after this then whats next
#✨│ai-help
smoky arrow
low shard
after this then whats next
you downloaded it
you should have a resfresh models button and find it here
smoky arrow
so no need to seperate the vocal and instrument?
low shard
so no need to seperate the vocal and instrument?
nope, this fork does it automatically
that's the reason why it was made
low shard
so whats the different then the first ones?
the first one is just a guide explaining how to make ai covers generally speaking on other programs like Applio
the 2nd is a page dedicated to this fork which makes life easier
low shard
i though i was like singing
generally speaking = used in other programs like applio
low shard
okk thanks for the help
yw
hallow thistle
for the voice model is it the pth file ?
The "pth" file is voice model, while index file is index.
smoky arrow
The "pth" file is voice model, while index file is index.
okk
also there are some parts, in the song, where the original vocal kinda exist
hallow thistle
also there are some parts, in the song, where the original vocal kinda exist
What does this mean?
!howtoask
vernal bone
hallow thistle
While I can't pick a specific male voice model for you, you can find one in #1175430844685484042.
Did you follow a tutorial from YouTube?
smoky arrow
What does this mean?
this ai vocal, should remove the orignal vocal right, so when the model sing a word, there are still some parts that the orignal of the vocal
sing the same word
hallow thistle
Don't follow tutorial videos on YouTube; they only telling you to install and use older W-Okada versions.
hallow thistle
this ai vocal, should remove the orignal vocal right, so when the model sing a w...
Which program/website are you using? I don't remember asking you about this.
smoky arrow
hallow thistle
A typical RVC program itself doesn't separate original audio into stems like what UVR5 does, although some RVC forks may bundle with UVR5 in itself.
knotty moth
there are too many regular voices to recommend
so just search any random ones and try until you find which is the best
hallow thistle
Did you follow this guide? This is where you'll find the better W-Okada version. https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#download-nvidia-on-windows
knotty moth
oh i see
just like we can't recommend you which formal suit to wear
it's just up to you
hallow thistle
You can't expect one voice model to be great in quality.
knotty moth
that's just nothing strange for anyone to want the spongebob voice
hallow thistle
If I use W-Okada, I would have occasionally use it for funny and never meant to be that far serious. 
hallow thistle
also is there a way to tweak a specific model voice,
The thing is, you can't. However, merging two voice models into one is possible.
Why did you say like that? I didn't mean to be that serious lmao. I was just saying my own point, not yours.
knotty moth
it's "wrong software" unless it's the latest deiteris fork wokada as in the guide
hallow thistle
knotty moth
don't recall either of us said about any different software, as I know you just want a voice changer for your purpose
hallow thistle
W-Okada, MMVCServerSIO and "realtime voice changer" refer to the same program, just called in different ways. Deteris' fork W-Okada and original W-Okada versions are kinda the same but developed differently by different authors.
smoky arrow
The thing is, you can't. However, merging two voice models into one is possible.
how do you do that?
hallow thistle
The difference between Deiteris' fork and old versions from YouTube is mostly performance and minor changes, while most functions are pretty much the same.
hallow thistle
how do you do that?
You want to do that?
smoky arrow
well...i tried to make my own voice but the results are weird
latent cypress
does applio have tf32?
knotty moth
you'd better first get the voice changer working with any voice model, then you can go search any models you want
smoky arrow
then i try to find the model provided, there are some that i like but when i put it into the song, its not really what i expected
@hallow thistle for these voice model which are best for singing?
knotty moth
also is there a way to tweak a specific model voice,
there's no way to modify any existing models, other than merging it (but it's mostly hit or miss)
to make a model by your own, you'd need to prepare a dataset
knotty moth
<@561807329977434122> for these voice model which are best for singing?
try using only UVR de-echo aggressive instead of two separate de-reverb & de-echo models
smoky arrow
try using only *UVR de-echo aggressive* instead of two separate de-reverb & de-e...
like this ?
i dont really have much options to choose from though
knotty moth
like this ?
no idea but for original UVR, it'd be like this (you'd ofc first need to do melforormer separation)
brazen parrot
can somebody help me?
smoky arrow
there's no way to *modify* any existing models, other than merging it (but it's ...
how about if a model then i want to change the pitch abit
brazen parrot
?
knotty moth
how about if a model then i want to change the pitch abit
there's pitch setting in the inference, and again it doesn't modify the model itself, only the inference output
hallow thistle
can somebody help me?
!howtoask
hallow thistle
What are you looking for?
hallow thistle
How do I connect my VBcable to my discord
Oh no. What do you use W-Okada for? Trolling your friends or catfishing someone for your own benefit? And what is your PC GPU?
smoky arrow
!howtoask
well because before i wwas told if i want to make the ai cover only , its better to used the one i used right now
hallow thistle
well because before i wwas told if i want to make the ai cover only , its better...
That command was meant for another user not you.
brazen parrot
Oh no. What do you use W-Okada for? Trolling your friends or catfishing someone ...
I want to use it to troll people, my CPU is Ryzen 5600
hallow thistle
I want to use it to troll people, my CPU is Ryzen 5600
What troll? And, AMD Ryzen is CPU, not GPU (graphics processing unit).
To check your PC GPU, open Task Manager or press Ctrl + Shift + Esc on your keyboard.
brazen parrot
What troll? And, AMD Ryzen is CPU, not GPU (graphics processing unit).
AMD Radeon RX 6600
hallow thistle
AMD Radeon RX 6600
Now what troll? And did you follow a tutorial video from YouTube?
brazen parrot
Like i wanna troll some friends,yeah i watched the tutorial.
hallow thistle
Like i wanna troll some friends,yeah i watched the tutorial.
Follow and download the better W-Okada from this guide instead. https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#download-amd-intel-and-cpu-on-windows Tutorial videos on YouTube are outdated.
Make sure to use Virtual Audio Cable lite instead of VB-Cable.
analog obsidian
does applio have tf32?
yep
just add
torch.backends.cuda.matmul.allow_tf32 = True
torch.backends.cudnn.allow_tf32 = True
to train.py
latent cypress
yep
just add
torch.backends.cuda.matmul.allow_tf32 = True
torch.backends.cudnn....
okie, thanks
simple ore
okie, thanks<:cat_sunglasses:1359900720924590281>
makes 0 difference
one of those is true by default
pastel oak
Whats your GPU and version of the voice changer you got
This is too bad to run the voice changer
If you have decent internet speed run the online hosted, needs phone number verification and can be more complciated to set up, heres a guide
https://docs.aihub.gg/realtime-voice-changer/cloud/deiteris-w-okada-fork-kaggle/
No
If your internet is decent itll work yes
Else need to upgrade your pc
Thats more than good enough
On a web browser
No reduced quality, just check in here again for recommended settings when you got everything set up
quiet axle
Hi guys, i want to asked. Im trying out the newer version W okada but when i activated it i realize theres no GPU option? im stuck on CPU only?
pastel oak
I think the kaggle notebook page already gives you some recommendations, but start with
F0 rmvpe
Chunk 180
Extra 2.7
Then later send a screenshot in here of the voice changer while its running
pastel oak
Hi guys, i want to asked. Im trying out the newer version W okada but when i ac...
Whats your gpu and the version you downloaded
quiet axle
The newest one from august 14, my gpu is 3050 laptop
pastel oak
The newest one from august 14, my gpu is 3050 laptop
I dont know anything about a new version from yesterday, can you be more specific
Do you mean vonovox?
pastel oak
@low shard can you have a look at this, I think you added this download?
Person can't see their gpu (rtx 3050) on nvidia all cards version
quiet axle
For more information, tried installing it on my main PC with 5060 TI, still only cpu option
low shard
@pastel oak @quiet axle the spin version of wokada deiteris fork has been temporarily hidden for a bug
quiet axle
<@965636299509882980> <@1313830361800904717> the spin version of wokada deiteri...
okay ty
bronze elk
hi i got it working, are you still in need of help?
@glacial sleet care to join me in the voice chat?
this is the setting you must be tweaking for that issue
can't uplaod images yet
pastel oak
scarlet wedge
why i can't see bot commands messages?
every chat is working fine except bot commands
leaden saffron
what f0 det shoudl i use with this
AMD Radeon(TM) RX 6550M
scarlet wedge
it pinged me but i can't see the chat
low shard
what f0 det shoudl i use with this
AMD Radeon(TM) RX 6550M
Rmvpe onnx
hallow thistle
it pinged me but i can't see the chat<:misc_cry:1176674698629750975>
Try close and open Discord again. Discord can sometimes lag on mobile.
leaden saffron
Rmvpe onnx
it says n/a
low shard
it says n/a
Rmvpe onnx, not rmvpe
hallow thistle
rmvpe_onnx
leaden saffron
Rmvpe onnx, not rmvpe
what chunk and extra do you recomend
hallow thistle
what chunk and extra do you recomend
leaden saffron
its choppy
hallow thistle
its choppy
Try reduce extra down to 2.5 s or increase chunk number. Additionally, you can post your screenshot of while the program is processing.
scarlet wedge
Try close and open Discord again. Discord can sometimes lag on mobile.
i did it but it unfortunately didn't work
-colab
hallow thistle
i did it but it unfortunately didn't work
That's unfortunate huh. You can try clear the cache of Discord app, check your internet or report the bug to Discord support.
quick hazel
hello im totally new to this. im looking into using rvc ai cover features but it seems they struggle to hit higher notes so it just sounds like an awful voice crack.. is there a fix?
latent kettle
@viscid moss sorry to bother you but when i'm trying to use Mel-Roformer-Denoise-Aufr33-Aggr model in UVR UI i'm getting "unexpected error" i'm sending logs. please take a look
viscid moss
<@274566299349155851> sorry to bother you but when i'm trying to use Mel-Roform...
