
- Tutorial Used: Link to the guide you were following.
- Screenshot: A picture of the full error message.
To maintain a legal, safe & ethical community, we will NOT provide help for:
- ANY illegal activities.
- NSFW/Porn.
Requests for these topics may be ignored, not helped and result in moderation action.
- Be Polite & Patient: Our helpers are volunteers. You may ping the
Helpersrole once. - English Only: Please keep all conversations in English.
- Don't Ask To Ask.
ye but what is the best version to download of rvc
Last update: August 9, 2025
is it real time ?
sorry but which is what you actually want: "how to train a voice" or the realtime voice changer (see below)?
-rt
Guides for Programs that use RVC Models in Realtime for Calls/Games
A Realtime Voice Changer with similar performance to Wokada Deiteris Fork, with extra features, but supported only for Nvidia GPUs on Windows. and without cloud options GUIDE
A personal fork (modified version) of Wokada Deiteris Fork, it just adds some Quality of Life improvements to it like supporting Spin Embedder and Audio Effects. Don't expect too much about it since the creator made it originally as a personal project. GUIDE
Most suggested WebUI with the best general support for many platforms. GUIDE
For Windows Nvidia, Both Wokada Deiteris fork and Vonovox have similar performance & quality. Users should read the pros and cons for both and choose based on their differences, such as UI and Vonovox's paid effects.
Read Wokada Deiteris Fork Pros&Cons & Vonovox Pros&Cons
These options are not recommended for use.
Not suggested, older versions in youtube tuts are even way worse. GUIDE
The program is worse compared to the ones above, and much less updated. GUIDE
what happend with Ilaria_RVC?
There are two Ilaria RVC forks: one is the mainline version with full features, while the another is a lightweight RVC fork optimized for run on Hugging Face space. The mainline has last maintained since Apr 17th, 2024, while another has been broken on Hugging Face.
Can someone help me. I used mangio RVC 23.7. Last time i used prob 2 months ago and since yesterday, suddenly it wont start to work. I'm super newbie on coding or ai so maybe i need more layman-term help. Thank you
holy ancient technology
Do i need to download latest version or something?
mangio is old
I told you i'm super newbie in this. Just following some youtube videos and it worked before.
So, what's the brand new, factory fresh one?
did you ever find where to download the .onnx models? I too am looking for those, not RVC
Mangio the RVC fork has long outdated and never maintained.
Some tutorial videos from YouTube can also tell you to install some old RVC forks, which don't have to be only just W-Okada realtime voice changer.
I don't understand what fork means. I downloaded the mangio RVC-INFER one. Is that just the same?
Better forget that term anyways, but if you wanna know what that word mean you can Google search that.
Not all RVC is Mangio.
Applio RVC is more recommened for doing AI cover and even training a voice model.
!howtoask
- Check Docs & Guides: Your answer may already be in the AI Hub Docs or the https://discord.com/channels/1159260121998827560/1159513888199540817 channel.
- Search the https://discord.com/channels/1159260121998827560/1192011222023950368 : Look for existing posts that solve your issue. Do not invade someone else's post.
Tell your:
- Full GPU Name: (e.g.,
NVIDIA RTX 4060 8gb vram desktop) - Operating System: (e.g.,
Windows 11) - Detailed Description: What were you trying to do and what went wrong?
- Tutorial Used: Link to the guide you were following.
- Screenshot: A picture of the full error message.
To maintain a legal, safe & ethical community, we will NOT provide help for:
- ANY illegal activities.
- NSFW/Porn.
Requests for these topics may be ignored, not helped and result in moderation action.
- Be Polite & Patient: Our helpers are volunteers. You may ping the
Helpersrole once. - English Only: Please keep all conversations in English.
- Don't Ask To Ask.
RVC doesn't always mean "realtime voice changer". Applio RVC is not a realtime voice changer either, but rather a program where you can make an AI cover and even train an RVC voice model. It can be possible if in future Applio versions would support processing audio in real time. W-Okada and Vonovox are actual realtime voice changers, where they both use RVC voice models.
So like other staff members said, it's best to explain about your needs/goal and your PC GPU and operating system, rather than asking simple questions.
did you ever figure this out? I need to convert to piper .onnx as well for use with Piper
so
what would be the best settings to use for the voice changer, for my rtx 3050 and a ryzen 3 pro 2200g?
i also want to use it in a roblox game but it lags

?
Did you follow any tutorial or guide before this?
no
i used to use this before
i rem a guy from here
forgot his user but he helped me with my gtx 1650
before i upgraded
but what would be good settings
If you've ever tried Deiteris W-Okada or the "original" W-Okada before, there's a newer W-Okada for that, it is Tg Develop W-Okada (b2377). You can try this one. https://github.com/tg-develop/voice-changer/releases/tag/b2377 https://docs.aihub.gg/realtime-voice-changer/local/tg-develops-w-okada-fork/
The best known settings for GeForce RTX 3050 on Deiteris/Tg Develop W-Okada are something like these:
Chunk: around 110 - 130 ms; if you ever expierence lag you can try increase up until audio stable
Extra: 2.7 s
GPU: NVIDIA GeForce RTX 3050
F0 Det: rmvpe (not rmvpe_onnx)
The latest version of Deiteris W-Okada was released on December 7th, 2024. The "original" version of W-Okada is not recommended to use because it's more outdated.
Is there an option with Vonovox to hear yourself? I have it setup and working properly, and can listen to the output through recording with Audacity
Voicemod hear myself option, use it with Vonovox
how can i make it so i can clear my throat without it doing anything weird
just get hydrated
Drink water or lemonade water. 
thats the funny part
you dont
theres no other solution than to mute 🤷
If you're talking about RVC realtime voice changers, they can't do much realistically all non speech sounds, so nothing
Be sure to also not use YouTube tuts for those btw
Last update: September 6, 2025
Mangio RVC is a 2 years old outdated fork (modified version) of original/mainline RVC. Don't use it
@echo frost don't use YouTube tuts for RVC, elaborate your specs
!howtoask
- Check Docs & Guides: Your answer may already be in the AI Hub Docs or the https://discord.com/channels/1159260121998827560/1159513888199540817 channel.
- Search the https://discord.com/channels/1159260121998827560/1192011222023950368 : Look for existing posts that solve your issue. Do not invade someone else's post.
Tell your:
- Full GPU Name: (e.g.,
NVIDIA RTX 4060 8gb vram desktop) - Operating System: (e.g.,
Windows 11) - Detailed Description: What were you trying to do and what went wrong?
- Tutorial Used: Link to the guide you were following.
- Screenshot: A picture of the full error message.
To maintain a legal, safe & ethical community, we will NOT provide help for:
- ANY illegal activities.
- NSFW/Porn.
Requests for these topics may be ignored, not helped and result in moderation action.
- Be Polite & Patient: Our helpers are volunteers. You may ping the
Helpersrole once. - English Only: Please keep all conversations in English.
- Don't Ask To Ask.
