same exact code
#✨│ai-help
rare gobletBOT
Ayo? @rigid dirge level 8 !!! 
nocturne mural
nocturne mural
Possibly your ram is too busy to try to make inference with a too long audio, remember that the audio is loaded in your ram.
try a smaller audio or wait for your ram to calm down.
proud elbow
that case usually either throws out of memory error or prob crash. anyway Applio has split audio feature to save time and memory usage when there are silences, though you'd better noise gate the input audio before.
rigid dirge
oh as well, does anyone know how to bypass this im trying to split vocals using x-minus
really?
thanks
still getting memory error but now when generating the audio
my memory apears to be fine tho thats the wierd part
il troubleshoot for a bit
give me few mins
autumn glade
-colab
proud elbow
oh as well, does anyone know how to bypass this im trying to split vocals using ...
you should subscribe to x-minus pro for accessing some best models like melband roformer and BVE v2
or for instead, use mvsep with a free account, though the size limit for a file is 10 mins and 100 MB
hexed kiln
im using the web ui colab and i have no clue how to use it
caus like im uploading a voice over audio and returning a whole chorale of voices
not just one
what am i doing wrong
oblique crescent
stark depot
why it says File exists, proceed with creation of creds and client
it doesnt download on easygui
help
azure marshBOT
Suggestions for @stark depot
stark depot
Ohfkkk thats why thanks
lunar rampart
what's the latest RVC for AI covers? do we still use this one https://huggingface.co/spaces/TheStinger/Ilaria_RVC
peak tusk
what's the latest RVC for AI covers? do we still use this one https://huggingfac...
applio
peak tusk
depend on what you want to do
lunar rampart
Thanks i'll try
rare gobletBOT
Ayo? @lunar rampart level 1 !!!
remote epoch
Yooo I'm new here
lunar rampart
I see
dire wraith
What should I do if I get this error at the second stage:
JSONDecodeError Traceback (most recent call last)
<ipython-input-17-228b32c5e2ef> in <cell line: 31>()
31 if os.path.exists(config_path):
32 # File exists, proceed with creation of creds and client
---> 33 creds = Credentials.from_service_account_file(config_path, scopes=scope)
34 client = gspread.authorize(creds)
35 else:
5 frames
/usr/lib/python3.10/json/decoder.py in raw_decode(self, s, idx)
353 obj, end = self.scan_once(s, idx)
354 except StopIteration as err:
--> 355 raise JSONDecodeError("Expecting value", s, err.value) from None
356 return obj, end
JSONDecodeError: Expecting value: line 1 column 1 (char 0)
azure marshBOT
Suggestions for @dire wraith
remote epoch
-colab
rare gobletBOT
Ayo? @remote epoch level 1 !!!
remote epoch
Ayo? <@1212492797303722069> level **1** !!! <:lfg:1159355870119993496>
YOOO
azure marshBOT
Suggestions for @remote epoch
lunar rampart
I tried and the voice sounds distorted what do I do?
it never happened before
oh yeah i changed model now it works
sinful brook
why cant i choose a model?
brittle wing
Am I better off using the rvc realtime client or using the collab one
I have a gtx 1660 super
slate compass
Yo guys my rvc doesn't work with discord propely
wanton sleet
im using W okada and some of the ai voices here that says RVC don't have the index file
rare gobletBOT
Ayo? @somber finch level 1 !!!
naive matrix
guys i can not hear myself and my output is working
brittle wing
Should I ignore this
The yellow text
Does anybody has the download link for the rvc dev version
The one with WASAPI
pastel oak
Does anybody has the download link for the rvc dev version
i repackaged the latest rvc versions from april
so u dont need dev patch thing that i made a while ago
then just follow this for wasapi
https://rentry.co/Realtime-less-delay
brittle wing
i repackaged the latest rvc versions from april
Oh so the "Retrieval-based-Voice-Conversion-WebUI-dev" thing is outdated now?
pastel oak
Oh so the "Retrieval-based-Voice-Conversion-WebUI-dev" thing is outdated now?
Yea, no need for that anymore!
just get one of the links from the first link i sent
brittle wing
just get one of the links from the first link i sent
Thanks
I don't have to do any folder dragging again right
Just like what the old one told to
pastel oak
I don't have to do any folder dragging again right
no need to folder drag. you just download the "rvclatest" file either nvidia or amd and run that
brittle wing
no need to folder drag. you just download the "rvclatest" file either nvidia or ...
For nvidia GPU's do I have to enable cuda
because I recently just replaced my amd gpu with an NVIDIA gpu
Had to reinstall rvc again so it works on my nvidia gpu
pastel oak
Had to reinstall rvc again so it works on my nvidia gpu
it should automatically work already. run the go-realtime-gui.bat and check if it detects your gpu
else you have to run a command to install it but thats not too difficult
brittle wing
it should automatically work already. run the go-realtime-gui.bat and check if i...
