#🧬│ai-chat
1 messages · Page 128 of 1
is 2023 ro
ion even know
is there a certain version of rvc i should download?
i been banned for like 8 months
i was in the original server before it exploded
Almost 24
i just use the colab, ion have a graphics card
20 more days to 2024
idk but i used to bully ppl on the og server
then Qo banned me
who tf will go to countdown
my train in the country litterally broke down in the middle of the tunnel
Right now?
2023 countdown
broooo 1ski needs to give me my roles back
Ayo? @crisp verge level 2 !!! 
Roles have changed quite a bit
Level roles are gone and you need to earn them back
i used to train models all the time and i remember the colabs started bugging out, is there one that works 100% of the time?
afk im bored
Model Maker works the same way
yes but AI devotees and that higher ranks didnt chnage
They did a bit
i used to have model maker as well
Do /resync
omg someone on insta named gohfamily777 followed me and his post is full of... woman pissing in toilet
same
Do /resync
prolly wont work but it's thinking
Or you can show one of your models in #1159514067187277865 using /submit
its jus thinking bruh
Or show other proof
idk kalomaze gave it to me or idk who
Do /submit in #1159514067187277865 to submit your model to get the role
will it get posted tho
RVC models only
yooo maki ligon i know him
Once you get accepted you have to post it
i remember him having it and also the role that contributed to the ai hub album
ai hub to p9rn hub
u remember me lil bro
or no
it's been long icl ion even remember most of the people here
almost no one's og anyways
PACILHub community
most of them are gone
besty gif
about time you return 
1skiiii my brotherr
i'm good man how about u g
everyone frm the old server disappeared right
i just recognize u, maki ligon and some other ppl
is there a way i can get my roles back




is the colab supposed to be training my model fast asl?💀
i mean i could but dont know what roles you had back then
no it's probably corrupted
what colab should i use
model maker, and lvl till ai fanatic
i wont be active as before tho, i jus want them cah they look cool
i use this one
i trained 3 models w this one
💀
is the run time shit fixed on here?
what 😭
it works just as good bro right??
The one he linked doesn’t work anymore
yk which ones do?
i used to work with this like 1 month ago or sum
-colab you have to use RVC Disconnected for training
- AICoverGen-NoWebUI, by Ardha Google Colab
- RVC Disconnected, Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
Credits to Faze Masta and Antasma for compiling these links.
You can find more info on the #1159513888199540817 channel. If you can't find your answer, feel free to ask for help in #✨│ai-help.
damn ion even know these things
alright thanks
are they easier than the og ones
i was gonna train locally but i was already seeing errors during ther feature extraction lol
or they more difficult
my laptop couldnt handle it either
No
yeah crazy
training locally on a laptop 💀
A bit harder
😔
ima jus use the one i linked
it works good for me i think
gl with training
give a sec with the roles
nah ion make models anymore, i js use them
training and making models is hell for me
used to be awake 24/7 to finish one 😭
gl doing more then 2 inferences before google takes away your runtime
I'm sorry, the user you mentioned does not have any models uploaded in my database.
i noticed that indeed
Ayo? @crisp verge level 3 !!! 
ill give you verified roles for now
appreciate it
- mm
Yeah that’s why you have to use the other colabs that automaze linked
when using disconnected do i gotta split my audios up into hella files or can i just put the full raw audio in there?
alr, so it works the same ?
like i jus put the zip and it'll work?
snm g love
Kinda!?
alr thanks
-colab
- AICoverGen-NoWebUI, by Ardha Google Colab
- RVC Disconnected, Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
Credits to Faze Masta and Antasma for compiling these links.
You can find more info on the #1159513888199540817 channel. If you can't find your answer, feel free to ask for help in #✨│ai-help.
First one is the inference colab
ima try it rn
can "santa illaria" stop pinging bruh
Ayo? @spring maple level 1 !!! 
no i want updates just not from them, they pinging my ass hourly
i forgot abt blocking act6ually
guys, where can I find the "this is a sample audio for you, do you like the voice model?" audio
You got the role that pings you for ai updates so you get pinged for such
they will ban u if u do the bare minimum
mb
imagine 50s receipts but the opposite version of it
bro fin so restrictive now
Ayo? @storm meadow level 76 !!! 
cant say nthn
1942
it's cause he's scared of getting termed again
bro he getting termed anyway
i should make my own server and lifetimers to join only
If you say it in #✦│chat you have less risk 
should i do it
man i'll jus do it fr
onb
finna invite cloudy icono and all these mans
inv everyne
Ayo? @formal quiver level 1 !!! 
snm g
yuh
This discordian hood life ain't easy 💯. Fake ass MFs who could snitch just to get unbanned 😔. Watch out for snakes in Discordia. 🙏
icono why r u here brother
This discordian hood life ain't easy 💯. Fake ass MFs who could snitch just to get unbanned 😔. Watch out for snakes in Discordia. 🙏
u invited me
LOOOL
that gif makes no sense here bro
Ayo? @teal girder level 1 !!! 