Hmmm lemme take a look and test on my end
elfin nebula
Is it worth providing a full 30 min recording for a live tts, when making your own model?
latent kettle
Hmmm lemme take a look and test on my end
Does it worked ?
viscid moss
Does it worked ?
yeah it's working for me on HF and locally
Can u pls wait like an hour. I'll publish the new update so u can get it
How did u installed it?
Precompiled or normal?
latent kettle
Precompiled or normal?
.zip precompiled
viscid moss
.zip precompiled
okke
latent kettle
is there any new features for new updates
im the one who requested some features such as automatically save stems to different folders (vocals to vocal folder and BGM to instruments folder) and check boxes to select stems etc 😭
viscid moss
is there any new features for new updates
For now, just new models and now u can fully update even with precompiled version so u don't need to download the new version every single update
just once
viscid moss
im the one who requested some features such as automatically save stems to diffe...
not yet, sorry
I was very busy
latent kettle
I was very busy
ohh, no problem don't say sorry
viscid moss
I'm working onrequests adding them one at a time
latent kettle
I'm working onrequests adding them one at a time
thank you soo much have a good day and bug free code : )
viscid moss
thank you soo much have a good day and bug free code : )
also be sure to get the dev updates role so u'll be pinged on #📰│dev-updates
latent kettle
also be sure to get the dev updates role so u'll be pinged on <#1159380240271953...
got it.
viral mason
Tbh it'd be better if you used Kaggle if you're looking to train a voice model
polar parrot
can you pass me the link?
viral mason
30 hours beats the small 2-4 I think it is that Google collab gives you
viral mason
can you pass me the link?
Sure! I can give the link and I made a video showing how to use it very easily
It's up to date of this year
viral mason
I'll give the links and stuff in a sec because I'm busy rn, DM me tho in case I forget
I have kinda a bad memory
polar parrot
I'll give the links and stuff in a sec because I'm busy rn, DM me tho in case I ...
Ok
I'll send you a point in the DM
viscid moss
got it.
Update is ready. Pls update running UVR5-UI-updater.bat. If u face an error that says "epository isn't found or something" run the updater again. Also if u face warnings like: xxxxxx isn't on the PATH just ignore it
viscid moss
I sent u the video and two links u need for it all
Update is ready 🔥
Will be available on HF in a few, curretly just colab/kaggle/lightning.ai/locally
viral mason
Will be available on HF in a few, curretly just colab/kaggle/lightning.ai/locall...
peak
tell me, how much better is fv6 compared to fv4-5
viral mason
Nice
leaden saffron
my thinG just stoped working not sure how to fix it i dont hear sound anmd the vc isnt working
low shard
You could try blending/merging models but still it will sound like 2 voices
low shard
my thinG just stoped working not sure how to fix it i dont hear sound anmd the v...
Show a screenshot
leaden saffron
leaden saffron
@low shard
low shard
Uncheck sup1, always check the triangle, play on lowest graphics 1080p 60fps cap, set the chunk always higher than perf
leaden saffron
Uncheck sup1, **always check the triangle**, play on lowest graphics 1080p 60fps...
turned sup1 off no difference what is the triangle
i did that still no change
low shard
turned sup1 off no difference what is the triangle
Hover the triangle with your mouse
You need to always do it
low shard
how can i download the ai thing
This is a General AI Server, AI has many fields, so we can't know your issue with little info
Please Elaborate:
- your PC GPU (Such as Nvidia RTX 3060, AMD RX 9070 XT..)
- your operating system (Like Windows 10, Windows 11, Linux, MacOS 26..)
- what you want to do? There isn't a program that does everything, there's a program for each thing:
- AI Covers
- Train RVC Models
- E girl trolling
- TTS
- Roleplay in VC
- Roleplay in Games
- etc
- what tutorial link are you using
- a screenshot of the program
fleet trail
Not realtime voice changwr
i want to download the realtime voice changer
fleet trail
This is a General AI Server, AI has many fields, so we can't know your issue wit...
rtx 4070 i7-13700k
32gb
win11
Roleplay in VC and trolling
low shard
Roleplay in VC and trolling
E girl trolling?
leaden saffron
Hover the triangle with your mouse
under the perf right i justy did that and nothing change di change the chunk to be higher than it too
low shard
maybe
That is carfishing and illegal, you have been warned and won't get help. Continuing to ask help will lead to further actions like ban
fleet trail
That is carfishing and illegal, you have been warned and won't get help. Continu...
no no
leaden saffron
so what do i need
please dont egirl troll like what nick said
if i were u id js use the vc for shits and giggles
fleet trail
E girl trolling is catfishing by definition
Sorry, I didn’t mean anything illegal or against the rules. I misunderstood and I won’t ask about that again
I just mean making voices like characters for clips
fleet trail
please dont egirl troll like what nick said
nah dw
@low shard im sorry
I only meant fun character voices for clips, nothing serious or against the rules. Sorry again
low shard
under the perf right i justy did that and nothing change di change the chunk to ...
No, I meant the triangle next to your GPU
Did you do what it says?
fleet trail
could you help me download and set up RVCI just want to make some fun character voices for my clips
leaden saffron
No, I meant the triangle next to your GPU
yeah I did that but it didn't fix it at all
swift jungle
Is there a way to use w-okada with a meta quest 3s headset on games? And is there a high hardware requirement (like a 3070 or higher ish)
pastel oak
Show full screenshot
select gpu
and f0 rmvpe
and extra 2.7
You are not using kaggle, you are using the normal NVIDIA-GPU local version
You dont have an nvidia gpu
So why did you download the nvidia download 😭
Are you?
All versions are browser
kaggle gpu is nvidia so idk
@low shard can you help with kaggle
Bro be my guest and download the amd version from the docs guide I cant stop you
You wont have the lowest delay and cant run anything on your pc next to it
Wait for nick, youre doing something wrong
Kaggle is online hosted, so youre accessing a gpu online, it should show up on the nvidia one for you
its not using your own, it doesnt matter if you have nvidia or amd
I never used kaggle so i dont know where it went wrong
Should be
pastel oak
theyre correct, vb cable is just garbage with labels
Ye
You can try increasing chunk to 1000 and see if the red "perf" number becomes green
You asked this already I told you to do it if you want to
wraith elk
Please Elaborate:
Nvidia RTX 3060
Windows 10
Ai Covers
I just used the guide in this server
With the actual softwares replay it sounds nice and clear but when i try using it somewhere with audio cable it sounds really muffled why?
low shard
could you help me download and set up RVCI just want to make some fun character ...
You won't get help as you got already warned for catfishing few seconds before asking this
are you trying to do e girl trolling?
low shard
yeah I did that but it didn't fix it at all
did u set the chunk to an higher value than the perf value at top left? can u show me again?
low shard
Is there a way to use w-okada with a meta quest 3s headset on games? And is ther...
be sure to not be using video tutorials, but yes you can use it with wokada deiteris fork and ofc it requires more power
if u want i can check ur settings
swift jungle
be sure to not be using video tutorials, but yes you can use it with wokada deit...
is there a way to plug the pc into the vr and use the software seperately from there or smth? Like having 2 pcs 1 for streaming and 1 for gaming
low shard
uninstall vb audio cable, what's ur pc gpu? have u checked in task manager? and what tutorial link are u using, what u looking to do?
low shard
Please Elaborate:
Nvidia RTX 3060
Windows 10
Ai Covers
I just used the guide in ...
can you show a screenshot of the other program settings where it sounds muffled?
swift jungle
yes https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#op...
no like do that but for a vr and a pc, I wouldnt be concerned about performance if I had 2 pcs on standby 
low shard
there are thousands of voice models, you could try searching just "male" but u could maybe try merging models 
wraith elk
can you show a screenshot of the other program settings where it sounds muffled?
I can't send any screenshots here
low shard
!give-media-perms 1h @wraith elk
low shard
no like do that but for a vr and a pc, I wouldnt be concerned about performance ...
wdym? didn't you say about using 2 different pcs?
wraith elk
I just tested it on discord, and yeah sounds muffled on there too
swift jungle
wdym? didn't you say about using 2 different pcs?
forget that, I mean just a vr and a pc
low shard
uninstall vb audio cable
are you trying to do e girl trolling?
wraith elk
No
wraith elk
uninstall vb audio cable
are you trying to do e girl trolling?
Should i reinstall it?