I didn't install Pytorch with cuda
Do I have to seriously reinstall python and pytorch again 
pastel oak
I didn't install Pytorch with cuda
Did you first try out if it runs
or did an error like that show up
brittle wing
Did you first try out if it runs
Yes it runs very well when rvc is enabled
Cuda is detected
pastel oak
Then you dont need to do anything else if it runs
brittle wing
Hmm I just worried if it would cause any stuttering issues if I didn't install Pytorch with it's CUDA package
I have a Gtx-16 series card
rare gobletBOT
Ayo? @brittle wing level 4 !!!
pastel oak
Hmm I just worried if it would cause any stuttering issues if I didn't install P...
you dont have to install them, the package already utilizes them
it comes with python
brittle wing
you dont have to install them, the package already utilizes them
Ohh
I was just about to send you this lol
I did some research and the GTX 1660 super has CUDA compute campability of 7.5
Just like the RTX 2060
pastel oak
I did some research and the GTX 1660 super has CUDA compute campability of 7.5
Look as long as you have something about found gpu nvidia you dont have to mess around with all those things
brittle wing
brittle wing
Look as long as you have something about found gpu nvidia you dont have to mess ...
Ah alright thanks
pastel oak
if you have realistic results and your voice changer doesnt take 20 seconds to convert voice, youre fine
brittle wing
I'm pretty clueless about this Pytorch stuff lol
dense abyss
how to fix this? "Invalid index. You must use added_xxxx.index but not trained_xxxx.index!"
proper shale
how to fix this? "Invalid index. You must use added_xxxx.index but not trained_x...
index used is probably the wrong one, try using a different model or try to find the added index in case you trained it
queen flame
i have the same issue, but it get the error for every model that i use despite using the respective added_ index
rare gobletBOT
Ayo? @queen flame level 1 !!!
proper shale
where's this happening btw?
queen flame
console log. It doesn't crash and i can still use it tho
proper shale
what are you using exactly tho
queen flame
using this with the amd/intel version idk if thats what you mean https://rentry.co/rvcrealtimeguide
proper shale
Oh I see...
For realtime, I don't really see index as necessary tbh
queen flame
oh okay, ty. I don't really the need the realtime aspect, i just want to create short audios with the model. Is there something else instead of RVC that is better fitting for this purpose?
torpid siren
Hi guys can someone tell me how to make my own voice model? Or send a tutorial
proper shale
oh okay, ty. I don't really the need the realtime aspect, i just want to create ...
yeah, you could use applio for that
queen flame
will take a look into it, ty 🙂
proper shale
Hi guys can someone tell me how to make my own voice model? Or send a tutorial
https://docs.aihub.wtf/ has everything you need to know about training models
frank abyss
can somebody help me I'm using Ilaria RVC and the program is saying out of memory
its trying to use 37 GB of VRAM
torpid siren
Thanks a lot
proper shale
will take a look into it, ty 🙂
yw!
proper shale
Thanks a lot
yw!
frank abyss
Model:
i'll send more just one sec
and also i'm uploading a file that is 83 MB anda 2 hours long
changing the file duration doesnt really affect it i tested it
proper shale
and also i'm uploading a file that is 83 MB anda 2 hours long
mmmmm
frank abyss
changing what model i use does tho
neon osprey
why are some voice models gone?
proper shale
changing what model i use does tho
might be the model then
subtle thicket
someone can help me i can hear myself in but not with my voice changed it's still the same pls
proper shale
why are some voice models gone?
server got deleted in october
weights.gg has most of em tho
neon osprey
ok ty
proper shale
someone can help me i can hear myself in but not with my voice changed it's stil...
passthru might be on, disable that
frank abyss
now its asking for 57 GB bruh
proper shale
yes but when its not on i hear some strange noise
okay okay, let's get more details here
what's your GPU?
subtle thicket
how do i know ? :'x
proper shale
now its asking for 57 GB bruh
i have a feeling it might be the 2 hour audio, too
proper shale
how do i know ? :'x
task manager > performance tab
subtle thicket
3060 Ti
rare gobletBOT
Ayo? @subtle thicket level 1 !!!
