the output window looks like this
#✨│ai-help
latent knoll
ok
simple ore
and it shows preprocess completed and it loaded 30 seconds of audio - all good
if tha number of seconds matches your dataset, you go to the next step
latent knoll
bytes means megabytes???
@simple ore is there a chance i can send you the audio files to train the model and youll train it for me?🙏
latent knoll
<@155030383648440320> is there a chance i can send you the audio files to train ...
or you @hallow thistle
hallow thistle
bytes means megabytes???
Byte simply means byte unit; if it doesn't say MB or GB, it's not megabyte or gigabyte.
latent knoll
or you <@561807329977434122>
can you?
simple ore
i'm training different things right now and electricity is not cheap
hallow thistle
<@155030383648440320> is there a chance i can send you the audio files to train ...
As of now, I don't train a voice model. To request a "model maker" to train one completely for you, there's #1159289738314919936.
What is your PC GPU? And what is this program or which W-Okada version?
drowsy sorrel
is it possible to get a stable real time voice changer for amd gpu
hallow thistle
is it possible to get a stable real time voice changer for amd gpu
There is Tg Develop W-Okada fork DirectML.
drowsy sorrel
There is Tg Develop W-Okada fork DirectML.
well I tried that But in real-time I’m still getting weird echo/ghost sounds not normal echo but random distorted audio that I didn’t actually say. even when I’m not speaking at all, some sounds still come through.
hallow thistle
well I tried that But in real-time I’m still getting weird echo/ghost sounds not...
What is your PC GPU?
!howtoask
drowsy sorrel
What is your PC GPU?
rx 6600
hallow thistle
rx 6600
Chunk: around 80 ms
Extra: 2.7 s
GPU: AMD Radeon RX 6600
Pitch detection: rmvpe_onnx
drowsy sorrel
Chunk: around 80 ms
Extra: **2.7 s**
GPU: AMD Radeon RX 6600
Pitch detection: rm...
Okay so the sound much stable than before and I tried some random models but idk why its my voice or what but it still sounds very robotic
sullen sinew
NVIDIA RTX 4060 TI 8gb vram pc
Windows 11
I need help with my voice, I tested all official models and used some others, but my voice sounds really robotic (W-Okada) (Basically, the same problem of Brave)
hallow thistle
NVIDIA RTX 4060 TI 8gb vram pc
Windows 11
I need help with my voice, I tested al...
Are you using Tg Develop W-Okada fork? If not, better try one. https://docs.aihub.gg/realtime-voice-changer/local/tg-develops-w-okada-fork/
sullen sinew
Alright, thank you for now, but can I return if it still sounds robotic?
hallow thistle
Alright, thank you for now, but can I return if it still sounds robotic?
Use this settings in your Tg Develop W-Okada:
Chunk: around 60 - 80 ms
Extra: 2.7 s
GPU: NVIDIA GeForce RTX 4060 Ti
Pitch detection: rmvpe
Input: microphone
Output: Line 1 (Virtual Audio Cable)
Monitor: optional, set this to your speakers/headphones to hear the program.
sullen sinew
Okay, thank you again
steep bison
Hey, sorry to bother you, I did everything according to the tutorial but I'm getting no sound output.
odd roost
can anyone help me?????
wanton oxide
what is a good ai for text to speech where you can use the .index and .pth models
viral mason
Why not just use it live or ecord your voice and out the ai over it
wanton oxide
question for me?
viral mason
question for me?
Yea
low shard
Hey, sorry to bother you, I did everything according to the tutorial but I'm get...
hello, this is a general ai server, please elaborate
!howtoask
runic ermine
my voicechanger is a bit laggy and ive changed almost every setting. can anybody help me ?
steep bison
hello, this is a general ai server, please elaborate
my bad thx
Full GPU Name: Radeon RX 7800 xt
Operating System: Window 11
Detailed Description: I've fixed my old problem, but I have a new one: I can't select "Echo Cancellation" or "Noise Suppression" in RCV Voice Changer.
Tutorial Used: https://docs.aihub.gg/realtime-voice-changer/local/tg-develops-w-okada-fork/
Screenshot: i can't post picture i think i don't have permission
radiant cypress
btw is there a way to send code on dc
without it missing up the code
one of friend is teaching me c
soo
yea
viral mason
What gpu is good enough for applio?
applio can be used on a browser as well as on ur pc locally
idk what is needed for minimum tho gpu wise
low shard
Full GPU Name: Radeon RX 7800 xt
Operating System: Window 11
Detailed Descriptio...
are you using server mode perhaps?
low shard
one of friend is teaching me c
either send the files, or use discord's formatting codeblock like:
print("Hello World!")
thanks
viral mason
are you using server mode perhaps?
I would assume they're using server mode as client mode has been reported on both wokada tg fork and deiteris even for me to just not work anymore
frigid star
Hi, I would like to get a consultation.
Russian Russian models I tried to use, mostly Russian, because I speak Russian and I assume that this should make it more realistic.
The problem is as follows:
When changing the voice in real time, random words are eaten if you speak quickly. Very often, certain letters, r, w, h, n and others are not pronounced. In general, speech and voice using even Russian models are always very far from the minimum acceptable level.
I use Tg Develop.
GPU - RTX 3060
I've read all the guides a thousand times, tried all the settings. The result has always been very far from "realism"
And yes, I'm writing this through a translator.
Thanks in advance for any answers.
simple lichen
quick and simple question, how do i delete a voice model in the slot config?
nvm, i found this https://github.com/w-okada/voice-changer/issues/440
viral mason
nvm, i found this https://github.com/w-okada/voice-changer/issues/440
this looks old
viral mason
<@1023278814752677918> hey what was the voice changer thing you said was better ...
all of them are, but what gpu do u have
there's 3 realtime voice changers (wokada)
depending on what gpu ur pc has will depend on which one u should use
red loom
all of them are, but what gpu do u have
It’s for a friend, you mentioned it the last time we talked
One that uses ai voice modules and it’s free
My computer doesn’t have enough space for it
viral mason
One that uses ai voice modules and it’s free
They all do that and are free, what gpu does your friend have?
red loom
They all do that and are free, what gpu does your friend have?
One sec
red loom
They all do that and are free, what gpu does your friend have?
Sorry it took so long, here
Nvidia GeForce GTX 1660 Super
viral mason
Sorry it took so long, here
Nvidia GeForce GTX 1660 Super
It's ok, your friend will be able to use wokada deiteris but it won't run too well in games unless the graphics are super low
Should work fine on discord calls and stuff tho
It's the third guide
-rt
red loom
It's ok, your friend will be able to use wokada deiteris but it won't run too we...
Where would I download it?
viral mason
It's the blue guide button right there next to wokada deiteris fork
faint schooner
Question, for Lightning Ai i keep getting this popup saying [SIO] reconnection failed error: xhr post error. Any fixes?
viral mason
Question, for Lightning Ai i keep getting this popup saying [SIO] reconnection f...
watcha using lightning ai for?
faint schooner
watcha using lightning ai for?
For deiteris w okada fork
steep bison
are you using server mode perhaps?
Ah yes, indeed, my bad.
upper dock
is there a way to use the rvc in other apps like discord and not the browser 🤔
viral mason
what I do
stable kestrel
Sup guys, i'm running an Nvidia RTX 3090 with Windows 10 and i've been using the newest Applio 3.6.0 version for my trainings(locally) and alot of the models i've trained look like trial and error even when using the same exact dataset, i wonder what i can do to have better and consistent results. Anyone willing to help? I haven't followed any tutorials, just text guides on how to prepare a dataset.
alpine lotus
i get this error when i try to install applio on kaggle
No solution found when resolving dependencies:
╰─▶ Because faiss-cpu==1.7.3 has no wheels with a matching Python ABI tag
(e.g., cp312) and you require faiss-cpu==1.7.3, we can conclude that
your requirements are unsatisfiable.
hint: You require CPython 3.12 (`cp312`), but we only found wheels
for `faiss-cpu` (v1.7.3) with the following Python ABI tags: `cp37m`,
`cp38`, `cp39`, `cp310`, `cp311`
any help fixing it?
bc last time it worked fine
viral mason
i get this error when i try to install applio on kaggle
No solution found when...
same just started
red loom
@lone ruin here
nocturne mural
i get this error when i try to install applio on kaggle
No solution found when...
uhhh
I suppose the same method as in Colab will be required: forcing the Python version.
alpine lotus
but then again applio on colab works just fine
nocturne mural
but then again applio on colab works just fine
I meant that the Python version isn't forced in the Kaggle notebook, but it is in Colab (if you don't understand me, it's the translator's fault).
alpine lotus
ohhh
nocturne mural
!apt update -y
!apt install -y python3.11 python3.11-distutils python3.11-dev portaudio19-dev
!update-alternatives --install /usr/bin/python3 python3 /usr/bin/python3.11 2
!update-alternatives --set python3 /usr/bin/python3.11
from sys import path
path.append('/usr/local/lib/python3.11/dist-packages')
yeah it's just kaggle's very annoying to use
for me
have to make sure everything went smoothly because i recently just lost all data of the model i was training
alpine lotus
```# Install older python
!apt update -y
!apt install -y python3.11 python3.11-d...
kaggle's literally looks like this
alpine lotus
i had that on
nocturne mural
have to make sure everything went smoothly because i recently just lost all data...