Would it fix it
low shard
ok but where do i search? I don't know where to search because everywhere i search seems kinda bad
You can search rvc ai voice models at:
- https://discord.com/channels/1159260121998827560/1175430844685484042
- In https://discord.com/channels/1159260121998827560/1163592055830880266 , Do /find with @earnest musk
- https://weights.com/ (login required)
- https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
- https://voice-models.com/ (just scrapes the huggingface link from the voice-models channel)
- https://thevoicemodels.com/ (for Turkish Models, login required with discord and level 2 on their server)
also doesn't merging models going to sound like two people talking?
it will make a 'new' voice based on the charateristics of the models merged
earnest muskBOT
> ok but where do i search? I don't know where to search because everywhere i se...
:wave: @low shard, How can I help?
Available Commands:
• @weights find <query> or /find <query> - Search for RVC Voice Models
low shard
it can be also more than 2, to make it more unique
low shard
The difference between Server and Client Audio in Wokada Deiteris Fork:
- Client: easy to use, can use echo, sup1, sup2 cancellation
- Server: harder to use, can significally reduce delay if used with Wasapi or Asio
yw 🔥
low shard
forget that, I mean just a vr and a pc
you just need to set the vr input to line 1
want me to check your settings?
be sure to not use vb audio cable
swift jungle
want me to check your settings?
Im using a 3050
swift jungle
you just need to set the vr input to line 1
is there some kind of tutorial specifically for it?
not the 2 pc one u sent ofc, prob somewhere in there tho maybe
low shard
Im using a 3050
could you show me a screenshot of your current settings without cropping? i feel like you are using the wrong setup
!give-media-perms 50m @swift jungle
swift jungle
!give-media-perms 50m <@786924005360795648>
I doubt I am and I feel like you're looking for something else than just checking for "wrong" setup but u got it 
low shard
I doubt I am and I feel like you're looking for something else than just checkin...
I'm not sure what you're talking about, I thought you were using vb audio cable since you didn't seem to understand what i meant when I said line 1, but you seem to have it installed already
Be sure to set chunk higher than the perf value after clicking start, be sure to also play with the pitch
I'm not sure if your advanced settings are correct too, but in case they aren't, on wokada deiteris fork, you can **optionally **use more advanced settings for benefits:
- Advanced Settings -> Force FP32 mode: on (THIS IS OFF BY DEFAULT! Turning this on improves stability. Increases VRAM usage by 200 MB)
- Advanced Settings -> Disable JIT compilation: off for faster loading speed of the program, on for slightly better performance (10-15 ms) for Nvidia only)
- Advanced Settings -> Crossfade Lenght: Controls how smoothly the AI stitches different processed parts "chunks" of your voice back together. 0.1 for fastest voice, 0.15 for improved quality but increases delay by ~50 ms
- Reduce the delay on Windows significately via the Wasapi / Asio Guide
about the VR, you can:
- follow the optional advanced settings above to help with performance
- play on lowest graphical settings, 1080p and 60fps cap (like the worst ever you could get)
- then set the vr game's input to line 1 and it should work all fine
Let me know! :D
fleet trail
You won't get help as you got already warned for catfishing few seconds before a...
SO WHAT CAN I DO
low shard
SO WHAT CAN I DO
We don't help catfishers
swift jungle
I'm not sure what you're talking about, I thought you were using vb audio cable ...
Where are the advanced settings
low shard
Where are the advanced settings <:cat_huh:1359900596500431050>
yeah, that's why i always ask for a screenshot, you never know if the other has corrected all settings lol
bottom left
swift jungle
bottom left
ive watched like 20 guides on this including the one that introduces nvidia fork thing, why isnt it on there for the improve performance stuffs
Yeah I went on the website u mentioned to download nvidia-b2332 a longg time ago
low shard
<:yt_nails:1159569314848972891> ive watched like 20 guides on this including the...
https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#advanced-settings, it's in this exact section
swift jungle
And I did the optimizations in task manager
low shard
Yeah I went on the website u mentioned to download nvidia-b2332 a longg time ago
I'm not sure what you're talking about, but when there's a "#" at the almost end of the link, it's because i'm mentioning a specific part/section of the guide (it's not the same as just sending the link)
swift jungle
I'm not sure what you're talking about, but when there's a "#" at the almost end...
I mean like this part
I think I was searching up ways to optimize wokada or smth idk
been like a whole year
low shard
I mean like this part
yeah, i know that, from the screenshot you can check the version you're using
which you're using the latest wokada deiteris fork
fleet trail
We don't help catfishers
i don't catfish
fleet ruin
hello
low shard
i need help
Please Elaborate:
- your PC GPU (Such as Nvidia RTX 3060, AMD RX 9070 XT..)
- your operating system (Like Windows 10, Windows 11, Linux, MacOS 26..)
- what you want to do? There isn't a program that does everything, there's a program for each thing:
- AI Covers
- Train RVC Models
- E girl trolling
- TTS
- Roleplay in VC
- Roleplay in Games
- etc
- what tutorial link are you using
- a screenshot of the program
fleet ruin
i have all of that
low shard
i have all of that
voice changer is too vague, it can mean different programs and versions, and different ways to use programs
swift jungle
hello, do you need any help?
this is good I suppose
low shard
please elaborate what I asked
fleet ruin
roleplay in games
low shard
this is good I suppose
if you aren't going to use an index file, then yeah
low shard
roleplay in games
please elaborate everything asked in #✨│ai-help message , not just a single part
swift jungle
if you aren't going to use an index file, then yeah
Should I just make it like 0.1s then?
low shard
Should I just make it like 0.1s then?
it would be the best you leave it to 0.33
swift jungle
it would be the best you leave it to 0.33
not much different so far as I can see but ig it kicks in at critical situations when the voice changer decides to break in half or smth
or rather my gpu ig
empty nest
hello
knotty moth
well tbh anonymous egirl and eboy voice models are no longer allowed to list
also not many ppl publicly share models
viral mason
why would you wanna be just some dude
fair
I have no idea as I use game characters like from star wars
knotty moth
are you okay with sounding like an old man?
anyway you can post in #1159289738314919936 but only free ones are allowed
nah it actually includes youtubers
so you gotta make a request or try making by your own using some random youtube contents
I'd recommend taking only single speaker voice for consistent quality
fusing a few models may be an option but it would mostly be hit or miss
well I think you can even fuse several models of well-known youtubers & celebs, so wish you good luck
the popular youtuber ones to search may be pewdiepie, ishowspeed, technoblade, asmongold, etc
^
no it doesnt belong to me lol
if you cant even name several known youtubers, just browse in youtube as much as you can
geez you sound like dictating me to search those craps for you
wide oasis
i was about to get him
empty nest
.
marsh stratus
Is there currently a way to get pytorch to use amd gpus on windows?
i tried to use pytorch rocm on wsl but cant seem to get it to work
latent kettle
Update is ready. Pls update running UVR5-UI-updater.bat. If u face an error that...
Alright let's try
sage rapids
what app should i use for a rvc ai model?
wheat pewter
Yeah is it okay if I send it in the morning though?
swift jungle
you just need to set the vr input to line 1
I'm trying to do it right now, do I just plug it in with like a 10ft cable and put vr as input?
spark mesa
hello! i'm unfamiliar with these infos on how i can tell which among these models are the best when it comes to extracting vocals.
can i ask which one among these are the best one to extract vocals?
small sorrel
when is TITAN large coming out?
simple ore
when is TITAN large coming out?
small sorrel
<:misc_lets_fucking_go:1159355870119993496>
wdym
simple ore
not happening
knotty moth
i tried to use pytorch rocm on wsl but cant seem to get it to work
rocm doesnt currently support WSL
maybe in the future version 7
zluda is the best bet you could try to work on
knotty moth
when is TITAN large coming out?
doubt if they are actually working on
small sorrel
doubt if they are actually working on
that sucks
knotty moth
I'm trying to do it right now, do I just plug it in with like a 10ft cable and p...
install virtual audio cable lite to get line 1
I'd suggest the lite version, not trial
get the link here https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork
simple ore
rocm doesnt currently support WSL
what lol?
small sorrel
i made a quandale dingle voice model, trained on ov2super, 40k, 3:53 dataset. can you recommend some ways on how to improve it?
also 300 epochs
simple ore
Is there currently a way to get pytorch to use amd gpus on windows?
follow the AMD's guide how to install its drivers on wsl ubuntu
unfortunately some things are working terribly slow
knotty moth
@small sorrel publicly sharing datasets is NOT allowed
small sorrel
<@845532411482996766> publicly sharing datasets is NOT allowed
why?
chilly furnace
minor isle
is there any websites that convert texts inside a video to a text file for easier copying text
swift jungle
install virtual audio cable lite to get line 1
yeah I already have that set up and nvidia-b2332, so I'd need a cable right, I don't got that yet
next breach
Is it possible to play Spotify music from the microphone so that others can hear it?
north burrow
hello, can someone guve me the link to download the voice changer for win_cuda?
viral topaz
what r the best settings to sound good asf
while using a model
like for the in and out
honest sparrow
yo guys is the best sounding epoch the one with the lowest loss because its only epoch 163
hallow thistle
what r the best settings to sound good asf
!howtoask
thorny plaza
Guys, in RVC v2 Disconnect how to properly set up the custom pretrain, say KLM5 from hugging face, i copied links to G and D files, pasted them in right place & set thing to custom but when i run training it did not detected the pretraining. So dumb question, but what kind of links does it want or is KLM5 not supported on RVC v2 disconnect. ( i wanted 1st time to use diff pretrain than those included, i had generated models in past & yes, i had read the guide)
hallow thistle
Guys, in RVC v2 Disconnect how to properly set up the custom pretrain, say KLM5 ...
