-docs or #1159513888199540817
#🧬│ai-chat
storm meadow
rare sorrelBOT
-docs or <#1159513888199540817>
Suggestion for @left hornet
left hornet
thanks
feral haloBOT
Ayo? @left hornet level 1 !!! 
feral haloBOT
Ayo? @torpid spire level 1 !!!
amber light
3 dots then download model
twin fractal
freemium mid
flint grotto
guys
whata a good relic voice changer easy i could use guy whos his high piched voiced to sound like a girl
flint grotto
is okada the the in github
amber light
yes
amber light
if you need help with okada #🔍│help-w-okada
royal vigil
hey
flint grotto
smees like a lot of people are incative
signal cairn
freemium mid
best paid tts with easy cloning is 11labs, cant deny their quality
signal cairn
smees like a lot of people are incative
past midnight
storm jay
im new here
i want to create an ai voice model
and gives me a google drive link
what do i do?
radiant harness
ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
chex 0.1.85 requires numpy>=1.24.1, but you have numpy 1.23.5 which is incompatible.
torchdata 0.7.0 requires torch==2.1.0, but you have torch 2.0.1 which is incompatible.
torchtext 0.16.0 requires torch==2.1.0, but you have torch 2.0.1 which is incompatible.
torchvision 0.16.0+cu121 requires torch==2.1.0, but you have torch 2.0.1 which is incompatible,
How can this problem be solved? Do you know how to fix it?
random forge
Hallo
austere saffron
is there any kishore kumar voice model?
amber light
is there any kishore kumar voice model?
austere saffron
<#1175430844685484042> or weights.gg
tried
civic tapir
what tts program can i use with the models i made
feral haloBOT
Ayo? @civic tapir level 2 !!!
snow silo
flat plaza
halo
tepid basin
why my voices sounds weird
Can you describe what youre hearing from Okada?
It could be poor microphone quality
Ideally your real voice should be clear, with no background noise in the microphone
floral aurora
hi
cedar elbow
How i can make a AI cover by arisu from blue archiver
Please help me
lean beacon
hi
lean blaze
.wk
indigo pikeBOT
jade shore
colab keps crashing, man
earnest ravine
bruhh
valid prism
colab keps crashing, man
do it locally on your own computer
valid prism
what do i do
Are you trying to do voice conversion or images
valid prism
Sorry, I was talking about ai voices for Mansi
bold saffron
jade shore
Sorry, I was talking about ai voices for Mansi
I have an AMD GPU so that ain't happening anytime soon sadly
terse oak
お願いします
earnest ravine
oh
feral haloBOT
Ayo? @earnest ravine level 3 !!!
terse oak
大丈夫
valid prism
I have an AMD GPU so that ain't happening anytime soon sadly
What do you mean, you've tried locally and your GPU is the reason why it's not working?
feral haloBOT
Ayo? @valid prism level 1 !!!
valid prism
LESGO LVL 1
terse oak
誰か日本語を話せますか
jade shore
What do you mean, you've tried locally and your GPU is the reason why it's not w...
It is working locally, but it's very slow because it's training on the CPU, because it doesn't recognize my GPU
valid prism
It is working locally, but it's very slow because it's training on the CPU, beca...
Ok I see, annoying
devout palm
The bot physically cannot send files larger than 25mb
If you can't be bothered to use a shorter audio file or convert it to a FLAC then IDK what to say to you, that's a you issue
jade shore
NVIDIA sucks man
jade shore
I do not speak japanese
terse oak
しませんか
oh i dont speak japanese too
i was just randomly saying random english words
kinda sounded japanese
jade shore
Me waiting for AMD GPU support for RVC2
terse oak
Me waiting for AMD GPU support for RVC2
hit it with a hammer
it usually works
if you hit your GPU with a hammer it makes it go faster
you gotta hit it hard though
odd knot
that didn't age well, now i have an rtx 4050 laptop
odd knot
???
feral haloBOT
Ayo? @elder willow level 1 !!!
elder willow
hi
covert lake
do u use the ai voice thingy
check our docs for guides https://docs.aihub.wtf
elder willow
why cant i hear myself in the voice changer
jade shore
hit it with a hammer
or just commit extreme schizophrenia and trick RVC2 that indeed my AMD GPU is actually from NVIDIA
feral haloBOT
Ayo? @jade shore level 3 !!!
timid musk
bạn có chắc không
@peak dome
plucky lark
bongo
silk robin
Hello!
timid musk
hello
silk robin
how is everyone
fervent tiger
is there a way to use a voice model as a voice changer?
silk robin
😮
feral haloBOT
Ayo? @silk robin level 1 !!!
prime pelican
Guys how can i promote my drake ai song can I post the link here?
feral haloBOT
Ayo? @prime pelican level 2 !!!
craggy cove
is there a way to use a voice model as a voice changer?
I mean, ye
nova jungle
s
elder willow
man cmon
twin fractal
My birthday is today
Happy birthday
inland sand
hello, anyone know how to deepfake full body locally?
something like this:
https://vm.tiktok.com/ZGexK4m5t/
tall zephyr
hi
sacred fog
is anyone really good at the piano?
elder willow
Ok then
steel coral
elder willow
<:drakepain:1159570745068888174>
chorom gorom
feral haloBOT
Ayo? @elder willow level 1 !!!
elder willow
ei ghorom e khara hoy nah
steel coral
hae asholei
serene tinsel
hello
I remember that a few months ago I was reading from a website guiding on how to train ai voice models, which included how to edit the vocals in audacity
I think it came from this server, can anyone suggest its location? I can't seem to find it
novel cipher
Hi hi, where should I get started on learning some more about current generative AI solutions and how to apply them?
I have been using Stable Diffusion to generate images on my own so far but haven't been in the loop for a while now
So would appreciate any news or help in getting up to the current standards xD
As well as any helpful tools or tutorials you guys consider useful ;3
novel cipher
Do I just copy my message there or hop into pins and hope one of them is what I'm looking for?
forest quarry
Has anyone tried using Zluda with RVC?
Not sure
winter nimbus
hello
jade shore
So ROCm wouldn't work?