Are you talking about the Ilaria RVC Hugging face Space ZeroGPU? https://huggingface.co/spaces/TheStinger/Ilaria_RVC it seems to have been paused after it had issues
@proven hill hey ila, btw do you have any news on this ? Just asking since I didn't really check it since this summer
anyobody know how to resolve the "To create a public link, set share=True in launch()" problem?
guys how do i remove reverb from a vcals
It all works fine now, finally. Thank You very much for the help @viral mason
When i use applio on kaggle all of the final links direct me to tensorboard instead of the UI
Have you tried restarting multiple times?
@nocturne mural even tho we can't seem to replicate this issue, this is the 2nd user that it happens to, the first one was @viral mason
Yeah I have. I tried to delete the notebook and uploading it again but the same thing happend
are you perhaps using Ngrok tunnels?
yeah, even tho it's recommended, it seems like a super rare issue is happening only to you and another user in all this time lol,
what happens if in the 2nd start cell you change the code from:
elif Tunnel == "Ngrok":
!pip install -q pyngrok
from pyngrok import ngrok
ngrok.set_auth_token(ngrok_authtoken)
p_tunnel = ngrok.connect(6969)
t_tunnel = ngrok.connect(8077)
f_tunnel = ngrok.connect(9876)
clear_output()
print(f"Applio Public URL: {p_tunnel.public_url}")
print(f"Tensorboard Public URL: {t_tunnel.public_url}")
print(f"FileBrowser Public URL: {f_tunnel.public_url}")
to:
elif Tunnel == "Ngrok":
!pip install -q pyngrok
from pyngrok import ngrok
ngrok.set_auth_token(ngrok_authtoken)
p_tunnel = ngrok.connect(6969)
time.sleep(7)
t_tunnel = ngrok.connect(8077)
time.sleep(7)
f_tunnel = ngrok.connect(9876)
time.sleep(7)
clear_output()
print(f"Applio Public URL: {p_tunnel.public_url}")
print(f"Tensorboard Public URL: {t_tunnel.public_url}")
print(f"FileBrowser Public URL: {f_tunnel.public_url}")
try changing the exact code I said, then run it again, let me know if it works
I'll try this thanks!
any idea for the best config for Vonovox considering I'm using a 5090?
show your current settings so i can check what's wrong and what's right
!give-media-perms 1h @cerulean spire
let me know, i just made so it takes more time between making tunnels, but I can't test if it fixes it since I can't replicate this issue
Almost a plug and use, I dind't change much
(I tried to change a couple things and didn't notice much of a difference, so I reverted them back)
- crossfade duration: 0.15
- extra time: 5.0
- check AP BWE 48k Upscaler
- Use RNNoise Reduction if you're having any noise issues
Do I keep Smart SINE as well as AP, or only AP?
both
you could also try using the pitch smoothing factor to 10, and the swift f0 method, but those aren't sure to help or not
after that, you just need to play with the pitch
wdym? please elaborate, that's not an issue
!howtoask
- Check Docs & Guides: Your answer may already be in the AI Hub Docs or the https://discord.com/channels/1159260121998827560/1159513888199540817 channel.
- Search the https://discord.com/channels/1159260121998827560/1192011222023950368 : Look for existing posts that solve your issue. Do not invade someone else's post.
Tell your:
- Full GPU Name: (e.g.,
NVIDIA RTX 4060 8gb vram desktop) - Operating System: (e.g.,
Windows 11) - Detailed Description: What were you trying to do and what went wrong?
- Tutorial Used: Link to the guide you were following.
- Screenshot: A picture of the full error message.
To maintain a legal, safe & ethical community, we will NOT provide help for:
- ANY illegal activities.
- NSFW/Porn.
Requests for these topics may be ignored, not helped and result in moderation action.
- Be Polite & Patient: Our helpers are volunteers. You may ping the
Helpersrole once. - English Only: Please keep all conversations in English.
- Don't Ask To Ask.
The voice seems cleaner with those settings tbh
Anything in advanced settings?
don't touch these at all
even the creator of the program suggests to not modify them
oh, gotcha! fawuhfuaw
also swift might be worse than rvc and rmvpe for rvc, so yeah just keep rmvpe
the last thing tho, regarding the embedder, i think i've read that it's better to use contentvec, but there were few suggestions for spin, any thoughts on that?
why is the file a .zip.001
i cant extract it
right click, winrar -> extract to folder
extract 001 and 002 to the same folder
De-reverb by anvuew or however you spell it, and de-echo normal, but if the de-reverb destroys the voice like basically erases it from existence use de-echo -w- reverb
wait i have to install 001 and 002
that's not a thing you can change, that depends on what embedder was the model trained
- contentvec: the default one for RVC, most used
- spin: newer experimental embedder which might be suggested for realtime usage and softer, there's also spin-v2 which is more optimized for RVC
if there's 001 i tend to believe there's a 002 as well, and yes, they're to be extracted in the same folder
oh, gotcha! appreciate the help nick!
btw, you're on an rtx 3050 on windows 10 or 11 right? Vonovox might be more suggested performance wise for your setup compared to wokada deiteris or tg-develop fork
but it's your choice on what to use
you're welcome
I don't wanna have to do this every single time, at this point just make a new applio appy file or whatever it's called
hey
I'm talking about room reverb I already tried that and it's not that good
Yk what let me send recordijg
i've been using the site version of okada but im using the ms on 280-400 to get green graph. Is the app better than site?
do you still experience the ngrok 3 exact same tunnels on Applio Kaggle?
That might be a fix for that, I can't add it since I can't even replicate the issue, is just a theoretical test, you need to test if it even works
There are different versions of wokada
please show your versions and settings
What is Melody's rmvpe?
!give-media-perms 1h @hasty iris
@low shard Well it did something because now all the links direct me to File browser
Oh yea the 3 kinks direct you to one of the three, file browser, applio, or tensorboard
Because broken
I don't really remember the version, because i downloaded the site version of wokada back in Early 2024 from this server
but i will send a screenie of it
You don't have to download it you just have to click the play button
seems like an experimental rmvpe for fixing voice cracks
I haven't heard of it much
Interesting
So there is no fix at all?
Just swap to Gradio local tunnel every single time
It's ok bc you don't even need to sign up for anything
It's all given in the training thing
could you show a screenshot of your current code and output?
even after changing to what I said, it didn't fix anything?
it's another tunnel method, please read about it in the guide and in the notebook, specifically gradio + localtunnel
it was just to check if you correctly changed the code
it's better you do that
right above the cell code
Last update: September 30, 2025
Sweet! Thank you so much!!
If I use Gradio should I still use the version of the cell you gave me?