Ah yes, that happens if you refresh the page. You could say you lose the storage "state." It is fixed if you try to execute a cell and then stop it; that will stop the training but save your files
alpine lotus
it's just i forgot to save the version of the notebook i was training the model on
viral mason
u got unlucky then 
alpine lotus
ykw i'm gonna start over while also trying to fix the current problem i have rn
pasting this onto it solved the issue
!apt install -y python3.11 python3.11-distutils python3.11-dev portaudio19-dev psmisc
!update-alternatives --install /usr/bin/python3 python3 /usr/bin/python3.11 2
!update-alternatives --set python3 /usr/bin/python3.11
viral mason
pasting this onto it solved the issue
!apt install -y python3.11 python3.11-dis...
where'd ya paste it
alpine lotus
kaggle's literally looks like this
here
just replace whatever this is with the one above
viral mason
I need a screenshot of exactly where bc I'm slow
nocturne mural
!uv pip install -q -r /kaggle/working/program_ml/requirements.txt --extra-index-url https://download.pytorch.org/whl/cu128 --index-strategy unsafe-best-match --system
%cd /kaggle/working/program_ml```->
viral mason
my eyes aughhh
nocturne mural
my eyes aughhh
from IPython.display import clear_output
rot_47 = lambda encoded_text: "".join(
[
(
chr(
(ord(c) - (ord("a") if c.islower() else ord("A")) - 47) % 26
+ (ord("a") if c.islower() else ord("A"))
)
if c.isalpha()
else c
)
for c in encoded_text
]
)
new_name = rot_47("kmjbmvh_hg")
findme = rot_47(codecs.decode("pbbxa://oqbpcj.kwu/Dqlitvb/qurwg-mtnqvlmz.oqb", "rot_13"))
uioawhd = rot_47(codecs.decode("pbbxa://oqbpcj.kwu/QIPqaxivw/Ixxtqw.oqb", "rot_13"))
!git clone --depth 1 $uioawhd $new_name --branch 3.6.0
clear_output()
!pip install uv
!apt update -y
!apt install -y python3.11 python3.11-distutils python3.11-dev portaudio19-dev psmisc
!update-alternatives --install /usr/bin/python3 python3 /usr/bin/python3.11 2
!update-alternatives --set python3 /usr/bin/python3.11
!uv pip install -q -r /kaggle/working/program_ml/requirements.txt --extra-index-url https://download.pytorch.org/whl/cu128 --index-strategy unsafe-best-match --system
%cd /kaggle/working/program_ml
!python core.py "prerequisites" --models "True" --exe "True" --pretraineds_hifigan "True" > /dev/null 2>&1
!sudo curl -fsSL https://raw.githubusercontent.com/filebrowser/get/master/get.sh | sudo bash
!filebrowser config init
!filebrowser config set --auth.method=noauth
!filebrowser users add "applio" "applio123456" --perm.admin
clear_output()
print("Finished")```
crystal mountain
applio can be used on a browser as well as on ur pc locally
I cant open applio on google cola mb for some reason
idk if i should buy google colab pro
viral mason
I cant open applio on google cola mb for some reason
idk if i should buy google...
ew no paying
go to kaggle
crystal mountain
Oh it has a kaggle?
crystal mountain
ew no paying
How do i access the kaggle
viral mason
crystal mountain
https://www.kaggle.com
No i mean kaggle for applio
viral mason
viral mason
@alpine lotus what browser should I use then that comes prebuilt with a good ad blocker
alpine lotus
<@817243603801473024> what browser should I use then that comes prebuilt with a ...
any browser could work + ublock origin
nocturne mural
any browser could work + ublock origin
Yes yes
knotty moth
any browser could work + ublock origin
chrome only allows ublock lite due to manifest v3 migration
terse halo
I can't hear myself with the AI voice.
knotty moth
I can't hear myself with the AI voice.
!howtoask
tame oracle
!howtoask
fierce crown
Does anyone know which AI is used for ai Cat dancing videos?
latent knoll
what does this mean
jaunty coyote
sos, i need help, is there anybody can support me? i don't know how to use w okada voice changer😭
cyan mantle
Hi ! I want to run Tensorboard but the console says "the path is not found", can you help please ?
I've put it inside the folder "Applio"
nocturne mural
-rvc
glass igloo
Hello, friends! I haven't worked with speech conversion models in a long time, and the last time I did, Apollo was the best option. Now I need to create a model, and I decided to ask if there is anything better or if Apollo is still relevant.
cyan mantle
Hi again, I can't find my index files, it says "succesfully created" but I can't find it on logs, can someone knows why ?
prisma merlin
My ms is so completly high & the voice is delayed, im in the right version for my gpu so im confused why
runic ermine
can somebody help me ? My ref time is at 12k and the voices are very laggy
prisma merlin
can somebody help me ? My ref time is at 12k and the voices are very laggy
literally my same problem
my res is at 30k ms
runic ermine
my res is at 30k ms
yeah mine sometimes too, but i cant find anywere a solution
prisma merlin
im on the right gpu too and i reduced the index extra and chunk but it hasnt seemed to work
i put everything on the lowest to minimize performance and its still terrible
viral mason
i put everything on the lowest to minimize performance and its still terrible
What gpu do u have?
Same for u @runic ermine
runic ermine
Same for u <@675702816072794112>
Amd RX 7700 XT
viral mason
It's the second guide
runic ermine
It's the second guide
okay thanks and this should help to 100% ?
viral mason
viral mason
okay thanks and this should help to 100% ?
Not sure, I'm just giving you the best voice changer for ur gpu to see if it helps
runic ermine
You'd need
https://github.com/tg-develop/voice-changer/releases/download/b2397/v...
ah okay, is there any type of tut video ? or just this guide ?
viral mason
ah okay, is there any type of tut video ? or just this guide ?
Just the guide, but all u do for this one is just run mmvcserversio
It launches in browser
runic ermine
okay, the current one i use is this one MMVCServerSIO_win_onnxdirectML-cuda_v.1.5.3.18a.zip
runic ermine
Just the guide, but all u do for this one is just run mmvcserversio
is there any differnce ?
viral mason
okay, the current one i use is this one MMVCServerSIO_win_onnxdirectML-cuda_v.1....
Oh yeah that's uh, old
viral mason
Yea it's not brand new but it's much more optimized and up to date code wise and has extra features like voice effects u can apply for free
Pretty cool stuff
runic ermine
ah okay nice thanks. This guide seems complicated but ill try
runic ermine
You'd need
https://github.com/tg-develop/voice-changer/releases/download/b2397/v...
but what is this for ?
viral mason
ah okay nice thanks. This guide seems complicated but ill try
Like I said all u gotta do is just download the two things, the voice changer and vac lite (just vb cable but better for windows)
And open mmvcserversio
For the voice changer
Then for vac lite run setupx64
Not complicated at all
runic ermine
okay ill try and ill text u if i need help yeah ?
viral mason
okay ill try and ill text u if i need help yeah ?
Sure! I won't be too busy today
wispy glen
where can i download this?
brittle wing
how do i know when its good?
chatgpt is saying that the sound or it wont be any better or good
gilded barn
-rt
gilded barn
-dt
midnight bloom
can I get help about\ [Voice Changer] Pipeline is not initialized.
[Voice Changer] Waiting generate pipeline...
viral mason
how do i know when its good?
Test each time it saves an epoch like if u have it where it saves every 10 test around like 50-170
Same for save every 5
knotty loom
Test each time it saves an epoch like if u have it where it saves every 10 test ...
hey i followed every step but no voice is coming out my mic and even though the app
viral mason
hey i followed every step but no voice is coming out my mic and even though the ...
What steps?
If u got the voice changer off a YouTube video it's outdated no matter if the video is like a couple months old
What gpu do u have
fiery wing
any idea on how to fix high "res" delay in the voice changer
fiery wing
What gpu do u have
are you able to let me know whats best for my gpu
fiery wing
You can check it using task manager if u dunno what it is
i believe that is my gpu right?
viral mason
I wouldn't know for sure what gpu ur pc has in particular but if that's the one u have I'd say use wokada tg fork
viral mason
You'd need
https://github.com/tg-develop/voice-changer/releases/download/b2397/v...