Nope
Not unless if I just upgrade my PC or start medicating my ADHD
elder willow
Unmute
coarse vessel
Not unless if I just upgrade my PC or start medicating my ADHD
upgrade to a decent Nvidia gpu if u have enough budget
jade shore
upgrade to a decent Nvidia gpu if u have enough budget
Thinking about the RTX 3080
coarse vessel
Thinking about the RTX 3080
you'll be more likely to find the used 10 gb variant, or 4070 12 gb that has similar performance but with less power draw, a bit more price, AV1 support and some stuffs
jade shore
Okay i believe it's time to let go of one of my kidneys
coarse vessel
3060 12 gb is also fine for such AI stuffs
sharp path
atomic wasp
hello
atomic wasp
Does anyone have midjpurney subscription?
I have a character, can you make it facing 90 degrees ahead, unfortunately I don't have a subscription.
shy night
Hey there
jovial mural
hi everyone
shy night
When I try to download an AI model it through huggface it says "invalid username and password"
how can i fix it?
frail drum
.
rare sorrelBOT
grizzled bane
-train
fallen jasper
hi
weak beacon
hi
dreamy crypt
hi
safe sky
hi
rocky stone
hi guyse how to make myself sound like kendrick lamar? i wanna make a brainrot 6:16 in la remix but i dont know shit about ai and when i paste a link to the model in the google doc shit i get an error
woven spade
can you request a AI voice models
feral haloBOT
Ayo? @woven spade level 1 !!!
potent oar
forest quarry
can you request a AI voice models
forest quarry
hi guyse how to make myself sound like kendrick lamar? i wanna make a brainrot 6...
use applio:
https://colab.research.google.com/github/iahispano/applio/blob/master/assets/Applio.ipynb
if u want a guide it's on https://docs.aihub.wtf/ on the RVC > Cloud > Applio Colab section
grizzled bane
is training ai voice on weights.gg a good idea?
forest quarry
is training ai voice on weights.gg a good idea?
if you want easy training, yeah
reef forge
ok
jade shore
Okay, I used the RVC Disconnected colab, it didn't crash
plucky goblet
I don't know if anyone knows if there is a project to convert audio from AI that can be used to do something like this?
I think it could be put to more use.
https://www.youtube.com/watch?v=QaY3eirY94w
feral haloBOT
Vote Booster: Vote now for a 10% boost. https://arcane.bot/vote
✨ Tip: Use /card to modify your rank card
forest quarry
I don't know if anyone knows if there is a project to convert audio from AI that...
i mean... there isn't something exactly like that, afaik
you could always use SynthV and convert with RVC though
hard spear
yo guys, all my voices in my voice changer (okada) are a bit robotic, they are not that believable. any idea how to get a better one?
sonic tundra
HAHAHAHA
forest quarry
yo guys, all my voices in my voice changer (okada) are a bit robotic, they are n...
increase extra and use better models, ig
there's not really a lot that can be done
this model seems to be pretty good for example
lacks a bit in phonemes but eh quality wise it seems clean
rancid token
hi
jade shore
Question, according to y'all, what would be a good amount of epochs for a 10 minute dataset?
forest quarry
Question, according to y'all, what would be a good amount of epochs for a 10 min...
it varies
grizzled bane
weights.gg didnt train good
forest quarry
my personal recommendation, set a goal to 500 and train until that. if it isn't overtraining continue to 750 or 1000
grizzled bane
my ai model sounds bad like it only got trained with 10 epoches
forest quarry
weights.gg didnt train good
thats probably your dataset then
grizzled bane
thats probably your dataset then
i used official acapellas
forest quarry
oooor the thing didnt train properly
grizzled bane
Yea
forest quarry
i used official acapellas
have you taken the time to de-reverb them and denoise
if needed of course
jade shore
my personal recommendation, set a goal to 500 and train until that. if it isn't ...
It gets overtrained by 400
most likely a little bit before
forest quarry
have you taken the time to de-reverb them and denoise
dont forget to trim silences too
grizzled bane
have you taken the time to de-reverb them and denoise
i used lead vocals with no delay, autotune, and reverb
forest quarry
honestly, you should try training on colab then
grizzled bane
Can you recommend me a website?
grizzled bane
Thanks
jade shore
Actually yeah RVC Disconnected is good
I used applio before and it kept timeouting me
forest quarry
Actually yeah RVC Disconnected is good
be careful tho
stop training after like 1 or 2 hours
and export model
so you dont lose stuff
and can train more later
jade shore
I'm gonna get a PC upgrade later on so I don't have to mess with collabs
gonna get an RTX 4070
crimson basin
how do yall guys change the voice though?
jade shore
how do yall guys change the voice though?
What voice
crimson basin
What voice
the ai models
jade shore
50s series are coming though
For a ridiculously high price I am unwilling to pay for yea
forest quarry
you just need a good gpu
tight cypress
like welp, if everything go as expected you can have a 5060
crimson basin
you just need a good gpu
nah i got a potato in the name of pc
tight cypress
like crypto mine no longer causing huge spike of the price like it was before
jade shore
like welp, if everything go as expected you can have a 5060
No way it'll still cost as much as what the 4090 was at release
I'm sticking with the 4070 for now
Absolutely no reason to buy any of the 50 series
feral haloBOT
Ayo? @jade shore level 4 !!!
tight cypress
hm
jade shore
I don't have the money
tight cypress
yeah not going to stop you
forest quarry
nah i got a potato in the name of pc
jade shore
rn I have an RX 570 with an I5-9400F
tight cypress
but that 4070 cost in average 650 USD right?
crimson basin
https://colab.research.google.com/github/hinabl/voice-changer-colab/blob/master/...
tysm
storm meadow
keywords are the whole PC
What is a whole pc for you
tight cypress
1 thousand is pretty reasonable
storm meadow
What is a whole pc for you
Does that count the monitor/keyboard/mouse?
tight cypress
What is a whole pc for you
case, power supply
jade shore
Does that count the monitor/keyboard/mouse?