When I use Gradio+local tunnel it works as intended
nah you don't need to
now what
It's working now. Thanks everyone for the help!
now that you have them downloaded, right click 001 then winrar then extract to (name of the file)
Oh I didn't know that the Spin embedder is an option now
I thought that I would have to select custom
if you want, Vonovox has an easier setup
he's not using vonovox
yupp, he just wanted to extact the files but seems he doesn't know how
ignore the file name
he seems to be using wokada tg-develop fork
The file name is of an outdated software is it not?
there's a gif showing how to extract them in the git repo afaik right?
done but what now
same with the other file, you need both the contents of zip 1 and 2 in the same folder you just created. but as nick said, vonovox might be easier and better considering you're using windows
vonovox still gets performance updates
we aren't even sure what you're using right now
it might be an old version of wokada
no
what exact guide or program are you using then?
i got it from him @hallow thistle
wait give me sec
this is the link he sent me https://github.com/tg-develop/voice-changer/releases/tag/b2377
@hallow thistle please suggest Vonovox to Nvidia GPU + Windows 10/11 users, it might be better for them, rather than wokada deiteris fork which last update was 7th december 2024
oh thats wokada tg-develop fork, i mean it's not a super active fork, but it doesn't get performance updates like vonovox, just QOL ones
are you really sure you want to use that instead of vonovox?
Last update: September 6, 2025
i dont know how to get decent settings for it
all the settings for vonovox are explained in the docs, and you can ask here for help
yeah vonovox could be better for performance
don't expect to run marvel rivals 4k high settings while using vonovox ofc
VAC Lite (Virtual-Audio-Cable by Muzychenko) is suggested in all the 3 rvc realtime voice changing programs guides in the AI HUB Docs, because some users reported random issues with vb audio cable
wokada tg develop fork is a fork of the deiteris fork of wokada, it adds some QOL and UI Updates and recently got them, but not performance related ones
Why does it come with a file like that, kinda misleading tbh
Looks like one of the old stuff when it isn't
even wokada deiteris fork does the same file name, wokada and its forks use an automated compiling process via github actions
Ah
I used it too when I forked wokada tg-develop to make users be able to use Spin via a github actions release, do you remember?
Nah I don't have the best memory
Wokada deiteris tho only uses mmvcserversio instead of the full file name
If my the sample rate of my audio is 44khz is it smarter to use 40 or 48khz?
Idk I always train 32k
im trying to get a public link but it gives me that phrase when im using the apolio collab
oh how come?
Pretty much all my audio comes from YouTube and I think that all yt videos are around that sample rate
- there are not many 48k sources 2) 32k is more than enough for speech
- higher SR is slower to train
even if some parts of the dataset peak at 44khz?
please elaborate firstly your PC GPU and Operating system
then, I'm guessing you're following https://docs.aihub.gg/rvc/cloud/applio-colab/
Last update: August 7, 2025
Imagine not using applio Kaggle, Collab is so silly
im using a Mac mini Apple M2 8 GB
What gpu does he have
tell him to join here and make a help post in #1192011222023950368
@slate pulsar
yea
please make an help post in #1192011222023950368 and elaborate your pc gpu and operating system
I'm guessing you want a realtime voice changer
ah, that's not really good for RVC AI Local model training, so yeah it's better to use cloud for you
Applio UI Collab current default tunnel method is localtunnel, so it's normal it says that, you just need to use the localtunnels instead
btw, Applio Kaggle or Lightning.AI would be more suggested for RVC training for free, since it gives more FREE GPU Time even tho they are a bit harder to use, you sure you want to use google colab?
which one should i use Applio Kaggle or Lightning.AI?
@slate pulsar please make an help post in #1192011222023950368 , it's easier for us instead of telling your friend
Kaggle gives more free gpu time but is riskier and has more frequent issues, lightning.ai is more stable and safe
it's your choice
check https://docs.aihub.gg
Last update: August 5, 2025
he has an nvidia gpu and is probably running windows 10 or 11, he just needs to use https://docs.aihub.gg/realtime-voice-changer/local/vonovox/
Last update: September 6, 2025
i finished installing vonovox but how do i open it
i did setup
already
when i do start up it shows hits
this
nvm
it just takes really long
Riskyyy? How so
yeah, just the first time
they technically do not allow web uis in the free tier, unless you buy a colab pro and link it iirc
yeah it depends on the wifi
wasn't it for your friend ?
my pc just fully froze right as i click start in the voice changer
my mouse wouldnt move
i had to restart
what's your pc gpu and operating system again?
I thought this was for @slate pulsar
win 11
also yes but now he doesnt want it
and pc gpu?
huh
he changed his mind
f0 method: rmvpe
extra time: 2.5
add a model and try now
lil better it isnt freezing now but like my mouse it just delyed and everything is elayed
delayed
close everything else in background
are you playing any games too?
I can't really VC
have you checked the background services here too?
you can send images tho
i use razer cortex to free up ram and close stuff
i debloated my pc to
i think i should try something else
does sample rate do anything
never heard nor used that program before ngl
show a screenshot of your current vonovox settings
whats the difference between uvr and eddgy uvr
f0 method: fcpe is lighter but a bit less precise, your choice
use extra time 2.0
and blcok size 0.40
@viscid moss 's UVR usually gets the newer models and more updates
is it better
that depends on the model you use
same model
thanks its a bit less laggy now but my mouse is just delayed and everything is just a tiny bit delayed
be sure your pc is properly cooled
it is
@zinc spindle did you check this too?
avg temps in game are like 55 to 60
yes
i debloated my pc
i have 16gb ram if that changes anything
i think its my cpu tbh
its a 4 core
but yes
is there any better voice changers
that doesnt make my pc slow
and can work without lagging in game
not really, vonovox is the only rvc realtime voice changer that still gets some performance updates
lower extra time to 1.5
play on lowest graphics 1080p 60fps cap
i do that already
dont expect too much, an rtx 3050 isn't that good
that gives more potential risks like heat, noise and a shorter lifespan 
should i try the one that @hallow thistle sent
ye i know
but
i dont care tbh
im gonna upgrade soon
that one was wokada tg develop fork, it doesn't get any performance updates and won't be any better
if you really want to try it, do it, but you still can't expect too much
well when i use orignal it doesnt slow my pc
and doesnt really lag it alot
its just the voices that lag when im in game
do you have good wifi speed? you could try using cloud (remote good pc) running wokada deiteris or tg-develop fork, with limited free gpu time
So the game runs on your hardware, and your realtime voice changer will run on a remote pc, ofcourse if you don't have a good connection you will still lag
can you do a speed test to check?
you could try http://docs.aihub.gg/realtime-voice-changer/cloud/tg-develops-w-okada-fork-kaggle/ or http://docs.aihub.gg/realtime-voice-changer/cloud/deiteris-w-okada-fork-lightning-ai/
Last update: September 6, 2025
Last update: September 6, 2025
what do you mean 30 hours weekly for free
lightning.ai gives around 80 hours of T4 gpu monthly, even tho it depends on the gpu you choose here
is it paid?
yeah, it's cloud computing, remote good pc running the code/program
this way your pc can run games without being stressed by the realtime voice changer
the original wokada is worse, you said that one lags too
ye when playing games
i have to make the extra and chunk more slowing just for it to not lag
slower
yeah that's the same exact thing you're doing on vonovox
try extra time to 1.0 and block size to 5.0
if you have AP BWE 48k upscaler enbaled, disable it
does sample rate cause lag
if it does
should i lower it down
that's the output sample rate, you could try lowering it but i'm not sure of the differences
what game are you trying to run btw?
