okok
#✨│ai-help
uneven wraith
ion get the index
viral mason
Just steal one from another model
uneven wraith
oh ok
uneven wraith
Just steal one from another model
how do i make my voice smoother?
tulip quiver
Hi, I need help. I'm using the voice modulator, but it's very slow and crashes too much.
viral mason
Hi, I need help. I'm using the voice modulator, but it's very slow and crashes t...
What gpu do u have, and did you download it from a YouTube tutorial
I'll tell ya now it's outdated by years if u got it from YouTube
tulip quiver
What gpu do u have, and did you download it from a YouTube tutorial
I don't know what GPU
and I can't send screenshots
tulip quiver
ok ready and then?
tulip quiver
Sometimes it is at 100 and other times at 0
tulip quiver
what's urs say
ok
viral mason
what
viral mason
wasuup
viral mason
forrrr what exactly
viral mason
vonovox only works with Nvidia, but Wokada tg fork and deiteris works with amd as well as nvidia
viral mason
if that's what u meant
analog ferry
And it isn't working for some reason
viral mason
that's like
have u been to egypt
that's where u got that
from the tombs
super old and outdated friend
analog ferry
LMAO
viral mason
what gpu do ya have
viral mason
I gotr it from some dudes video
makes sense, all yt videos use outdated stuff
viral mason
radeon 6800
radeon is amd right?
analog ferry
ye
analog ferry
makes sense, all yt videos use outdated stuff<:pepe_sad_sit:1424312993092472906>
do i ddelete this ?
viral mason
second link
viral mason
do i ddelete this ?
ye it's old
analog ferry
apprec
viral mason
np
analog ferry
hopefully i dont encounter anything
viral mason
read the guide for tg fork and just ask me questions if u got any
analog ferry
got you
tulip quiver
well what's ur gpu lol
cpu is terrible
viral mason
then prob not done yet
analog ferry
bronze veldt
CAN SOMEONE Help?
viral mason
CAN SOMEONE Help?
with what
analog ferry
Its been like this for 5-6 minutes
viral mason
this should open the app soon
analog ferry
got you
viral mason
if it doesn't do anything try pressing enter on the command promt
idk if that helps
but sometimes it does smth to make it remember it's running
viral mason
u don't
bronze veldt
wdym
viral mason
basically you're out of luck if u have no gpu
gotta get your wallet out and go buy one off amazon
viral mason
alr that's the voice changer
bronze veldt
basically you're out of luck if u have no gpu
beatrice model
analog ferry
Its asking me for a module
viral mason
Its asking me for a module
a what
analog ferry
where do i get them from @viral mason
viral mason
beatrice model
nobody makes beatrice stuff
analog ferry
im trying to start a server
viral mason
where do i get them from <@1023278814752677918>
oh a model
viral mason
download one from herehttps://discord.com/channels/1159260121998827560/1175430844685484042
search for whatever u wanna use
analog ferry
search for whatever u wanna use
ok i downloaded it where do i go from here @viral mason
Its a rvc
and its in a zip
viral mason
unzip the model into a new folder
analog ferry
did
viral mason
I don't exactly know how to upload it to that voice changer in particular as I haven't used it
it might be in the guide
analog ferry
i found it out
all you have to do is litearlly drag it
its asking for my input
and a output
would this just be my headphones
or do i make it my virutal cable
viral mason
would this just be my headphones
input would be your regular microphone for your headset/headphones and output is line 1
analog ferry
viral mason
change your cpu to what gpu u have and change rmvpe onnx to regular rmvpe
analog ferry
how do i download stuff
of the weights app
@viral mason
this is my last question
and for some reason i can't use my headphones when using the server thing
viral mason
I wouldn't really recommend downloading models off weights because you have to rename the models since they will all be called "model"
analog ferry
and virutal cable isnt working with my discord
analog ferry
I wouldn't really recommend downloading models off weights because you have to r...
thats fine with me
i just wanted to test it
viral mason
thats fine with me
this right here
analog ferry
viral mason
and virutal cable isnt working with my discord
try this on discord
input: line 1
output: headphones
analog ferry
its not giving me the file
analog ferry
i did it its doing this
analog ferry
^
viral mason
that's for searching to make covers
analog ferry
oh i click download anyways
viral mason
analog ferry
listened to all the previews
they sound good
then i use the module and it just sounds like a robot
viral mason
scroll down maybe? again I haven't used it before
so I'm unsure about it all
here's my weights profile if u wanna look through my stuff
analog ferry
ok ill use your profile
the little miss boomie one
it'll sound robot for me for some reason bro
ill gi9ve it one more chance till i wrap things up
viral mason
change your cpu to what gpu u have and change rmvpe onnx to regular rmvpe
the settings right here, chunk size, extra processing time
match these
analog ferry
i did those
i set it to my gpu
ye it sounds robotic
unless i have to change these settings
viral mason
match these
that is a pic of wokada deiteris but the same settings will work on the one u opened
analog ferry
alright
thanks for everything
i lowkey give up tho
taking too long to setting and just simply isn't working for me
appreciate the help again @viral mason
silver zodiac
I want to train my voice and create cover songs can someone help me how to do it
formal wind
how do i re-apply for model maker, i used to post in this server but took a break
the channel isnt there anymore so now im lost loll
fickle bough
anybody know how to get it to work with fl studio? or any videos out there for that by chance?
viral mason
anybody know how to get it to work with fl studio? or any videos out there for t...
ye
viral mason
I can help in dms
viral mason
could u help me with that by chance
if that's ok with u
fickle bough
if that's ok with u
yeah bet ill dm you
viral mason
@pseudo zenith
tawny radish
is there nay way i can make male voicemodels sound good? all of them i tried sound so bad
hot lagoon
What’s the minimum requirement for UVR and rvc
fossil sage
whats the best version of gpt sovit
raven phoenix
anyone know how to stop the random ass lip smacking sounds after i say smth in a vc
modest terrace
i have a amd gpu a rx7600 and the voice changer does nothing to my voice after i start it can someone help me setup?
steady vigil
yoyoyo
i need major
help
like
like my cable output for some reason
pics up my mic
before the voice changer even reacts
makes no sense
:((
hallow thistle
makes no sense
!howtoask
hallow thistle
i have a amd gpu a rx7600 and the voice changer does nothing to my voice after i...
Which W-Okada DirectML version do you use?
modest terrace
Which W-Okada DirectML version do you use?
im using the std version
hallow thistle
What’s the minimum requirement for UVR and rvc
NVIDIA GeForce RTX 20 desktop GPU is typically a suggested minimum for RVC, UVR5 and even Stable Diffusion, but RTX 30 and greater can be preferred for speed.
hallow thistle
im using the std version
I don't remember any W-Okada version is made by that name, so I think you might try this one from this guide link. https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#download-amd-intel-and-cpu-on-windows
Let me know if you have any issue running the program or looking for settings.
fleet marsh
where is the pages of the rvc to make covers?
hallow thistle
where is the pages of the rvc to make covers?
You might be looking for this one. https://docs.aihub.gg/essentials/how-to-make-ai-cover/
Not needed. 
fleet marsh
You might be looking for this one. https://docs.aihub.gg/essentials/how-to-make-...
no, the list of pages, i dont remember the command for it
like the applio page and stuff
hallow thistle
no, the list of pages, i dont remember the command for it
-rvc
hallow thistle
This list has a URL to Applio website, if you mean so.
fleet marsh
This list has a URL to Applio website, if you mean so.
where are the google collab thingies?
hallow thistle
where are the google collab thingies?