No
jade shore
Just the GPU, CPU, RAM, Storage, Motherboard, PSU and the case
tight cypress
usually most people choose the most expensive part, GPU, CPU first then the others like cooler, storage
GPU in most case just cost half of the price of the entire build
jade shore
CPUs are unironically cheap
tight cypress
CPUs are unironically cheap
Xeon E5? 
jade shore
Unless if you are buying a threadripper
crimson stratus
~~Xeon E5?~~ <:lusbertmoment:1159804751924432906>
🔥
tight cypress
and can run heaps of stuff
but hey the issue is that how do you find a board that is compatible with that Xeon chip
jade shore
just buy a ryzen 7 5800x
tight cypress
unless you decided to buy and use some funky modified board
tight cypress
just buy a ryzen 7 5800x
yeah of course, that's just for tight budget
jade shore
yeah of course, that's just for tight budget
For the same price I can buy 2.5 2TB HDDs
tight cypress
... even I am not even recommending running 2016 Xeon chips on tight budget case.
I probably will suggest a Core i3 i5 12th gen?
jade shore
for CPU, just buy anything AMD has
for GPU it's NVIDIA, unless if you are on a tight budget then go on AMD there as well
tight cypress
like that i3-12100F for 100 bucks
tight cypress
4070 is pretty much a safe bet for most AI application now
jade shore
I wanted to have a 3090, up until I looked at the power consuption then I noped out
tight cypress
Although I heard its gaming performance is kinda funny compared to 4080
actually even better if someone choose 4070 Ti
jade shore
Ehh
I'll stick with the 4070
Also I'm gonna admit
having unmedicated ADHD is not good
high beacon
is there really 45 people using astra labs 😭
forest quarry
yep
high beacon
i’ve never seen that many people use it
quiet mango
Any recommendations on a model to remove Echoe + Reverb with UVR ? Thanks.
forest quarry
Any recommendations on a model to remove Echoe + Reverb with UVR ? Thanks.
de-reverb de-echo in vr arch
quiet mango
de-reverb de-echo in vr arch
Thanks 👍
Just tried it, works really well for my use case. Perfect !
forest quarry
i’ve never seen that many people use it
ooor maybe the servers had problems
like what happens on MVSep sometimes
fading meadow
Hello everyone. Can you help me? There is a neural network that makes the face in the photo sing. I lost the link
forest quarry
Hello everyone. Can you help me? There is a neural network that makes the face i...
honestly idk
fading meadow
ty
forest quarry
check #1159513888199540817 or #📰│dev-updates to see if there's anything like that
high beacon
how long will it take for my ai cover to generate if there is 46 people queued 😭
forest quarry
how long will it take for my ai cover to generate if there is 46 people queued �...
frankly, we don't know
but it might take a while
if it's a server problem it should take even longer 
high beacon
ah i see
muted fable
guys I wanna make a unique voice model from two different sample sounds. can i do that? even if I can do that , will I face issue with YouTube monetization?
forest quarry
guys I wanna make a unique voice model from two different sample sounds. can i d...
yes
idk
muted fable
> yes
> idk
what can i do at this case ? please i really need your help
forest quarry
frankly man, idk if you're gonna be monetized or not
elder willow
hello
forest quarry
that's between you and YouTube
grizzled bane
guys I wanna make a unique voice model from two different sample sounds. can i d...
you mean voice blending?
muted fable
you mean voice blending?
yea yea what shall i do bro
forest quarry
where can i get a free voice model that is free for commertal use?
and that's the thing
idk
promo on #1159290752195633273 guys
storm meadow
guys I wanna make a unique voice model from two different sample sounds. can i d...
If the song your covering is copyrighted yes
And when a video is marked as ai you get less money
forest quarry
If the song your covering is copyrighted yes
no they're probably making actual content and not ai covers
forest quarry
And when a video is marked as ai you get less money
and yeah this
storm meadow
no they're probably making actual content and not ai covers
Oh
next spade
Yo
quartz roost
where can i get a free voice model that is free for commertal use?
To profit you will need a license (typically paid) or written permission from the voice provider. This is no different than voice actor contracts.
muted fable
If the song your covering is copyrighted yes
no no , ill make scripts to read. not music
quartz roost
And when a video is marked as ai you get less money
I don't have to tag teto as ai she is legally an instrument.
zenith hamlet
If the song your covering is copyrighted yes
Mine was copyrighted,it's partially blocked from every country on earth
next spade
How many of you support andre tate
grizzled bane
no no , ill make scripts to read. not music
Make ai with your own voice
muted fable
To profit you will need a license (typically paid) or written permission from th...
what about blending? will this solve the problem? cause ill use two diffrent voices and make a unique one
quartz roost
what about blending? will this solve the problem? cause ill use two diffrent voi...
Like this?
zenith hamlet
How many of you support andre tate
Andrew tate?,im ashamed that we're biologically both XY
muted fable
Make ai with your own voice
yes yes ill record my voice then want to change it , can i do that?
storm meadow
How many of you support andre tate
No
next spade
Andrew tate?,im ashamed that we're biologically both XY
Meaning that you didnt like him at all
storm meadow
no one besides little grubby iPad kid “sigma” 9 year olds
forest quarry
@muted fable ngl you could just make an AI model of your own voice
zenith hamlet
Meaning that you didnt like him at all
Why would and should i 
muted fable
https://youtube.com/shorts/9KKjcb7Gy90?si=Omze_cq99HeJDmZj
i can do this on fl studio, im expert on that, no i mean creating a new voice model
quartz roost
yes yes ill record my voice then want to change it , can i do that?
Yes this is highly encouraged!
It's just when using other people's voice you need to check first
muted fable
<@1143251222149730424> ngl you could just make an AI model of your own voice
really? ooo okay but can it be modify like how i want it to sound?
next spade
Ok I didnt ask you to?
Can anyone tell me what can we do in ghis server
grizzled bane
no one besides little grubby iPad kid “sigma” 9 year olds
Im sigma🗿🔥
feral haloBOT
Ayo? @next spade level 1 !!!
forest quarry
really? ooo okay but can it be modify like how i want it to sound?
ehhh not a lot
muted fable
ehhh not a lot
ahh thats a bit harder
zenith hamlet
Ok I didnt ask you to?<:huggingface:1179800228946268270>
Can anyone tell me wha...
Pretty much everything ai related
forest quarry
Ok I didnt ask you to?<:huggingface:1179800228946268270>
Can anyone tell me wha...