roblox
just roblox
really
cause i would try it in other ganes
games
but it will just lag even if i change the settings
the only thing that actually may be the problem is my cpu
i see, did you also notice your gpu having any issues recently? because i remember some people running vonovox with roblox on a gtx 1650 ti enough decently
no
never
i got it a couple months ago
never bad really happenend
put extra time to 1.0?
btw i cant cause lowest is 1.5
does smart sine do anything
@low shard
?
oh nvm, your cpu is actually pretty bad, it seems to be bottlenecking your gpu
yes
really really bad
the only thing that can make it better is just opitimizing it
leave it on
its been on anyways
yeah it's on by default
what gpu delay do you get?
mouse and keyboard delay
wait 1.5 adn block size on 0.50 is good
it isnt lagging alot now
check this after clicking start, it's your gpu delay in vonovox
i dont see that
you didn't click start
set block size to 0.60
i don't think there's anything much more that you can do other than:
- buying a better cpu (and possibly gpu)
- using cloud with limited free gpu time (kaggle: 30 hours weekly | lightning.ai: T4 80 hours monthly)
- testing ASIO (it isn't documented much for vonovox yet, but you could check the asio / flexasio part of: https://docs.aihub.gg/realtime-voice-changer/local/tg-develops-w-okada-fork/#reduce-more-delay-windows-only)
well the block thing is working really well
im gonna go into the game and see what block size is perfect for it
alright, let me know
wawit
its actually doing pretty good at max
just delayed
i gonna lower it don
just to see how low i can get
sometimes it doesnt input my mic
but 0.65 to 0.70 is really good
for game
and theres no delay
with my mouse
and keyboard
thank you i mean
be aware that RVC can't do all non speech sounds, such as laughs and sneezing, realistically
you're welcome
but one thing is
what sample rate do you use
i wanna try it
48000
hm
oh it was the sample rate
that caused my lag a lil bit
when i put it to 38000 its now lessed delayed
do you guys reccoend some good rvc models . cuz most models i have suck
sigh
are u gonna do catfishing like some loser
found some good perfect settings with no lag on the voice
0.75 and 2.8
really good
for my specs
there is
i can show some
okay
idk why you need it but here https://huggingface.co/ZokaxDesu/e-girl/resolve/main/e-girl.zip?download=true
use 9 for pitch
its perfect for it
or 7 maybe 8 if your voice is a bit deeper
i know that model
the thing is that is sometime it work amazaing and some times it sound like shit
well here's another one i use https://huggingface.co/pendmg/Models/resolve/main/egirl.zip?download=true
can i see you use it possible
possibly
i wanna know what you mean by "shit"
i m not on my pc now
what echo are you talking about? 🤔
if you're able to post images here sure but if not dms are fine
How do i download Okada?
Where do i download it
the terminal pops up for a half a second btw if anyone can help
what gpu do u and @past pier have
rtx 4090
i have 4060
Guides for Programs that use RVC Models in Realtime for Calls/Games
A Realtime Voice Changer with similar performance to Wokada Deiteris Fork, with extra features, but supported only for Nvidia GPUs on Windows. and without cloud options GUIDE
A personal fork (modified version) of Wokada Deiteris Fork, it just adds some Quality of Life improvements to it like supporting Spin Embedder and Audio Effects. Don't expect too much about it since the creator made it originally as a personal project. GUIDE
Most suggested WebUI with the best general support for many platforms. GUIDE
For Windows Nvidia, Both Wokada Deiteris fork and Vonovox have similar performance & quality. Users should read the pros and cons for both and choose based on their differences, such as UI and Vonovox's paid effects.
Read Wokada Deiteris Fork Pros&Cons & Vonovox Pros&Cons
These options are not recommended for use.
Not suggested, older versions in youtube tuts are even way worse. GUIDE
The program is worse compared to the ones above, and much less updated. GUIDE
first guide
which one do i download
def outdated by now
is it open source
Vonovox is not open source, it's completely free tho with an optional thing u can pay for
i cant find the download mb im dumb
same
u found it
im still too dumb to download ngl
i mean
i downloaded it
idk how to run it
there is no start.bat
it makes models sound more realistic and better, it has improved optimization and other stuff
🥀
run setup.bat then run start.bat
i downloaded the wrong thing 💀
what
can i download it from hugging face
can i download it from hugging face
why?
just wanna
his ping on games is higher than his iq
btw how do i download
ratio
do i have to jsut wait
or what
nvm
bro
yea just wait till it says press anything to continue
what
I can help you
ok bet
send me a screenie
that's not the one I recommended
of what
application
. @viral mason
i thought it was better
it shows debug
i want that
@viral mason how make it work in vcs
virtuall cable
i didnt ask u
input: microphone
output: cable output
other programs:
input: cable input
output: headphones
where do i do these settings
vonovox
oh mb
and in discord use the second thing
other programs:
input: cable input
output: headphones
this
https://github.com/dr87/Vonovox/archive/refs/tags/v1.6.9.zip
this one is the current best, if u ask if u can download it off huggingface again I will not be helping u
What’s up people I’m a music producer. I’m looking for someone to help me train a voice, (I already made the dataset) but my AMD is not suitable for training in return I offer you a completely free production so that your iA voices sound of quality here I attach a sample https://drive.google.com/file/d/1LThZy3FpNAxaB2qjZihAEAV9DPRCIZHj/view? USP=drivesdk
If you are interested, leave me a message and we will gladly talk 🔥
Just train on applio on kaggle, it's like Google Collab but it gives 30 hours to free users and completely free
Do you have the link? It would help me a lot 🙏🏻 I really thought that the notebooks to clone voices had already died
!howtoask
- Check Docs & Guides: Your answer may already be in the AI Hub Docs or the https://discord.com/channels/1159260121998827560/1159513888199540817 channel.
- Search the https://discord.com/channels/1159260121998827560/1192011222023950368 : Look for existing posts that solve your issue. Do not invade someone else's post.
Tell your:
- Full GPU Name: (e.g.,
NVIDIA RTX 4060 8gb vram desktop) - Operating System: (e.g.,
Windows 11) - Detailed Description: What were you trying to do and what went wrong?
- Tutorial Used: Link to the guide you were following.
- Screenshot: A picture of the full error message.
To maintain a legal, safe & ethical community, we will NOT provide help for:
- ANY illegal activities.
- NSFW/Porn.
Requests for these topics may be ignored, not helped and result in moderation action.
- Be Polite & Patient: Our helpers are volunteers. You may ping the
Helpersrole once. - English Only: Please keep all conversations in English.
- Don't Ask To Ask.
What's an rvc fork
-colab
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.
by IA Hispano
Google Colab
by Hina
Google Colab
by Eddy
Google Colab
by Eddy
Google Colab
by Deiteris & Hina
Google Colab
by Shiro & Eddy
Google Colab
by Nick088
Google Colab
by Nick088
Google Colab
by Jarredou & Makidanye
Google Colab
Does applio also has the thing with batch size where 2 is actually 4 because of the 2 gpus?
there is no link, all u need is kaggle and ngrok
only on kaggle
an edit of either an ai voice changer or of a voice training software
it is batch size 2 each gpu
So just like kaggle when I select 2 it's actually 4?