subtle thicket
nvidia geforce rtx
proper shale
hm
send a screenshot of the voice changer gui, you might be using the wrong version for your gpu
proper shale
that explains it
use rvmpe as your f0, increase chunks to 112/128
subtle thicket
sorry but how do i do that :'I
proper shale
sorry but how do i do that :'I
just click on the dropdowns and select the options i told you
proper shale
yep
proper shale
Mess with tune btw
subtle thicket
thanks
proper shale
You're welcome :)
proper shale
IT STILL ISNT WORKING
? what's up
torn nebula
i ran my ai vc
with start_http
at first
it worked
once
and then i was like ima use this later
and then closed
but later came
and this shit is just being a whitescreen
proper shale
o yeah, delete stored_setting.json and restart
should work
torn nebula
ah ok thx
torn nebula
o yeah, delete stored_setting.json and restart
should work
now i need js the best settings for this
proper shale
now i need js the best settings for this
i think that's decent tbh, increase extra if you want a lil bit extra quality, will give extra delay tho
torn nebula
ah extra is quality?
rare gobletBOT
Ayo? @torn nebula level 3 !!!
torn nebula
i thought it was just random numbers that was included for something based on your gpu
proper shale
ah extra is quality?
not really, it's like it's giving an amount of extra time for the gpu to process the conversion
but yeah it does give better quality in the end
rare gobletBOT
Ayo? @fervent oar level 1 !!!
proper shale
I'm not sure how you'd get something similar tbf. Either use ElevenLabs or... record yourself and convert the voice with RVC, using some anime voice model
ionic oak
I have a 5 minute voicemail, what should I do next pleasee?
daring heath
maaan wtf how do i fix this shit
pastel oak
maaan wtf how do i fix this shit
you link /content/dataset
not the ZIP file directly
and you would probably have to put all your mp3/wav/flac whatever you have inside that zip inside that folder
worn fable
which software do you all use guys?
worn fable
you link /content/dataset
welche software ist das was er/sie benutzt?
pastel oak
welche software ist das was er/sie benutzt?
findest alles hier
https://docs.aihub.wtf/
unter RVC wenn du ne gute NVIDIA gpu hast dann guck in local rein, sonst über cloud
worn fable
findest alles hier
https://docs.aihub.wtf/
unter RVC wenn du ne gute NVIDIA gp...
hab 3060 reicht das?
proper shale
I have a 5 minute voicemail, what should I do next pleasee?
Now you just read the Mainline guide, specifically the training part:
pastel oak
hab 3060 reicht das?
ja
proper shale
not so sure, check mel and kl to see if the model is still improving
proper shale
still improving imo
boreal sage
can i refresh colab without it stopping
proper shale
probably yeah
proper shale
g/mel ye
rare gobletBOT
Ayo? @worn fable level 1 !!!
pastel oak
welche der 3 softwares empfiehlst du ?
mainline
boreal sage
it's a pretty clean dataset too
rare gobletBOT
Ayo? @boreal sage level 2 !!!
boreal sage
insta live recordings of him talking
boreal sage
you could probably keep going
btw how do i detect overtraining on mel and kl
thin stump
btw how do i detect overtraining on mel and kl
if kl rises but mel keeps going down then keep training. It's bound to drop
boreal sage
if kl rises but mel keeps going down then keep training. It's bound to drop
huh both mel and kl are going down
i don't get it
proper shale
it's still going
thin stump
huh both mel and kl are going down
You can stop if all two of them are sharply rising up
boreal sage
You can stop if all two of them are sharply rising up
oh okay
worn fable
https://weights.gg/de/models/clwtlgea900onn3u0x94h5kqp
I downloaded this model but where is the audio example file?
thin stump
https://weights.gg/de/models/clwtlgea900onn3u0x94h5kqp
I downloaded this model ...
Listen to the weights sample. It doesn't tell you if its a AIhub model anymore 
proper shale
basically, stop training, export model
after this, just do all the steps as before with the same settings inserted, but skip preprocessing/feature extraction, run the 2 steps for resuming, then just start training once you've got everything set up
if you're using RVC Disconnected that is
boreal sage
it stopped itself since i ran out of gpu space lol how do i export model
proper shale
you lost your model
boreal sage
😭
proper shale
😭 im sorry
boreal sage
naw i exported it at like 165 epochs
proper shale
ah
boreal sage
its not lost
proper shale
you can resume then
proper shale
basically, stop training, export model
after this, just do all the steps as bef...