AI stuff
quartz roost
really? ooo okay but can it be modify like how i want it to sound?
Are you making music with it? There is a free open source editor
storm meadow
Ok I didnt ask you to?<:huggingface:1179800228946268270>
Can anyone tell me wha...
Make ai models and use them for covers and realtime
muted fable
Are you making music with it? There is a free open source editor
no not music ill make it talk, read. just like that
next spade
Like? Business, video editing, content writing, etc.
quartz roost
no not music ill make it talk, read. just like that
Gpt sovits
storm meadow
Like? Business, video editing, content writing, etc.
More or less for fun
muted fable
Gpt sovits
is it free? and realistic like elevenlabs?
zenith hamlet
Like? Business, video editing, content writing, etc.
Not that type of ai but ye i think
storm meadow
is it free? and realistic like elevenlabs?
FOSS
next spade
I wondered coz i am new here
What do you guys do here?
muted fable
Like? Business, video editing, content writing, etc.
will psg win?
forest quarry
I wondered coz i am new here
What do you guys do here?
again, ai stuff
rn we're focused more on voice model stuff
next spade
Against real madrid ?
If yes then never
muted fable
Free and Open Source Software
oo so its free for commertial use?
forest quarry
hope they lose everything
storm meadow
oo so its free for commertial use?
Idk if the license
jade shore
How many of you support andre tate
Support who?
muted fable
hope they lose everything
😂 noway
storm meadow
Idk if the license
I know RVC is MIT
jade shore
Oh you mean insignificant bald man
storm meadow
not sure about GPT so Vits
muted fable
not sure about GPT so Vits
risky
forest quarry
prioritizing mbappe when bro is gonna leave, for free
muted fable
not sure about GPT so Vits
but hey let me try runway
forest quarry
they will deservedly lose against borussia
muted fable
prioritizing mbappe when bro is gonna leave, for free
i wish he come to united
muted fable
GPT so Vits also mit
so free for commertial use ?
forest quarry
i wish he come to united
that's the end of united right there
i cannot respect mbappe
way too selfish
muted fable
that's the end of united right there
i cannot respect mbappe
england premier league isnt easy
zenith hamlet
so free for commertial use ?
It's free,depends on the person who you're voice cloning
If yourself then yes mate
twin fractal
so free for commertial use ?
You can do anything but the creator of the software wont take blame if the software does any damage by the user or the software itself
next spade
Oh you mean insignificant bald man
Bro dont show hate, It will nor do nothing goods for you neither for him? I know he might be bad but at real time, all he wants i a world of winners. He has never did anyyhing wrong and doesn't agitate against any of the yhings he just wants a world full of winners as good made us
muted fable
It's free,depends on the person who you're voice cloning
If yourself then yes m...
i will try it bro , tnx btw
muted fable
Bro dont show hate, It will nor do nothing goods for you neither for him? I know...
mpape is a big fan of ronaldo
zenith hamlet
Bro dont show hate, It will nor do nothing goods for you neither for him? I know...
By doing a pyramid scheme
forest quarry
Bro dont show hate, It will nor do nothing goods for you neither for him? I know...
he's just an average dude
it's ok to have the winner mentality but don't discriminate against anybody who doesn't want that
or whatever
forest quarry
By doing a pyramid scheme
and this too
oblique marsh
hello
coarse vessel
so free for commertial use ?
basically yes but depends on the voice owner you want to make model of
muted fable
basically yes but depends on the voice owner you want to make model of
let me try it bro
next spade
Bro its your choice whom you support but being averse against a person who has no effect of anyone's hate. You're just waating your time
*wasting
oblique marsh
what should i do after downloading a model from here?
muted fable
basically yes but depends on the voice owner you want to make model of
what about i paid some here on model - master- shop?
zenith hamlet
Bro its your choice whom you support but being averse against a person who has n...
Tbh that man's fanbase harassed a lot of girls mate,it did affect quite a lot lol
But he does what he does,im just a hater because his fanbase is on my fyp
next spade
Bro that's part of his misogynist mentality
oblique marsh
what should i do after downloading a model from here?
?
next spade
We should embrace the good, never the bad
zenith hamlet
Shouldn't we move on ai stuff instead of a baldy
forest quarry
what should i do after downloading a model from here?
now you can convert vocals
https://docs.aihub.wtf/ for converting vocals
https://rentry.co/VoiceChangerGuide for realtime
next spade
Ok ok
zenith hamlet
Bro that's part of his misogynist mentality
And that's no good
oblique marsh
now you can convert vocals
https://docs.aihub.wtf/ for converting vocals
https...
tysm
zenith hamlet
now you can convert vocals
https://docs.aihub.wtf/ for converting vocals
https...
Gonna do rickroll ai cover today just because of the gif
coarse vessel
unreal lake
guys i have a pth file for voice changer seems good but still need some touch what i can do about it ? how i can make it perfect less robotic more girly?
forest quarry
guys i have a pth file for voice changer seems good but still need some touch wh...
mess with tune, index and extra settings
muted fable
guys since when runway tts gets so realistic
feral haloBOT
Ayo? @muted fable level 7 !!!
unreal lake
mess with tune, index and extra settings
im doing my best since days but i couldnt find yet
is there any ways to improve that one
muted fable
Ayo? <@1143251222149730424> level **7** !!! <:lfg:1159355870119993496>
ayyyyyyy u on level1
forest quarry
is there any ways to improve that one
as i said
mess with those settings, see what makes it better
feral haloBOT
Ayo? @unreal lake level 1 !!!
forest quarry
because tbf I can't tell you how to make it better because idk what model you're using and it varies a lot
so yeah
tight cypress
what about a screenshot?
unreal lake
can you suggest me any good girl voice ?
tight cypress
... oh here we go
unreal lake
i can screenshot
forest quarry
can you suggest me any good girl voice ?
maybe this one, idk
unreal lake
cant post pic here yet
tight cypress
maybe this one, idk
https://discord.com/channels/1159260121998827560/1236217350...
hey, a suggestion to put some "good girl voice" into a guide.
in case anyone somehow wanted it
forest quarry
hey, a suggestion to put some "good girl voice" into a guide.
oh yeah
unreal lake
im using file not from here idk where did i get but its most realistic so far i have use
name is "nerd"
fallen ruin
Aight,lets put an end to this question...
who is more likely to impress Gordon Ramsay..
tight cypress
https://cdn.discordapp.com/attachments/1116204005853507584/1236116131362639962/i...