Just download these two, the voice changer, and vac lite
fiery wing
Just download these two, the voice changer, and vac lite
ok perfect, do i need to uninstall the current version of the voice changer i have right now?
fiery wing
I'd recommend it tho
perfect thanks so much
viral mason
You're welcome! If u have any questions just ask me or some of the mods or hekpers
viral mason
i used a vid in june my gpu is 7800xt
Yeah any yt video uses outdated voice changers, you should use wokada tg fork the download is here
#✨│ai-help message
knotty loom
Yeah any yt video uses outdated voice changers, you should use wokada tg fork th...
how do i set it up
i downloaded both
viral mason
For vac lite run setupx64
And wokada tg fork run mmvcserversio
knotty loom
For vac lite run setupx64
And wokada tg fork run mmvcserversio
i cant seem to find anything named wokada
fiery wing
The voice changer you downloaded
do you reccomend any settings within the voice changer?
viral mason
i dowloaded 2 things mmcserveriso and vac
Mmvcserversio is an exe file you're supposed to run within the wokada tg fork folder, and for vac lite did you run setupx64
knotty loom
Mmvcserversio is an exe file you're supposed to run within the wokada tg fork fo...
yes iv done both and im in a website
viral mason
do you reccomend any settings within the voice changer?
Tbh it depends on your PC itself, for chunk I'd say somewhere in the 300s is ok for amd and extra time at 2.7
knotty loom
Tbh it depends on your PC itself, for chunk I'd say somewhere in the 300s is ok ...
the website one is real time yeah sorry i was confused
viral mason
the website one is real time yeah sorry i was confused
That's ok, I've got work soon so idk if I'll be able to keep helping rn
knotty loom
That's ok, I've got work soon so idk if I'll be able to keep helping rn
oh ok all good
is it the same steps for the model?
viral mason
To put a model into it just press the plus button on the left side, download them from herehttps://discord.com/channels/1159260121998827560/1175430844685484042
knotty loom
yup i did it thanks
viral mason
You're welcome
knotty loom
is this real time like in discord
viral mason
Yea it should work on discord if you have the mic settings set up correctly
knotty loom
Yea it should work on discord if you have the mic settings set up correctly
whaat should be my outpuy
viral mason
voice changer:
input: headset/headphone microphone
output: line 1
other programs:
input: line 1
output: headphones
knotty loom
voice changer:
input: headset/headphone microphone
output: line 1
other program...
i went to discord and it keep spamming the same thing
viral mason
so how do i know its working
There's a thing on the right side that would with be green, red, or yellow
When it's running
If it's green it's good
If not change your settings around till it's good
knotty loom
If it's green it's good
are u reffring to performnce stats
viral mason
I believe so
viral mason
There's a way to hear yourself with vac lite but I cannot remember as I don't use the normal setup most people use
knotty loom
There's a way to hear yourself with vac lite but I cannot remember as I don't us...
alr but when im in discord i connects to line 1 and nothing is being played
viral mason
Have you pressed start server in the real-time voice changer?
knotty loom
yes and it keep reapting the same thing over and over for some reason
knotty loom
Have you pressed start server in the real-time voice changer?
ive got it to work but its verry laggy
knotty loom
Try messing with the chunk and extra time
what should be my audio driver
hallow thistle
what should be my audio driver
Virtual Audio Cable lite or Realtek HD Audio?
Intel/Realtek High Definition Audio is an integrated sound card; Virtual Audio Cable is a third-party software that gives virtual audio line and works similar to "Stereo Mix". When you mention "audio driver", it sounds more like the HD Audio. 
solar jetty
How do I fix Failed to Load rmvpe. fallback to rmvpe_onnx (using deiteris kaggle)
warm mason
what are ephochs
digital saddle
hey i have a lot of lags with voices idk how to fix it
terse halo
Hey, I have a question. My friend downloaded something from this site, is it trustworthy?
https://huggingface.co/Shadicti/deiteris-Fork/blob/main/voice-changer-windows-nvidia-b2332.zip
distant hamlet
what are ephochs
number of epochs = number of times the model looked thru your dataset (audio)
minor tundra
DOES ANYONE HAVE GOOD SETTINNGS FOR A GIRL
misty hornet
anyone know any tips to clear up the voice glitching some or having the audio randomly chirp? The microphone is good and my mic is quiet, if there are settings to fix this can someone help?
frank plume
tiny shard
why can i hear m y on voice on the voicechanger but the one i put
iron widget
hey
So I download this model
" Deep male voice " or whatever it.was uploaded by razer I think?
The voice always come so bad and different from the Sample he sent I don't really understand the settings of it
Like
Batch Size:** 6
Dataset:** 30 Minutes
Hop Length:** 32
Pretrain:** Titan
Sample Rate:** 32k
Like are these?
I can't really see then on
applio
glacial grotto
You'd need
https://github.com/tg-develop/voice-changer/releases/download/b2397/v...
@viral mason i want nvidia one
do i use vonovox?
knotty loom
anyone know how to use the rvc voice changer
viral mason
anyone know how to use the rvc voice changer
didn't I already help u?
viral mason
do i use vonovox?
ye
knotty loom
didn't I already help u?
yes but nothing is coming out of my mic
brittle wing
what are the most accurate girl voices?
viral mason
why are u like this 
viral mason
yes but nothing is coming out of my mic
do u have the settings right?
knotty loom
do u have the settings right?
yes exactly
stray copper
my fucking cmd is acting up
im not the admin on this pc btw
rtx 4060 ti
stray copper
my fucking cmd is acting up
im not the admin on this pc btw
rtx 4060 ti
@simple ore
ok i got it to work
i just had to put in
py -m pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118
knotty loom
do u have the settings right?
hey i made it work but not it just sound like a robot instead of girl
simple lichen
this looks old
Oh, what’s the newer method to my question then?
viral mason
hey i made it work but not it just sound like a robot instead of girl
why u being a girl
knotty loom
why u being a girl <:cat_sadcat:1359900998587650219>
lol its really robit sounding
amber merlin
I downloaded w okada and im trying to figure it out, just unsure about it
brittle wing
why does the voice get mechanic or u can hear it´s ai sometimes? how can I fix that
arctic monolith
Sorry to Interrupt, But I have another request, Can you do Elmer Fudd voice by Billy West please, once again, I found on weights the Looney Tunes Back in Action video game for PS2, but it doesn’t sound good enough for the actual PS2 and GCN version, It’s sound like a DS version. So can you do it with the RVC please. Here’s the video and photo that I showed you https://share.icloud.com/photos/0fbAgqlVIY3jexXB5TzRWCRbA https://share.icloud.com/photos/072eIwsjQlL2IqWcEhojGzyHw
brittle wing
what is the best rvc to use?
iron widget
Nobody helped me 
quasi condor
my fucking cmd is acting up
im not the admin on this pc btw
rtx 4060 ti
why do u have a photo of epstein 😭
balmy rain
my fucking cmd is acting up
im not the admin on this pc btw
rtx 4060 ti
epstein.jpg
silent stratus
i got this output on 50 epochs using refinegan does anyone know the reason for this? im guessing this is a pretty common issue because i did the usual setup for training a model but i clicked refinegan
im guessing it has to do with the pretrain but im not sure
stray copper
i got this output on 50 epochs using refinegan does anyone know the reason for t...
set the embedder when you’re converting the audio to spin v2 instead of content vec in advanced settings
in the inference
quasi condor
<@1019743420786421851> <@813786967280058378>
i was asking why but okay
silent stratus
set the embedder when you’re converting the audio to spin v2 instead of content ...
still sounds the same
stray copper
show me your settings
silent stratus
show me your settings
wait i just checked and they only have the 32k pretrain for refinegan and not the 40k
that might be it
stray copper
lol
quasi condor
i got this output on 50 epochs using refinegan does anyone know the reason for t...
why are u still using refinegan
that shi old
stray copper
refinegan sucks ass
quasi condor
refinegan sucks ass
i mean its not that bad
silent stratus
so is it better to just use hifi and spinv2 then
stray copper
ofc
silent stratus
ofc
if hifi is still objectivley the best why does applio say "HiFi-GAN: Default option, compatible with all clients.
MRF HiFi-GAN: Higher fidelity, Applio-only.
RefineGAN: Superior audio quality, Applio-only."
stray copper
don’t listen to it
155 is og and 150 is refinegan
og has better pronunciation and can yell better but refinegan has better quality
silent stratus
hmmm
stray copper
i like og more
silent stratus
i got an official acapella of evil jordan at 40k and its a pretty whispery song
stray copper
use og
silent stratus
okay will do
stray copper
happy to help
silent stratus
lowkey ill do both at the same time
balmy rain
<@1019743420786421851> <@813786967280058378>
Thanks bro
flat idol
bro why mine a bit too delay for like 2 sec im using von models
knotty moth
bro why mine a bit too delay for like 2 sec im using von models
please first mention which voice changer version you're using and your GPU spec or colab/kaggle notebook you're using
flat idol
please first mention which voice changer version you're using and your GPU spec ...