BRO DELETE THAT
Embed fail
😭 😭 BRO DELETE IT ICONO
i said it and everyone went crzy bruh
hii
why r u even talking here
Ayo? @crisp verge level 4 !!! 
ik i'm level 4 lil bro
bro if i had embed perms
ahahha
this is why you are worth NOTHING in here gangy, ill cook u
yea ive had sex with him i would know
ok so slide embed and verified
Does anyone have RVC v2 disconnected? and can you send the link
bruh
Suggestion for @floral kestrel
- AICoverGen-NoWebUI, by Ardha Google Colab
- RVC Disconnected, Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
Credits to Faze Masta and Antasma for compiling these links.
You can find more info on the #1159513888199540817 channel. If you can't find your answer, feel free to ask for help in #✨│ai-help.
thanks
tf did i say
You told him to roblox himself
k1ll himself?
Mhmm
https://discord.com/channels/1159260121998827560/1183596956254220288
I will be very sorry
Hi, I just joined
in Applio's Train tab, what "target sample rate" should I pick between 32k, 40k and 48k
Sup, I rejoined after a few months nothing special tho
hi, i joined because of lethal company
Ayo? @gloomy halo level 1 !!! 
Brazilians are either trans, hardcore favela members or football players
Speaking from experience
That's how they got r9
wut
Honestly I think you're 100% sigma
I think it's time to pay your fanum tax
bruh wut
Need some input from the community here: does anyone know what the deal is with the AI legislation the EU passed specifically in regards to voice models? Also - in regards to copyright stuff, where does that line fall with voice synthesis? Someone could claim a voice clone as a violation (if they didn't give permission) but what about a voice model made by amalgamating several voices? Isn't the final product no longer the likeness of anybody? So where's the line? Any thoughts? I need to know for commercial purposes
I think it really depends on the context. I’m no legal man but a lot of the time it should probably fall under parody if used for jokes and such
What about for original content?
Not really sure
Depends on the voice I guess 
If it’s a mashup of 2 or 3 voices probably fine but if your using just someone’s voice for a commercial project prob not
Take this with a huge grain of salt since I don’t really know the legistics
That's the tricky part - there's no real legal framework around this stuff. Cloning is easy for people to claim because it literally sounds like them, but what about amalgamating many voices? That's the idea - and how many voices would be needed to ensure it's truly a new voice?
idk much about the legal stuff, but keep in mind that voice synthesis ≠ voice conversion/changer
That would probably be ok but I’m not very sure 
True - someone showed me you can use the voice changer software though with a voice profile made from MANY voices - so it still applies a tone to your vocal performance but it's a new voice tone not held by any living person
what if said individual personally asked you to use their voice?
If you have permission then it's no problem and no question whatsoever
What's that?
ai voices
legal
In a STORE?
multimillion dollar industry
Or are those vocaloid?
they sound better
I spy katakana characters - don't the Japanese samples have trouble with English?
I've tried the vocaloids before and they struggle sometimes
tell me more senpai
Where do you get these?
So it is just vocaloid then? It can't apply to a voice actor performance? I need it for more than just music
hollywood is so far behind
Was that applied to a vocal performance or done via MIDI? It sounds ok - little bit of artifacting
no kidding - but that's true for SO many reasons
that was a raw input, can clean it up by hand if you want
Have you ever made someone from this server buy teto?
yes
Is there a site for these models? I can at least research it more
Ayo? @half river level 3 !!! 
@proud eagle them?
MoMone
thanks - I have homework now - I'll investigate and see if this can work for my purposes
use a legal voice for voicechanger work, there are many ethical ones out there
Are they all Japanese based though?
with permission you can sell and monetize it
That's the idea
no, the newest one is from new york
fucking bullshit damn Saros is so cunt and fits this song so well
SVP: https://www.dropbox.com/scl/fi/9c7fz4dgwikwmsm7dbz77/galacticmermaidunedited.svp?rlkey=rv2z4waaekyrhobffbtz1rsa6&dl=0
Original Media: https://www.netflix.com/title/80992137
(Episode 9)
Based off of this musecore: https://musescore.com/user/14925586/scores/6639367
i forgot ...
vocaloid 6 ai, cevio, neutrino, diffsinger, ace studio. all other ai voice software
rvc is very useful and versatile as it works on every sound and open source
I see - I'll puruse the catalog see if I can find a voice that'll work
This exists made with 4 voices, their name is ai hub chan
The website you sent for the voices - is there a version in english? I can't seem to find that here
Oh we got those translation tools too
https://dreamtonics.com/
https://www.vocaloid.com/en/products/show/v6std_mphdl_en
Gumi ai is on both of these
For voice acting on rvc, you can do it free with written permission
I like Gumi a lot but the songs I've heard her in do have a bit of artifcating which won't be good for voice acting
Wait - Gumi can work on RVC?
『Synthesizer V AI Megpoid』デモソングです。
デフォルト/Ballade/ Cute/ Soft/ Vividのボーカルスタイルをコントロールしながら声質を変化させています。
「Synthesizer V AI Megpoid」は、人間らしくリアルな歌声での歌唱を可能とする最新のAI技術を使用して開発された「Synthesizer V」専用の歌声データベースです。
歌手・声優「中島愛」の声をベースに制作しており、中島愛の声質で透明感のある自然な歌い方を再現します。
歌声合成ソフトウェア「Synthesizer V Studio Pro」を使えば、日本語 / 英語 / 中国語 / 広東語で歌せることが可能です。
◉ Synthesizer V AI Megp...