Senshi...?
although it may shock Gordon Ramsay with his own cuisine
safe holly
guys, i used the phone guy speech, and i made the saddest speech ever with it
wanna see it?
unreal lake
which one is better? w okada or rvc?
forest quarry
which one is better? w okada or rvc?
you mean rvc built in voice changer?
unreal lake
yea
forest quarry
it depends
W-Okada is easier to use, but RVC realtime gives more options and is more tweakable
unreal lake
as i know there is 2 different voice changer app people using here
unreal lake
it depends
W-Okada is easier to use, but RVC realtime gives more options and is...
that one might be i need
im searching for dl link for that one
more tweak could make more realistic my own file
forest quarry
more tweak could make more realistic my own file
mhhh idk
worn geyser
hi
gilded hearth
guys i have a question, can someone help me?
gilded hearth
i have a photo of a room, with a tv cabinet ( ig thats the correct name) and id like the ai to generate the different provisions that i can make
i dont know if its all clear
forest quarry
storm meadow
i have a photo of a room, with a tv cabinet ( ig thats the correct name) and id ...
So you want to see how different tvs fit with the cabinet????
forest quarry
thats beyond my knowledge ngl
gilded hearth
So you want to see how different tvs fit with the cabinet????
no, i have to install the cabinet that supports the tv, and it has many pieces, and i would like to see the best options of how i can install it
plain zealot
hello there
gilded hearth
but i cant share a photo so that you can have an idea
forest quarry
hello there
ello
storm meadow
no, i have to install the cabinet that supports the tv, and it has many pieces, ...
So you want ai to generate instructions to install the tv?
gilded hearth
So you want ai to generate instructions to install the tv?
not the tv, idk how its called in english
the thing that everyone has in the living room😂
feral haloBOT
Ayo? @gilded hearth level 1 !!!
storm meadow
the thing that everyone has in the living room😂
Couch
storm meadow
like the wardrobe ahaha
Cabinet
gilded hearth
exactly
forest quarry
im searching for dl link for that one
https://rentry.co/RVCRealtimeGuide should have it
gilded hearth
and it comes with many pieces, its a modern one, id like to see many ways i can install it
how can i share a photo?
grizzled bane
the thing that everyone has in the living room😂
But i dont have a cabinet in my living room
storm meadow
and it comes with many pieces, its a modern one, id like to see many ways i can ...
So it’s already built, you just want to know how it would look in your room?
gilded hearth
So it’s already built, you just want to know how it would look in your room?
can i send a photo in dm so you can understand? my english isnt that good sorry
unreal lake
https://rentry.co/RVCRealtimeGuide should have it
i really dont understand people are really using realistic voices but when i try its always so bad
whats wrong with me i dont know
i have 40+ voice file but none of them good
frigid walrus
Will dis help?
forest quarry
naaah
feral haloBOT
Ayo? @frigid walrus level 1 !!!
forest quarry
i really dont understand people are really using realistic voices but when i try...
it's probably config problems and the models maybe....
or it could be ur mic which i highly doubt
forest quarry
U make me sad
Why
frigid walrus
Why u delet my message
forest quarry
Why u delet my message
Because to be honest we don't need meth cooking tips
unreal lake
it's probably config problems and the models maybe....
or it could be ur mic wh...
wish i could send pic but i have tested tune -50 to +50 all gain mods and index also chunk extra etc
feral haloBOT
Ayo? @unreal lake level 2 !!!
forest quarry
Do you have better recipe or what
No we rather you NOT sharing that lmfao
forest quarry
wish i could send pic but i have tested tune -50 to +50 all gain mods and index ...
You should normally use tune at 0-24, normally you'll keep from 0-12
unreal lake
im using at 8
forest quarry
index is for accenting so if you want the accent closer to the model yeah
forest quarry
im using at 8
if it's for a woman voice yah its fine
unreal lake
yes woman voice
quartz roost
i really dont understand people are really using realistic voices but when i try...
The synthesized robotic voices are far more popular
unreal lake
sadly i need realistic
south atlas
hey guys
brisk hollow
hii guys!
south atlas
i am having like a 1 minute data set only, are there any guide to generate a 10mins dataset from that file ?
ripe mango
I am back after a temporary break.
forest quarry
i am having like a 1 minute data set only, are there any guide to generate a 10m...
i dont think you can extend datasets like that
but you could probably train even if it was 1 min
if the quality's good it'll be decent
forest quarry
I'd say
train on batch size 4
It might lack on pronunciation though so it might be a mixed bag
south atlas
I'd say
train on batch size 4
😮 may i know how can i choose batch size base on dataset lenght ?
i mean i thought we should choose as big batch size as possible to maximize the speed
elder willow
Hello
forest quarry
for small sets I'd just recommend 4
like less than 5 mins
but, experiment with different batches
2 and 6, 8, etc
south atlas
😮
elder willow
😮
forest quarry
see what gives a better result in the end
elder willow
How are you?
south atlas
see what gives a better result in the end
ty ><
forest quarry
How are you?
pretty good, wbu?
frigid walrus
🥲
bleak river
Hahahahah I'm having a hard time to remove the accent from a model hahahhaha
high nymph
guys whats rvc
elder willow
A program to convert vocals to AI vocals with AI voice models of characters
To be simple, making AI covers
forest quarry
yeah, voice conversion
coarse vessel
guys whats rvc
learn more in https://docs.aihub.wtf/
forest quarry
promos on #1159290752195633273 pls
south atlas
@forest quarry ayyy sir what crepe hop length should i set for 1 min dataset model
i usually use 64 for like 10mins model
but 1 min idk
forest quarry
<@622856015444049937> ayyy sir what crepe hop length should i set for 1 min data...
if you use rvmpe, it doesn't change anything
forest quarry
yw :)
subtle fox
south atlas
@forest quarry sir
feral haloBOT
Ayo? @south atlas level 6 !!!