-colab
fleet marsh
oh thx
hallow thistle
This is the list for usable Google Colab notebooks.
tawny radish
is this new? i never seen all of this
hallow thistle
is this new? i never seen all of this
That is pretty much the same list of Colab notebooks as much I remember, with a few of them rearranged in order.
formal wind
is fork w okada still the best voice changer for amd or is there a better one
i needa catch back up on the ai knowledge
viral mason
is fork w okada still the best voice changer for amd or is there a better one
Check out w-Okada tg fork
Second one
-rt
no thanks
quiet axle
hey i was wondering if theres anyway to use vc at your phone.
i doubt it can run in local, but isnt there any other way like cloud
viral mason
<@1023278814752677918>
I was asleep
quiet axle
i got some weird question. is there anyway to make the voice changer sounds like it came from a bad phone mic
viral mason
If u use voicemod or fl studio with it but that requires a computer
viral mason
It would have to be one of the voice changers that are provided in the voicemod app
I don't use it often so I'm not sure
quiet axle
oh snap i thought you meant voicemeeter sorry got it messed up lol
hmm using voicemod and wokada, is that even possible
viral mason
hmm using voicemod and wokada, is that even possible
Yup!
viral mason
oh snap i thought you meant voicemeeter sorry got it messed up lol
The setup i use with voicemod and fl studio has voice meeter in it, as well as vb cable and vac lite
sleek edge
this is the start bat and i did it 10 times
viral mason
At this point go ask Dr87 in the Vonovox server
He's the creator and could help u better than anyone else
The discord should be in the guide
-rt
hollow moon
can any1 elp me decide whether to go for chatgpt plus or gemini pro?
balmy vale
does someone maybe have a solution if I use a AI voice changer my normal voice is still hearable?
viral mason
does someone maybe have a solution if I use a AI voice changer my normal voice i...
Are you trying to make it where it's only the ai voice?
balmy vale
Are you trying to make it where it's only the ai voice?
yuh
wicked dome
https://www.youtube.com/watch?v=lRCP2GYrsAs how to make pulser visualizer ?
wispy lake
hello i wanted to know how do i know which pretrain model is best for specific stuff like singing/rapping for artist with very unique tones
viral mason
yuh
What do your voice settings look like in the voice changer?
Sorry I kinda fell asleep
balmy vale
What do your voice settings look like in the voice changer?
uhhh
w
can I send a picture here?
elder pagoda
Is it normal to see a 4 hour queue? idk if I'm bugged, or if it's just taking this long, but it's taking a long while
viral mason
Is it normal to see a 4 hour queue? idk if I'm bugged, or if it's just taking th...
what's this for
elder pagoda
what's this for
It's for Weights
viral mason
It's for Weights
are you training a model?
if u are, stop because weights is the worst thing ever to train on, u should use applio
elder pagoda
are you training a model?
No, I'm converting a file into the shown voice
I know not to use weights to train voices 💀
viral mason
thank god
elder pagoda
But the queue is taking forever. 4 hours of waiting
elder pagoda
Oh
viral mason
so they're forcing credits
elder pagoda
Oof
viral mason
just download replay to avoid the issue
elder pagoda
Replay?
viral mason
it's weights if it was good and on your computer
viral mason
how are you on that-
viral mason
ohh u use laptop, not sure how well it will work on a laptop
elder pagoda
Yeah
viral mason
I guessss u could try still, although idk what the link would be I got the download from Nick I believe
viral mason
good luck Big the Cat
gaunt plaza
yo i need help
knotty moth
I don't think replay works on Windows 8 tho 💀
win 8 is actually dogshit, just get win 10, even a decade old laptop can still run it
but still, if your pc spec is too old (older than GTX 1060/RTX 2060), you can't run locally unless you afford slow processing in cpu mode
simple ore
I ran win8.1 for quite a long time, too bad AMD did drop the driver support
switched from win7 mostly for iso mounting reason
formal wind
everytime i click on any of these links in kaggle's applio it opens one of those links at random,
Lemme explain more. So lets say I click on the top link. It'll either bring me to applio, tensor board, or file browser. Its just random so it'll be different at times. I've tried clicking every link to see what happens and its just the same page that was randomly picked at first. At times it'll change to the right one but then it'll go back to the wrong one.
I've tried restarting the notebook, creating a new acc, and trying a new ngrok token
torpid phoenix
whats actually the best ai with almost no rate limit: like gemini that we can talk to it in like an full day
but that is actually good as an coding tool
hallow thistle
win 8 is actually dogshit, just get win 10, even a decade old laptop can still r...
I use Windows 11. 
knotty moth
I use Windows 11. <:kokonacry:1142151932199317594>
they prob have an old pc, even if they find a way to bypass TPM, the win 11 bloatware would suffocate it unless considering tiny11 image
winged moat
Does anyone know a free online Ai cover maker
sturdy dawn
anyone know how to fix pipeline generations for the AI?
hushed mauve
for kaggle w-okada is it files only of both variables and files? cuz aihub and kaggle both say smth different
golden walrus
guys. can i ask where can i download eddie's UVR? 
craggy bough
guys. can i ask where can i download eddie's UVR? <:SCWWcrying:63289394927029454...
golden walrus
https://github.com/Eddycrack864/UVR5-UI/releases/tag/v1.8.4
thank you kind sir
twin ice
can anyone help me get some good settings
craggy marsh
does anyone has guide on finetune tts?
hallow thistle
can anyone help me get some good settings
What is your PC GPU? Did you follow any tutorial before this?
!howtoask
twin ice
What is your PC GPU? Did you follow any tutorial before this?
amd gpu
windows 11
no tutorial
amd 7800xt
hallow thistle
amd gpu
To check your AMD Radeon GPU, open Task Manager, go to Performance tab, spot GPU 0 or GPU 1 in the left panel and click any of these to reveal its full name on the right panel.
twin ice
amd 7800xt
.
hallow thistle
Which W-Okada DirectML version are you using by the way?
hallow thistle
https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/
On your Deiteris fork W-Okada, here's your settings:
Chunk: around 60 - 80 ms; lower than this can be possible if your GPU can handle that
Extra: 2.7 s
GPU: AMD Radeon RX 7800 XT
F0 Det: rmvpe_onnx
https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#finding-my-own-settings-for-chunk
twin ice
ty
hushed mauve
On your Deiteris fork W-Okada, here's your settings:
Chunk: around 60 - 80 ms; l...
why does my ping keep increasing in the online version?
whoops wrong ping
hallow thistle
why does my ping keep increasing in the online version?
You have now asked the right person, just wrong message lol. Which W-Okada Kaggle notebook are you using currently? 
hushed mauve
You have now asked the right person, just wrong message lol. Which W-Okada Kaggl...
the one in the docs https://www.kaggle.com/code/suneku/voice-changer-public
hallow thistle
the one in the docs https://www.kaggle.com/code/suneku/voice-changer-public
In Suneku (Deiteris) W-Okada Colab notebook, there's a list of few recommened settings said in the first code cell. Some people just ignore this and go on to focus on the app itself.
hallow thistle
If you have set your notebook session to run with "T4 x 2" right now, on your W-Okada frontend, make sure to set the "GPU" to a T4.
hushed mauve
If you have set your notebook session to run with "T4 x 2" right now, on your W-...
okay thanks, also does lowering the chunk reduce delay or is it the opposite?
hallow thistle
okay thanks, also does lowering the chunk reduce delay or is it the opposite?
Reducing chunk number to reduce audio delay is possible if the GPU (T4) can handle that and the perf/audio is stable.
hushed mauve
Reducing chunk number to reduce audio delay is possible if the GPU (T4) can hand...
oh ok, also do i need to re run everything everytime i use the voice changer? or is there an easier way
hallow thistle
oh ok, also do i need to re run everything everytime i use the voice changer? or...
If you disconnect and delete the current runtime session and start a new one, yes.
hushed mauve
If you disconnect and delete the current runtime session and start a new one, ye...
oh okay tysm. also last question 
, i was confused as to put the persistence to files only or both. cuz it says 2 different things on both websites
hallow thistle
oh okay tysm. also last question <:skull_sob:1159372531992645662> , i was confus...
Any choice would work, but better stick with the ones that say "files".
hushed mauve
Any choice would work, but better stick with the ones that say "files".
okay thanks again
hallow thistle
You're welcome. 
unreal karma
what are the absolute best settings for rvc web ui (have an nvidia gpu and intel processor)
for quality I mean
hallow thistle
what are the absolute best settings for rvc web ui (have an nvidia gpu and intel...