I know about the packs - but how can that work with RVC?
(And I'm still fairly new to Japanese so I still can't read most anything on the website 🫠 )
どこですか - lol best I can do
Pls help
It says "Pipeline is not initialized" and this "waiting generate Pipeline"
Ayo? @clear steeple level 1 !!! 
nakadashi
well alrighty then - just looked that up and now I have to add that to my Japanese vocabulary - I didn't intend to but here we are
look up Yua Aida
You guys keep up the great work with voice models 👍
Yua Aida is a person
ah
Has anyone tried training MusicGen?
Very easy to train your own voice models on rvc
I know - but you need the audio samples to train them
Am I to take it I should just buy a vocaloid and then use that to train the RVC? Seems quite a work-around - and may not even be that good of quality
bukkake
Ayo? @tight lark level 4 !!! 
Vocaloids and rvc dont mix well, and is against some ToS
Some already have rvc models made "officially"
Which ones have RVC - officially? Are they any decent?
What are you looking for exactly
hang on I can send a sample
https://www.youtube.com/shorts/iaTGfjELPqI This was done using another voice ai - I can't use it for commercial projects but this at least gives you an idea of the TYPE of voice
Maybe? Hard to tell how it would perform in english voice acting
Tone sounds decent
@hollow tangle
mh?
Got more examples of ilaria rvc?
my voice or the program itself?
Wait - is that voice available for commercial use?
I'm working on starting up an animated series - that's the goal here
Thats up to them
gonna delete them in a bunch of minutes
@quartz roost i think you never heard those
thoughts on using the zoom h5 as my main mic? rn im using the mic on my headset
Ok - so wait, is this YOUR voice? Like YOU did the samples yourself or what?
Ayo? @half river level 4 !!! 
i wonder
Zoom mics are good but get a condenser mic for vocals
🤨
real question, why?
Because I need a voice and one that I have permission to use in a commercial project
what is a condenser mic? i just have a zoom h5 laying around so i wanted to know if it was better than my headset
oh yea in that case you need the voice of someone who voiced the thing
and you know
people ask money for that
like i wouldnt give my voice away, especially if i know if its a commercial project
By a transfer of money it becomes "legit" and consensual
I was gonna pay someone already and haven't heard back
Teto is free and has a lot of flexibility
why not just look for an online q n a or interview? thats what i did for a lot of my models
the voice you heard is a tweked voice of mine, so its higher, basically a voice that doesnt exist but i own still the copyright
shoot a price
But its an exaggerated character voice
and you will be the only one that have it besides me lol
I'll DM you
okie dokie
Thats how teto is made
yea exactly
i followed basically that
and it turned out a great thing
It does have a really good tone
Theres guidelines on how to write the legal terms of use if you wanted to market your voice.
Bruh the Google Collab stops after 5 mins of runtime.. why is it happening?
Overload maybe
It happens everytime I try to use Google Collab for rvc
Ayo? @tacit quarry level 1 !!! 
From how much he talks about it he has to have
.
yuh uh
hi
im on tha good kusxh an=d alchool
google wasn't very helpful, can someone explain Epoch count? is higher better?
-colab @half river
Suggestion for @half river
- AICoverGen-NoWebUI, by Ardha Google Colab
- RVC Disconnected, Kit Lemonfoot Google Colab
- easyGUI, by rejects Google Colab
- Ilaria RVC, by thestingerx Hugginface Spaces
- RVC-HF, by r3gf Huggingface Spaces
- AICoverGen, by r3gf Huggingface Spaces
- RVC v2 Huggingface version, by Clebersla Huggingface Spaces
- Advanced RVC Inference, by neuclya Huggingface Spaces
Credits to Faze Masta and Antasma for compiling these links.
You can find more info on the #1159513888199540817 channel. If you can't find your answer, feel free to ask for help in #✨│ai-help.
I JUST HAD THE MOST MIND BOGGLING IDEA
WHAT IFINSTEAD OF AI ART BLABLA INSTANT IMAGE
WE USE LIKE
A PAINT PROGRAM LIKE MS PAINT
FEED THE AI INFORMATION ON HOW TO DRAW
THEN MAKE IT DRAW
🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠
ai hub gotta hire me fr
is there a way to stop voice cracks if my voice model goes to high?
There's nothing we can do ~ Napoleon Bonaparte
No, voice cracks happen because that model is not supplied with that high range
you can fix this by transposing down
need help with real time vc setup
which one
send me an ss of what the gui looks like
just send me a screenshot of what the application looks like when you open it
i havent even downloaded anything type new my fault, some guy just t told me to join this discord for help
scroll up a bit
until you find the downloads for diff GPUs
yea i got a 4070 ti so i downloaded the nvidia one
it's downloaded?
ping me when done
k
its done
open
k
then open MMVCSERVERSIO
yep
then scroll down until you find start_http
ight, should we vc or dm so we dont spam chat or nah
open it and DONT touch ANYTHING until it fully opens
and for important stuff likr that lmao that i coudlve missed
When I press
It says window prevent me from something
Ayo? @soft portal level 2 !!! 