Still using RVC v2 Disconnected? This specific RVC fork notebook has been long outdated. #📰│dev-updates message There are other RVC notebooks as alternatives.
hallow thistle
why W-Okada is not working on firefox browser
What do you use W-Okada for? Did you follow a tutorial from YouTube? And what is your PC GPU?
viral topaz
i js dont know what to put into the gain and i nand out and index settings
idk which ones r good numbers
may u please tell me daddy namari
hallow thistle
may u please tell me daddy namari
Read "!howtoask" or help guidelines again.
viral topaz
!howtoask
viral topaz
i js need good sets
hallow thistle
Now answer my questions. What do you use W-Okada for? Trolling for funny or catfishing for your benefit? Which W-Okada version are you trying to use? And what is your PC GPU?
viral topaz
Trolling but it sounds chopped
hallow thistle
Otherwise, I can't help with that.
hallow thistle
Are you sure you use it for trolling?
viral topaz
50% sure
hallow thistle
viral topaz
hallow thistle
There's a better W-Okada version that works with RTX 50 series GPU.
viral topaz
i aint doing allat i js need numbers being told to me broski
hallow thistle
No, I ain't finishing my typing.
viral topaz
YUR I NEED NUMBERS TOLD TO ME BRO
EVERYTHING WORKS PERFECT
I JS NEED
A GOOD IN AND OUT AND GAIN SETS
hallow thistle
Calm down. No need to fucking hurry.
hallow thistle
Still.
hallow thistle
Since you don't follow me, I ain't giving you the settings.
viral topaz
lock it in bro
hallow thistle
If you think you know what you're doing, there you go.
thorny plaza
Still using RVC v2 Disconnected? This specific RVC fork notebook has been long o...
What would be the best one(s) to use nowadays then (that does not run locally, no gpu for that)
hallow thistle
Because you kept asking me for the same thing of your ego, it won't be that deep, I guess. 
hallow thistle
What would be the best one(s) to use nowadays then (that does not run locally, n...
-colab
viral topaz
hallow thistle
There's Applio.
hallow thistle
why?
Any copyrighted material, including your raw audio dataset, is prohibited to upload in here. Otherwise, you can get this entire server banned.
Also, make sure to read server's rules and help guidelines before you start asking anything here.
hallow thistle
hello, can someone guve me the link to download the voice changer for win_cuda?
What do you use W-Okada for? Trolling or catfishing? And what is your NVIDIA GeForce GPU?
hoary verge
What do you use W-Okada for? Did you follow a tutorial from YouTube? And what is...
Yes, I'm using rtx 2060 super
It works in chrome browser but when I switched to Firefox it doesn't
hallow thistle
Yes, I'm using rtx 2060 super
Now what do you use W-Okada for? Just wanna make sure you won't use it in an offensive situation.
odd roost
hey
hallow thistle
Why can't I hear my voice?
!howtoask
odd roost
why i cant send a pic
hallow thistle
why i cant send a pic
Image permission is for level 5 user, so you may answer as words instead.
What do you use W-Okada for? Trolling or catfishing? Did you follow any tutorial video from YouTube before? And what is your PC GPU?
thick ferry
whats the best rvc for rtx 4060 i512400f?
hallow thistle
whats the best rvc for rtx 4060 i512400f?
Huh? What do you use RVC program for?
thick ferry
Huh? What do you use RVC program for?
trolling my friends
hallow thistle
trolling my friends
So which RVC? The normal RVC or realtime voice changer?
thick ferry
So which RVC? The normal RVC or realtime voice changer?
whats the difference?
hallow thistle
nah for run , yes i followed a tutorial , i have nividia gtx 1660 super
Download and use the better W-Okada from this guide instead. https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#download-nvidia-on-windows
hallow thistle
whats the difference?
The difference is that, if you didn't know, RVC is a program where you make AI cover, while realtime voice changer (W-Okada and Vonovox) converts your voice in real time. Also, RVC doesn't always mean realtime voice changer, so these are different.
thick ferry
oh i need realtime ig
odd roost
Download and use the better W-Okada from this guide instead. https://docs.aihub....
it work?
hallow thistle
oh i need realtime ig
For W-Okada, there is the best W-Okada afterall. https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#download-nvidia-on-windows
odd roost
alright
thick ferry
For W-Okada, there is the best W-Okada afterall. https://docs.aihub.gg/realtime-...
thanks
hallow thistle
Try another browser or screenshot your terminal window.
thick ferry
Try another browser or screenshot your terminal window.
what does terminal window mean?
like cmd or shell?
hallow thistle
like cmd or shell?
CMD.
hallow thistle
what is cmd exactly?
Something like this. https://cdn.discordapp.com/attachments/1159290139609137264/1400500146529107978/image.png
thick ferry
oh so i have to keep my terminal open while using the vc?
hallow thistle
Don't mind what in CMD, that was not W-Okada. But this is what Windows Terminal (Windows 11) or CMD looks like.
hallow thistle
oh so i have to keep my terminal open while using the vc?
Yes, when you have W-Okada running in the background.
analog obsidian
I already knew the model maker from before and I knew he had a male model
formal jackal
somehow the outdated version of wokada works nicely for me but the new one kinda suppresses my voice
can anyone help me set up the new version by comparing my settings in the outdated version?
hallow thistle
somehow the outdated version of wokada works nicely for me but the new one kinda...
The difference between Deirtis fork W-Okada and the outdated original W-Okada is mostly performace, while most features all function the same.
What do you use W-Okada for? Trolling or catfish? And what is your PC GPU?
formal jackal
The difference between Deirtis fork W-Okada and the outdated original W-Okada is...
i use it in games to mostly cover my accent
formal jackal
What do you use W-Okada for? Trolling or catfish? And what is your PC GPU?
rtx 4050
hallow thistle
rtx 4050
Did you follow this guide? https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#download-nvidia-on-windows
formal jackal
Did you follow this guide? https://docs.aihub.gg/realtime-voice-changer/local/de...
yup
i've tried various settings for the new version but everything felt the same and i have the perfect settings for the outdated version
hallow thistle
Chunk: 110 - 130 ms
Extra: 2.7 s
GPU: NVIDIA GeForce RTX 4050
These settings for fork W-Okada.
F0 Det: rmvpe
formal jackal
on the old one i have
chunk on 112
extra on 65536
rmvpe
formal jackal
Chunk: 110 - 130 ms
Extra: 2.7 s
GPU: NVIDIA GeForce RTX 4050
yes i use these
doesnt work better than the outdated version
i tried increasing chunk to around 252, no change
hallow thistle
If the audio still in lower quality, it's pretty much that, it can be the voice model itself.
hallow thistle
is there any specific format the voice model should be in?
What format? Do you mean RVC pth file?
odd roost
Yes, it works. Any issue installing or running the program?
thanks you it work but i have a question it work on instgram ?
thick ferry
@hallow thistle is higher chunk better or lower?
formal jackal
What format? Do you mean RVC pth file?
like spin or convec
hallow thistle
<@561807329977434122> is higher chunk better or lower?
Lower chunk number makes less delay, while higher makes more delay. The extra number indicates the audio quality and how much data should process at the time.
unique dew
where can i download RVC?
hallow thistle
thanks you it work but i have a question it work on instgram ?