W-Okada the realtime voice changer or some RVC GUI forks?
unreal karma
I am using this for the first time so I just donwloaded pinokio and the RVC app there 😅
hallow thistle
unreal karma
I just need it to be good enough for this one project haha
hallow thistle
Do you remember the name of that RVC fork you use? Is it Ilaria RVC or Applio?
!howtoask
unreal karma
it's literally just rvc.pinokio.git
I can't upload a pic here but I have transpose -12->12; resample 0->4800; volume envelope scaling 0->1; median filtering to harvested pitch results 0->7, search feature ratio 0->1; pm vs harvest vs crepe vs rmvpe
and protect voiceless consonants and breath sound 0->0.5
hallow thistle
The "rvc.pinokio" doesn't look like a typical RVC fork, but rather a launcher kind of extension for an RVC program.
hallow thistle
I found out some other engineers/mods were advising against using Pinokio with compiled Applio because the Pinokio thing was somehow causing Applio to bugged, as Applio RVC program itself could have just run directly by double click on the batch file.
hallow thistle
That's pretty much I know.
unreal karma
I found out some other engineers/mods were advising against using Pinokio with c...
oh I see. Can you please briefly explain which of those settings would give me the best results because I need to finish this pretty soon? 😅
I might mess around with other rvc later on
hallow thistle
For the actual Applio, consider this guide. https://docs.aihub.gg/rvc/local/applio/ There are some other RVC forks (also including the original RVC mainline), but Applio is more preferred.
hallow thistle
oh I see. Can you please briefly explain which of those settings would give me t...
- the transpose is where you set the pitch (-24 is lower pitch, 0 is default, while +24 is higher) of the audio to specific pitch.
- resample (32000, 40000 and 48000 Hz) is the sample rate of audio; the higher sample rate (48000 hz) can get at higher frequencies, but most RVC voice models are trained at 40000 Hz so the audio might output as what it is.
- pm, harvest, crepe and rmvpe are F0 models that used to do pitch extraction from an audio. rmvpe is more preferred because it's better than any other F0 model, while others could process faster but also give worse results.
Any other settings than these (like volume envelope scaling, median filtering, search feature ratio and protect voiceless), I have no idea what they are supposed to, so my best is just to leave them default unless necessary.
subtle tusk
Okay guess I'll ask here
I'm a vocalist, I do metal vocals and I can't sing for shit
I've recently made a cover of Tears Don't Fall, using one of the female voice models for the singing and it worked really well.
I'm here to ask what voice models would be good for singing, RVC models as I use w-okadia
I already figured out how to sing to get the voice changer to handle it which is just mono tone with some pitch change but nothing major or it has a stroke.
Looking for male and female voice models recommendations as I don't have the memory to find stuff, nor am I cultured, I'm a white american male, so I'm bout as dumb as it gets.
tawdry needle
what is the best female voice model
viral mason
what is the best female voice model
what are you up to
formal wind
what is the best female voice model
Are you perchance wanting to catfish
viral mason
there's two options, either they wanna be a girl for money/trolling, they wanna be a girl because they're trans, or they wanna be a girl because they don't like their voice (the second and third one is the only correct option)
formal wind
there's two options, either they wanna be a girl for money/trolling, they wanna ...
I thought you said 2 options not 3
You're readbaiting me rn
viral mason
shh I edited it mid typing and forgot to change that first part
formal wind
Why do people even catfish
No one's ever catching me being a girl
Unless it's a character from a game then sure but I normally tell people I'm a guy when people ask
viral mason
Why do people even catfish
because they're losers
formal wind
Frr
viral mason
I had a huge crashout over one that I do not regret one single bit
inner harbor
Does anyone’s know where I can use my RVC model and throw a mp4 over it so it can like sing a song or something for test samples?
viral mason
Does anyone’s know where I can use my RVC model and throw a mp4 over it so it ca...
no mp4s but mp3s work fine without having to clean the audio of bg music
https://www.weights.com
just make an acc and u can upload your model ther or just use applio, applio tho u have to clean any song/sample audio u use of bg music ect
kaggle applio is broken again I've deleted like 5 notebooks changed ngrok token idk what's up with the code but it's fucked, jumping off a bridge time
I hope a fix for this is made soon
it was literally working earlier today
silver ether
i downloaded the voice changer but its lit just my normal voice it not changing even on the basic models can anyone help me
subtle tusk
i downloaded the voice changer but its lit just my normal voice it not changing ...
w-okadia?
Well im bout to be heading to bed
If you are using w-okadia:
Upper right, under the bar that has your voice models, you should see Tune, and a bar.
If you have a deeper voice than the model you are using, try 12 for the tune.
If you have a higher pitched voice than what the modal you are using has, go into the negatives.
If you're still having trouble finding the tune bar, its right next to the picture of the model, and right under the bar that shows your models
If you need more help, I can do so tomorrow, or someone else can help.
Theres also youtube
viral mason
Well im bout to be heading to bed
If you are using w-okadia:
Upper right, under ...
don't use yt tutorials they use outdated software
subtle tusk
RTX 2060
I mean I can send you a dm of something I did using the voice changer
viral mason
-rt
viral mason
use wokada deiteris fork
third link
don't use the original old outdated wokada
deiteris runs better and is better optimized
subtle tusk
Uh huh
Well my software works soooo
knotty moth
i downloaded the voice changer but its lit just my normal voice it not changing ...
don't read the @subtle tusk 's hallucination, you just need to click passthru button to change to green
also could I ask what is your gpu and the version being used?
limpid depot
Heyo does anyone know if there any ai autotune that can runs live ?
viral mason
Uh huh
Well my software works soooo
u don't want better stuff? sounds like a personal problem then
subtle tusk
don't read the <@1273348557054673029> 's hallucination, you just need to click p...
Rtx 2060
Don't know what version but it works great for me
Sounds good.
subtle tusk
u don't want better stuff? sounds like a personal problem then
Well im sorry not everyone can understand coding and other shit like that
knotty moth
Rtx 2060
Don't know what version but it works great for me
Sounds good.
no not you
subtle tusk
Fuck you too then
I know how to use the damn software
This is why I'm slowly just not helping people anymore
I get bitched at and told off
viral mason
Well im sorry not everyone can understand coding and other shit like that
It has nothing to do with coding, it's just better optimized, and not outdated tech, but you have already proved you don't care so I'm sorry you suck
subtle tusk
Fucking came here for voice models and then decided to help out with shit
Honeslty yall can just give me a special role that makes it so I can't talk.
That way you dont have my bitch ass to deal with and I can have my voice models
subtle tusk
It has nothing to do with coding, it's just better optimized, and not outdated t...
Also i didnt fucking know about this softwares existentance I found it from youtube
subtle tusk
Looking at the damn thing, it looks like i do in fact have the right version.
subtle tusk
Yeah
Though I have the Nvidia one as I have a rtx card
viral mason
my bad I grabbed the wrong link
btw @subtle tusk your wokada looks like this right? minus the pictures and stuff
subtle tusk
A touch different as you changed the graphics on it
Im not getting up to go grab a screen shot of mine as im about to go to bed
viral mason
I saw u were heading to bed before all this, I don't wanna keep you up
viral mason
A touch different as you changed the graphics on it
wdym changed the graphics btw, like made it dark mode or smth
subtle tusk
I may be a version older, but I have that layout
subtle tusk
wdym changed the graphics btw, like made it dark mode or smth
Dark mode and the bars are different
viral mason
mk, sleep well sorry about being annoying if I was
subtle tusk
But I mean it works well
The female vocals are the voice changer
subtle tusk
mk, sleep well sorry about being annoying if I was
Eh sorry for being an asshole
I got a hell of a temper
viral mason
Eh sorry for being an asshole
I got a hell of a temper
you're good man, it happens
subtle tusk
But I mean it works well
The female vocals are the voice changer
Tbh I was expecting this to be something that would not sound good at all
But I uh, I did something actually kinda good
Anyways goodnight
viral mason
graceful shore
help 😭
also hello everybody
i have no clue how to work this otaku realtime voice changer on collab
the second step is busteedd
any help is greatly appreciated
glass estuary
if i have a amd gpu and i download tg-devlop fork what file do i down load in the github
viral mason
i have no clue how to work this otaku realtime voice changer on collab
is your pc not strong enough to handle it locally? :(
graceful shore
very much so
i went ahead and tried the deities w okada fork colab thing
i think it working
except i cant hear a thing
viral mason
that's odd, what mic settings do u have set
graceful shore
currently: input- communications headset output-headphones
both are on my earbuds
viral mason
currently: input- communications headset output-headphones
hmm I'm not entirely sure if this will work but most of the time for okada it's used with vac lite to put it though other games or calls, but I dunno if that would work on the browser version
graceful shore
ah so should i test it on lets say discord mic test and see if it works then? i did download that virtual mic that was on the forums
*docs
viral mason
sure
tulip cloak
yo @simple ore any tweakings i need to do in settings of the applio no ui when training? cus when i trained it sounded nothing like the dataset, dataset isnt the problem that i know
simple ore
yo <@155030383648440320> any tweakings i need to do in settings of the applio no...