hi
Hi
hola, soy Dora

hello
hi
what does rmvpe 900 epochs mean exactly? Which setting is that?
hello
I want to download a file I saw in the past, but I can't find it even after searching. What should I do? I checked many times for spelling mistakes.
why am i getting disconnected to a runtime
Which notebook are you using
can I ask a question
Sure!
chu chu lovely muni muni mura mura
Anyone got any idea how to reduce like choppiness of the voice when in game
Free
You can ask any questions. And can solve your doubts
Ayo? @queen kernel level 10 !!! 
what do u mean?
most* if not all software here is open source
Is there a reason for after using a model for a bit in game it like starts to change tune and sound wrong
Hey Movie Clips Can Be In CPB??
bad pc
gpu gets tired
not really bad pc
Hey Movie Clips Can Be In CPB??
Ayo? @visual axle level 1 !!! 
what r u talkin about?
just straining the gpu with two intensive tasks
Fair enough
dedicated NVIDIA is best
rtx 30+ recommended
tikotok beta broo? with edited movie clips
i have a 1070 and i can leave mine on for days at a time
I got a 4070 ti
i got no clue what ur talking about
lobotomy
hm, try limiting the framerate of the game and running it in low settings
I am 60 frames at low settings
hopefully then it won't strain the gpu as much
i sadly gotta play most games on lowres
It’s also the unoptimized mess of vrchat
oh
But I’m playing desktop
doesn't matter
vrchat just sucks for rvc
Damn
Ayo? @acoustic pewter level 3 !!! 
u gotta put all ur setting down
Like the main game I wanted to use it on
i recommend running rvc on a server
How do I do that
u can use collab
I’m brand new, I just got it running today
I don’t know what the best settings are for rvc
@molten stone
can u help this dude set up collab?
@acoustic pewter check pins in #🔍│help-w-okada
theres a link to a working collab there i think
its on my clipboard lol
nice
theres also a kaggle ver but meh
yeah i fixed it wokada collab
want the kaggle ver also?
yeah
noice
meow
nya
yeah thats fair
kek imma go make a model now
yay
just gonna go sit for 30 mins looking for a good dataset
found dataset
ong
but my download speed is shit
hi
Hi
Does anyone know the state of the art in TTS models?
Hi
hi
hello
ilaria tts works great for me
ElevenLabs is great but all the tech is private
@idle light u have a link right?
XTTS is open source i think
ilaria tts is open source aswell
Is it better than TalkNet 2?
never heard of that💀
Guide made by Julia
Works on mobile too! Just make sure you have a stable Internet connection.
How To Use
Start by making an account in Hugging Face here.
Go to the Hugging Face space of Ilaria RVC. It's recommended to duplicate the space to make it faster and to avoid errors in the system. For t...
TalkNet is pretty great but generates very stilted speech, it sounds robotic
I've heard from one of the members that XTTS is pretty good
Can you train it on custom voice samples? I have a model I'm trying to evaluate for RVC but it requires a certain style of speaking that i cant find a lot of samples for
@tardy hare how to make tts cover
Yup, you can
Actually encourages you to
Only need 6 seconds
And you can even make emotions
Very cool and open source
Coqui is going in the right direction but still lacks the quality. I wish there was an option to give it more data
So wait XTTS on the github page links Bark
is XTTS an interface for other transformer based models?
Their commercial models I'd say are on par with ElevenLabs
Ayo? @plucky sundial level 5 !!! 
Im still holding out for english voicepeak
I literally thought frimomen was real
I'm not sure the tech would transfer over to english though
Would you say XTTS is Open SOTA?
I asked them on twitter one day and they said "we do not have an ETA yet" so they are working on it
I have not tried SOTA
Oh no sorry SOTA = State of the art
Oh i misread that
Xtts has a lot of potential
At the current moment i dont need an english tts since i can speak it myself. I use a japanese tts since im still studying
I can speak it myself too the issue is RVC heavily favours accents heard in the recording 😦
Look in talqu3PRO ai
Yo
Also can you not self host Ilaria?
hi
Using Applio, I tried to run the preprocess step on 293 audio files and i never got the "End Preprocess" message as stated by the Applio tutorial despite waiting for hours. Is this because 293 files are too much?
there is actually
search for it
Anyone have any advice for getting a P2 instance from AWS?
Submit a ticket and ask, they usually are willing to give them if you make the effort of making a ticket
Hi Uwu
hi
Hello 🙂
hello guysss 🙂
My vc is stuttering and not working with a 4070 but my friends is working fine with a 3060 and we have the same settings and everything
no idea if its just applio but using mainline seems to work just file with 700 audio file
They may have CUDA acceleration set up with cuDNN
Hello
Nah I was the one who showed them how to set it up
We did everything the same
Is there anything I could’ve done wrong in the setup to cause this
Ok bruh
My friend with a 1070 can even use rvc in vrchat without it stuttering
Anyone knows how to better the low quality audio in kits ai cover songs?
Ayo? @blazing folio level 1 !!! 
Tried with UVR but it cuts some parts of the voice with Kim Vocal or Karaoke
OKAY XTTS Is pretty fucking amazing..