AMD rtx and king von or juice wrld
flat idol
please first mention which voice changer version you're using and your GPU spec ...
idk what is kaggle
new to this thing
knotty moth
AMD rtx and king von or juice wrld
voice changer version not your juice wrld model
flat idol
**voice changer version** not your juice wrld model
oh how do ik if it supports me
knotty moth
AMD rtx and king von or juice wrld
also please be more specific, like RX 6600 or RTX 3060?
flat idol
also please be more specific, like RX 6600 or RTX 3060?
6600
flat idol
is that exactly yours or perhaps higher?
mine wait lemme double check i
no doubt i js check it, it is rs 6600
knotty moth
oh how do ik if it supports me
I still don't know which version you're using, but I'd recommend tg-develop fork as listed in the following
-rt
knotty moth
the deiteris fork would still work for you despite being a bit older, but still better than the original wokada
flat idol
oky ty
half flame
What rvc model would you recommend me for the natural feminine voice? 🙂
Not looking for some anime or 2 minute dataset training.
Quite new in this rvc world, willing to explore 🙂
I am practicing trans voice lessons, and I noticed that it works better if you make tone softer? 😅😅
Maybe a bit weird, but that's what I have noticed
half flame
Hmm. Wokada is bad?
simple schooner
how to cvonnect the voice changer with discord
quasi sparrow
what should I do if my voice doesn't appear in the game?
hallow thistle
what should I do if my voice doesn't appear in the game?
What is your PC GPU? And did you follow any tutorial or guide before here?
!howtoask
hallow thistle
Hmm. Wokada is bad?
Not all W-Okada voice changer versions are bad, though the original W-Okada version (like v.1.5.3.18a) is often not recommended because it's an old and outdated version.
hallow thistle
how to cvonnect the voice changer with discord
simple schooner
ty
hallow thistle
no doubt i js check it, it is rs 6600
More like "AMD Radeon RX 6600", an AMD GPU. When you say you're new to all things, it's best to learn along. AMD doesn't make RTX GPU; AMD makes Radeon, while NVIDIA makes their RTX GPU, although there are some very rare cases where GPU got mixed up as something like "GeForce Radeon RTX 9060 XT" from their respective manufacturer. Kaggle is a website that hosts Jupyter Notebook where you run Python-related software on their site.
stray copper
why are u like this <:cat_sadcat:1359900998587650219>
viral mason
It's the second guide
supple cloak
T-T
Ion need the guide
I've done this so many times
I main linux
I reinstall every week or so
I get how to use github
minor tundra
hi
supple cloak
T-T
supple cloak
Since u got amd just use this version
https://github.com/tg-develop/voice-chan...
This won't hel0
viral mason
Wdym
minor tundra
what should i change?
supple cloak
I need to use docker
minor tundra
Chunk set to 600 or so ms
640?
supple cloak
Also @viral mason it's cuz I'm epic
minor tundra
okay i did it
minor tundra
yes it has
viral mason
I use arch linux
U told me u use all AMD but alr, there's a Linux build I believe in here somewhere
https://github.com/tg-develop/voice-changer/releases/tag/b2397
minor tundra
Does it sound alright
uh a little hold on js really delayed like very
supple cloak
U told me u use all AMD but alr, there's a Linux build I believe in here somewhe...
Is this the main page
viral mason
Is this the main page
It is yes
minor tundra
Is this the main page
do i change anything to do with the in and out or tune?
supple cloak
do i change anything to do with the in and out or tune?
I need a hear it
viral mason
I actually had the wrong link I just changed it to the actual main page
I'm for sure this time
supple cloak
I'll check when I come back from the restaurant
supple cloak
I'm for sure this time
Okiii
supple cloak
Since u got amd just use this version
https://github.com/tg-develop/voice-chan...
This is just the straight up link to the windows x64 CPU varient wth
supple cloak
ok im logging in
What?
minor tundra
is it buns?
viral mason
This is just the straight up link to the windows x64 CPU varient wth
Wdym?
minor tundra
vram okay wait
viral mason
Wdym?
Ig once u get home u can scroll through the guide bc it has multiple different links to use for downloading the different versions of tg fork
minor tundra
What?
is this it
bro its so buns istg
so what can i do?
bru
i see okay
oh
so it has nothing to do with the sets right
minor tundra
maybe bc ur using a demo client
so what should i use then
viral mason
they all work for nvidia
supple cloak
I know
minor tundra
alr let me read this hold on
hidden sphinx
on gpu i have the options cpu, gpu0, gpu1, gpu2, gpu3, but i have no way to select my gpu, how do i fix this?
it says gpu (dml) then the options i said
my gpu is amd
supple cloak
on gpu i have the options cpu, gpu0, gpu1, gpu2, gpu3, but i have no way to sele...
How many GPUs you got
Oh
minor tundra
they all work for nvidia
so is the 4060 bad for that or is it sufficent
supple cloak
Open task manager or poor man's btop
minor tundra
?
supple cloak
so is the 4060 bad for that or is it sufficent
It works but it could be the outdated launcher
minor tundra
i see
supple cloak
Get vonovox or another one
minor tundra
Get vonovox or another one
im getting vonovox
minor tundra
Get vonovox or another one
which one?
i got both
okay
okay
1 sec its downloading rn
viral mason
which one?
for nvidia u need the 001 and 002 zip
supple cloak
for nvidia u need the 001 and 002 zip
I already said it in vc
viral mason
👍
viral mason
It's not titled Nvidia btw
ik, but it's in case some ppl aren't aware of what that means
supple cloak
With 7z
supple cloak
ik, but it's in case some ppl aren't aware of what that means
Okay but some people aren't aware that Nvidia version is called cuda
hidden sphinx
Open task manager or poor man's btop
what about now?
supple cloak
what about now?
Open task manager
viral mason
Okay but some people aren't aware that Nvidia version is called cuda
that's why I said to download those two for Nvidia lol
bc it says cuda and most ppl won't know what cuda means
supple cloak
that's why I said to download those two for Nvidia lol
You said download nvidia
It's titled cuda
How are they going to understand
R u dumb
hidden sphinx
Open task manager
i did but what do i do now
minor tundra
okay they both extracting right now @supple cloak
supple cloak
okay they both extracting right now <@835519549105242122>
They need to extract as 1
It's part 1 and part 2 which are done in the 7z menu
minor tundra
They need to extract as 1
should i screenshare im confused
viral mason
How are they going to understand
this is pretty easy to understand I think "for nvidia u need the 001 and 002 zip" I said for nvidia which means if u use nvidia get those two zip files
minor tundra
They need to extract as 1
which vc? i cant do it here
supple cloak
this is pretty easy to understand I think "for nvidia u need the 001 and 002 zip...
Yes
But if you look at the screenshot
It says cuda
hidden sphinx
Send it over
i cant send photos here tho
supple cloak
And not nvidia
supple cloak
i cant send photos here tho
Oh
viral mason
It says cuda
which is why I said which two zips to download
they're the only ones with 001 and 002
supple cloak
stray copper
whats going on
minor tundra
where is that im lowky slow i apologise
i heard
where do i add the model
OH
its loaded now
start server?
oh
okay
okay thank u btw
cya
supple cloak
whats going on
Tryna help this dude through vc
tame oracle
is this dude a moron
not everyone has a strong or even a weak understanding of computers
hidden sphinx
why is my voicechanger so laggy it doesnt record my voice pratically and when it does it comes with a bunch of cuts and the voice sounds broken
also takes a lot to give the output
i have the chunk at 832 and it takes double the time its supposed + it keeps cutting and its impossible to understand what its saying
viral mason
did u get the voice changer off a youtube tutorial?
supple cloak
...
viral mason
u should get wokada tg fork, here's the amd download
minor tundra
@viral mason can u help me
viral mason
then that'll work
viral mason
do yk how to get this to work
I do yea
viral mason
I have a 507ti nvidia card and these settings work fine with it and sound great, try them
viral mason
does it matter what model i use or
nah, only ones that won't work are any refinegan models
there's only 4 or 5 of those tho here
minor tundra
nah, only ones that won't work are any refinegan models
for input should i use my virtual cable or just mic?
normal one
viral mason
for input should i use my virtual cable or just mic?
regular headset mic for input and output as virtual cable
and opposite for using it on discord ect
virtual cable for mic in games/discord and output just whatever headphones ur using
minor tundra
virtual cable for mic in games/discord and output just whatever headphones ur us...