I can only suspect you've done something wrong
tulip cloak
I can only suspect you've done something wrong
which part
simple ore
show your work 🙂
viral mason
yo <@155030383648440320> any tweakings i need to do in settings of the applio no...
what's ur current setting look like?
tulip cloak
show your work 🙂
what work
tulip cloak
what's ur current setting look like?
default
as it is
i didnt touch anything in settings
i would have trained on rvc disconnected on usual but it doesnt work anymore bc of some missing code or something
viral mason
i didnt touch anything in settings
prob why, training requires specific settings for each model
glass estuary
anyone now what file i should download on the tg-devlop github
tulip cloak
prob why, training requires specific settings for each model
oh damn
how would i know what to use though
glass estuary
what does specified sample rate is not supported by all selected audio devices. Available sample rates: [Input]: [16000, 32000, 44100, 48000, 96000, 192000] [Output]: [] [Monitor]: [16000, 32000, 44100, 48000, 96000, 192000]
simple ore
what does specified sample rate is not supported by all selected audio devices....
show tghe screenshot
glass estuary
show tghe screenshot
cant send in here
viral mason
how would i know what to use though
just look at what the dataset size is, length wise, in the first part of applio always open advanced settings and turn off or on the stuff in the photo
nimble blaze
I'm not home currently so I can't exactly look but I have downloaded multiple RVC things to use with edge TTS, but all of the RVC stuff seemingly doesn't have this RVC_infer that is apparently needed. I believe I'm using Python 3.10, I might have a number wrong but it's whatever version is required to avoid dependency conflicts or whatever the heck. I just need to find the correct RVC thing to download. I already have a pre-trained model with an index and path file
Where exactly can I find RVC things to go through to have the correct version with an RVC_infer file. I know there's a better word for thing but I can't think right now
Edit: forgot to mention this is for use in Python
hallow thistle
I'm not home currently so I can't exactly look but I have downloaded multiple RV...
"Applio RVC fork" is known to have Edge TTS built in within the app itself, while some other RVC forks may require you to manually install Edge TTS into them, including coding, which could get bugged easily. See Applio guide. https://docs.aihub.gg/rvc/local/applio/
nimble blaze
"Applio RVC fork" is known to have Edge TTS built in within the app itself, whil...
I'll definitely check it out. I'm just wanting to piece together a custom AI assistant as I'm learning Python. In the process of dealing with All of the dependency issues I was told to start using a virtual environment to contain things. Which I definitely understand now
simple ore
cant send in here
not what I asked, show the application UI
outer isle
Artifact
covert dome
yo guys, how y'all make an ai model voice ? i wanna make one
radiant rapids
what if my changer wont load
full pond
Hello! i'm currently running RVC on my laptop i have a few questions
1st- do i have to keep the localhost tab running for the RVC to work
2nd- some optimal settings (nvidia gpu 2050, 8gigs ddr5)
simple ore
Hello! i'm currently running RVC on my laptop i have a few questions
1st- do i h...
RVC <> Realtime Voice Changer
full pond
RVC <> Realtime Voice Changer
huh-
silk wave
hey guys does anyone know how to train rvc model with the nvidia 50 series?
im using https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI
and get the "Unfortunately, there is no compatible GPU available to support your training"
when choosng the gpu
i got a 5060 8gb
simple ore
hey guys does anyone know how to train rvc model with the nvidia 50 series?
im u...
how tf you find a project with the last update date two years ago and decide oh man, i'm gonna use it with my spanking brand new 5000 series card?
hallow thistle
huh-
RVC doesn't always mean realtime voice changer. There are two different programs; RVC typically means Retrieval-based Voice Conversion, which includes Applio, while the realtime voice changer is a voice changer (W-Okada and Vonovox for example).
full pond
voice change, mainly for voice over/roleplay shenanigans
full pond
RVC doesn't always mean realtime voice changer. There are two different programs...
ouh- okok sorry for that
im using realtime
voice changer
simple ore
the reason you have browser window with UI is because the voice changer has server and a client components
the UI is done in the browser because devs did not want to mess with standalone app for windows/linux/mac
and the server part can run on a remote server with a good GPU technically
full pond
i only have a single machine so, i guess i gotta keep running the browser in the bg
but regardless, what settings are like the normal. considering the gpu i have
simple ore
hey guys does anyone know how to train rvc model with the nvidia 50 series?
im u...
get applio 3.5.0 here https://huggingface.co/IAHispano/Applio/tree/main/Compiled/Windows
hallow thistle
LOL
That was sarcastic. The "RVC" GitHub you found is the original mainline RVC GUI. The latest version of Applio RVC is known to work with RTX 50 GPU, if you didn't know.
silk wave
That was sarcastic. The "RVC" GitHub you found is the original mainline RVC GUI....
ok im gonna give it a try, thanks
mystic salmon
how do i instll
acoustic forum
hey can anyone help me
vast timber
when i talk it will talk for like 1 second then stop then talk again for 1 second
low shard
when i talk it will talk for like 1 second then stop then talk again for 1 secon...
please elaborate
!howtoask
low shard
hey can anyone help me
yes, if you elaborate
low shard
how do i instll
install what?
viral mason
install what?
I'm assuming voice changer, I'm betting $0
If u win u get 0 if I win I keep my 0 
empty leaf
Any voice changers working on Snapdragon Laptops??
low shard
Any voice changers working on Snapdragon Laptops??
nope, realtime or not, unless you use cloud (remote good pc)
the most supported GPU for AI generally is a dedicated nvidia gpu
fringe blaze
i ran the start_https and MMVC_Rest and VoicechangerManager have been taking a while to initialize , anyone know why ? the rest finished instantly
empty leaf
nope, realtime or not, unless you use cloud (remote good pc)
Aw man that's sad. Not even clownfish?
low shard
Aw man that's sad. Not even clownfish?
we do not suggest such programs
empty leaf
All I want to do is lower my pitch so it sounds a bit deeper. For singing and stuff
low shard
All I want to do is lower my pitch so it sounds a bit deeper. For singing and st...
you want a voice changer for pre-recorded audio / songs, right?
empty leaf
All I want to do is lower my pitch so it sounds a bit deeper. For singing and st...
Don't even want to change my voice to something specific, simply lowering the pitch by -6 or so
low shard
there is a difference between realtime voice changer and voice changer for pre-recorded audios
empty leaf
you want a voice changer for pre-recorded audio / songs, right?
I'd like it real-time
low shard
Don't even want to change my voice to something specific, simply lowering the p...
those are AI Voice Changers, they use an AI Trained voice models
they don't just simply lower your pitch
fringe blaze
i ran the start_https and MMVC_Rest and VoicechangerManager have been taking a w...
?
empty leaf
they don't just simply lower your pitch
How do I find ones that do just that for snapdragon?
viral mason
How do I find ones that do just that for snapdragon?
just download voicemod
viral mason
might need an upgrade 
empty leaf
Upgrade how?
viral mason
new device
empty leaf
I bought this device 3 days ago
Everything works perfectly
Every app works nicely
Battery lasts me the whole day on the laptop
The single thing that doesn't work is voice changers 
viral mason
oop
I mean u coulllld still get your money back, laptops are pretty bad for any kinda of voice changing stuff, i'd always suggest getting a pc over latpop
empty leaf
Might need to buy the voicemod dongle then...
empty leaf
I mean u coulllld still get your money back, laptops are pretty bad for any kind...