@tight mason do you have to Octoling Boy + Girl model? Because the Restore Bot didn't upload them.
Ayo? @steep spire level 1 !!! 
Don’t use kits
It absolutely sucks
Does anyone have a Octoling Boy + Girl model?
Thanks for the advice
yall how can i use the voicechanger like on a call or in a game it wont work for me
Hello? I said, does anyone have a Octoling Boy + Girl model?
bro can someone tell me how i can use the voice changer in a game and on a call
Virtual Audio Cable #🔍│help-w-okada
Ayo? @elder willow level 1 !!! 
Okada/Real Time Voice Changer Client Demo
ok ty
Does no one wanna answer me?
Check weights.gg
i still cant use this voicechanger on discord or any game
im gonna start hating that site for the reason that they need you to make an account to download a model you can find on huggingface for free without an account
For HuggingFace most people post all of their models as 1 model and you don’t know if it’s a RVC model or not
can u maybe explain it better i still cant use it on discord or any game
You need to install a virtual audio cable to use in games or discord
oh ok
After installing set your output on Okada to it and same with your input on the game your playing
#✨│announcements message I lost my 3DS
pp

Ayo? @waxen grove level 48 !!! 
It's on my desk as we speak
make a Caillou cartoon voice please
Pls don’t
it still dose not work
Did you install it
yes
So you extracted everything and then run the setup exe
yh
That’s why
NOOOOO
ONLY IF I DIDNT LOSE MY MODDED 3DS
I'm not claiming anything btw, just for the fun of it
it dose not show line 2
Ayo? @elder willow level 2 !!! 
What are the options?
If tea's prize still stands it's for whoever bothers next
only line 1
Click it
i did
Test it
now my mic dose not work
Is your output on Okada set to the cable
And is your Okada even working
noooo
try no ui
colab most definitely won't do
What's left ? Ssh'ing to a system and starting from CLI ? That's cheating
the colab UI is just not gonna render lmfao
a lmao
A mere 3ds vs the modern web
Hi, there is any guide how to train a new voice model?
Boys, I am in a server where they banned me from talking about AI. How can I convince the server owner that AI is good and makes original, creative art?
Hey what's the best local TTS interface/model/ui, I've downloaded Ilaria RVC but it seems to only run on huggingspace
Pleaseeee likeeeee
https://youtu.be/wBCzOYbQ5Yw?si=5CNLDNJ-v4cyJWY0
This is an ethical legal commercial ai voice ORIGINAL song
Maaaaaan hace mucho tiempo que había traducido esto, pero soy bien perezosa xdddd
Link:
Romaji: no lo hice D:
#neru #teto #kasane #synthv #重音テト
Everyone is so fucking stupid this shit is not new and has been well established for decades
The owner keeps yapping that we are stealing people's artwork to train the AI
This voice provider personally ASKED me to use her voice
And likes what i have made with it
Or image generation its up to the end user to make sure the data used is approved.
Show them this
60 bucks?! Dear gosh that's expensive
Cheaper than vocaloid
And oh, I will show
You can SELL and copyright your music
I am aware of that
Its defined as an instrument
This tech is fucking 40 years old
How are people this dumb
I ask the same thing, bud...
But this server owner that banned me from talking about AI just gets ignorant and pessimistic
Ayo? @spiral junco level 3 !!! 
Meanwhile we host live concerts and they go hard
Is it just me or was the old way of finding models easier?
Kept showing examples of AI being used for good things, like AlphaFold, and other medical and industrial applications, but ignorance kickd hard
I would like to go, but I hate loud sounds
Ai concerts are not the same as live band. They dont pretend to be. They are at electronic dj clubs.
Well yea, but they still play music really loud. Which is why I never go to concerts or clubs
Welcome to SynthV Teto! Very happy to have you here!
Wanted to test her English right away, and decided to release the SVP to comemorate her arrival. She's a real powerhouse even in SynthV!
Original Song by Demi Lovato
SVP, Tuning & Mixing by Ariotsu:
bit.ly/41Jcyzo
You guys have fun
This is how it works
No way around this part sadly, but its a sad reality in service of all the good shit that AI can do
They spent over a year perfecting this voice in a professional studio
Its all derived from weather research algorithms lol
Ello
i do ! i completely forgot abt this server though i can reupload them in a bit
Haii :3
The basis for a lot of this research was decades ago trying to model weather patterns
But it isnt for 98% of other people here
Yeah but if you look around that doesnt appear to be the primary fuction today
If I share that screenshot you sent me, I will get definitely banned off the server, but it's a risk I am willing to take. 
Im using her to prove a point that everyone perceives ai wrong
One edge case
Hooray
The cons exist but the pros are just that much better
Human calculators lost their jobs when computers came around
Do it, meanwhile im using ai to open a window in the language barrier
They didnt really lose their jobs. It just shifted to a new market
That too thats true
We need coders more than ever
A woman hand sewed copper wire to land men on the moon
Not really. Look at college students, they still need to be human calculators otherwise our instructors will fail us if we use a calculator
hello
A calculator is useless if you dont understand What you are calculating
Tell him the problem with ai is PEBKAC
You mean they learn math stuff to understand the concepts they are using and to cultivate the critical thinking skills to apply that knowledge well
And that's just the greatest moment ever.