Hi everyone, I'm having a problem with the voice changer. I downloaded the voice changer from this site: https://github.com/IllIlIlIllIl/voice-changer/releases/tag/b2335 since I have a 5070ti, but whatever I do I just hear my normal voice. Does anyone know how to fix this?
hello guys do u know how to minim the delay
Great, thank you, you helped me a lot
vonovox
Are there any videos about vonovox?
like how to download and set it up
nah all u need is to download it from the guide as well as vac lite
then run setup.bat then start.bat
Could you just give me the download link for Windows
what gpu do you have
it's not always about voice changer
please explain what download link and something
!howtoask
- Check Docs & Guides: Your answer may already be in the AI Hub Docs or the https://discord.com/channels/1159260121998827560/1159513888199540817 channel.
- Search the https://discord.com/channels/1159260121998827560/1192011222023950368 : Look for existing posts that solve your issue. Do not invade someone else's post.
Tell your:
- Full GPU Name: (e.g.,
NVIDIA RTX 4060 8gb vram desktop) - Operating System: (e.g.,
Windows 11) - Detailed Description: What were you trying to do and what went wrong?
- Tutorial Used: Link to the guide you were following.
- Screenshot: A picture of the full error message.
To maintain a legal, safe & ethical community, we will NOT provide help for:
- ANY illegal activities.
- NSFW/Porn.
Requests for these topics may be ignored, not helped and result in moderation action.
- Be Polite & Patient: Our helpers are volunteers. You may ping the
Helpersrole once. - English Only: Please keep all conversations in English.
- Don't Ask To Ask.
Could you give me the download link for the NVIDIA voice changer for Windows
then this is the answer #✨│ai-help message
how to download the Windows voice changer should be provided
whats this then?
yes
What are the best settings for W-Okada deitris fork?
please elaborate your setup
!howtoask
- Check Docs & Guides: Your answer may already be in the AI Hub Docs or the https://discord.com/channels/1159260121998827560/1159513888199540817 channel.
- Search the https://discord.com/channels/1159260121998827560/1192011222023950368 : Look for existing posts that solve your issue. Do not invade someone else's post.
Tell your:
- Full GPU Name: (e.g.,
NVIDIA RTX 4060 8gb vram desktop) - Operating System: (e.g.,
Windows 11) - Detailed Description: What were you trying to do and what went wrong?
- Tutorial Used: Link to the guide you were following.
- Screenshot: A picture of the full error message.
To maintain a legal, safe & ethical community, we will NOT provide help for:
- ANY illegal activities.
- NSFW/Porn.
Requests for these topics may be ignored, not helped and result in moderation action.
- Be Polite & Patient: Our helpers are volunteers. You may ping the
Helpersrole once. - English Only: Please keep all conversations in English.
- Don't Ask To Ask.
get vonovox in the guide links provided by the other helpers
RVC means Retrieval-based-Voice-Conversion, not realtime voice changer
if you're talking about the old go-realtime.bat in original/mainline rvc known as rvc gui, that's outdated and worse
there's no updated video about rvc realtime voice changers
be sure you're on windows 10 or 11 to use vonovox
-realtime
Guides for Programs that use RVC Models in Realtime for Calls/Games
A Realtime Voice Changer with similar performance to Wokada Deiteris Fork, with extra features, but supported only for Nvidia GPUs on Windows. and without cloud options GUIDE
A personal fork (modified version) of Wokada Deiteris Fork, it just adds some Quality of Life improvements to it like supporting Spin Embedder and Audio Effects. Don't expect too much about it since the creator made it originally as a personal project. GUIDE
Most suggested WebUI with the best general support for many platforms. GUIDE
For Windows Nvidia, Both Wokada Deiteris fork and Vonovox have similar performance & quality. Users should read the pros and cons for both and choose based on their differences, such as UI and Vonovox's paid effects.
Read Wokada Deiteris Fork Pros&Cons & Vonovox Pros&Cons
These options are not recommended for use.
Not suggested, older versions in youtube tuts are even way worse. GUIDE
The program is worse compared to the ones above, and much less updated. GUIDE
above there's the guide link
- Check Docs & Guides: Your answer may already be in the AI Hub Docs or the https://discord.com/channels/1159260121998827560/1159513888199540817 channel.
- Search the https://discord.com/channels/1159260121998827560/1192011222023950368 : Look for existing posts that solve your issue. Do not invade someone else's post.
Tell your:
- Full GPU Name: (e.g.,
NVIDIA RTX 4060 8gb vram desktop) - Operating System: (e.g.,
Windows 11) - Detailed Description: What were you trying to do and what went wrong?
- Tutorial Used: Link to the guide you were following.
- Screenshot: A picture of the full error message.
To maintain a legal, safe & ethical community, we will NOT provide help for:
- ANY illegal activities.
- NSFW/Porn.
Requests for these topics may be ignored, not helped and result in moderation action.
- Be Polite & Patient: Our helpers are volunteers. You may ping the
Helpersrole once. - English Only: Please keep all conversations in English.
- Don't Ask To Ask.
nevermind, im using vonovox now. i did no changes on vonovox and my pc has 4060 ti amd ryzen 7 5700x
win 11 no tutorial
Could you just give me the download link for Windows
yeah that's good
f0 method: rmvpe
crossfade duration: 0.15
extra time: 3.0
play with the pitch
What could i do to make the voice not cut out when it speaks and for the delay to be normal
the delay is a little longer then it says
What could i do to make the voice not cut out when it speaks and for the delay to be normal <@&1159293204038955078>
-audio
So I’ve been using real-time voice changers, but now I want to know, how can I convert my voice into an AI voice using a program or something, like not in real-time?
Do you mean making your own voice into an ai?
What gpu do u have Nvidia or amd
Yoo
Yo
You have any e girl voice's
I'm gonna end my life
Noo nyaa ~ you gotta send me the e girl voice first
Until then i won't let you die my kitten o:-)
I don't help with catfishing e-girl/e-boy stuff so have this lizard as compensation, go find someone else who's willing to break the law
-colab
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.
by IA Hispano
Google Colab
by Hina
Google Colab
by Eddy
Google Colab
by Eddy
Google Colab
by Deiteris & Hina
Google Colab
by Shiro & Eddy
Google Colab
by Nick088
Google Colab
by Nick088
Google Colab
by Jarredou & Makidanye
Google Colab
And we all know that you must have 100s of models of e girl' hidden
You can look at what Lyery has made in their voice models just go to the search bar at the corner and copy paste this
in: https://discord.com/channels/1159260121998827560/1175430844685484042
From: whatever Lyery's current name is
I can't seem to find them rn
I have never made one of those because genuinely I hate them because of people who call the voice changer the "e-girl voice changer" or people like this
These things are why I don't help with it or make them
!howtoask
- Check Docs & Guides: Your answer may already be in the AI Hub Docs or the https://discord.com/channels/1159260121998827560/1159513888199540817 channel.