I do have a beefy desktop device back at home. I moved because of university and required a laptop. I'll eventually build another desktop pc, but the laptop was more pressing for uni
I even have voicemod premium, always used that and wokada on my desktop pc
But neither work on snapdragon laptops
Shouldn't be an issue once I build a desktop but until then I won't be able to use it
viral mason
I do have a beefy desktop device back at home. I moved because of university and...
that's fair honestly
I couldn't live without my pc (I have no irl friends and only have work bc highschool traumatized me so badly me that I'm never going back) so if I had a laptop that could run all the stuff I need it to I'd be good even if I had to be away
empty leaf
I'm kind of perplexed that no simple program exists that serves the only purpose of changing your microphones pitch. Making whatever sound goes in deeper or higher. That's all I need
viral mason
kaggle applio is broken again I've deleted like 5 notebooks changed ngrok token ...
btw @low shard when will applio be fixed
empty leaf
I couldn't live without my pc (I have no irl friends and only have work bc highs...
I got no irl friends either but sadly the laptop runs everything I need but voice changers lol
If it did that I might not even need a normal desktop pc
viral mason
I got no irl friends either but sadly the laptop runs everything I need but voic...
we should be friends :D
btw I'd probably switch to a hp laptop or something more wellknown if u got the money as I haven't heard of snapdragon until now
empty leaf
I got no irl friends either but sadly the laptop runs everything I need but voic...
I'm crashing out over this voice changer compatibility issues im losing it
empty leaf
we should be friends :D
Aye
viral mason
only snapdragon I know is this guy
empty leaf
btw I'd probably switch to a hp laptop or something more wellknown if u got the ...
I got the Microsoft Surface Laptop 7
It's the rival of the MacBook air
Only downside is the chips being only 1 year old and having some compatibility issues
But it has stupendous battery life
Normally laptops would only get ya like 3 hours
empty leaf
only snapdragon I know is this guy
Xd
viral mason
I got the Microsoft Surface Laptop 7
I use whatever this thing is lol
viral mason
Ye only downside is the fact I need a dongle or whatever that is to connect a mouse to it because of how thin it is
empty leaf
I just wanted one with very good battery life, a nice screen, touch, haptic touchpad and a nice build quality and after searching the surface 7 ticked all of those for me
empty leaf
Ye only downside is the fact I need a dongle or whatever that is to connect a mo...
Does it not have built in ports?
This is the Surface 7
viral mason
Ye I need touchscreen and decent battery life as I game and draw on my laptop
empty leaf
Nice
viral mason
Does it not have built in ports?
Nah they're all type c and no mouse does that
Unless I went to wireless
Which could be glitchy
empty leaf
Nah they're all type c and no mouse does that
I just use an adapter for my mouse
empty leaf
You should give it a try xd
empty leaf
Oh the one im using came with my microphone (Zealsound K66)
But any USB A to USB C adapter should work fine
viral mason
I'll try and find one sometime since the cord is honestly really silly because how long it is
Unnecessary for just a mouse
empty leaf
Yea
low shard
I bought this device 3 days ago
Everything works perfectly
Every app works nice...
Snapdragon are ARM64 chips, not X64 chips
ARM64 is more battery optimized
X64 instead is more, vulgarly said, performance optimized
the majority of mobile devices is arm64
The majority of desktop devices run on x64 usually
So companies, products and people focus on deploying programs for the majority
That's why
plus, AI tasks require usually an expensive computer with a modern Nvidia dedicated GPU, AI is pretty intensive
low shard
kaggle applio is broken again I've deleted like 5 notebooks changed ngrok token ...
this happens only when you're using the ngrok tunnel? you tried the other tunnels? or have you tried lightning.ai applio ui?
empty leaf
Yeah. I just wanted good battery. I don't really game on laptops because to get decent performance I'd have to spend 2k+. Surface 7 does all I need. Except a program to pitch the voice ofc lol
Photoshop works like a dream on the snapdragon laptop
All I really need
viral mason
this happens only when you're using the ngrok tunnel? you tried the other tunnel...
I'm unsure, I could try with horizon
I'm never switching back to lightning
low shard
I'm never switching back to lightning
why not? it seems more stable than kaggle for applio ui, especially recently
viral mason
why not? it seems more stable than kaggle for applio ui, especially recently
Too confusing and I can't even turn it off so it runs down my usage even after I close the webpage
low shard
Too confusing and I can't even turn it off so it runs down my usage even after I...
I can't even turn it off so it runs down my usage even after I close the webpage
it automatically goes to sleep 10 mins after inactivity (that you closed the page, this doesn't affect when code is running) + you can just click the "processor" icon and turn it off, that's all
@viral mason yes, that's how you would setup the horizon tunnel
viral mason
ty
low shard
i deleted your screenshot because your censoring wasn't censoring well, so to protect your privacy and token
viral mason
Ah crap
low shard
Ah crap
you should check https://www.reddit.com/r/CensoringIsHard/
viral mason
you should check https://www.reddit.com/r/CensoringIsHard/ <:misc_trolley:115946...
what's this btw
oh nvm it dissapeard
and I got the same error
why does it hate mee
is there something wrong with the code
low shard
what's this btw
downloading/updating npm + horizon
low shard
and I got the same error
the issue seems related to Applio, not the tunnels
I will check it soon
@nocturne mural you might want to check this too
viral mason
Ty, I shall be waiting patiently
Is this error happening to anyone else or am I the only one getting it so far
simple ore
Is this error happening to anyone else or am I the only one getting it so far
it is probably the realtime component of sorts missing
you can edit app.py and exclude the realtime tab which does not work anyway
edit app.py and delete line from tabs.realtime.realtime import realtime_tab
viral mason
I aint doing that every time
low shard
you can edit app.py and exclude the realtime tab which does not work anyway
wait what? the realtime tab got broken?
viral mason
there should be a new one made that removes that since it's broken
simple ore
names is a variable defined in tabs.inference.inference
not sure why it fails to import
twin estuary
how can i lower the ram usage of the real time voice changer ?
@viral mason how can i lower the ram usage of the real time voice changer ?
viral mason
<@1023278814752677918> how can i lower the ram usage of the real time voice chan...
what voice changer are u using, and what's your gpu
also important question, did u get the voice changer from a youtube tutorial
viral mason
phew
twin estuary
what voice changer are u using, and what's your gpu
its like u launch and it lead u to site
viral mason
its like u launch and it lead u to site
sounds like w-okada deiteris fork
twin estuary
what voice changer are u using, and what's your gpu
3060 ti
viral mason
what settings do u have on the voice changer?
twin estuary
wait
@viral mason
its like when i launch it and then launch another game with like high ram usage my pc start lagging i cant fix that ?
viral mason
<@1023278814752677918>
oh yea, so try turning down the graphics of the game to the lowest u can first and see how well it works
twin estuary
oh yea, so try turning down the graphics of the game to the lowest u can first a...
graphics are low
any other ideas to lower the lag ?
if not its okay
itsjust bc sometimes this goes like to 2000
NOW looks fine
what i i change it to cpu
will that be better ?@viral mason
NVM gpu better
perf went to 500
viral mason
what i i change it to cpu
maybe ur voice changer settings, I have mine like this but I do have a 5070ti
try that out
twin estuary
okay thank uuu
viral mason
okay thank uuu
np ^^
nocturne mural
<@989772388508000306> you might want to check this too
nocturne mural
and I got the same error
?
viral mason
try to change this when starting the machine
nocturne mural
I recommend that in persistence you only keep the files, and then create a separate cell just to reinstall the dependencies.
Do you have anything important saved in that notebook?
viral mason
Do you have anything important saved in that notebook?
nope it's a new one
viral mason
I recommend that in persistence you only keep the files, and then create a separ...