i see #
let me test it out one second
brittle wing
BRAZIL SERVER PLEASE
minor tundra
virtual cable for mic in games/discord and output just whatever headphones ur us...
forgot to ask u what do i put as monitor
device
viral mason
forgot to ask u what do i put as monitor
should be ur headphones
I don't use the monitor setting since I have a different setp from normal use
minor tundra
should be ur headphones
for sum reasonm i cant hear myself
minor tundra
I don't use the monitor setting since I have a different setp from normal use
bru
it dont work
can someone
help
?
gusty shuttle
is there a rvc google colab link thingy so i can make an ai cover
i cant find a website that works
viral mason
is there a rvc google colab link thingy so i can make an ai cover
u can use applio, btw kaggle is much better than collab since it gives 30 hours free
instead of like 2-4
-rvc
viral mason
no idea what that even is tbh
gusty shuttle
i have a voice recorded back then
viral mason
u would need pth and index file to make it work
nocturne mural
no idea what that even is tbh
.index files always started with the prefix “added”
brazen gorgeBOT
nocturne mural
uh
viral mason
.index files always started with the prefix “added”
used to yea
in current applio I haven't seen that
nocturne mural
in current applio I haven't seen that
in applio that prefix was removed for simplicity
gusty shuttle
how do i put my voice model to applio
i read the tutorial but didnt understand anything
i already got a voice model and a index file but i cant import it in there
craggy brook
How can I create voice dialogues with realistic voice tones? For example, how can I use a shy voice? Elevenlabs has very few character models.
unreal pumice
Heya, I have an M2 Macbook Pro and I've been wanting to use RVC. I found the original wokada but trying to figure it out has been awful. Someone linked me to this server saying i could get help for mac and wokada
inland summit
can someone tell me how to get a better version of rvc?
frigid star
Does Tg Develop have any tools to deal with the "robotization" of voice?
viral mason
Does Tg Develop have any tools to deal with the "robotization" of voice?
the wha
hallow thistle
Does Tg Develop have any tools to deal with the "robotization" of voice?
That's an awkward question. Tg Develop's W-Okada doesn't have any known filter that could fix the audio in realtime. Instead, the audio sounding robotic has to have being your settings (extra being set to 0.5 s, while the recommended extra is 2.7 s) or the voice model itself.
hallow thistle
Sorry to Interrupt, But I have another request, Can you do Elmer Fudd voice by B...
As what the other moderator warned you in #1192011222023950368, not even one time, the #1159289738314919936 is the only proper channel to request anyone to train a voice model as what you really expected. You have stated to not spamming your requests in other channels, like #1450361836468834325 message, yet you still make the same mistake while you learn nothing after all. To get a voice model, you should learn to train one by yourself, otherwise just don't complain about your requests elsewhere.
hallow thistle
i have a old asset named added_ivf591_flat thing
No idea, though I assume it's simply an RVC voice model file (pth) which would still work in any RVC fork. 
hallow thistle
for sum reasonm i cant hear myself
Try Virtual Audio Cable lite instead of VB-Cable one.
hallow thistle
BRAZIL SERVER PLEASE
Look up "AI Hub Brazil" on Google, if the Discord server exists, it exists.
By the way, @minor tundra, if you need more help about W-Okada, you can ping @ helper role, and don't always expect or force those non-helper members to help you about that.
!howtoask
hallow thistle
Heya, I have an M2 Macbook Pro and I've been wanting to use RVC. I found the ori...
RVC (Retrieval-based Voice Conversion) and W-Okada are two different programs. Tg Develop's W-Okada fork (b2364) is the only known version to work with Apple Silicon Mac; the b2377 and newer version removed support for Mac. https://github.com/tg-develop/voice-changer/releases/download/b2364/voice-changer-macos-arm64-cpu.tar.gz Applio RVC fork never has any of their version compiled for Mac, there being only Linux and Windows.
ancient trench
i need help with voice chanager i set up the aubio virtual hooked up to discord but it doesnt work
@tame oracle help me pls
tame oracle
<@1416622683163983902> help me pls
please read the help guidelines above before asking, and i can't help rn as im going to bed, im very sorry, but you can ping the helper role and wait and someone will help you
viral mason
<:PrayToGod:911331363200856075>
Goodnight fellow femboy
modern basin
<@&1159293204038955078> hey, can anyone tell me how to stop hearing myself when using the voice changer client demo? When I click start, everything works perfectly but I don't wanna hear myself
quasi condor
<@&1159293204038955078> hey, can anyone tell me how to stop hearing myself when ...
stop using monitor on okada
that's how you avoid hearing yourself
modern basin
yeah okay, i'm dumb thank you sm
modern basin
that's how you avoid hearing yourself
and if you don't mind- i hear some kind of background noises, it's very minimal but still annoying. any setting that i could tweak to mitigate that or is it just purely my mic?
quasi condor
turn on sup2 on okada
modern basin
turn on sup2 on okada
alr man thank you, that seemed to do the trick
craggy brook
modern basin
<@&1159293204038955078> kinda came across another issue, my voice plays twice cause it's taking input from the sound coming from my headphones too for some reason. So if there's music playing in the background it picks that up as well. Any help is appreciated
Actually nvm it isn't picking music but it picks up other peoples voices and my own as well and plays that
hallow thistle
i need help with voice chanager i set up the aubio virtual hooked up to discord ...
What is your PC GPU? And did you follow any tutorial video or guide before about W-Okada voice changer?
hallow thistle
<@&1159293204038955078> hey, can anyone tell me how to stop hearing myself when ...
"Voice changer client demo" sounds like an old version of W-Okada. What is your PC GPU?
modern basin
"Voice changer client demo" sounds like an old version of W-Okada. What is your ...
RTX 3050
That's what I saw on github, I'm not sure maybe I got the versions wrong
hallow thistle
RTX 3050
Try Tg Develop's W-Okada fork. https://docs.aihub.gg/realtime-voice-changer/local/tg-develops-w-okada-fork/ https://github.com/tg-develop/voice-changer/releases/tag/b2397
quasi condor
@unborn briar i can't help u but they can cuz im just a model maker not a staff
hallow thistle
<@681416611240083456> i can't help u but they can cuz im just a model maker not ...
Who is "they"? 
quasi condor
Who is "they"? <:Drem:1104354663492485192>
u and the staff 🗿
hallow thistle
@unborn briar use #✨│ai-help for help about W-Okada realtime voice changer instead of expecting a non-helper member to help you in DM.
quasi condor
<@681416611240083456> use <#1159290139609137264> for help about W-Okada realtime...
he's using the outdated one 
meager laurel
hallow thistle
https://tenor.com/view/high-five-fail-missed-high-five-high-five-fail-dumb-and-d...
Embed fail.
quasi condor
https://tenor.com/view/high-five-fail-missed-high-five-high-five-fail-dumb-and-dumber-gif-1387319524678409638 u need lvl 5
remote trellis
guys how can i clone my voice
hallow thistle
guys how can i clone my voice
!howtoask
eager adder
embed fail.
viral mason
how do u import the model into the client?
What voice changer do u have, there are 3 (and then the outdated demo client that u shouldn't be using if u got it off YouTube or smth)
ancient trench
Oh yeah I figured it out I had to do system not client I’m pretty sure
ancient trench
well i just logged back on from last night and it doesnt work at all for my mic
ashen musk
Yo! I have a subscription to Kling AI and ive been trying to generate text to video ai generated videos but even though the quality is very good, the video just seems to not be following my prompt and now im scared of using more credits. Who can help?
viral mason
does anyb know the link to the voice changer
what gpu do u have
gleaming bough
amd
viral mason
amd
take these, for the first one run mmvcservsio, for the second one run setupx64
https://github.com/tg-develop/voice-changer/releases/download/b2397/voice-changer-windows-amd64-dml.zip
and here's the guide if u need it
https://docs.aihub.gg/realtime-voice-changer/local/tg-develops-w-okada-fork/
gleaming bough
tysmm
lyric moat
what does index do in a voice model
daring raft
i am trying to use pretrain Snowie 40 or v3.1 40 and i get this every time even tho my dataset is 40khz
why is this happening?
abstract comet
Hey guys, I’ve recently downloaded RVC and trying to make my own voice model, but it doesn’t sound as great as the online models like kits, audimee etc etc
Wondering if I’ve screwed up some settings? I have 30 mins of fairly clean audio and trained it at 500 epochs
Is this too much?
viral mason
i am trying to use pretrain Snowie 40 or v3.1 40 and i get this every time even...
Snowie is a very old pretrain
daring raft
what does that mean for the process?
viral mason
Wondering if I’ve screwed up some settings? I have 30 mins of fairly clean audio...