- Search the https://discord.com/channels/1159260121998827560/1192011222023950368 : Look for existing posts that solve your issue. Do not invade someone else's post.
Tell your:
- Full GPU Name: (e.g.,
NVIDIA RTX 4060 8gb vram desktop) - Operating System: (e.g.,
Windows 11) - Detailed Description: What were you trying to do and what went wrong?
- Tutorial Used: Link to the guide you were following.
- Screenshot: A picture of the full error message.
To maintain a legal, safe & ethical community, we will NOT provide help for:
- ANY illegal activities.
- NSFW/Porn.
Requests for these topics may be ignored, not helped and result in moderation action.
- Be Polite & Patient: Our helpers are volunteers. You may ping the
Helpersrole once. - English Only: Please keep all conversations in English.
- Don't Ask To Ask.
I got a problem when I play my fans speed goes like 50% (can’t change it) and the voice changer moans or something like that and someone knows how to fix it, even when I talk normally it will do that and won’t sound like a girl, I’m tryna troll people on Roblox vc but it’s not believable since the voice changer sounds like shit and bad. My specs are AMD Radeon RX 7600 7500F CPU Kingston FURY Beast RGB DDR5 6000MHz 32GB 2TB
This shit is a pain in the ass
(I watched this video as tutorial: https://youtu.be/SxdnGxicJOg?si=ILcIG8aqgWmTsLle )
That's because you have an outdated program 💥
Never use YouTube tutorials for the voice changer
Help me please
Will do
Btw
Guides for Programs that use RVC Models in Realtime for Calls/Games
A Realtime Voice Changer with similar performance to Wokada Deiteris Fork, with extra features, but supported only for Nvidia GPUs on Windows. and without cloud options GUIDE
A personal fork (modified version) of Wokada Deiteris Fork, it just adds some Quality of Life improvements to it like supporting Spin Embedder and Audio Effects. Don't expect too much about it since the creator made it originally as a personal project. GUIDE
Most suggested WebUI with the best general support for many platforms. GUIDE
For Windows Nvidia, Both Wokada Deiteris fork and Vonovox have similar performance & quality. Users should read the pros and cons for both and choose based on their differences, such as UI and Vonovox's paid effects.
Read Wokada Deiteris Fork Pros&Cons & Vonovox Pros&Cons
These options are not recommended for use.
Not suggested, older versions in youtube tuts are even way worse. GUIDE
The program is worse compared to the ones above, and much less updated. GUIDE
Make sure you download the AMD version when you get the chance, there are 3 versions of it one for Nvidia, one for specifically 50 series gpu for Nvidia, and one for amd and Intel
No rush 👍
Okay thanks
But is RTX 4060 better than my amd
RX 7600
Because I got RTX 4060 in my basement
Oh if you have an rtx use the Nvidia download instead!
But is it better?
Okay I will try to find it
👍
I haven’t found it last time I checked
Hopefully you do because you could even use Vonovox with it which is the current best AI voice changer only available for now for Nvidia
Yea that one wouldn't work well, the AMD card u have would work better than that lol
Yeah but you said nvidia is the best for ai so it got me thinking
It's true that Nvidia is better but a tiny Nvidia card like that won't work so well, I originally had a 1660 and that struggled with using the voice changer
I am home sorry for the wait
I had to do groceries
all good
Alright what do you want me to do
Also I was looking for my other graphic card and couldn’t find it
Wait
alr if u can't find the 4060 just stick with the amd one
alright
screenshare how?
discord call
I guess?
ok
-realtime
Guides for Programs that use RVC Models in Realtime for Calls/Games
A Realtime Voice Changer with similar performance to Wokada Deiteris Fork, with extra features, but supported only for Nvidia GPUs on Windows. and without cloud options GUIDE
A personal fork (modified version) of Wokada Deiteris Fork, it just adds some Quality of Life improvements to it like supporting Spin Embedder and Audio Effects. Don't expect too much about it since the creator made it originally as a personal project. GUIDE
Most suggested WebUI with the best general support for many platforms. GUIDE
For Windows Nvidia, Both Wokada Deiteris fork and Vonovox have similar performance & quality. Users should read the pros and cons for both and choose based on their differences, such as UI and Vonovox's paid effects.
Read Wokada Deiteris Fork Pros&Cons & Vonovox Pros&Cons
These options are not recommended for use.
Not suggested, older versions in youtube tuts are even way worse. GUIDE
The program is worse compared to the ones above, and much less updated. GUIDE
@uncut haven go to the guide for vonovox, the first one, find the download for vonovox and then run setup.bat first then start.bat second
not really at all no, u should be using rmvpe not crepe, but anyways this is an outdated ai voice changer and u should get the new one that wil lrun much better ^^
This?
yup! just donwload the 1.6.9 or whtever the newest version is
the source code.zip file
Do you think the cuts like when it doesn't pick up my voice are because of the webcam? I don't have a real microphone
btw settings for it are a little diff from the normal wokada
crossfade duration: 0.15
extra time: 0.30 or 0.27
block size: 0.60–0.70 try messing with it, it controls your delay like chunk in wokada
not sure, as long as the quality of the mic on the webcam doesn't suck it should be fine
Ik for the model I was using it was 11 pitch
it would be 11.0 for that then
no need for index btw
I mean keep it at 0 nut still install it to vonovox
I've added index but set it at 0 and what about sample rate and f0 method and pitch smoothing Factor
extra 0.27? I can max 1.5
idk I got the info from Nick088
try 3.0 or 2.7
Is WASAPI better than ASIO?
wasapi is less confusing, idk what asio does tbh
is sample rate 48000 right and rmvpe?
f0 method do rmvpe, sample rate go for 48000/48k
also ur using the wrong cable
it should be cable input not the 16 inch thing
all voice changers have delay but try turning down the block size some until it's better for u
maybe 0.30?
.
if u go too low tho it'll be choppy or just straight up bad so u have to balance delay with being able to speak and understand what you're saying
controls delay
@uncut haven u get it working btw?
Yes I've some delay issues like 128ms sometimes
But it working better than before
I'm testing around
And I found out that my webcam is around 16k rate and the program is set at 48k
I've to change my web cam
👍
Hello, what is the best version of the voice changer app? Or is there a new update?
voice-changer-windows-nvidia-b2332
MMVCServerSIO
??
what new
what gpu do u have?
Where can I get a dataset full of wiffle balls or just pictures of balls with holes in them
This might be a basic question but I need some help. I basically made some characters using wisk and gemini, and I need the file to be much larger resolution for a large print. I had to alter the faces in photoshop to look more like the real people they are based on, but I can't seem to find an upscaler that will properly reinterpret the image without changing the faces too much. I downloaded upscayl, but each result looks not so great. Any advice?
how can I make the app work on discord ive tried everything
Where can we find high-quality voice models? I tried searching for models of game characters and girls/boy models I wanted but after testing them, it seemed like they only made the voice sound higher or lower.