I legit have no clue how to do that and prefer that it just, work like it should :(
nocturne mural
nope it's a new one
Press this button
and run the cells again, both for installation and to start the application.
viral mason
will try that 👍
low shard
will try that 👍
it works fine for me too
viral mason
Press this button
pressing that did not work, can a new one be made without realtime code at all, just for training
I did re run the cells after pressing it btw
same error
nocturne mural
pressing that did not work, can a new one be made without realtime code at all, ...
Try this.
viral mason
if I have to stop training my files will be sent to hell
nocturne mural
umm
nocturne mural
if I have to stop training my files will be sent to hell
!sed -i '/from tabs.realtime.realtime import realtime_tab/d' app.py
!sed -i '/with gr.Tab(i18n("Realtime")):/{N;d;}' app.py```?
nocturne mural
Create a new cell, run this, and it will remove the realtime.
viral mason
that needs to be imported to a diff applioapy
nocturne mural
viral mason
that needs to be imported to a diff applioapy
I meant a diff version is made just with that in it automatically
bc the glitch is annoying
nocturne mural
I meant a diff version is made just with that in it automatically
That simply modifies the app.py file, so you can run the cell that starts applio without any changes.
viral mason
I'll have to do this with each time I import the notebook now :(
hopefully next update fixes this
nocturne mural
I'll have to do this with each time I import the notebook now :(
No, since you have persistence configured to save your files, this ‘change’ will simply remain saved.
viral mason
No, since you have persistence configured to save your files, this ‘change’ will...
I delete the notebook each time a model is complete and I have downloaded the files. it's so I have storage for the next model when I train a new one
btw do I put it before cell 1 or two
nocturne mural
I delete the notebook each time a model is complete and I have downloaded the fi...
To avoid doing that, simply do this: !rm -r /kaggle/working/program_ml/logs.
viral mason
To avoid doing that, simply do this: `!rm -r /kaggle/working/program_ml/logs`.
that's confusing
nocturne mural
that's confusing
That command is used to delete the Applio repository, so you save the time of deleting and recreating the notebook.”
viral mason
would having files only save the index and pth files without the extra stuff kaggle makes?
nocturne mural
would having files only save the index and pth files without the extra stuff kag...
I didn’t understand what you said. If you’re referring to persistence — no, that’s used to save all the files located in the /kaggle/working path. That means it will save all the files from the Applio repository, including your model or any dataset you’ve uploaded.
Additionally, you can run these three commands:
!rm -r /kaggle/working/program_mlto delete the entire repository,!rm -r /kaggle/working/program_ml/logsto delete the logs folder, including your trained model!rm -r /kaggle/working/program_ml/assets/datasetsto delete the datasets you’ve uploaded.
viral mason
I think i'd rather just suffer with copy pasting that command to delete realtime every time I make a new notebook bc I don't get it
so do I put the command before cell 1 or 2
viral mason
so do I put the command before cell 1 or 2
@nocturne mural srry for the ping, just need to know if it matters where the code goes that removes realtime
analog obsidian
That command is used to delete the Applio repository, so you save the time of de...
edit: yeah the suggestion below is better
just make applio realtime its own thing, uh like mainline realtime gui
low tangle
looked for an answer to this and didnt see a conclusive one...training on applio colab and it says "model trained successfully" like a few seconds after pressing start training. tensorboard isnt working so i assume it's not doing anything. then i looked back at the colab page and even though everything in the applio ui said it went successfully, im seeing these errors repeatedly:
/usr/local/lib/python3.11/dist-packages/torch/utils/data/dataloader.py:626: UserWarning: This DataLoader will create 4 worker processes in total. Our suggested max number of worker in current system is 2, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary.
warnings.warn(
Not enough data present in the training set. Perhaps you forgot to slice the audio files in preprocess?
An error occurred extracting the index: need at least one array to concatenate
If you are running this code in a virtual environment, make sure you have enough GPU available to generate the Index file.```
i have about 30 minutes of data and used the recommended settings in the aihub applio tutorial. sorry if this question is redundant, just having trouble figuring this out. i can put it on the forum if neededyou've failed preprocess step
probably by providing a file instead of a path
low tangle
i put the path like /content/drive/MyDrive/.../dataset.wav, should i remove the filename at the end? i left it cuz they had it that way in the tut
low tangle
is it an issue that it says on 00:25:06 seconds of audio when i have a lot more than that
low tangle
yea 👍 wasnt sure if it was milliseconds
low shard
<@911742715019001897> not sure who made this
it was by Eddy last year, I should just replace the audio file with extension with only the dataset folder right?
simple ore
it was by Eddy last year, I should just replace the audio file with extension wi...
yep, just a folder path
viral mason
I have fixed the kaggle issue by putting the code in after cell 1🧖♂️
simple ore
cell order does not matter, the order you run them does
viral lodge
Is there a system architecture diagram of RVC?
strong crypt
hey i kinda have a problem like i selected my mic and my headphones but this aint working
idk whats up
could someone help
dusky ravine
I selected my mic as input and in game as the vac audio cable and there’s no sound coming through the audio cable.
simple ore
Is there a system architecture diagram of RVC?
viral lodge
https://gudgud96.github.io/2024/09/26/annotated-rvc/
Perfect, thank you!
gloomy lynx
Every time I try and start training I keep getting this error, and Idk what any of it means
Initializing the generator with 109 speakers.
Using HiFi-GAN vocoder
Using AdamW optimizer
Using Single-Scale Mel loss function
Using Float16 for training.
Starting training...
Loaded pretrained (G) 'rvc\models\pretraineds\custom\G_LSpeechV2.5.pth'
The parameters of the pretrain model such as the sample rate or architecture do not match the selected model.
Error(s) in loading state_dict for Synthesizer:
size mismatch for dec.ups.0.parametrizations.weight.original1: copying a param with shape torch.Size([512, 256, 20]) from checkpoint, the shape in current model is torch.Size([512, 256, 24]).
size mismatch for dec.ups.1.parametrizations.weight.original1: copying a param with shape torch.Size([256, 128, 16]) from checkpoint, the shape in current model is torch.Size([256, 128, 20]).
size mismatch for dec.noise_convs.0.weight: copying a param with shape torch.Size([256, 1, 64]) from checkpoint, the shape in current model is torch.Size([256, 1, 80]).
size mismatch for enc_q.pre.weight: copying a param with shape torch.Size([192, 513, 1]) from checkpoint, the shape in current model is torch.Size([192, 1025, 1]).
simple ore
Every time I try and start training I keep getting this error, and Idk what any ...
you need to delete the model folder and preprocess at 32k
because the pretrain is 32k and not 40k as you prepared
gloomy lynx
Like you mean delete the entire models folder?
simple ore
Like you mean delete the entire models folder?
yeah, since your files are sliced incorrectly
gloomy lynx
yeah, since your files are sliced incorrectly
Do I gotta redownload my pretrain model then?
simple ore
Do I gotta redownload my pretrain model then?
no
analog obsidian
just do the whole preprocessing again
with another name, and select 32000 in sample rate
gloomy lynx
Alright im pretty sure its working, thank you guys
hallow thistle
I selected my mic as input and in game as the vac audio cable and there’s no sou...
!howtoask
hallow thistle
hey i kinda have a problem like i selected my mic and my headphones but this ain...
Read help guidelines before start asking. If you're looking for W-Okada the realtime voice changer, what are your PC GPU and operating system? Did you follow any tutorial before this? And which W-Okada version you tried to run? Or anything else? 
lime canopy
i have a bit of a question, so ive been making rmvpe models for sometime now and well, sometimes when it comes to girl voices they can sound a bit airy, could it be that i leave in the breath in and out sounds from after i process the acapellas to be completely dry? or the samples i sometimes leave in that sound a little compressed?
opaque pecan
hey i wanted to ask someone know an uncensored ai? like i pay for chatgpt monthly and that have a lot censore ship and i want to find an uncensored ai where i can do what ever i want pls help me guyssss chatgpt is soo pisses me offfff
craggy marsh
hey i wanted to ask someone know an uncensored ai? like i pay for chatgpt monthl...
openrouter also have a uncensored model, i have yet tried it though
craggy bough
hey i wanted to ask someone know an uncensored ai? like i pay for chatgpt monthl...
buy a gpu and run a model locally
hallow thistle
hey i wanted to ask someone know an uncensored ai? like i pay for chatgpt monthl...