Desktop or mobile phone? W-Okada won't gonna work on mobile phone, but with OBS Studio on desktop is yes.
bitter cloak
just running the start batch file after installing, previous version didnt do this
odd roost
Desktop or mobile phone? W-Okada won't gonna work on mobile phone, but with OBS ...
desktop yeah how can i do it?
hallow thistle
Did you run as administrator? The program is supposed to run as not-admin (normal user) mode.
bitter cloak
Did you run as administrator? The program is supposed to run as not-admin (norma...
nope, just opening it regularly
:(
hallow thistle
nope, just opening it regularly
Did you run it on CMD or double click directly?
bitter cloak
Did you run it on CMD or double click directly?
double click
i tried dragging it into cmd and its the same problem
hallow thistle
Which Applio version did you use? And have you reinstalled it fron start?
bitter cloak
Which Applio version did you use? And have you reinstalled it fron start?
got it from here https://github.com/IAHispano/Applio via Code->Zip
and yeah i ran the install batch first
analog obsidian
run install worked but run applio didnt?
bitter cloak
run install worked but run applio didnt?
yep lol
analog obsidian
uhhh @simple ore help
bitter cloak
yeah im stumped
bitter cloak
are you using a custom windows version?
i dont think so but i changed hardware recently and used massgrave because it broke my activation
analog obsidian
its trying to run applio as admin for some unknown reason
bitter cloak
ya weird
hallow thistle
like spin or convec
Deiteris fork W-Okada only accepts RVC voice model. Other voice model types go to original W-Okada.
analog obsidian
Deiteris fork W-Okada only accepts RVC voice model. Other voice model types go t...
spin and cvec are embedders
simple ore
1st start takes a few minutes
matplotlib making fonts or something
not sure
you cant start it as admin any more
you get a notice you're dumdum
hallow thistle
spin and cvec are embedders
Really? How do these work in W-Okada?
analog obsidian
wait @bitter cloak you ran the install script as admin?
bitter cloak
wait <@758039921322885312> you ran the install script as admin?
no i just double clicked it without doing anything before that
simple ore
if they ran install as admin, god help them
analog obsidian
Really? How do these work in W-Okada?
a helper that doesnt work what embedders are, classic ai hub
edit: Leo, the other helper, knows what they are tho
hallow thistle
a helper that doesnt work what embedders are, classic ai hub
edit: Leo, the othe...
analog obsidian
cvec works with wokada but spin only in vonovox
hallow thistle
What you say makes me look useless. Sorry if I interpreted it like that.
bitter cloak
wait <@758039921322885312> you ran the install script as admin?
i copied the old run batch and that worked lol
analog obsidian
if they ran install as admin, god help them
he ran the install script as regular user
bitter cloak
idk whats different
analog obsidian
i copied the old run batch and that worked lol
lol
simple ore
then there may be a black window for a few min
dunno what itis doing.. downloading models maybe
bitter cloak
got it working, thank you all 👍
analog obsidian
hallow thistle
Still.
While some people say I know almost every about RVC and W-Okada, it doesn't mean that I know literally everything on earth even the codes inside. I'm just like that, I do my best I can do.
hallow thistle
<@561807329977434122>
Use Virtual Audio Cable lite instead of VB-Cable.
thorny plaza
Can you fool appolio with mklink? or is it not necessary and it will work on other drives and that part in "before downloading" is only a spook? (https://docs.aihub.gg/rvc/local/applio/)
formal jackal
which is better ? client or server?
unique dew
-howtoask
low shard
Can you fool appolio with mklink? or is it not necessary and it will work on oth...
@simple ore is applio fixed with the other drives?
simple ore
You can put it on any drive
formal jackal
what are u talking about?
w okada
simple ore
as long as you're not stupid enough to use spaces, ( ) or other bs in the path name
no OneDrive for sure, .../Desktop/ is a stupid place to store it
low shard
w okada
always be sure to elaborate, this is a general server rather than an ai voice one now
also, you sure you got wokada off the docs, and not video tutorials? because all video tutorials are outdated and give worse performance
low shard
You can put it on any drive
alr thx
simple ore
most people just have C and nothing else
thorny plaza
You can put it on any drive
Aha! Good, thanks, i asked only because of "Make sure that you place Applio inside a folder on C drive." that is stated in manual
formal jackal
always be sure to elaborate, this is a general server rather than an ai voice on...
yeah docs, im using the dec 2024 one
so my question is which audio setting gives better performance, client or server? or are both the same?
simple ore
Aha! Good, thanks, i asked only because of "Make sure that you place Applio insi...
you can symlink logs or assets/audio folders if you want, that should work
low shard
Aha! Good, thanks, i asked only because of "Make sure that you place Applio insi...
it has been fixed and that statement has been removed
simple ore
also I'd recommend setting
low shard
yeah docs, im using the dec 2024 one
so my question is which audio setting gives...
- Client: easy to use, can use echo, sup1, sup2 cancellation
- Server: harder to use, can significally reduce delay if used with Wasapi or Asio
server can give less delay, but only if you use it with wasapi or asio
simple ore
so it does not store million files in C:\users\user\appdata\local\temp
formal jackal
- Client: easy to use, can use echo, sup1, sup2 cancellation
- Server: harder to...
i see, thanks! and are they same performance wise, like cutting out and stuff?
hallow thistle
I install Applio and other Python related programs on E drive.
low shard
i see, thanks! and are they same performance wise, like cutting out and stuff?
performance wise, server is better if used with wasapi, or even better with asio
there's a section about using server with wasapi or asio in the docs
simple ore
I install Applio and other Python related programs on E drive. <a:cat_disco:1359...
Automatic1111 so outdated
hallow thistle
I have Forge zip, but I haven't extracted to there.
simple ore
comfyui if you wanna fiddle with different workflows
sd.next if you dont, it has power
hallow thistle
I also even have ComfyUI zip. This is the whole new folder structures as I formatted the hard drive by accident.
formal jackal
there's a section about using server with wasapi or asio in the docs
for asio, in the docs its mentioned to download a .NET Desktop runtime but the one which is mentioned is outdated... do i go for the outdated one or the latest one?
simple ore
you dont need .net updates, windows does it for you
you may need VC++ redist
but there are some weird issues sometimes with microsoft releasing one version with libomp dll and one without
win11 has this issue
anyway, just try installing the last one, if it says it is already installed you're good
formal jackal
you dont need .net updates, windows does it for you
i see, ty!
high spoke
Hi i wanna put the mommy girl but idk how to put it bec i only have the model pth and i dont have the model.index, (I have too the metadata)
latent kettle
@viscid moss hello, sorry again, but after updating UVR5-UI I'm still not able to use that model.
viscid moss
<@274566299349155851> hello, sorry again, but after updating UVR5-UI I'm still n...
just that one?
Mel-Roformer-Denoise-Aufr33-Aggr I mean
latent kettle
just that one?
i have not tried other but FV4 model And De-reverb models are working fine
viscid moss
Maybe that model got corrupted somehow
go to /models folder and delete it their config file too
latent kettle
delete this
latent kettle
go to /models folder and delete it their config file too
lemme try
viscid moss
yeah pls, that model is fully working for me on colab/hf and locally
viscid moss
go to /models folder and delete it their config file too
denoise_mel_band_roformer_aufr33_aggr_sdr_27.9768.ckpt
denoise_mel_band_roformer_aufr33_aggr_sdr_27.9768_config.yaml
those files
latent kettle
done
viscid moss
nice! Now try again, model will be downloaded again
latent kettle
nice! Now try again, model will be downloaded again
ill let you know if it works or not
viscid moss
yes pls
latent kettle
one thing i want to ask which modle is good for de-echo, i used de-reverb model but still i can hear echo
viscid moss
U mean this one: MelBand Roformer | De-Reverb by anvuew
latent kettle
U mean this one: MelBand Roformer | De-Reverb by anvuew
yes i used it
but still i can hear echo, i want to remove echo, the delay kind of thing, that goes like hell0.....hello...... hello
viscid moss
yes i used it
Hmmm that's still the best one for reverb
viscid moss
but still i can hear echo, i want to remove echo, the delay kind of thing, that ...
in that case try a de-echo model
like: MelBand Roformer | De-Reverb-Echo Fused by Sucial
thorny plaza
If you want RVC for talking W-Okada https://huggingface.co/wok000/vcclient000/tree/main and if u win with nvidia gpu select vcclient_win_cuda_2.0.78-beta.zip
viscid moss
That model should remove both
latent kettle
That model should remove both
what if i want to remove echo only ?
viscid moss
what if i want to remove echo only ?
MelBand Roformer | De-Reverb-Echo by Sucial or MelBand Roformer | De-Reverb-Echo V2 by Sucial
latent kettle
@viscid moss so here we go again
viscid moss
<@274566299349155851> so here we go again
hmmm same error?
latent kettle
viscid moss
Can I ask... Where are u from? U can reply on my DMs if u don't want to say it here
latent kettle
Can I ask... Where are u from? U can reply on my DMs if u don't want to say it h...
yep. so i DM you
low shard
If you want RVC for talking W-Okada https://huggingface.co/wok000/vcclient000/tr...
@uncut pollen no, don't use that one, it's an old version of original wokada,
never use video tutorials for realtime voice changers
you shouldn't use that one at all
low shard
please elaborate what's not working, show screenshots too
thick ferry
please elaborate what's not working, show screenshots too
not working on discord
low shard
alr
be sure to reply to everything
@uncut pollen we can move this convo to https://discord.com/channels/1159260121998827560/1406311449013387334 if you want
you can reply to all i asked in there
low shard
not working on discord
show a screenshot of ur program settings
thick ferry
thick ferry
show a screenshot of ur program settings
low shard
Hi i wanna put the mommy girl but idk how to put it bec i only have the model pt...
are you trying to do e girl trolling?
swift jungle
I'm trying to do it right now, do I just plug it in with like a 10ft cable and p...