Do you guys have any recommended websites or good places to find models?
What settings are good for a egirl voice model
my GPU AMD radeon RX 6600
i use this MMVCServerSIO_win_onnxdirectML-cuda_v.1.5.3.18a.zip
but wont work
is it wrong
“For high-res prints without distorting faces, use an upscaler built for portraits: Topaz Gigapixel AI (Face Recovery mode), Upscayl, or VanceAI. Upscale in smaller steps (2–4×) to preserve details, then do light touch-ups in Photoshop if needed. Export at 300 DPI for print—this usually keeps faces accurate and ready for large prints.”
“Looks like that build is mainly optimized for CUDA (NVIDIA GPUs). Since you’re on an AMD Radeon, it won’t work properly. You’ll need a version that supports DirectML or AMD GPUs specifically.”
“Try a slightly higher pitch, some breathiness, and moderate energy. Use playful or cute tones without overdoing it.”
Well,“Check out ElevenLabs or Resemble AI for high-quality voices. Look for emotion and style settings, not just pitch changes, to get natural character voices.”👍
can someone help me plz with the outputs im too stupid for this..
guys guys guys
stable diffustion web ui and comfy ui
which is better
what is the best free ai software nowaday?
@civic tiger that answer was automated using AI, take it with some grain of salt to find if it is useful or not
and also to you @vital thistle
- Check Docs & Guides: Your answer may already be in the AI Hub Docs or the https://discord.com/channels/1159260121998827560/1159513888199540817 channel.
- Search the https://discord.com/channels/1159260121998827560/1192011222023950368 : Look for existing posts that solve your issue. Do not invade someone else's post.
Tell your:
- Full GPU Name: (e.g.,
NVIDIA RTX 4060 8gb vram desktop) - Operating System: (e.g.,
Windows 11) - Detailed Description: What were you trying to do and what went wrong?
- Tutorial Used: Link to the guide you were following.
- Screenshot: A picture of the full error message.
To maintain a legal, safe & ethical community, we will NOT provide help for:
- ANY illegal activities.
- NSFW/Porn.
Requests for these topics may be ignored, not helped and result in moderation action.
- Be Polite & Patient: Our helpers are volunteers. You may ping the
Helpersrole once. - English Only: Please keep all conversations in English.
- Don't Ask To Ask.
explain what you want to ask first
allat just to ask a question...
tl;dr^
hey im usin the owakada voice changer or what ever its called whats the input and output for can someone explain pls ❤️
i think i have the same issue
what issue do u have?
idk wich one to use in the output and input
do u mean this https://docs.aihub.gg/realtime-voice-changer/local/tg-develops-w-okada-fork/#discord--games
Last update: October 26, 2025
ye
idk wich microphone to use
also im using it for the first time so idk what to do and press
your real one
open the Sound control panel to find it
"failed to import model because beatrice toml is not found"
whats the solution of this issue when i want to upload my vc model
my question was for the volume of out and in. Like if my output volume is low does it mean my volume ppl will hear me is gonna be low too?
@viscid minnow you've got such nasty AI answer cuz it thinks you want to do sort of catfishing 🥀
Ahh okay thank you. I did try to try one of those it mentioned, but there was no trial and you had to commit to a year without trying it out first. Crazy! I found one that was only $12/mo and it worked well enough for what I needed (finally! :))
i cant upload my ai zip
you can find some upscaler tools here https://huggingface.co/spaces/enzostvs/zero-gpu-spaces
note that they may not be perfect to always preserve all the details
https://www.youtube.com/watch?v=6YxwSYAoyQA hey guys does anyone know if has an AI voice of a pen writing sfx like this? like a toothbrush ai stuff
is there something better then loveable for free to build website apps?
is RVC GUI safe / a tool I should use?
wich one do i use
i use my mic in input
do i use microphone array?
hmm idk which one ur mic is
i rly have no idea about stuff like that haha
!howtoask
- Check Docs & Guides: Your answer may already be in the AI Hub Docs or the https://discord.com/channels/1159260121998827560/1159513888199540817 channel.
- Search the https://discord.com/channels/1159260121998827560/1192011222023950368 : Look for existing posts that solve your issue. Do not invade someone else's post.
Tell your:
- Full GPU Name: (e.g.,
NVIDIA RTX 4060 8gb vram desktop) - Operating System: (e.g.,
Windows 11) - Detailed Description: What were you trying to do and what went wrong?
- Tutorial Used: Link to the guide you were following.
- Screenshot: A picture of the full error message.
To maintain a legal, safe & ethical community, we will NOT provide help for:
- ANY illegal activities.
- NSFW/Porn.
Requests for these topics may be ignored, not helped and result in moderation action.
- Be Polite & Patient: Our helpers are volunteers. You may ping the
Helpersrole once. - English Only: Please keep all conversations in English.
- Don't Ask To Ask.
!howtoask
- Check Docs & Guides: Your answer may already be in the AI Hub Docs or the https://discord.com/channels/1159260121998827560/1159513888199540817 channel.
- Search the https://discord.com/channels/1159260121998827560/1192011222023950368 : Look for existing posts that solve your issue. Do not invade someone else's post.
Tell your:
- Full GPU Name: (e.g.,
NVIDIA RTX 4060 8gb vram desktop) - Operating System: (e.g.,
Windows 11) - Detailed Description: What were you trying to do and what went wrong?
- Tutorial Used: Link to the guide you were following.
- Screenshot: A picture of the full error message.
To maintain a legal, safe & ethical community, we will NOT provide help for:
- ANY illegal activities.
- NSFW/Porn.
Requests for these topics may be ignored, not helped and result in moderation action.
- Be Polite & Patient: Our helpers are volunteers. You may ping the
Helpersrole once. - English Only: Please keep all conversations in English.
- Don't Ask To Ask.
that's an over 1 year old version of original wokada, don't trust youtube/video tuts for rvc realtime voice changers. and also CUDA is for Nvidia GPUs, not AMD GPUs
I'm guessing you're atleast on windows 10 or 11
-realtime
Guides for Programs that use RVC Models in Realtime for Calls/Games
A Realtime Voice Changer with similar performance to Wokada Deiteris Fork, with extra features, but supported only for Nvidia GPUs on Windows. and without cloud options GUIDE
A personal fork (modified version) of Wokada Deiteris Fork, it just adds some Quality of Life improvements to it like supporting Spin Embedder and Audio Effects. Don't expect too much about it since the creator made it originally as a personal project. GUIDE
Most suggested WebUI with the best general support for many platforms. GUIDE
For Windows Nvidia, Both Wokada Deiteris fork and Vonovox have similar performance & quality. Users should read the pros and cons for both and choose based on their differences, such as UI and Vonovox's paid effects.
Read Wokada Deiteris Fork Pros&Cons & Vonovox Pros&Cons
These options are not recommended for use.
Not suggested, older versions in youtube tuts are even way worse. GUIDE
The program is worse compared to the ones above, and much less updated. GUIDE
try wokada tg-develop fork instead
how do i download the models?