There's an image of a female coder who was stabding hers to a stack of books as tall as her with the code of Saturn V
Tf is PEBKAC?
Problem exists between keyboard and chair
Oh
Industry terminology
I see
Gotta sound professional
At that it works perfect
@quartz roost do youknow what a feature search database is?
I dont do programming anymore. I cant rot my brain for 14 hours a day on a screen
@idle light whats the link to your guide
i'm just trying to understand all the advanced features, like f0 curves, is there any way currently to "remove" an accent from an audio recording
Guide made by Julia
Note: It works on mobile too! If your device is slow and can't run RVC locally, don't worry. This software doesn't use your device's resources. All you need is a stable internet connection.
READ THE GUIDE TO THE END. It will be worth it!
~Table of Contents~
How To Use
Settings...
@plucky sundial there ya go :b
Phoneme strength and duration (different software)
Check chapter two
Yeah I can't use Ilaria as it's not local, does this apply to Applio too?
I mean, it's discord, professional speech isn't mandatory
Yeah!
WHAT
Fantastic thank you 🙂 i will read this but are you cool if I dm you a question real quick?
BRO WHAT
Im in a dev server for a japanese tts software. Im as professional as possible
Oh you can tell me here, it's fine
what do you even mean?!?
Do so in that server...
I use professional speech only when communicating with my professors and employees at industry, but here on discord I am really laxed on speech
Weird
Your speech still better than 99% of average discord users
100%
The bar is pretty low.
⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⡿⡿⢿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿
⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⡟⣫⠱⣬⣴⣾⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿
⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣟⠨⠅⣾⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿
⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⡄⠡⠻⣿⠿⠟⣛⣛⢛⡛⠿⠿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿
⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⡿⢇⣑⢢⠞⣙⢦⡙⣎⢝⡺⡔⢤⢈⡛⢿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿
⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⡿⢋⡔⣫⢌⢧⡛⡜⢦⡹⣌⠞⡴⣙⢆⢣⠘⣣⡙⢻⣿⣿⣿⣿⣿⣿⣿⣿⣿
⣿⣿⣿⣿⢟⡩⡕⡲⢤⠈⠍⡴⢃⡜⢦⣋⢖⡩⢚⡥⢳⣌⠻⡴⣩⠞⡤⢓⠠⢏⡥⢈⠭⠩⣍⣛⡻⠿⢿⣿
⣿⣿⣿⡏⢮⡱⢭⡱⢃⠎⡸⢁⠶⣩⠆⠘⡲⢄⢫⡜⣣⠈⢳⢥⠳⡌⢲⢩⢆⠩⠜⢢⠘⡀⠼⣡⢏⡝⡲⡌
⣿⣿⣿⡗⠤⢂⠥⢢⡁⠀⢡⠞⡑⢆⠢⡥⢙⡌⠶⡙⢦⠑⣆⢌⡳⡱⢈⡳⢎⠄⢀⠃⠂⣀⢻⡐⡆⢨⠑⣁
⣿⣿⣿⣷⡘⠄⢪⡕⠎⠱⠋⠀⡜⢆⡹⣼⣅⠚⡤⢙⢢⣹⡎⣦⠱⡓⡄⠹⢎⠐⢪⡜⡁⢀⠣⡝⡼⣡⠎⣸
⣿⣿⣿⣿⡇⢌⢦⡨⠔⡡⠀⠄⡛⠄⣡⡀⠹⣷⡄⢃⢆⡿⡩⠄⠀⠙⡄⠣⠌⣈⠒⠬⢁⠆⡐⡒⠰⠁⢺⣿
⣿⣿⣿⣿⣇⠊⠢⠑⠃⢁⠡⠀⡍⣰⡁⠀⢀⣿⣿⣆⢼⣿⡇⠀⠀⣀⡌⡐⠡⡀⠆⡘⢠⠘⠤⣙⠱⠈⣼⣿
⣿⣿⣿⣿⣿⡏⠔⠦⣌⡐⠨⠁⠄⢻⣿⣷⣿⣿⣿⣿⣾⣿⣿⣷⣾⣶⠇⡐⢁⠐⠠⠈⠂⠉⠰⠤⠃⠇⣼⣿
⣿⣿⣿⣿⣿⣿⠊⠴⠠⠐⢀⠃⡀⡸⢿⣿⣿⡿⢿⣿⡿⢛⣿⣿⣿⡟⠀⢀⠂⠌⢀⣱⡈⢌⡡⠐⠠⣹⣿⣿
⣿⣿⣿⣿⣿⣿⣆⠰⠐⠢⢤⣾⡄⠙⣮⡻⢿⣿⣶⣶⣾⣿⣿⡟⢏⣱⣾⣷⣶⣶⣾⣿⣷⢈⠡⢉⣢⣿⣿⣿
⣿⣿⣿⣿⣿⣿⣿⣄⠃⢤⣿⣿⣿⣼⡄⡙⣷⣍⠛⠿⠻⡙⠦⡘⣾⣿⣿⣿⣿⣿⣿⣿⣿⣄⠃⣠⣿⣿⣿⣿
⣿⣿⣿⣿⣿⣿⣿⣿⣀⣼⣿⣿⣿⣿⣿⠾⠛⡉⢀⢂⠱⣈⢖⣇⠛⠿⣿⣿⣿⣿⣿⣿⣿⣿⣶⣿⣿⣿⣿⣿
⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⡿⠿⠋⠄⡆⠡⠐⡈⣎⣧⡝⠚⣨⠐⠠⢀⠙⠻⣿⢿⣿⣿⣿⣿⣿⣿⣿⣿⣿
⣿⣿⣿⣿⣿⣿⣿⣿⢛⣭⣇⠐⠠⢁⠸⠤⠑⡐⠀⡝⢂⠔⡣⢄⢃⡡⠂⡐⠠⢀⠂⡨⣙⣽⣛⡻⣿⣿⣿⣿
⣿⣿⣿⣿⣿⡿⡿⣱⣿⣿⢃⠌⡐⠠⠘⠢⡅⢂⠰⡈⠰⠩⢔⠢⢊⠀⠔⠠⢁⠂⡜⣴⣿⣿⣿⣿⡜⣿⣿⣿
⣿⣿⣿⡏⣭⠑⢲⡻⣟⢮⡈⢒⠄⠡⡈⠱⢀⢢⠃⡔⡡⣁⠊⡐⠤⠈⠄⡁⢢⠰⣼⡿⣟⣿⣻⣿⡗⣿⣿⣿
⣿⣿⠩⣆⠘⢿⠆⣛⠡⠊⢀⠩⡘⠄⡐⠡⠀⠎⡄⠠⠑⡬⢑⠌⢂⠡⠌⢰⡁⢧⣟⡳⢮⡜⣝⢯⢳⣿⣿⣿
So I have a recorded model that's based on a specific english accent, I can't find enough audio online to transpose it with, I'm debating going back to something like TalkNet, what are my options, what is the current SOTA in dealing with accent removal/transposal (sorry for the repost just wanted to reply so it's tagged if you can respond later if youre super busy)
No idea what transpose, TalkNet and SOTA is 😭😭
If youre using rvc, the accent doesnt especially matter (ish?)
For example any model trained on any language can infer audio of any other language perfectly due to how audio inference works