There are free LLM models availabe to run either locally (on your PC) or online host, each have their own pros, but these models won't be the same level as OpenAI ChatGPT or Google Gemini.
low shard
yep, just a folder path
fixed
low shard
hey i kinda have a problem like i selected my mic and my headphones but this ain...
!howtoask
dim jewel
how to add voice models to w-okoda?
hallow thistle
dim jewel
when I click upload, nothing happens
hallow thistle
In my screenshot, the one I use is Deiteris W-Okada. If you're still using the original W-Okada which opens its own separate window, better switch to a better one.
!howtoask
primal steeple
I cant open the voice changer, it just says this is the cmd: Please open the following URL in your browser.
http://<IP>:<PORT>/
In many cases, it will launch when you access any of the following URLs.
any1 help?
low shard
I cant open the voice changer, it just says this is the cmd: Please open the fo...
!howtoask
primal steeple
GPU: rtx 5060 ti 16gb
Windows 11
I was just trying to download the voice changer, and at first it worked but now that I try to open it the cmd just says:
" Please open the following URL in your browser.
http://<IP>:<PORT>/
In many cases, it will launch when you access any of the following URLs."
The voice changer opens, but its just white and doesnt show anything. I cant send a screenshot in this channel for some reason but like I said, the client is just white and nothing pops up.
knotty moth
GPU: rtx 5060 ti 16gb
Windows 11
I was just trying to download the voice changer...
try one of the following first three options depending on your gpu spec and there shouldn't be issue like that
-rt
uncut atlas
can someone help me
spare wren
n8n error Problem in node ‘HTTP Request‘
viral mason
-rt
viral mason
@brittle wing since u have AMD I recommend Wokada deiteris fork, the third guide
brittle wing
Hi, Im using RVC Client and most of the models except one (I have tried 6) seem to sound cutting the mic off all the time, but Im clearly speaking to the microphone
Oh, I see. Im going to try that
viral mason
read the guide carefully and download the amd version of the voice changer since there are diff versions, just scroll down a bit to find the downloads for the different versions ^^
brittle wing
okey ^^
viral mason
if u need any help just ask here or me
brittle wing
so, I have kinda same issue, voice cuts off seems working bad
Apparently I watched a video as well and it says its works good when the red number that says 394 its on green and below 100 ideal, but not sure how can I improve it
@viral mason do you know if some setting will improve that thingy?
idk but on yt video works good when he gets it like this:
I think its this but idk how to improve it
viral mason
so, I have kinda same issue, voice cuts off seems working bad
change your cpu to gpu
brittle wing
viral mason
your pitch is probably too high as well
brittle wing
pitch is on 11 right now
hallow thistle
@silver walrus for W-Okada, better ask in here.
brittle wing
im male, trying female voice
hallow thistle
GPU: 0: AMD Radeon RX 6600M
viral mason
im male, trying female voice
most female voices sound good at 3, 6, or 10 at highest
brittle wing
oh works better with GPU, number got green
okey it got better
even now with green thing and trying the different pitches, it sounds robotic sometimes
like it legit voice sometimes dies off in middle of sentence
hallow thistle
brittle wing
i mean yeah better than before, still sucks tbh but better 😄
maybe model I use also not so good
is there any good model you guys recommend?
hallow thistle
Try raise extra number up to 2.7 s.
brittle wing
I changed some settings idk
extra to 2.7s, okey ima try that
yeah got a bit better
viral mason
I changed some settings idk
try these
brittle wing
whats annoying me is that the voice is like, sometimes breaking a bit as if it didnt recognise my words properly. But my english is good and Im doing very simple sentences Im reading somewhere, should not be complicated
okey ima try that
oh no, numbers got red because my perf is about 250 ms most of time
so in my case it seems chunk needs higher than my perf so cant do 232ms, i can do like 300, 350, seems bit better
doesnt seem too bad now
yeah pretty decent
viral mason
that should work good
brittle wing
I want to try a good model, I just picked some random one
viral mason
look for any of mine
brittle wing
thats probably not very well trained maybe I could do better model, you guys got any decent one?
oh
wait how do i find you on weights.com?
like where do i find your models?
@viral mason
viral mason
right here
remember to change the names of the files so they're not all called "model"
hallow thistle
but most are male 😢
If he has most of male voice models uploaded, it means he has most of male voice models.
viral mason
you got any cute egirl voices?
no I don't like those things, go to Lyery :3
I make good models of characters I like not catfish material
brittle wing
no I don't like those things, go to Lyery :3
what is Lyery?
viral mason
a person lol
brittle wing
oh
viral mason
idk their weights if they have one
brittle wing
@analog obsidian
oh like their weights profile name is called Lyery too?
their user?
hallow thistle
Neither do I. 
viral mason
what is a "femboy" voice
brittle wing
what is a "femboy" voice <:yt_nails:1159569314848972891>
average discord user voice I think
that one
😄
viral mason
😭
hallow thistle
or femboy too
That's getting too ridiculous.
viral mason
just way u wanan catfish so we can get rid of you :(
brittle wing
not catfish, I can be trans woman
viral mason
femboy is just feminine guy, whats wrong with that
I am one, and you're confusing me bc idk what they sound like anyone can be one
brittle wing
and want to sound femenine
rough tangle
yo for some resin it just doesnt work the voice changer just doesnt pick up my mic or some how can i fix it?
brittle wing
I am one, and you're confusing me bc idk what they sound like anyone can be one
you know, male femenine voice, for me its like impossible to sound femenine
even if try I sound awful
hallow thistle
yo for some resin it just doesnt work the voice changer just doesnt pick up my m...
!howtoask
viral mason
yo for some resin it just doesnt work the voice changer just doesnt pick up my m...
did u get it from a yt link
rough tangle
im not gone lie i cant remember
viral mason
that's ok, what gpu do u have
brittle wing
just way u wanan catfish so we can get rid of you :(
dont be homophobic or we can get rid of each together, very easy to close a discord server based on homophobic behaviour, i am trans woman and you are telling me im a catfish, thats toxic behaviour of you.
hallow thistle
rough tangle
that's ok, what gpu do u have
um ngl i got a gtx 1070
brittle wing
I took screenshot of conversation just in case.
rough tangle
dont be homophobic or we can get rid of each together, very easy to close a disc...
ewwww
viral mason
dont be homophobic or we can get rid of each together, very easy to close a disc...
hey asshole, I am not homophobic I am with a dude, fuck off
end of conversation
I will NOT be helping you since you're rude
brittle wing
hey asshole, I am not homophobic I am with a dude, fuck off
no need to "get rid of me" then, why dramatic, you started telling me this
viral mason
um ngl i got a gtx 1070
your best bet is to get wokada deiteris fork it's this third one
-rty
hallow thistle
um ngl i got a gtx 1070
You can either go for Deiteris fork W-Okada or Vonovox, but Deiteris is easier to use. https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#download-nvidia-on-windows Vonovox can give better audio quality but its UI is less friendly and more professional than other W-Okada forks. https://docs.aihub.gg/realtime-voice-changer/local/vonovox/
rough tangle
what
viral mason
You can either go for Deiteris fork W-Okada or Vonovox, but Deiteris is easier t...
vonovox only works with I believe 30 series and up I thought, Nick said that at least
hallow thistle
idk what yall talking about
Did you just not pay attention?
rough tangle
idk
brittle wing
Thank you for your report regarding hateful speech and / or violent behaviour. We will review this profile. Thank you for contributing to making Discord a safer community.
viral mason
Thank you for your report regarding hateful speech and / or violent behaviour. W...
if u reported me I'll never let this go
rough tangle
You can either go for Deiteris fork W-Okada or Vonovox, but Deiteris is easier t...
man idk what the fuck to do
viral mason
I hate you now, because of your personality, not bc who you are
rough tangle
wtf is this
viral mason
it's loading the voice changer
hallow thistle
wtf is this
I don't think either Deiteris or Tg Develop W-Okada would look like this.