Hello- I had some issue with the client and server part of the AMD-DML fork in the RVC program, where I was first using it just fine, then suddenly one day audio through client stopped working, I can still use it through server, but without voice supression it sucks since my every breath can be heard.
I tried the Wokada Deiteris Fork and I had the same issue
Is this where help questions about www.DreamTavern.ai can be asked?
An issue on DreamTavern.ai... I can't find my 'streak' of writing anymore. It is the number of days I've written in a row, and I was coming close to a year of consistent writing . The number no longer appears, suggesting the streak is broken, but not by me. It shows I wrote yesterday but the streak count is gone. I understand it sounds trivial but building the streak kept me writing every day. No idea who or where to ask about this. It is one of the reasons I use the site, www.DreamTavern.ai.
I'm thinking why is the noise suppression even have to be bundled with the voice changer... I use an external one that's always enabled for the mic (voicechanger or not).
hadn't made a voice model for a year, is it the same? remove background noise, isolate vocals and into a .wav?
making my own dataset, whats the best background noise remover?
i did the tutorial and evberything worked fine but when i hearmyself talk i can hear that it keeps like cutting out is that just on my end or can they hear it cut out too and if so how do i fix that
i7-14700f
??
Yes it's safe, are you speaking about the realtime voice changer or about training an ai voice
If that is Intel you can't run any ai voice changers or training locally
Use gabox's vocal fv4, and as well as using applio to train your models
It comes with it for ppl who don't have noise suppression on their mic already
Yuck, go away catfisher
Yeah, I actually tried to fix that by using an external noise supression, but- Server works directly by skipping all software and uses the hardware, so I actually wasted half an hour downloading an external noise supression.
Still need help with this:
Hello- I had some issue with the client and server part of the AMD-DML fork in the RVC program, where I was first using it just fine, then suddenly one day audio through client stopped working, I can still use it through server, but without voice supression it sucks since my every breath can be heard.
I tried the Wokada Deiteris Fork and I had the same issue
Hey so I was training a model and I kept seeing "new best epoch is __" and I reached 400 epocs.. does that mean the model can and should be trained more or do I stop?
yo the voice is so laggy like it keeps cutting out
I would look at the graphs on Tensorboard, and also listen to a few of the checkpoints by ear.
Try turning up the chunk/block size.
is the buf being high bad? becuase changing the chunk didint work
Last week I didn't need to put up the Chunk higher than "150". Is there something going on? I am on AMD
thanks!
You turned the chunk up? When you refer to buffer, you're talking about the performance in the top left box, right?
i tried turning it up and down and both didint work and yes
What did you mean by it not working? What in particular isn't working?
like it didint fix the voice cutting out
What's your extra at?
whats that
It's the meter underneath chunk.
131072
How did you install W-Okada? Did you follow an old YouTube tutorial or the guides in the server?
You have a super old version of the program. I recommend installing one of the newer apps.
ohhh
Check out this page for more information on that.
Last update: September 6, 2025
This is most recommended for beginners.
And overall easy to use.
okay and it works for amd?
Yeah, just make sure you download the AMD version and not any other version.
Keep in mind that many of these programs aren't optimized as well for AMD as they are for NVIDIA.
you should have first asked what voice changer version and the gpu spec he was using
Could tell it was an old version just by his description.
As for the GPU specs, I suppose I should have asked what it was specifically. However, the Deiteris fork is more optimized than the older versions, so if his specs are good enough to run a voice changer, he'll probably have a better time running that either way.
turn off overtraining detector, it's not really helpful and you still have to test the model in several different epochs besides the tensorboard graphs
I use a translator. My language is Komi.
Does anyone know about TTS technology? I use the program "Balabolka". But it's very old.
Is there a program that would allow me to easily and easily produce text in real time into a microphone via a virtual cable, and could use some RVC voices for more originality and a more lively sound?
I've heard good things about GPT-SoVITS.
Does it have the features described above? Does it put a heavy load on the system?
you'd have to cope with the TTS delay in "realtime" purposes
Obviously, there will be a delay, since text isn't written instantly. But I need it so that, say, I write a sentence, press "Enter," and the sound immediately starts playing in the virtual audio cable
typing also takes more time than simply speaking
so much more delay than the RVC STS
https://github.com/RVC-Boss/GPT-SoVITS/
Check out this repo. It might suit your purposes. Not entirely sure, though.
I have a microphone from a military control room and all my models still produce unpleasant "Artifacts", which makes my hearing very bad when I use the RVC
I think I've used it before, but it doesn't seem to have the functionality to send sound directly to a virtual audio cable. You have to generate the sound, save it, load it into a soundpad, and then do it all, which is time-consuming. It's difficult to implement in a game
again he meant for realtime, that wouldn't be feasible to use
Wasn't sure. Never used it. Just trying to be helpful.
you can't be helpful unless you have enough knowledge on it
Welp, alright. I had heard good things about SoVITS, so I just wanted to mention it as an option.
guys what is this option? "voice texture"
I haven't seen you offer an option other than "use a voice changer instead of TTS" yet.
That's not exactly helpful either.
again and again, he meant for realtime use, gpt sovits doesnt suit for it
also the "realtime" TTS is not really realtime that much
idk what that "new best epoch" is, it's not the best epoch just go off what the model sounds like each time it saves
don't mess with any of those settings there
I can send a video to more accurately demonstrate what I am describing. But in this chat you can not send video
thx ❤️
before considering to get help, make sure you have been using a recommended version of voice changer as well as the correct settings in following guides
-rt
Guides for Programs that use RVC Models in Realtime for Calls/Games
A Realtime Voice Changer with similar performance to Wokada Deiteris Fork, with extra features, but supported only for Nvidia GPUs on Windows. and without cloud options GUIDE
A personal fork (modified version) of Wokada Deiteris Fork, it just adds some Quality of Life improvements to it like supporting Spin Embedder and Audio Effects. Don't expect too much about it since the creator made it originally as a personal project. GUIDE
Most suggested WebUI with the best general support for many platforms. GUIDE
For Windows Nvidia, Both Wokada Deiteris fork and Vonovox have similar performance & quality. Users should read the pros and cons for both and choose based on their differences, such as UI and Vonovox's paid effects.
Read Wokada Deiteris Fork Pros&Cons & Vonovox Pros&Cons
These options are not recommended for use.
Not suggested, older versions in youtube tuts are even way worse. GUIDE
The program is worse compared to the ones above, and much less updated. GUIDE
otherwise post your thread in #1192011222023950368
My topic is so specific that it is unclear which guidelines I should follow.
Аттьо
I dont speak russian, only english here
Its komi 
My republic is in the federation, but few people know about it
This should help. It's a list of the best AI TTS tools in the AI Hub docs. https://docs.aihub.gg/tts/tts-tools/
Last update: Dec 12, 2024
yo, my language is turkish but moslt ai voices are english and they dont work very well how could i fix that
What is the most.. Realistic... Female voice changer y'all know about...
shoo you catfisher, get outta here
What the hell kinda insult
Local worm I see u on like every voice model
If anyone's a catfisher
It's u
Like?