@low shard
low shard
monitor: headphones optionally to hear urself
uncheck sup1
did you do the recording and playback tab? https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#virtual-audio-cable
thick ferry
monitor: headphones optionally to hear urself
uncheck sup1
did you do the record...
i did
low shard
i did
can you show a screenshot of it then?
like the recording and playback tab
to be sure you fixed the default devices
thick ferry
is this ok
thick ferry
like the recording and playback tab
rigid tartan
Hey guys can anybody help me thru the voice changer? I dont get a thing
viscid moss
quick question <@274566299349155851>
U should place the path not the link
Path seems like this:
/Content/drive/Mydrive/folder....
viral mason
ah ok
rigid tartan
Hey guys can anybody help me thru the voice changer? I dont get a thing
I mean i put the models and configured the settings but it just doesnt work when i use it on discord like my mic just doesnt work
And no im not doing catfishing im helping my friend for a game voicelines
rigid tartan
And no im not doing catfishing im helping my friend for a game voicelines
Also my gpu is 3050
viral mason
Path seems like this:
/Content/drive/Mydrive/folder....
that worked thx!
rigid tartan
Also my gpu is 3050
Nvidia Geforce RTX 3050 6GB
Also i was using this tutorial
https://youtu.be/SxdnGxicJOg?si=W9Kg_DA8TV9V0VlM
rigid tartan
I dont understand do i just follow this guideline and done?
low shard
And no im not doing catfishing im helping my friend for a game voicelines
you're confusing programs
low shard
And no im not doing catfishing im helping my friend for a game voicelines
RVC = Retrieval-based-Voice-Conversion, the best Few Shots Speech To Speech AI Models (on v2), Inferences (use models) pre-recorded audio (ai covers) and train (make) models. Technically, Mainline RVC does have a go-realtime.bat (aka RVC-GUI), but it's pretty messy and outdated so it's extremely not suggested for realtime. There also updated forks with extra features like Applio.
Wokada = uses RVC for realtime inference. There's 2 main versions, Original made by Wok, and the most suggested one is Deiteris Fork (modified version)
Vonovox = Another Realtime Voice Changer based on RVC, with similar quality and performance to wokada deiteris fork but other perks
your friend needs RVC, not Wokada, wokada isn't for making game voicelines
voice changer is not the same as realtime voice changer
RVC doesn't mean realtime voice changer
rigid tartan
your friend needs RVC, not Wokada, wokada isn't for making game voicelines
voice...
No we need a realtime one
Do i use wokada for realtime?
latent kettle
@red kayak hello, sorry to bother you, but I want to know how can I remove clipping from dataset
low shard
No we need a realtime one
huh? for making game voice lines, for a mod or video, you'd need a pre-recorded program, else you wouldn't be able to use it except in things like a discord call
I think you're confusing up voice changer with realtime voice changer, it's common
rigid tartan
huh? for making game voice lines, for a mod or video, you'd need a pre-recorded ...
We are doing a real time voice example for the funders
Sorry for the lack of my explanation
thorny plaza
We are doing a real time voice example for the funders
Wokada will do you both real time and from a file (voice recording)(those links are to the newer wokada fork)
"How to use
Running locally on Windows
Before you start
1 [If not installed] Download and install 7-Zip or WinRAR. (https://www.7-zip.org/)
2 [If not installed] Download and install VAC Lite by Muzychenko. (https://software.muzychenko.net/freeware/vac470lite.zip)
3 Navigate to the releases section." (https://github.com/deiteris/voice-changer/releases)"
And then you can use this discord (https://discord.com/channels/1159260121998827560/1175430844685484042) to get voice models, or use online tool like (https://colab.research.google.com/github/IAHispano/Applio/blob/main/assets/Applio_NoUI.ipynb#scrollTo=0pKllbPyK_BC) to make your very own voices (easy to use, no set up required) or you can install appolio locally (https://docs.aihub.gg/rvc/local/applio/) but that might not be easy for someone who is green
thorny plaza
Try PC sound options and set virtual output as your main pc output
Make the virtual output default device
thick ferry
Try PC sound options and set virtual output as your main pc output
js turns everything mute
thorny plaza
mmm, tbh i'm not qualified to trouble shot you on this, my friend had problem once with discord that when he was in a call he did not had any other sounds, and it was some windows call thing setting that was responsible. If being in discord call mutes everything, it might be that? I'm not sure
thick ferry
mmm, tbh i'm not qualified to trouble shot you on this, my friend had problem on...
they can hear my normal voice
not voice changer one
thorny plaza
Meaning the default device for PC is wrong and is a mic not the virtual out put, that would be system sound setting
thick ferry
Meaning the default device for PC is wrong and is a mic not the virtual out put,...
output or input
thorny plaza
Hmm, but line 1 do generate sound right?
you checked?
have some sort of, jumping green bar indicator that you got signal on line 1?
thick ferry
Hmm, but line 1 do generate sound right?
this?
thorny plaza
+also, apart form wokada you need to have turned on the VAC tool (software needs to be booted otheriwse virtual sound outputs dont work) (i bealive its called that)
thick ferry
i alr have it
thick ferry
+also, apart form wokada you need to have turned on the VAC tool (software needs...
thats the line 1 port
thorny plaza
Got something like this turned on apart from wokada?
thick ferry
no
thorny plaza
(i use diffrent tool for virtual output)
but same philosophy (i believe)
in recording
thick ferry
?
thick ferry
thick ferry
thorny plaza
as default communicating device.. not line, the VB audio virtual cable
that is the VEC i think
thick ferry
as default communicating device.. not line, the VB audio virtual cable
cant click on both
thick ferry
hmm
thorny plaza
had you restarded PC since you installed it? (in case of mine VAC it requred that)
thick ferry
had you restarded PC since you installed it? (in case of mine VAC it requred tha...
yes
thorny plaza
Had you tried the Cable in 16ch?
thorny plaza
Hmm
Im not sure why its not working for you, i use something diffrent :/
hmm
well
the problem would be that your Virtual Line one is not displaying in the system sound settings, that for one
thorny plaza
:/ You need to get help from someone who uses exact tool mate, sorry
misty pecan
Hey everyone. I've been using the "Deiteris' W Okada Fork" voice changer to troll in games, it has worked for like a month, now suddenly I get no voice output. I have reinstalled every piece of software, checked my input device, and everything seems fine, but still nothing. I've used this guide: https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/
Not sure what's up, but I can't send screenshots in here
snow lance
Hey peoples, i wonder: is there a w-okada fork that supports SPIN? (Trying to understand this i dont know if i am even asking this right)
snow lance
its like, idk if i say this right. but most moddels trained with hubert or contentVec as embedder, but SPIN is different one. and i found models that use that. i cant load them in w-okada. i wonder if there is a fork of w-okada that has this. i cant use vonovox since i dont have nvidea
for what i have heard yes, also anhances nonverbal sounds
low shard
Hey peoples, i wonder: is there a w-okada fork that supports SPIN? (Trying to un...
there is an experimental fork, but has issues compiling and isn't ready (meaning u can't install it), so it would be best you use vonovox for now
snow lance
there is an experimental fork, but has issues compiling and isn't ready (meaning...
i can not use vonovox, i have a AMD gpu and vonovox doesnt have a CPU fallback
hoary verge
Now what do you use W-Okada for? Just wanna make sure you won't use it in an off...
just making fun in games bro
low shard
i can not use vonovox, i have a AMD gpu and vonovox doesnt have a CPU fallback
I did compile tg-develop's deiteris' wokada fork before he added the audio effects and dependencies commits (so it was only spin and rtx 50 serie), but for it's weird dependencies no gpu would be ever detected
you could either test this on CPU only, meaning shit performance, or just wait
spin embedder type models are extremely new anyways, and very very few as of now
snow lance
you could either *test* this on CPU only, meaning shit performance, or just wait
i have been using my cpu for all time now. my cpu can do w-okada at 32 chunks so i should be fine i think?
misty pecan
Hey everyone. I've been using the "Deiteris' W Okada Fork" voice changer to trol...
Update ^^ got it working with another browser, not sure what's up with chrome all of a sudden
low shard
i have been using my cpu for all time now. my cpu can do w-okada at 32 chunks so...
hold on what CPU & GPU u have? it's also strange u rather use cpu rather than gpu 😭
u had previously wokada deiteris fork b2332 im guessing? rather than those video tutorials old ones
snow lance
hold on what CPU & GPU u have? it's also strange u rather use cpu rather than gp...
i tried to get it to detect my gpu but it does not work, i tried for hours, altereed stuff installed destalled reinstalledn othign worked.
CPU: AMD Ryzen 9 9950x 16-Core Processor
GPU: AMD Radeon RX 6900 XT
low shard
i tried to get it to detect my gpu but it does not work, i tried for hours, alte...
yeah i'm guessing you got the old original wokada off youtube
wokada deiteris fork has very good improvements for AMD GPUs
low shard
okada quited or somethign right
huh? if u want u can share the download/tutorial link u used
snow lance
i havent seen updates in forever
snow lance
huh? if u want u can share the download/tutorial link u used
i used one from like 2 years ago but i reinstalled it on my new computer about 1 months ago, got the latest version
low shard
i havent seen updates in forever
original wokada does still get updates, but they are User Interface or beatrice only (no performance involved related to RVC)
Wokada Deiteris Fork isn't receiving any updates since b2332 december 2024, but it's still the best option for most
here's the download to the wokada deiteris fork CPU only spin (which is an earlier version of tg-develop fork) https://huggingface.co/Nick088/Wokada-Deiteris-Fork-Spin/tree/main
The installation would be the same as following the wokada deiteris fork b2332 in https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/ (well, except the download link)
snow lance
So basicly the best i can do to get SPIN support is
either patch it myself into an open source version of w-okada.