500 is a bit much, most models sound good around 100-300
daring raft
Snowie is a very old pretrain
which russian pretrain should i use for singing (in Serbian)
abstract comet
500 is a bit much, most models sound good around 100-300
Yeah I’ve tried at 50-300 as well but the voice just doesn’t sound like me and it’s not pronouncing some things right
It sounds like it’s got a cold
I’ve uploaded the same data set to kits.ai and that worked way better
So there must be some setting or something I’m doing wrong
daring raft
So there must be some setting or something I’m doing wrong
maybe when doing intrfrireance try change index file settings
viral mason
which russian pretrain should i use for singing (in Serbian)
U should just use og pretrain, legacy core, or use refinegan if u want. Any other pretrains aren't made properly and cause harmonic distortions and issues in general with how it sounds
daring raft
U should just use og pretrain, legacy core, or use refinegan if u want. Any othe...
but how to get that Serbian/Russian accent then on those pretrains?
abstract comet
maybe when doing intrfrireance try change index file settings
Can I send you screenshots?
viral mason
I’ve uploaded the same data set to kits.ai and that worked way better
Kits.ai is a very old and honestly bad site for making ai models, have u tried applio?
daring raft
Can I send you screenshots?
yupp
abstract comet
Kits.ai is a very old and honestly bad site for making ai models, have u tried a...
No, but kits has worked fairly well
I’m using RVC because that’s supposed to be the best right?
I downloaded it from GitHub
viral mason
-rvc
viral mason
Try applio, it's where all good models come from as of recent time
It can be used online or on ur pc locally
abstract comet
Try applio, it's where all good models come from as of recent time
But aren’t all voice models using RVC ?
Like they’re all built on top of rvc
That’s what ChatGPT is telling
Me
viral mason
They use rvc yes but some things create models better than other things
Some have bad old code
Some keep it up to date
And Improve
Plus applio is completely free
abstract comet
So use aplio to train and then put into rvc?
Apologies if I’m not understanding. I haven’t slept much and have been trying to fix this all day
viral mason
Yup! The guide to it is right here if you'd like to read up on it
viral mason
Apologies if I’m not understanding. I haven’t slept much and have been trying to...
It's ok man, get some rest soon so u don't fall asleep mid training lol
abstract comet
Would you guys consider yourself experts on rvc
Training and such
I need to get this to work
And am willing to pay
Obviously I will read the guide
daring raft
U should just use og pretrain, legacy core, or use refinegan if u want. Any othe...
could you help me out, I m using Applio NoUI on colab
should i use custom pretrain at all and which one for Russian dataset that i wanna train?
daring raft
Would you guys consider yourself experts on rvc
i am not, i am still struggling all the time
viral mason
could you help me out, I m using Applio NoUI on colab
should i use custom pretr...
I'm not sure what pretrain would work better, but u should be using Kaggle to train using applio as it gives 30 hours to free users per week
daring raft
I'm not sure what pretrain would work better, but u should be using Kaggle to tr...
i tried and yeah it's better BUT it alwayes gives me error when i open op the RVC link
viral mason
That's odd
daring raft
it says something about ports and servers etc
this thing, and this also happened on every collab and other sites..
@viral mason
viral mason
this thing, and this also happened on every collab and other sites..
Did u insert your Ngrok token
daring raft
yupp
viral mason
Hmmmm
daring raft
i even tried to go to another Wifi
viral mason
Have you tried switching it to local tunnel instead of Ngrok?
viral mason
That's odd
daring raft
never had those problems year ago
viral mason
maybe try asking one of the helpers or mods, I wouldn't really be smart enough to tell u exacttly what is causing it
daring raft
maybe try asking one of the helpers or mods, I wouldn't really be smart enough t...
thanks g
daring raft
this thing, and this also happened on every collab and other sites..
@tame oracle do you maybe know ?
daring raft
this thing, and this also happened on every collab and other sites..
@wide oasis ?
pulsar tusk
i need help with my voice changer, when i test my mic the result just comes in and all i hear is the voice cutting off and it comes back for a fraction of a second but never a whole sentence idk if its my settings or smth
viral mason
i need help with my voice changer, when i test my mic the result just comes in a...
what settings do u have on the voice changer?
like chunk and extra time
pulsar tusk
what settings do u have on the voice changer?
256 chunk and 16348 chunk
pulsar tusk
are u able to send a screenshot of what program you're using?
what do you mean program
like my specs
i cant send screenshots here
pulsar tusk
u can send in dms if u like
yeah ok i sent in dms
pulsar tusk
Intel(R) Iris(R) Xe Graphics
pulsar tusk
u can get someone else to help u
what why me saying some people is my brother not online thats too cruel
brittle wing
i need some help bad i have been stuck on thsi for over 2 hour's
brittle wing
maybe try asking one of the helpers or mods, I wouldn't really be smart enough t...
hello can you help?
viral mason
i need some help bad i have been stuck on thsi for over 2 hour's
stuck with what
brittle wing
So I’ve been trying to follow this tutorial https://www.youtube.com/watch?v=tWd1SMRXRMw, it's to do with the https://huggingface.co/wok000/vcclient000/tree/main because I haven’t got a GPU for my computer. It should work like normal but for some reason it’s not working properly. I can show you where I’m struggling for it to work?
viral mason
So I’ve been trying to follow this tutorial https://www.youtube.com/watch?v=tWd1...
this is ollld
if u don't have a gpu to ur pc idk what to tell ya
brittle wing
if u don't have a gpu to ur pc idk what to tell ya
LOL OK thanks
viral mason
sorry i cannot help u :<
nova abyss
this is ollld
How do i fix my Vonovox? I hear my voice first then the model trailing behind it
viral mason
How do i fix my Vonovox? I hear my voice first then the model trailing behind it
do u have hte mic settings setup properly?
nova abyss
do u have hte mic settings setup properly?
think so. the input is the mic and the output is line 1 on Vonovox then for discord its line 1 as the input and output the headphones. and the model keeps on like breaking up
sharp turret
Hey there, I downloaded the latest version of W-Okada voice changer, but whenever I start any voice it hits me with this:
[Voice Changer] Pipeline is not initialized.
[Voice Changer] Waiting generate pipeline...
mystic tangle
i keep getting this error
simple ore
i keep getting this error
probably downloaded some oudated af code
stray steeple
Is it possible to make a rvc sound real sounding?? I have cpu only sadly.. surface pro 8, rip.. do u think I could maybe install a GPU on an emulator? Ava max rvc is what I am going for.
gusty shuttle
whats the best epoches for a 8 min audio
hallow thistle
Hey there, I downloaded the latest version of W-Okada voice changer, but wheneve...
!howtoask
hallow thistle
So I’ve been trying to follow this tutorial https://www.youtube.com/watch?v=tWd1...
If your PC doesn't have any GPU, your PC won't output any image to monitor, unless there's an "integrated GPU", which still won't be ideal for the voice changer anyway. Instead, try online option.
hallow thistle
How do i fix my Vonovox? I hear my voice first then the model trailing behind it
Try change block size.
hallow thistle
Intel(R) Iris(R) Xe Graphics
Intel Iris Xe is an integrated GPU, often found in 11th generation Intel Core CPU system and newer. If you attempt run any W-Okada version (like the DirectML variant) with an integrated GPU, the program would perform at the fraction of your PC CPU, so unfortunately. Running W-Okada with only CPU is possible, performs a bit step up from an iGPU, though not always recommended to run as some serious task and along a game.
hallow thistle
this thing, and this also happened on every collab and other sites..
The ngrok website itself either fail or got blocked access to elsewhere. Make sure to reset your ngrok token just in case if it fixes.
hallow thistle
Would you guys consider yourself experts on rvc
No. 
hallow thistle
what does index do in a voice model
In RVC voice model, an index file stores accent of that voice model, typically found alongside pth file, although some voice models might not include index from start.
hallow thistle
Oh yeah I figured it out I had to do system not client I’m pretty sure
That's more like a short-term solution rather than a detailed one. Make sure to read help guidelines before start asking, rather than telling your story of using a "program".
abstract comet
No. <:chihirokeikaku:1015914946934083665>
Well you’re def more expert than me 😄
fierce zodiac
hey if i have like a intergrated gpu in my cpu which is intel
which download shall i do
lyric moat
In RVC voice model, an index file stores accent of that voice model, typically f...
oooooh thanks a lot for clarifying
although for my pc it makes the voice sound wobbly if that makes sense
maybe cuz I'm using an intel iGPU (HD Graphics 530) or killing my CPU
hallow thistle
oooooh thanks a lot for clarifying
although for my pc it makes the voice sound w...