For example thats why a cover of freddie mercury singing a michael jackson song sounds a whole lot like MJ (or if the model is good, freddie doing a really good MJ impression) The cadence (and by extension accent I would assume) carries over
I mean like, taking audio and applying a voice to it, instead of generating it from text, and SOTA I just mean like.. whats the state of the art/most effective tool etc
If this answers your question, RVC does not distinguish by accent. You can make anyone from any language and accent talk or sing in any language
I am using RVC but the output is heavily bias towards the accent of the input audio
Lower the search feature ratio
Search Feature Ratio: This modifies the accent of the model.
A higher value makes the output sound more like the audio the model was trained on. If the model's dataset has background noise, reducing this value will lessen that noise in the output.
Lowering this value can also fix artifacting issues on some models.
Just said that
I tried higher because I thought that meant it would sound more like the training audio, I'll try lower
I did read it just now
Beat me to it >.<
A higher value makes the output sound more like the audio the model was trained on. If the model's dataset has background noise, reducing this value will lessen that noise in the output.
He ignored me its all g
I'm not ignoring you...
I read the guide, but what the guide says doesn't match what you're saying so I'm trying to understand
Gex
💀
Wow
I think I didn't understand what you said then
Are you sure you mean accent and not just how the voice is
So if I have an input audio that has a totally different accent, say an American accent and an english accent trained model, I want to make the output sound less american
Ayo? @plucky sundial level 6 !!! 
Is there any way to do that, or do I need to use some english accent TTS or similar
if I upload audio that's english accent it works great, but if I upload an american accent audio it sounds like an english person doing an american accent, is there any way to prevent that, or is that just a limitation of the RVC mechanism
How do we find the application in this zip file on Mac?
??
Is there any urgent answer?
Send the audio if you have it Im trying to understand better what you mean
Cause I have audio of my own
how to put link drive??
What's a link drive?
Bryant Myers
https://drive.google.com/file/d/1hjAJY8k7AuVoNpUo6fRhXS7qDqIO6tSW/view?usp=sharing
Author: Luzz (@modern perch)
Created: <t:1685735005>
Type: RVC
Algorithm: Rmvpe
Epochs: 250
From that link, how far do I cut?
Maybe I'm explaining it badly, I have a model trained on an english female accent, but finding audio to "convert" through the model that matches that accent is hard, so if I upload for example an audio clip of an american female accent the model output is the english female doing a bad american accent, is there any way to prevent that?
Beyler karı var mı
Ayo? @coarse jungle level 1 !!! 
If this is a limitation of RVC that's totally fine, I'm just trying to understand if there's anything I can do
julia
If you want to raise the accent of the model, raise the search feature ratio value
soy español no hables en ingles
Right, I tried that 0.9 and 1 (From reading your guide) and it still sounds like an american accent
Rvc can only copy the pitch that is inputted you will need to manually edit this for what you want to do
It's not the pitch though
Send an audio it would help
e probado a cargar modelos tipo zip, pero los de drive no se como cargarlos, como lo hago?
pues explicamelo en ingles
Ayo? @bold hare level 1 !!! 