forest quarry
set sample rate to 48k, run set training variables again and preprocess/feature extract again ig
coarse vessel
you must set the same sample rate as when preprocessed before
south atlas
😮 so i have to run new colab session right
sonic tundra
ALENA MALENA! 
forest quarry
😮 so i have to run new colab session right
no
south atlas
btw i remember read somewhere that i should change model at 40k sample rate, right ?
cuz uvr5 output is also 44.1 khz
forest quarry
btw i remember read somewhere that i should change model at 40k sample rate, rig...
ehhh
sonic tundra
i'm a true king for alena malena 👑
forest quarry
im more of a 32k guy
coarse vessel
cuz uvr5 output is also 44.1 khz
not relevant, check the spectrogram cutout in Spek or Audacity
sonic tundra
he turns a king 👑 🫅
hin't?
forest quarry
you could probably use 40k
south atlas
not relevant, check the spectrogram cutout in Spek or Audacity
😮 what is it
south atlas
you could probably use 40k
😮 so are there any when i should use 48k
forest quarry
😮 so are there any when i should use 48k
probably never, depending on the audio
sonic tundra
the models hard thristy :dancing:
forest quarry
32k for most stuff will already work just fine
sonic tundra
hahahaha its funny 
bleak river
32k for most stuff will already work just fine
32k or 40k? What's the differences between the two?
forest quarry
32k or 40k? What's the differences between the two?
sample rates
honestly google would do a better explanation than i can of sample rates
but, for the most part, in rvc, you should use 32k
south atlas
dammit i always use 40k, maybe that is why the quality is bad
forest quarry
dammit i always use 40k, maybe that is why the quality is bad
not necessarily
it also depends on ur dataset too
if your dataset has noise, reverb and echo... it wont be good
haughty python
Yo i need help i wanna UVR to extract 4 stems vocals drums base and melody so how can i do that pls help
feral haloBOT
Ayo? @haughty python level 1 !!!
coarse vessel
dammit i always use 40k, maybe that is why the quality is bad
go learn what is spectrogram if not know yet
forest quarry
Yo i need help i wanna UVR to extract 4 stems vocals drums base and melody so ho...
just use the Demucs models in UVR
htdemucs_ft specifically
south atlas
btw i remember that if my dataset is 44.1k so i should choose 40k target sample rate right ?
forest quarry
again, not necessarily
bold saffron
@coarse palm check you dms!
forest quarry
id just recommend 32k
coarse vessel
btw i remember that if my dataset is 44.1k so i should choose 40k target sample ...
again, you should look at the spectrogram. go show it here, your role even has pic perm here
bleak river
id just recommend 32k
If I trained a model with 40k and
I created audios(using inference). What happens if a using this audios to create a new model? Should I use 32k or 40k?
forest quarry
If I trained a model with 40k and
I created audios(using inference). What happe...
i suggest you don't make models out of inferences
because it's gonna make the quality worse in the end
south atlas
again, you should look at the spectrogram. go show it here, your role even has p...
i think every song i have is about 20khz, for example this one, according to the guide i should double it and choose 40khz
my data is split vocal from songs
bleak river
because it's gonna make the quality worse in the end
Ok
forest quarry
i think every song i have is about 20khz, for example this one, according to the...
yeah but this isn't vocals right?
coarse vessel
i think every song i have is about 20khz, for example this one, according to the...
how is the extracted vocals?
south atlas
how is the extracted vocals?
i use UVR
south atlas
here sir
south atlas
did you use Voc-FT? try BS roformer or MDX23C to keep it fullband
i used MDX23C-8KFFT-InstVoc_HQ
dammm today i learnt new things
gentle canopy
hi
elder willow
Hello
forest quarry
do someone know how to make these files with audio/
?
glass prawn
these files in #1175430844685484042
glass prawn
Bro thank you so much
forest quarry
yw :)
if you have any more questions ask em on #✨│ai-help
helpers will answer if im not there
glass prawn
Oh okay thanks for the help
feral haloBOT
Ayo? @glass prawn level 1 !!!
forest quarry
glad i could help
sonic tundra
a
feral haloBOT
Vote Booster: Vote now for a 10% boost. https://arcane.bot/vote
✨ Tip: Use /card to modify your rank card
haughty python
just use the Demucs models in UVR
its so slow is there faster option ?
sonic tundra
see the game im playing
feral haloBOT
Ayo? @sonic tundra level 5 !!!
haughty python
it's sadly the best option
faster doesnt matter if the quality is worse
forest quarry
faster doesnt matter if the quality is worse
tbh idk
dunno what models are available in UVR by default except for a few
haughty python
man UVR was easier
elder willow
hey @proper stone, i just want to say that your rival octoling model is great, i have a octoling oc, so i can make AI covers with her now.
proper stone
hey <@1059964884906946611>, i just want to say that your rival octoling model is...
You’re welcome
zenith hamlet
Yo i need help i wanna UVR to extract 4 stems vocals drums base and melody so ho...
Demucs4 FT on mvsep maybe?
proper stone
I’m planning on doing models for more Splatoon characters that aren’t the playable Inklings and Octolings, and the idols
haughty python
Demucs4 FT on mvsep maybe?
mvsep ?
proper stone
Already did Sheldon, Cap’n Cuttlefish, and the Rivak Octolings
haughty python
how can i find it
zenith hamlet
how can i find it
It's a website with a bunch of models
elder willow
Weights is a lifesaver, who would have thought that someday you could make john lennon sing a nirvana song!
feral haloBOT
Ayo? @elder willow level 1 !!!
haughty python
It's a website with a bunch of models
oh i thought were talking about UVR models
zenith hamlet
oh i thought were talking about UVR models
There's also uvr models there yeah
haughty python
There's also uvr models there yeah
how can i find in uvr ?
zenith hamlet
how can i find in uvr ?
Demucs4 FT? On UVR?,idk
haughty python
yah i got merci
proper stone
Weights is a lifesaver, who would have thought that someday you could make john ...
Or Shiver singing I Can’t Fix You
zenith hamlet
Maybe go to audio separation server and ask in questions channel,or just do it on non locally
zenith hamlet
htdemucs_ft
Nice
haughty python
Maybe go to audio separation server and ask in questions channel,or just do it o...
theres audio separation server ?
zenith hamlet
theres audio separation server ?