Intel HD and UHD Graphics are the Intel integrated GPU, typically found in most Intel Core CPU systems. You sure running W-Okada DirectML with an integrated GPU would give better audio quality than one with a dedicated GPU (like Intel Arc B series)?
serene wigeon
this isn't so much an AI question, but I have an issue obtaining audio samples for doing inference. Since youtube clamped down on copyrights, the app i used to use for downloading videos no longer works.
How can I get audio samples off of yt now?
hallow thistle
hey if i have like a intergrated gpu in my cpu which is intel
See #✨│ai-help message and #✨│ai-help message.
lyric moat
Intel HD and UHD Graphics are the Intel integrated GPU, typically found in most ...
i mean yeah i am and the quality is nicer when the index is disabled
it's choppy
thin chasm
Hii i need help for voice changer i download one here but idk what do do after
hallow thistle
i mean yeah i am and the quality is nicer when the index is disabled
it's choppy
What the hell?
hallow thistle
Hii i need help for voice changer i download one here but idk what do do after
!howtoask
hallow thistle
Never say you don't know about your PC GPU, in this situation. To check your PC GPU, open Task Manager, go to Performance tab.
wet hinge
Idk what version to pick, the 001 cude or the 002
viral mason
Are you looking at wokada tg fork?
wet hinge
yee
viral mason
If u have Nvidia u need both if u have amd get the one below that, it says amd
wet hinge
oh ok
wet hinge
Not the 001 or 002
the 002 doesnt let me extrat it ;/
viral mason
Hmmm
wet hinge
prob help if i could send pics its just a white file and not extactlbe like the 001
viral mason
Put it in the same folder with 001 and that might work, like don't try to extract just go ahead and run mmvcserversio
I had the same issue where it wouldn't extract
lunar crow
Upon trying to insert my .json file into the index section for Realtime Voice Changer Client, it says the following:
" Extension of file should be the following: index, bin "
Am I doing something wrong?
wet hinge
I had the same issue where it wouldn't extract
keeps saying select an audio input device but I have my mic selected ;/
viral mason
Do u have both input and output selected?
wet hinge
weird i had to switch it off deafault
viral mason
Weird
Btw if u wanna this in games u need either VB cable or vac lite
The second one is more recommended
wet hinge
then whats a good chunk size and extra proccesing and ill be good to go
i already have that set up
viral mason
then whats a good chunk size and extra proccesing and ill be good to go
Try for chunk 122.7
And for extra time 2.7
wet hinge
Try for chunk 122.7
And for extra time 2.7
for games?
wet hinge
just seems like itll be hard to respond on a 3 second delay
viral mason
I have a 5070ti and I really only use the voice changer in vrchat so I wouldn't know for regular flat screen gaming
wet hinge
i use a 4070
viral mason
That's similar
Probably will work fine, but u can test it tho
Different chink sizes
wet hinge
Different chink sizes
just move it up to more?
viral mason
Lowering it would reduce the amount of time it takes for the voice to come out
But lowering it too much would cause it to be choppy and bad
So u gotta find a middle ground between the delay and it sounding good
wet hinge
female voices use around 11-13 pitch?
viral mason
Actually it's a bit lower than that
Depends on what female voice it is but most are good at like pitch 3-12
Unless you sound like corpse husband
Lol
viral mason
im using the mae asmr one idk another good one lol
Oh :(
dense pebble
i cant hear myself on discord
viral mason
what the Epochs: 400?
Epochs aren't important tbh unless you're training a model, but even then steps are more important
viral mason
i cant hear myself on discord
Did u set the voice settings up correctly
sturdy forge
why is it so laggy
stray copper
what batch size should i do for 34 minutes of 41.1 Khz audio
crude pumice
flat flame
when i try do download a few selected voice models it says on hugging face "Invalid username or password"
after i trinkered with hugging face configs now it says "Repository not found"
Its this one im trying to download
wispy fable
it lags alot and voices are not so good like the showed ones
finite bay
my app is just blank
tawny radish
how do i fix this? its been happening to me and my friend
civic star
im not getting any sound from virtual aduio cable
viral mason
my app is just blank
What app
viral mason
i mean the voicemod thing
Voicemod? Or wokada voice changer
Those are two different things completely
finite bay
i think wokada
knotty moth
https://huggingface.co/ScruffyRVC/RFCRVCV2/resolve/main/vaggienew_e700_s22400.zi...
the huggingface repo has been either deleted or set private
knotty moth
how do i fix this? its been happening to me and my friend
both of the og wokada and deiteris fork have bug on deleting the model slot, try directly deleting the slot folder in model_dir
or just try tg-develop fork or vonovox as listed in the following
-rt
knotty moth
i think wokada
you should also try one of the recommendations as just above^
knotty moth
how do i downbload wokada
radiant cypress
any reason there is too many audio stuff?
also does make difference related to tis
viral mason
any reason there is too many audio stuff?
That's not too many u just downloaded VB cable so it added extra stuff
radiant cypress
okok
knotty moth
any reason there is too many audio stuff?
yeah why an apple falls downwards not upwards?
you haven't quite explained the problem tho
do you mean the vb-cable is not working for voice changer?
radiant cypress
yeah why an apple falls downwards not upwards?
you haven't quite explained the p...
sorry man just curious
am not good with tis stuff
radiant cypress
do you mean the vb-cable is not working for voice changer?
its working dw
neat patio
why tf does okada look like that?
viral mason
why tf does okada look like that?
Like what
tough compass
What pretrain models should I use if I select "refinegan" vocoder and "contentvec" embedder, for 40k sample rate?
If there are no any, should I use "hifigan" + "contentvec" or "hifigan" + "spin-v2"? Can't find any info about standart 40k pretrain.
(Applio 3.6.0)
knotty moth
What pretrain models should I use if I select "refinegan" vocoder and "contentve...
that option is only available for 32k
tough compass
that option is only available for 32k
okay, then what embedder should be used for Applio\rvc\models\pretraineds\hifi-gan\f0[G/D]40k.pth pretrain?
knotty moth
okay, then what embedder should be used for `Applio\rvc\models\pretraineds\hifi-...
contentvec is the default one
but you can also try spin-v2
tough compass
contentvec is the default one
Thank you
It's a bit weird that if I select refinegan it doesn't even say that no pretrain available and that it actually does scratch-training
knotty moth
It's a bit weird that if I select refinegan it doesn't even say that no pretrain...
just keep 32k sample rate and "pretrained" checked but not the custom pretrain
to change the sample rate, you have to delete the current preprocessed model and start over from preprocess stage
tough compass
just keep 32k sample rate and "pretrained" checked but not the custom pretrain
well I have 48k sources and target files too, 32k will be too much loss of voice details I think (usual speech)
knotty moth
well I have 48k sources and target files too, 32k will be too much loss of voice...
it will automatically resample to the target 32k, but for better quality you'd better resample it first by yourself
as said, there's no refinegan for 48k yet
that's why if you select the 48k option it will train from scratch
tough compass
yes I got it, thank you
knotty moth
yes I got it, thank you
even if you try to do that, note that no one has figured out some optimal configurations for 48k refinegan at this moment.
right now someone is attempting for 40k #🔊│ai-development message
tough compass
right now someone is attempting for 40k https://discord.com/channels/11592601219...
No access to the link
bold obsidian
I got an error occurred during voice conversion and the console said pipeline is not initialized, im using
GPU: NVIDIA GeForce RTX 3050 6GB Laptop GPU
OS: Windows 11 24h2
I followed the instruction at https://docs.aihub.gg/realtime-voice-changer/local/tg-develops-w-okada-fork/. How do i fix that?
knotty moth
@finite bay
finite bay
alright
hallow thistle
alright
Use #✨│ai-help or #1192011222023950368 for help about an AI program. There's no reason to go in someone's direct message for such thing.
hallow thistle
why tf does okada look like that?
Like who?
hallow thistle
also does make difference related to tis
With audio mode server and WASAPI selected in W-Okada, the sample rate generally locks and works fine at 48000 Hz (or in a very rare case, 96000 Hz) because of "shared mode"; when you set sample rate to any number (like 44100 in your screenshot), the program will fail because of mismatch sample rate. It's an issue in almost every program that primarily uses WASAPI in shared mode not just the voice changer.
radiant cypress
the program doesnt fail tho
i have been using this settings for days
alsooo it doesnt switch to anything else
like its stuck at 44100
radiant cypress
With audio mode server and WASAPI selected in W-Okada, the sample rate generally...
it reverts any change i make when i start the server
hallow thistle
like its stuck at 44100
It looks like the sample rate from both your speakers and microphone all being set to "44100 Hz", so you might not encounter any issue when using WASAPI in W-Okada even if the program's sample rate is 44100 Hz, although 48000 Hz is often more preferred over 44100 Hz because 48000 Hz gives a bit better audio quality.
chilly sluice
Does anyone know if there’s a new fork for okada
radiant cypress
It looks like the sample rate from both your speakers and microphone all being s...
this is set to 48000 tho 😭
nvm it, if it works dont touch it ig
hallow thistle
radiant cypress
lol
finite bay
mine doesnt connect with discord?
hallow thistle
mine doesnt connect with discord?