This is the actual Deiteris W-Okada program.
rough tangle
im not useing that
hallow thistle
You better download the one from this guide.
rough tangle
why
hallow thistle
Why? Because the original W-Okada version that has to run through start_http.bat is outdated. I'm telling you to go for better W-Okada version, and you weren't paying attention at all.
hallow thistle
viral mason
we need a hall of dumb people, just like the hall of shame
heady haven
we need a hall of dumb people, just like the hall of shame
Stop being toxic to others
Not cool pal
viral mason
I'm not really in a good mood due to that person from before so I think its best if I not chat here for some time
I apologize only to the person I called dumb just now, he didn't deserve that
devout remnant
wtf is this
virus
carmine spear
Look, I downloaded it and everything is fine, I just don't know how to use it on discord.
rotund breach
new toi this discord can someone help me get a voice that i can use? plz
i wanna get a voice changer
viral mason
what's your gpu
rotund breach
5060ti
viral mason
u just click the link where it says guide next to vonovox
it should tell you how to download and use it, but I will be here to help
rotund breach
For Windows
Download this: VAC Lite (Virtual-Audio-Cable by Muzychenko) i just download this
stiff goblet
Applio is using my CPU instead of GPU. Any way to fix it ? I've installed the CUDA but nothing has changed
viral mason
For Windows
Download this: VAC Lite (Virtual-Audio-Cable by Muzychenko) i just d...
yup
dreamy seal
Applio is using my CPU instead of GPU. Any way to fix it ? I've installed the CU...
Do you still have applio btw?
raven ginkgo
radiant rapids
yo someone help me! i cant load the voice changer
knotty moth
virus
then copilot and other bloatwares on win 11 are also "virus" that collect some data through telemetry
viral mason
I agree with that deleted message
knotty moth
sorry it was wrong person
knotty moth
Thank you for your report regarding hateful speech and / or violent behaviour. W...
I meant this person @viral mason
nova cairn
Is there a fix to fix the cutting out when im talking:? It messes up my sentence
hallow thistle
Look, I downloaded it and everything is fine, I just don't know how to use it on...
!howtoask
hallow thistle
It seems like it's the first time you come here for a "program", but you didn't explain on which "program" you were looking for. Is it W-Okada realtime voice changer, Vonovox or some other RVC forks? What is your PC GPU? And did you follow any tutorial before?
hallow thistle
im new to pc dont really know my way around it yet
Pretty much almost everyone would say that, but when you go along with it you'll getting used with it. 
elder walrus
I downloaded this model its sound normal though the monitor, but when I listen to it on obs or in a call the voice is super inconsistant and it sounds fake, but not when im using the regular monitor thing to hear my voice straight from the app
hallow thistle
I downloaded this model its sound normal though the monitor, but when I listen t...
Read help guidelines before start asking. What is your PC GPU? Which W-Okada version you tried to use. And did you follow any tutorial before this?
elder walrus
3060ti
MMVCServerSIO_win_onnxgpu-cuda_v.1.5.3.18a
yes i followed a youtube tutorial , ive used the voice changer plenty of time and it worked great but now the voice sounds very noticably staticy
hallow thistle
3060ti
MMVCServerSIO_win_onnxgpu-cuda_v.1.5.3.18a
yes i followed a youtube tutor...
Try Deiteris fork W-Okada from this guide instead of the one you used. https://docs.aihub.gg/realtime-voice-changer/local/deiteris-w-okada-fork/#download-nvidia-on-windows
There's another voice changer named Vonovox, this one can give better audio quality but its UI is less friendly and more professional than other W-Okadas. https://docs.aihub.gg/realtime-voice-changer/local/vonovox/
elder walrus
alright
hallow thistle
Let me know if you looking for best settings.
elder walrus
sure
I downloaded the deiteris
and like the voice sounds bad
what are the regular settings
hallow thistle
and like the voice sounds bad
Here's your settings:
Chunk: around 50 - 90 ms, less than this is possible if your GPU can handle
Extra: 2.7 s
GPU: NVIDIA GeForce RTX 3060 Ti
F0 Det: rmvpe
tame oracle
-rt
hallow thistle
When a member asks for "RVC", it can either mean the "realtime voice changer" or literal "RVC (retrieval-based voice conversion)", so I'd always ask them before letting them know where the link is.
misty blaze
oh
viral mason
a lot usually mean the realtime voice changer, but like Namari said always ask in case
misty blaze
im GPU is NVIDIA GeForce RTX 3060 Ti use vonovox or okada fork ?
hallow thistle
im GPU is NVIDIA GeForce RTX 3060 Ti use vonovox or okada fork ?
Either Vonovox or Deiteris/Tg Develop W-Okada fork. Vonovox can give better audio quality but the UI is less friendly and more professional than other W-Okadas.
misty blaze
hmm
hallow thistle
While Vonovox can launch without GPU, I won't be testing anything there.
hallow thistle
Will it be harder to use?
Yes, a bit harder than W-Okada forks. Although it's possible to translate a setting like chunk from other W-Okada forks to be compabible with Vonovox (block size).
Chunk 300 ms on Deiteris fork can be block size 0.30 for Vonovox, as much as it stated on Vonovox guide. https://docs.aihub.gg/realtime-voice-changer/local/vonovox/#settings
misty blaze
vonovox how to setting is more best?
hallow thistle
vonovox how to setting is more best?
Block size: around 0.60 - 0.90
Extra time: 2.00
Crossfade: 0.15s
misty blaze
Block size: around 0.60 - 0.90
Extra time: 2.00
Crossfade: 0.15s
discord how can use？？i have download vac
hallow thistle
discord how can use？？i have download vac
earnest vigil
im trying to learn how to train ai voices, i did a 8 minute sound to train but i dont really understand how :(
fringe ravine
anyone know how to fix okada Pipeline is not initialized?
hallow thistle
anyone know how to fix okada Pipeline is not initialized?
What is your PC GPU? Did you follow any tutorial before this?
fringe ravine
What is your PC GPU? Did you follow any tutorial before this?
my gpu is NVIDIA Geforce GT 710
yes i do follow some tutoriial before
hallow thistle
my gpu is NVIDIA Geforce GT 710
This GPU is old. Better go for an online option instead.
fringe ravine
https://docs.aihub.gg/realtime-voice-changer/cloud/deiteris-w-okada-fork-kaggle/
so i just instal this?
hallow thistle
Neither original or Deiteris/Tg Develop W-Okada would even work with GeForce GT 7xx series GPU.
fringe ravine
ah thanks for the information
knotty moth
my gpu is NVIDIA Geforce GT 710
that tutorial not mentioning spec requirements where yours doesn't meet is really a scam
fringe ravine
i see
stiff goblet
Do you still have applio btw?
Yes
sturdy maple
why does my own trained ai voice models all sound like robotic errors with no voice heard?
lean dune
-kaggle
sharp ore
https://www.youtube.com/channel/UCYVUm41HtniQ2c3nYFaxWEw
How can I do these videos guys?
nocturne mural
@viral mason Do you remember the error you had in the Kaggle notebook? It’s because in that notebook you were cloning Applio from the main branch instead of from a release. I’ve just pushed a commit, so it should be fixed now.
low shard
why does my own trained ai voice models all sound like robotic errors with no vo...
are you training them yourself?
low shard
im trying to learn how to train ai voices, i did a 8 minute sound to train but i...
Did you check https://docs.aihub.gg perhaps? :)
earnest vigil
Did you check https://docs.aihub.gg perhaps? :)
yup im doing some search
i cant get past installing the thing in the google drive
i have the dataset ready
just need to try and train it
low shard
so i just instal this?
that's cloud (remote good pc)
low shard
i cant get past installing the thing in the google drive
why are you using google drive? did you check if your pc gpu is good enought to train RVC locally?
low shard
i have a 1060 6gb
ehh, I mean that could work, but it's not recommended because it's pretty old and you don't have 8gb vram
earnest vigil
yep
low shard
You could, but cloud (remote good pc) would be better for you