Yeah
zenith hamlet
From xminus to gsep to mvsep
elder willow
i made spongebob and patrick sing boombayah using the duet feature, it’s pretty cool.
zenith hamlet
send me one
One sec
zenith hamlet
i made spongebob and patrick sing boombayah using the duet feature, it’s pretty ...
Duet feature?
grizzled bane
i made spongebob and patrick sing boombayah using the duet feature, it’s pretty ...
Duet feature?
proper stone
i made spongebob and patrick sing boombayah using the duet feature, it’s pretty ...
I’m making Deep Cut sing songs now
Btw I’m gonna retrain the Squid Sisters at some point but this time they will be submitted as separate models otherwise Weights will only have the first linked model (Callie)
elder willow
Hi
forest quarry
ello
bleak jetty
Hi
Dos rx 6600 gpu good for rvc training and inference?
forest quarry
Hi
Dos rx 6600 gpu good for rvc training and inference?
sadly it won't be able to do training
because AMD stuff ain't compatible with most AI stuff
and that includes RVC
grizzled bane
And intel?
forest quarry
same thing
grizzled bane
Thats sad
bleak jetty
Thank you for this information. This will help me a lot. I will not get it. I'm happy because I haven't bought it now. Okay, I have searched about similar one for NVIDIA, and I found it. NVIDIA RTX 3060. What about it? What's the average speed for inferencing and training? I know it's different of the voice lens or the data set and the training settings. But I mean, will I get faster operations?
forest quarry
Thank you for this information. This will help me a lot. I will not get it. I'm ...
you most likely will have a decent experience overall
so yeah
bleak jetty
Thankyou
feral haloBOT
Ayo? @bleak jetty level 1 !!!
bleak jetty
Yea
empty lynx
Some sunny day
We'll meet again
Don't know where
Don't know when
We'll meet again some sunny day
forest quarry
theres someting like Nvidia Broadcast but for AMD?
not sure, probably not
steep pebble
helllp
feral haloBOT
Ayo? @steep pebble level 2 !!!
steep pebble
my voice model sound not bad when talking
but sounds weird once I sing song
whyyyy
forest quarry
did you train it?
steep pebble
did you train it?
I have to？
🤯
I am not using my model
I use the default ones with the software
strange escarp
Hey i need help. Im in applio right now messing with the TTS. I have a voice model downloaded and im in the tts tab. Which TTS voice should I select if I want just a general american sounding male?
forest quarry
Hey i need help. Im in applio right now messing with the TTS. I have a voice mod...
scroll until u find the enUS models
strange escarp
Hey i need help. Im in applio right now messing with the TTS. I have a voice mod...
its gives me a list like this af-ZA-AdriNeuralm, af-ZA-WillemNeural etc etc. theres a lot.
thought i did that at first. blew right past them. thanks
forest quarry
can't i use the program if i have amd?
no
fallen ruin
people need to stop taking the names on the tarot cards at face value
strange escarp
I need to use applio via python or some other programming langauge to synthesise text en masse. How can I achieve this?
steep pebble
how to adjust voice model in wokada
grizzled bane
Can you train ai voice on iphone?
feral haloBOT
Ayo? @grizzled bane level 4 !!!
grizzled bane
Ayo? @grizzled bane level 4 !!!
urban crane
Ayo? @grizzled bane level 4 !!!
storm meadow
Ayo? @grizzled bane level 4 !!!
honest cloud
Hey besties
feral haloBOT
Ayo? @fallen ruin level 10 !!!
fallen ruin
Irumyuui or lyza are The Empress
grizzled bane
Ayo? <@860200729405554708> level **4** !!! <:lfg:1159355870119993496>
Im training my voice model on my phone
I'm hoping that it works
fluid basin
heyyy☺️ ☺️
surreal relic
my cock was stuck in a blender help
fallen ruin
although it may shock Gordon Ramsay with his own *cuisine*
Gordon ain’t gonna be happy when he serves a dish containing slime
elder willow
eager wing
elder willow
hello everyone!
unborn veldt
It's been so fucking long......please tell me a website or app to let me make AI models....I'm dying over here
Someone please tell me...please
maiden shadow
unborn veldt
https://google.com/
Wtf am I supposed to do with a fucking Google website
sour vigil
how are u a model maker if u cant make models
unborn veldt
how are u a model maker if u cant make models
I DONT KNOW I NEWD HELP 
dense heart
wtf are these like broken or is it like this (its a cpu soket)
timid karma
feral haloBOT
Ayo? @timid karma level 1 !!!
unborn veldt
cyan kelp
hi
echo minnow
Hello o/ new to MMVC and absolutely lost trying to set things up, where might I find the best guide/info? Thanks
forest quarry
echo minnow
https://rentry.co/VoiceChangerGuide
Thank you!
forest quarry
yw :)
dense heart
https://rentry.co/VoiceChangerGuide
rick assly
celest cargo
Wtf am I supposed to do with a fucking Google website
we saw the image the first time
unborn veldt
we saw the image the first time
echo pollen
feral haloBOT
Ayo? @elder willow level 1 !!!
hollow osprey
Heyyy
cinder kestrel
I'm new to ESRGAN and its upscaling capabilities. I'm seeking assistance in accessing a functional Google Colab notebook featuring the 4x_DigitalFrames_2.5.pth model. This is essential as I intend to deliver projects to my client. The provided Colab notebook should operate smoothly and ensure that media files, including images and videos, are rendered without any performance issues. Thank you in advance for your support.
feral haloBOT
Ayo? @cinder kestrel level 3 !!!
tepid mantle
Yall whats the best way to make an ai cover
feral haloBOT
Ayo? @tepid mantle level 1 !!!
storm meadow
I'm new to ESRGAN and its upscaling capabilities. I'm seeking assistance in acce...
Your making people pay for open source upscaling? 
timid musk
Yall whats the best way to make an ai cover
timid musk
I wish there was just a couple, but everyone wants their own thing lol
elder willow
hi
stable ledge
elder willow
is suno any good?
fervent moss
elder willow
does udio not give instrumental
I keep getting an error when trying to make AI Songs
adolf rizzler
elder willow
I keep getting an error when trying to make AI Songs
what you using
elder willow
I am using Easy GUI?
use applio
free?
yes
Download it?
nah its on google colab
link?