!howtoask
hallow thistle
Your question sounds too simple but that doesn't refer to any specific program.
finite bay
I watched the tutorial and understood everything, but somehow the voice changer client just won't connect. I have no idea why.
hallow thistle
I watched the tutorial and understood everything, but somehow the voice changer ...
W-Okada and voice changer are the same program. Whatever. What is your PC GPU?
chilly sluice
Your question sounds too simple but that doesn't refer to any specific program.
I have a 4060 and I’ve been using this voice changer for like 2 years, is vonovox better?
warped osprey
hey can someone help me
hallow thistle
I have a 4060 and I’ve been using this voice changer for like 2 years, is vonovo...
No idea, as I've never seen any benchmark between Tg Develop's W-Okada fork (b2397) and Vonovox on the same PC specs nor I've ever tested them myself, but some people have been saying Vonovox can give better audio quality than other W-Okada versions and that's really it.
hallow thistle
hey can someone help me
Make sure to read help guidelines before start asking anything.
warped osprey
when i try to launch mmvcserversio it doesnt work it just says "analyzing...done" but its not launched
hallow thistle
when i try to launch mmvcserversio it doesnt work it just says "analyzing...done...
What is your PC GPU? And did you follow any tutorial or guide before?
warped osprey
What is your PC GPU? And did you follow any tutorial or guide before?
Nvidia Geforce RTX 3050
i was following the tutorial of "novision" but it doesnt work
hallow thistle
Nvidia Geforce RTX 3050
i was following the tutorial of "novision" but it doesnt...
Try Tg Develop's W-Okada fork. https://docs.aihub.gg/realtime-voice-changer/local/tg-develops-w-okada-fork/ https://github.com/tg-develop/voice-changer/releases/tag/b2397
warped osprey
whats that
hallow thistle
whats that
A better version of MMVCServerSIO, which the name is basically the same as W-Okada or voice changer as you called.
warped osprey
thank you so much, is it free?
finite bay
W-Okada and voice changer are the same program. Whatever. What is your PC GPU?
i have a really bad one bc im still waiting for my new one to come
asus GTX 1060
hallow thistle
i have a really bad one bc im still waiting for my new one to come
Just say one.
Sure, with NVIDIA GeForce GTX 1060, it will still work with W-Okada though far below than RTX 20 series GPU.
For a newer W-Okada version, see #✨│ai-help message.
warped osprey
A better version of MMVCServerSIO, which the name is basically the same as W-Oka...
hey do u have a yt tutorial i can follow with the ai u sent me?
hallow thistle
but how do i connect it now with discord
https://cdn.discordapp.com/attachments/1159290139609137264/1446171264489357322/image.png?ex=694a15b2&is=6948c432&hm=55c76522e40a0231c9e94c1f8c5a3a81764d07c3e7caa6c5fb89cf3905787989& https://cdn.discordapp.com/attachments/1159290139609137264/1451594734962217062/image.png?ex=694a0a32&is=6948b8b2&hm=de90fe156223b06829c074adc1bb7ea04e0af73765ce984fc149e38accf37229&
finite bay
it worked thanks
hallow thistle
hey do u have a yt tutorial i can follow with the ai u sent me?
warped osprey
Oh its okay, thank u so much for the help tho
inner jasper
how to install
hallow thistle
how to install
Make sure to read help guideline before start asking anything.
inner jasper
rx 6600 xt
hallow thistle
rx 6600 xt
inner jasper
do i need to run as admin?
hallow thistle
do i need to run as admin?
No.
hallow thistle
That's an awkward question, and here's your settings:
Chunk: around 70 - 100 ms
Extra: 2.7 s
GPU: AMD Radeon RX 6600 XT
Pitch detection (F0): rmvpe_onnx
inner jasper
also should i use gpu or cpu for processing unit
hallow thistle
inner jasper
..?
hallow thistle
In Tg Develop W-Okada, the processing unit is basically "GPU", so when I say GPU you should set it to your GPU not CPU.
inner jasper
oki
finite bay
why is the voice so laggy
hallow thistle
why is the voice so laggy
Here's your settings:
Chunk: around 256 ms, always check the perf number at top right
Extra: 2.7 s
GPU: NVIDIA GeForce GTX 1060
Pitch detection: rmvpe
hallow thistle
Do you remember when I asked you about using the b2332 a month ago? There's Tg Develop's W-Okada fork (b2397) to try, but I doubt if you ever found out about it.
inner jasper
That's an awkward question, and here's your settings:
Chunk: around 70 - 100 ms
...
i used them and they worked fine, but suddenly my voice started to repeat and lag
hallow thistle
i used them and they worked fine, but suddenly my voice started to repeat and la...
Check your perf number at the top right of voice changer. https://docs.aihub.gg/realtime-voice-changer/local/tg-develops-w-okada-fork/#finding-my-own-settings-for-chunk-size-and-extra-processing-time
gusty cipher
on w-okada its not picking up my voice even though i gave it permission, my input microphone works fine as i tested, what do you think happened
inner jasper
Check your perf number at the top right of voice changer. https://docs.aihub.gg/...
okay what do i do with the perf number..
its like 193 ms of 141
hallow thistle
Take time to read the guide instead of expecting me to provide an answer. Increase chunk number up until your perf number is green.
hallow thistle
on w-okada its not picking up my voice even though i gave it permission, my inpu...
What is your PC GPU? And did you follow any tutorial or guide before about voice changer?
gusty cipher
What is your PC GPU? And did you follow any tutorial or guide before about voice...
i followed the tutorial by easeus and AIsearch, my gpu is 5060ti
hallow thistle
i followed the tutorial by easeus and AIsearch, my gpu is 5060ti
Try Tg Develop W-Okada. https://docs.aihub.gg/realtime-voice-changer/local/tg-develops-w-okada-fork/ https://github.com/tg-develop/voice-changer/releases/tag/b2397 The original W-Okada version didn't compile to work with GeForce RTX 50 series GPU.
gusty cipher
oh, i used the download one from huggingface, is that the original one?
hallow thistle
oh, i used the download one from huggingface, is that the original one?
This https://huggingface.co/wok000/vcclient000/tree/main is the original W-Okada; the original W-Okada all outdated and now superseded by newer voice changer versions made by different authors.
gusty cipher
okay, thank you
gusty cipher
This https://huggingface.co/wok000/vcclient000/tree/main is the original W-Okada...
i have an intel cpu and it wants me to download "voice-changer-macos-amd64-cpu.tar.gz" but i dont see it on the downloads area
hallow thistle
i have an intel cpu and it wants me to download "voice-changer-macos-amd64-cpu.t...
That's not the correct one. The correct ones are https://github.com/tg-develop/voice-changer/releases/download/b2397/voice-changer-windows-amd64-cuda.zip.001 and https://github.com/tg-develop/voice-changer/releases/download/b2397/voice-changer-windows-amd64-cuda.zip.002.
The "macos" means macOS, an operating system of Apple Mac, while "windows" is simply Windows. Your PC is not Mac.
gusty cipher
That's not the correct one. The correct ones are https://github.com/tg-develop/v...
when i open it winrar says "unexpected end of archive"
signal crater
If i train a voice model using weights.com (partner of ai hub), after training does it gives the model .pth and index file once its trained? so i can run on w-okada?
warped osprey
That's not the correct one. The correct ones are https://github.com/tg-develop/v...
hey so u recommended me tg develop w Okada fork and im just wondering how do i extract both files like the cuda zip 001 and 002 cuz when i click on them ii cant extract them
sharp turret
!howtoask
AMD Radeon RX 6650 XT, Windows 10, I start a voice and it says [Voice Changer] Pipeline is not initialized. [Voice Changer] Waiting generate pipeline...
versed hull
dense pebble
how to be able to hear myself in discord
zealous steeple
there was this on RVC app i used
it came up when i searched like MMVC
and it had squigly lines
resemlbing audio
i cant find it anymore , does any1 know what im talking ab
viral mason
Yes, but that's probably an old one what gpu do u have
zealous steeple
Yes, but that's probably an old one what gpu do u have
rtx 4060
im looking for it as its rlly good
viral mason
U should use Vonovox it's the current best at the moment for Nvidia gpus
The guide is the first one
-rt
viral mason
To set it up just run setup, then after that run start
zealous steeple
To set it up just run setup, then after that run start
nice, ty but dyk the one i was talking ab at first?