I can PM you it but I don't want to post it here

I am familiar with programs like Google Translate, but it would be really nice if any of the computer Scientists in this server to develop an open-source program that is able to translate languages in real time with the least amount of delay.
Basically the W-Okada voice changer, but it's a language translator.
I can make a group chat quickly if that's easier, I've previously used talknet which takes the text data (so transcriptions of the audio) and the audio, and links them up, but it sounds robotic, so I hoped RVC would allow more realistic sounding audio
Good luck witht that
Gonna need beefy hardware
Pones el archivo pth e index de tu modelo en un archivo zip. Vas a tu cuenta de GD y tocas el botón con el símbolo de Más ( + ). Y ahí seleccionas el archivo
Ayo? @idle light level 57 !!! 
You send audios in help rvc
That's why I am sharing this idea to you engineers and computer scientists
I don't want them shared publicly
I am not a CompSci
Ah
There's this supposedly
I'm happy to pay you both to help
That's interesting
I feel like your only solution here is to either get an audio sample with less of an american accent, or a model with a bigger English accent
Or both
I don't understand what the latter means
oh ye that one.. would be nice if they make it work with rvc models
Wait we need to listen to the audio first to confirm thats the issue
With a more pronounced English accent
Sure then
que es el index
Los modelos de voz suelen estar compuestos de dos archivos. Un archivo de formato .pth y un archivo de formato .index
I tried the seamless demo, it seems to require good input audio
temgo 2 index en la carpeta, cual cojo?
Do you think it is possible to train models via HF? And without a time limit like Colab
Hay uno que se llame "added"?
🙂
si
Usa el que se llama added
Im using this. I found whisper + deepL works a little better
Testki Do you think it is possible to train models via HF? And without a time limit like Colab
Waiting to try v2 seamless
gets intonation?
12 hours
Im only using it for accurate transcription
good enough
vale, ahora tengo esos 2 archivos en la carpeta, cual es el siguiente paso?
ahh ye then its fine
Yeah seamless doesn't seem to get the intonation either
It kind of does ? but not significantly
v2 should
Como dije antes, ve a tu cuenta de GD y toca el símbolo de Más
hello everyvone^^
What's the version the website demo runs ?
I get a lot of errors with seamless v1
This supports multiple models
Working on a kasane teto plugin :3
y en mas que hago?
https://www.youtube.com/watch?v=RKEFZ44YOcc v2 preview ig
Tocas el Más y buscas tu archivo zip
pero gd es google drive no?
Si
aber si va aser geometry dash
Hubieras buscado en google "como subir un archivo a google drive" xd
Whispering tiger also plugs into seamless
Its just a nice ui that combines everything all in one
Ehh
It didn't pronounce "m40" right in french
and the translation is probably not the best
deepl translation still the best i can find
I have a lot of issues with seamless translation. Its transcription is decent but whisper v3 is more accurate
cuando este subido que hago?
I want an entirely free locally run translator
But deepL works the best
i remember the time when i was still watching some random dude on youtube making his own waifu ai
now here i am XD
I am using teto voice as the output... Lmao
Thats why i want the plugin so its automatic
wake up f1lthy
@wintry sinew that's crazy

teto command? wut u want it to be?
sussy link
@tardy hare ban
hes timed out but make sure all the links are deleted
yo, is there a chat to ask general questions beyond rvc or okada? I am having trouble even installing python... lets just forget about pytorch... it seems impossible for me to even get it to work despite searching for tutorials on youtube
tnx for the notify
What's the sleep vc for?
idk
oh boy, I dont think my starter classes of university will get me this far. I am a little lost here... how do I check my current version?
sleeping
Open up a terminal windows or cmd.exe, whichever
wait, I have 3.12.1?
Ayo? @kind cedar level 1 !!! 
and type in python --version
yeah yeah, ty!
so this means I got the right one?
What do you need it for?
for pytorch
voice models... I want to sound like toad
Oh, which software do you plan on using for that ?
uhhhhh, gimme a second. I have no clue
Do you want your voice to get changed in real time ? Or do you want to convert files from one speaker to another ?
best free ai voices? I use the voice changer client demo
@quartz roost Solana is in the chat 
Solanoooo, el prime
w-okada... real time
Oh, you don't even need python installed to use that !
wait what?
Yeah, the releases provide their own python bundle
what do I need? because every tutorial I have watched so far told me to download pytorch
Yeah... most online video tutorials for it are... plain wrong
I've yet to see one with the right steps and info x)
is there a guide on this server how to install it?
can you help me to get a nice one?
ah, thank you dude!!!
oh thx
It's unfortunately a text guide, hope that isn't too long lol
Honestly to find models, just go through https://weights.gg/ or #1175430844685484042 and do trial and error
Thats not an issue! But thanks a lot. I appreciate it.
thx
;3
I'll come back to annoy you if I run into some trouble
what are they looking for?
im too busy getting stoked for the moomin park
Tove Jansson Akebono Children’s Forest Park in Hanno city.
note on weights.gg you need to signin for an account to get the models from there
hermano soy español, no me entero de na
go to search models and put /search
or smth like that
I dont know where to make models i need help asap