-colab
feral haloBOT
Ayo? @elder willow level 1 !!!
feral haloBOT
Ayo? @elder willow level 16 !!!
elder willow
what was it
I clicked "Start Appio" and i see this orange box thingy
elder willow
I clicked "Start Appio" and i see this orange box thingy
screenshot it and show m
me
post where?
regal crypt
a
zenith flower
Does anyone know what languages the TITAN Pretrain was trained on?
timid musk
is suno any good?
i like suno
timid musk
Does anyone know what languages the TITAN Pretrain was trained on?
i have no idea
signal kettle
is suno any good?
udio is like 10x better
elder willow
udio is like 10x better
is udio music copyrighted
signal kettle
is udio music copyrighted
naw
elder willow
welp, Suno is more well known
you gotta pay for suno
signal kettle
welp, Suno is more well known
udio is becoming more well known
feral haloBOT
Ayo? @signal kettle level 1 !!!
signal kettle
you could literally perfectly replicate an artists style of song and they had to nerf itself
elder willow
what i generated with udio
signal kettle
was insane at launch
elder willow
but this doesnt make Udio completely free
i can generate 600 times tho
tight cypress
actually I have check up the website, and it is pretty impressive as early launch
signal kettle
only problem with udio is i cant choose a specific rappers voice
cus its copyright
timid musk
udio is like 10x better
it can be better
elder willow
cus its copyright
it chooses for me
elder willow
theres alot of times where it gave me lil peep voice
signal kettle
theres alot of times where it gave me lil peep voice
yea its really easy with lil peep
it could do juice too
opaque breach
,
elder willow
it could do juice too
yeah but i cant choose tho
i wanted to do juice voice but i cant
devout palm
timid musk
i wanted to do juice voice but i cant
you can't choose the voice on any site
devout palm
elder willow
does anyone have a tool i wanna detect if something is ai or not
tight cypress
does anyone have a tool i wanna detect if something is ai or not
I did know a tool for AI image
elder willow
I did know a tool for AI image
nah i got audio. supposedly its a demo version of the spongebob theme song but it sounds weird
found it on internet archive
tight cypress
uhm... I dont know if there's any tool for that but
timid musk
nah i got audio. supposedly its a demo version of the spongebob theme song but i...
99% of people in #✦│chat can tell is something is ai
tight cypress
you can find trace when you drag the file into the audio spectrum
tight cypress
99% of people in <#1159265300915687534> can tell is something is ai
i am one of the verified bro
native locust
99% of people in <#1159265300915687534> can tell is something is ai
realest spill i ever heard
elder willow
99% of people in <#1159265300915687534> can tell is something is ai
i cant post copyrighted stuff in chat
tight cypress
you can find trace when you drag the file into the audio spectrum
( that thing called spectrum analyser or audio frequency viewer? )
native locust
i think ive made 100 models and never gotten one that can truly fool everyone i play it too
i got close once in a kits model when that first released but it was super muddy and i never got a model to sound like it again 
💔
elder willow
i was told its not ai
rugged heath
Should I put a image or a model in #1175430844685484042
storm meadow
i got close once in a kits model when that first released but it was super muddy...
I think your doing something wrong if your best model is a kits one
native locust
I think your doing something wrong if your best model is a kits one
nope bro it was the only one that converted accent properly it fooled just about eveyrone
nowadays models sound cleaner but the accent conversion is just mid
storm meadow
nowadays models sound cleaner but the accent conversion is just mid
it’s the index ratio
storm meadow
ive tried all different index ratios
Might be song choice then
near current
does anyone know why your voice echos while using the voice changer in okada?
storm meadow
Might be song choice then
Some models work better with certain songs then others
native locust
Might be song choice then
we tested the exact vox we used all the way back when kits first dropped on diff models
i think its the muddy dataset was able to hide some of the accent artifacts
my target audience is all people who listen to exclusively juice wrld for years everyone can tell when the mannerisms of the ai arent exactly the same 😭
and well i can tell too it just has never sounded just like jucie wrld even when i run literal studio session of juice wrld saying a line thru a juice wrld ai model trained on another song's raw vocals recorded directly before
just need to wait for svc innovations 
RMVPE isnt the best either
room for improvement is what i mean not that another one has given me better outputs
cosmic helm
Anyone got a guide for how to make your own models?
rugged heath
Anyone got a guide for how to make your own models?
-colab
frosty vigil
yo whats poppin
elder willow
Making a frank Sinatra AI cover right now.
final merlin
how to insert models on rvc?
storm meadow
how to insert models on rvc?
.pth file goes in weights (inside assets)
and the .index file goes inside logs
mighty isle
.wk
indigo pikeBOT
sour dirge
hi.
dim mountain
no clue if I already asked or not but I read kn some documentation that you can make a model without using a pretrain, I wanna try that but no colab or other training space allows me to do so. Idrc how long it takes or if I run out of space on GDrive space. If I have to jerry rig my phone just to do this I will
sour dirge
guys, i have something i wanna ask.
celest cargo
guys, i have something i wanna ask.
then ask it
sour dirge
I downloaded the Ai Voice Client.
But it's not working.
I mean, the voice-changer-native-client.
amber light
dim mountain
no clue if I already asked or not but I read kn some documentation that you can ...
Also slightly pissed off cause I wanna train in 48k and it's all being downsampled
tiny glade
is there another app to do ai cover? Im still using RVC gui 💀
crisp gorge
is there another app to do ai cover? Im still using RVC gui 💀
There’s weights.gg
tiny glade
There’s weights.gg
ty+
crisp gorge
Oh
feral haloBOT
Ayo? @crisp gorge level 1 !!!
crisp gorge
Idk
amber light
no like a application
just use locally
amber light
-guide
jade shore
hi
rugged heath
i uploaded the model
final merlin
how to insert models on rvc?
oh my god im such a fucking idiot i kept using the unzipped file
long tundra
Do anybody got the ai John carpenter song by xxxtentacion
nimble lion
hi
grave aspenBOT
glacial patrol
Hi
rotund narwhal
hi
supple timber
im so curious, what do ya'll use to make your covers?