Wait for AMD supprot in vonovox (they say comming soon)
Wait for deiteris fork on spin supprot to work
||Buy a nvidea gpu lmao||
snow lance
here's the download to the wokada deiteris fork **CPU** only spin (which is an e...
this is SPIN only with CPU?
low shard
this is SPIN only with CPU?
yeah, because of some weird dependency issue
snow lance
yeah, because of some weird dependency issue
so this will work basicly
snow lance
so u just saved me
thats not an issue for me lmao
havign that for 2 years now
kinda used to it
ima test this
low shard
havign that for 2 years now
shouldn't be the norm 😭
maybe u can have the spin cpu only and both b2332 deiteris versions, so with one you can have better gpu performance, and the other is only for spin models
but hey ur choice
snow lance
this would mean i can use this?
#1364953969389994004 message
snow lance
but hey ur choice
works fine for me xd
low shard
So basicly the best i can do to get SPIN support is
either patch it myself into...
either patch it myself into an open source version of w-okada.
here's the fork https://github.com/tg-develop/voice-changer, i also talked with the developer #🔊│ai-development message
if you can manage to fix the dependencies and update the workflow scripts, then sure
low shard
So basicly the best i can do to get SPIN support is
either patch it myself into...
Buy a nvidea gpu lmao
would suggest that only if u actually care much about AI, since nvidia will ofc be better for ai, but isn't cheap
snow lance
> Buy a nvidea gpu lmao
would suggest that only if u actually care much about AI...
i like using it but making stuff, ehhh i guess ill pass for now xd
but thanks for all this i will make an update (if i dont forget)
low shard
this would mean i can use this?
https://discord.com/channels/1159260121998827560...
no, that's a pretrain model, not a model
pretrains are just used as a base for training a model
you can find models in #1175430844685484042
but as i said, there's extremely very few spin models
exactly 2 public models (i say public bc people can just not upload their models)
unless you train your own private one
low shard
its like, idk if i say this right. but most moddels trained with hubert or conte...
@grand tangle
the default and original embedder type model for RVC is ContentVec
spin is a newer one and experimental, it could result in clearer and more accurate pronunciation and can help for realtime models
low shard
I did compile tg-develop's deiteris' wokada fork before he added the audio effec...
@grand tangle check from this convo to below
yeah, and unless u train models on ur own, ud only find 2 models trained with the spin embedder model publically lol
i expected people to make more spin models since lyery's spin pretrains, but seems like not, atleast not yet
snow lance
no, that's a *pretrain* model, not a model
so then what is the reasoning for this upload? Can no one use this then?
low shard
so then what is the reasoning for this upload? Can no one use this then?
you can use it only as a base for training your own models
it's not a normal model, it's a pretrain
- Pretrains models: used only as a base for training voice models,it contains multiple voices and tons tons of hours to train and also a larger dataset
- normal voice model: trained using a pretrain, requires less time and fewer dataset, it's used by the users
snow lance
that clears up so much confusion
i was so confused on the whole ppoint of that forums existance
low shard
i was so confused on the whole ppoint of that forums existance
oh lol
desert beacon
Hey guys, does anyone use Voiceflow here?
low shard
that clears up so much confusion
well, wether u want to use still spin or not, id suggest to atleast use wokada deiteris fork
snow lance
well, wether u want to use still spin or not, id suggest to atleast use wokada d...
im going to dive deeper into SPIN models and maybe on training one. Thank you for ur help.
low shard
im going to dive deeper into SPIN models and maybe on training one. Thank you fo...
you're welcome, and https://docs.aihub.gg got all info for model training and other things
novel tide
soo i got a question. on the weights website does buying a plan give you more of a maximum for audios when training a model
low shard
soo i got a question. on the weights website does buying a plan give you more of...
buying a plan offers more dataset lenght and more trainings
novel tide
buying a plan offers more dataset lenght and more trainings
how much does each plan add
cus i see it there but it dont make sense
viral mason
soo i got a question. on the weights website does buying a plan give you more of...
there are much better free options that give better results, I wouldn't recommend spending any money on weights
I could teach you how to train easily using applio
low shard
how much does each plan add
iirc for basic 10 mins, and pro 30 mins but i'm not 100% sure, for more info it's better to check discord.gg/weights
novel tide
oh mb i aint know there was a server for weights ill ask there thank you
gloomy sentinel
i got the wrong version
low shard
i got the wrong version
This is a General AI Server, AI has many fields, we can't know which you're talking about
Please Elaborate:
- your PC GPU (Such as Nvidia RTX 3060, AMD RX 9070 XT..)
- your operating system (Like Windows 10, Windows 11, Linux, MacOS 26..)
- what you want to do? There isn't a program that does everything, there's a program for each thing:
- AI Covers
- Train RVC Models
- E girl trolling
- TTS
- Roleplay in VC
- Roleplay in Games
- etc
- what tutorial link are you using
- a screenshot of the program
desert tide
when i run the system to open the application it just shows for one second and goes away and the application doesnt open any help?
marsh stratus
huh, i managed to get torch to use cuda on wsl but the preformance is as bad or even worse than if i ran it using my cpu natively on windows. maybe i should bite the bullet and boot linux on my desktop
forest cypress
hey is there any free tools in which you can change certian thing in an image ? like changing the color of clothes by circling it changing hair etc ?
latent kettle
hey is there any free tools in which you can change certian thing in an image ? ...
If you have good pc you can use flux fill or flux kontext
crude bolt
<@356887060503724032>
what is a voicechanger
grim dragon
how do i fix white screen in the voicechanger
simple torrent
is batch size 16 okay for a 23 minute dataset
simple ore
huh, i managed to get torch to use cuda on wsl but the preformance is as bad or ...
why do you want cuda on wsl?
hallow thistle
hey guys which voice changer do you recommend?
What do you use realtime voice changer for?
!howtoask
ember arrow
What do you use realtime voice changer for?
talk to my friends and experiencing new things?
rigid tartan
And then you can use this discord (https://discord.com/channels/1159260121998827...
after putting the models and downloading those what do i do???
hallow thistle
talk to my friends and experiencing new things?
Is that a question? Is it trolling or catfishing? And what is your PC GPU?
ember arrow
Is that a question? Is it trolling or catfishing? And what is your PC GPU?
not a question these are my goals
I have an RTX3050
hallow thistle
ember arrow
what?
hallow thistle
What do I mean? The realtime voice changer is W-Okada. Follow this guide if you wanna know more. https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#download-nvidia-on-windows Let me know if you have any issue installing or running the program.
ember arrow
okay thanks
hallow thistle
i got the wrong version
What's wrong?
gray timber
i notice a significant delay when speaking
*about 3-5 secs
i do have a 4070, how do i fix it
hallow thistle
i notice a significant delay when speaking
!howtoask
hallow thistle
What do you use W-Okada for? And did you follow any YouTube tutorial before?
grim dragon
how do i get less delay while still having no lag and cuts in like my voice
gray timber
i just want to have fun with it n test out the voices xD, ive found out about it with this video https://www.youtube.com/watch?v=KkDjqteIqBE
hallow thistle
how do i get less delay while still having no lag and cuts in like my voice
Read "!howtoask" before you start asking.
hallow thistle
i just want to have fun with it n test out the voices xD, ive found out about it...
Never follow any tutorials from YouTube, most of which only telling you to install old W-Okada versions.
gray timber
Never follow any tutorials from YouTube, most of which only telling you to insta...
i'm sorry 🙏 thats why i came here to know how to use this program
hallow thistle
i'm sorry 🙏 thats why i came here to know how to use this program
It's fine. Although there's a better W-Okada version to try.
gray timber
It's fine. Although there's a better W-Okada version to try.
i downloaded it from here https://huggingface.co/wok000/vcclient000/tree/main
it says its version 2.1.4 cuda
grim dragon
i know i have a bad gpu but i still need help pls
Full GPU Name: (nvidia geforce gtx 1650)
Operating System: ( Windows 11)
Detailed Description: im trying to get no delay but my voice cuts when i put less chunk can i get some help?
hallow thistle
i downloaded it from here https://huggingface.co/wok000/vcclient000/tree/main
it...
This is the original version of W-Okada which, as I said, it's outdated, especially 1.5.x.x and 2.1.x versions.
gray timber
This is the original version of W-Okada which, as I said, it's outdated, especia...
oh 🙏 i apologize once again. do you know which version is better
hallow thistle
Full GPU Name: (nvidia geforce gtx 1650)
Operating System: ( Windows 11)
Detaile...
NVIDIA GeForce GTX 1650 is not that bad, although it can be slower than some RTX 20 GPUs at certain situations. What do you use "W-Okada" for? Trolling or catfishing?
grim dragon
NVIDIA GeForce GTX 1650 is not that bad, although it can be slower than some RTX...
trolling and catfishing
😭
hallow thistle
trolling and catfishing
What? Let's be real.
hallow thistle
Catfishing simply means to fake yourself as someone's else and go talk other person to gather their private information, which is bad.
grim dragon
Catfishing simply means to fake yourself as someone's else and go talk other per...
no i dont
i troll
but i installed the vcclient_win_std_2.0.78-beta.zip
gray timber
This is the original version of W-Okada which, as I said, it's outdated, especia...
oh 🙏 i apologize once again. do you know which version is better
grim dragon
but i installed the vcclient_win_std_2.0.78-beta.zip
the newest one
hallow thistle
but i installed the vcclient_win_std_2.0.78-beta.zip
This is the original W-Okada version, the true latest version of it, but also outdated.
grim dragon
This is the original W-Okada version, the true latest version of it, but also ou...
then where do i install the best working one?
hallow thistle
could u send a link
Download the better W-Okada fork from this guide. https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#download-nvidia-on-windows
grim dragon
alright
ty
but im in the github site now
now what do i install
i clicked the Download NVIDIA RTX 5000-series on Windows
oh nvm
ty @hallow thistle
gray timber
what do i do after downloading it?
hallow thistle
i clicked the Download NVIDIA RTX 5000-series on Windows
hallow thistle
what do i do after downloading it?
Please be productive by explain about your progress instead of asking every little step.
grim dragon
oh alr
hallow thistle
After finish download the voice changer zip, you extract it to a folder by 7-Zip or WinRAR.
Let me know if you're looking for settings, or an issue when install or running the program.
gray timber
okay, great. i got it to work, but there's still a bit of delay. do you know how i can reduce it to the minimum?
hallow thistle
okay, great. i got it to work, but there's still a bit of delay. do you know how...
On W-Okada:
Chunk: around 50 - 80 ms chunk
Extra: 2.7s
GPU: NVIDIA GeForce RTX 4070
F0: rmvpe
hallow thistle
whats the difference between all the other f0 dets?
You wanna know all of these pretrain models?
"rmvpe" is for NVIDIA GPU and is the best of all pretrain models, while "rmvpe_onnx" is for AMD/Intel GPU but not as the same level as regular "rmvpe".
The other pretrain models are older ones. Some of which are worse than rmvpe.
rain epoch
im using audio cable but i cant hear myself in discord
hallow thistle
im using audio cable but i cant hear myself in discord
!howtoask
hallow thistle
Make sure to read "howtoask" or help guidelines before you ask anything here.
ember arrow
I have a problem while downloading
I followed the tutorial
But when I run "start_http.bat" it doesnt do anything
@hallow thistle
hallow thistle
But when I run "start_http.bat" it doesnt do anything
Fork W-Okada doesn't need a bat file to run the program.
If you still following the tutorial from YouTube and not aihub website, better stop.
ember arrow
If you still following the tutorial from YouTube and not aihub website, better s...
okay
oh nvm its working
just took time to start
hallow thistle
The typical fork W-Okada should look like mine, not the one that has a lot of files in the same folder. Fork W-Okada should open a browser instead of a separate app program.
rich moss
heyy should I train my model for a French voice with a pretrain or without one? Which gives the best results?
hallow thistle
Ill reset my whole system then
Delete the original and outdated one and install fork W-Okada from start, real quick. https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#download-nvidia-on-windows
rich moss
heyy should I train my model for a French voice with a pretrain or without one? ...
@hallow thistle pls ?
simple torrent
is batch size 16 okay for a 23 minute dataset
hallow thistle
heyy should I train my model for a French voice with a pretrain or without one? ...
What voice model?
rich moss
What voice model?
I’m training a French voice using Applio
hallow thistle
Pretrained voice model is very crucial for voice model training. I'm not sure which "custom" pretrained voice model is best for French vocals, but some said the original pretrains would give the better result, while other custom pretrains can be buggy.
next geyser
vac or other cable
hallow thistle
I don't really train a voice model in RVC, so I kinda forgot how it should work.
hallow thistle
vac or other cable
Virtual Audio Cable.
next geyser
Virtual Audio Cable.
is there a way to make voices sound less robotic and more natural
hallow thistle
is there a way to make voices sound less robotic and more natural
!howtoask
hallow thistle
What do you use W-Okada for? Trolling or catfishing? What is your PC GPU? And which W-Okada version do you use?
next geyser
What do you use W-Okada for? Trolling or catfishing? What is your PC GPU? And wh...
sorry for not knowing how to ask
I wanna use it for fun, not catfishing (with my friends) m,y pc gpu is 6700xt,
here are my settings!
and win 11
hallow thistle
Chunk: 70 - 100 ms
Extra: 2.7 s
GPU: AMD Radeon RX 6700 XT
F0 Det: rmvpe_onnx
next geyser
Chunk: 70 - 100 ms
Extra: 2.7 s
GPU: AMD Radeon RX 6700 XT
F0 Det: rmvpe_onnx
is this fork the best
or should i get emojikage
hallow thistle
You're using the right W-Okada fork version, which is a great.
thick ferry
does anyone have the download for vac?
hallow thistle
does anyone have the download for vac?
The thing is, both "Emojikage" and Deiteris are the same person, who developed his fork W-Okada.
thick ferry
https://software.muzychenko.net/freeware/vac470lite.zip
thanks
next geyser
The thing is, both "Emojikage" and Deiteris are the same person, who developed h...
where do i put the metadata.json
simple torrent
is batch size 16 okay for a 23 minute dataset
@hallow thistle
next geyser
hallow thistle
where do i put the metadata.json
You don't need to upload a metadata json file into W-Okada. A pth file is needed.
hallow thistle
<@561807329977434122>
Your perf number is red and unstable, try increase chunk up until the perf number is green and stable. https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#finding-my-own-settings-for-chunk
thick ferry
Your perf number is red and unstable, try increase chunk up until the perf numbe...
marvel rivals was causing it its fine now
hallow thistle
marvel rivals was causing it its fine now
This looks fine now. The Marvel Rivals is a graphic-intensive game for real.
While this might not needed, you can set "monitor" to your speakers/headphones to hear your W-Okada audio, although you can set this to "none" later.
thick ferry
This looks fine now. The Marvel Rivals is a graphic-intensive game for real.
yeah even with a 4060 i dont optimize it on ultra
@hallow thistle i tried it and its js my normal voicce
i started it nd everything
hallow thistle
No need to ping me again. I am around.
thick ferry
alright sorry
hallow thistle
Make sure you have "passthru" disabled by default, no blinking red from the button. Otherwise, try another voice model or hear yourself on Discord instead.
thick ferry
Make sure you have "passthru" disabled by default, no blinking red from the butt...
i can hear the echo on website js not on disccord
next geyser
namari it still kinda sounds robotic irono
wait
how do i get drid of the trial thing in virtual audio cable
nvm i got it
its supposed to be lite
hallow thistle
how do i get drid of the trial thing in virtual audio cable
You used the "trial" version of Virtual Audio Cable. Uninstall that one and install lite version instead. https://software.muzychenko.net/freeware/vac470lite.zip
hallow thistle
How about screenshot of your W-Okada?
gray dirge
how do i use this lol?
next geyser
How about screenshot of your W-Okada?
its still robotic ://
hallow thistle
its still robotic ://
Yes, it's still robotic. But like what about your screenshot of your W-Okada? A screenshot means an image captured from a screen.
next geyser
Yes, it's still robotic. But like what about your screenshot of your W-Okada? A ...
i know what a screenshot means :))))))
do you want a voice recording?
hallow thistle
Did you do this as a sarcasm? I wanted one, not every thing.
next geyser
Did you do this as a sarcasm? I wanted one, not every thing.
no, dont you need everything?
im sorry
hallow thistle
The perf number is yellow on your W-Okada. Try increase chunk number up until perf number is green and stable.
hallow thistle
If the audio is still low quality, it's usually the voice model itself, which was trained in lower quality from start. Try switch to a better "Megumin KonoSuba" voice model that is newer and better than the one you're using.
next geyser
If the audio is still low quality, it's usually the voice model itself, which wa...
its megumi jjk
hallow thistle
hallow thistle
Well, there's not much you can do about it.
short nacelle
it takess foreverrr to playback my voice.
low shard
it takess foreverrr to playback my voice.
!howtoask
forest cypress
If you have good pc you can use flux fill or flux kontext
can i get free alternatives ?
strange mountain
where the link newest version?
gleaming latch
why cant people hear me on discord 💔
low shard
where the link newest version?
this is a general server, you need to elaborate
low shard
why cant people hear me on discord 💔
Please Elaborate:
- your PC GPU (Such as Nvidia RTX 3060, AMD RX 9070 XT..)
- your operating system (Like Windows 10, Windows 11, Linux, MacOS 26..)
- what you want to do? There isn't a program that does everything, there's a program for each thing:
- AI Covers
- Train RVC Models
- E girl trolling
- TTS
- Roleplay in VC
- Roleplay in Games
- etc
- what tutorial link are you using
- a screenshot of the program
next geyser
Please Elaborate:
- your PC GPU (Such as Nvidia RTX 3060, AMD RX 9070 XT..)
- yo...
what program is there for rolaplay ijn vc