just do the steps i mentioned, same settings n all and you should be fine
boreal sage
i got the .pth is that enough or
proper shale
i got the .pth is that enough or
wait do you have the d/g files in your drive
boreal sage
where would that be
proper shale
inside the rvcDisconnected folder
boreal sage
nope
proper shale
then the folder with your model's name
boreal sage
lol
lunar quartz
ivo got some JSONDecodeError idk what to do can someone help pls
boreal sage
you lost the training files
i'll retrain it locally
do u think it'll take long on a 3700x and 5700 xt
proper shale
ivo got some JSONDecodeError idk what to do can someone help pls
outdated colab, use the newer ones in https://docs.aihub.wtf/
boreal sage
amd unsupported
lunar quartz
thx
boreal sage
is there a way i can train on colab without being gpu stopped
proper shale
is there a way i can train on colab without being gpu stopped
yeah no, you'll have a restriction
normally what i do is just train for only 1 hour, stop and export model
proper shale
HAHHAHA AMD IS AZZ
no it's just that NVIDIA monopolizes AI lmao
brittle wing
JSONDecodeError Traceback (most recent call last)
<ipython-input-20-3fbf21a2f779> in <cell line: 31>()
31 if os.path.exists(config_path):
32 # File exists, proceed with creation of creds and client
---> 33 creds = Credentials.from_service_account_file(config_path, scopes=scope)
34 client = gspread.authorize(creds)
35 else:
5 frames
/usr/lib/python3.10/json/decoder.py in raw_decode(self, s, idx)
353 obj, end = self.scan_once(s, idx)
354 except StopIteration as err:
--> 355 raise JSONDecodeError("Expecting value", s, err.value) from None
356 return obj, end
JSONDecodeError: Expecting value: line 1 column 1 (char 0)
sooo what i can do
mint island
Hey i Need help i would Like to Train my Voice which Tool or Website i can use ?
low shard
---------------------------------------------------------------------------
JSON...
Outdated colab, please use the new ones in https://docs.aihub.wtf/
low shard
Hey i Need help i would Like to Train my Voice which Tool or Website i can use ?
We use RVC v2 here
feral saffron
@proper shale Hey sorry for pinging but I have a question. What is batch size and how do I know what to set it to?
proper shale
<@622856015444049937> Hey sorry for pinging but I have a question. What is batch...
Batch size is basically how many samples it will train at once - normally RVC separates your dataset in 3 second samples.
Try to pick a value according to your dataset size... something like 8 should do well on 10-30 mins, 4 and below should do well on less than 10 mins
feral saffron
Batch size is basically how many samples it will train at once - normally RVC se...
What if my data set is over an hour?
fluid matrix
What if my data set is over an hour?
thats a bit much, if you really want to use that much make sure its actually cleaned properly
that matters way more
feral saffron
The acapellas im using r filtered however they r pretty clean
brittle wing
sorry to interrupt but i tried to start the audio conversion, but the program is not responding suddenly? this is my first time using this
rare gobletBOT
Ayo? @brittle wing level 1 !!!
marble forge
Anybody know how to make something like this?:
https://www.youtube.com/shorts/r9n_8gs3T1U?t=8&feature=share
Also how would i run something like suno locally for free?:
https://suno.com
proper shale
What if my data set is over an hour?
Go for 12 or 16
normal karma
proper shale
sorry to interrupt but i tried to start the audio conversion, but the program is...
check command prompt
brittle wing
check command prompt
it wasnt saying anything abnormal, just sitting there. im gonna reinstall just in case and see what happens! coz i read online that some people just reinstalled and it worked
proper shale
did you even run the dependencies step or... the installing part of everything? that shouldn't happen
normal karma
did you even run the dependencies step or... the installing part of everything? ...
Idk I haven't done it In awhile
proper shale
yeah you probably missed one of the initial steps
normal karma
I'm gonna watch it lol
https://youtu.be/hSxTLCR_95Y?si=RD448_PWgUXw5973
Oops wrong link
Yeah idk what to do
rare gobletBOT
Ayo? @normal karma level 8 !!!
normal karma
I'm stuck
proper shale
Anybody know how to make something like this?:
https://www.youtube.com/shorts/r9...
-
Suno or Udio
-
Not sure if there's any local alternatives for Suno or Udio
normal karma
I'm using easyGUI
proper shale
I'm gonna watch it lol
https://youtu.be/hSxTLCR_95Y?si=RD448_PWgUXw5973
This is outdated
Use the newer Colabs in https://docs.aihub.wtf/ - RVC > Cloud section
RVC Disconnected for training, Applio/Ilaria RVC for conversion
feral saffron
Go for 12 or 16
Thank you
normal karma
?
marble forge
1. Suno or Udio
2. Not sure if there's any local alternatives for Suno or Udio
well it sounds like real songs though thats the thing
proper shale
well it sounds like real songs though thats the thing
Suno does that tho
unkempt coral
Any collab link working for AI covers?
proper shale
Any collab link working for AI covers?
Yeah just check https://docs.aihub.wtf/ RVC > Cloud section, either Ilaria RVC or Applio will work
marble forge
Suno does that tho
how?
marble forge
Suno does that tho
also any free alternatives or any way to bypass the paywall? (any workarounds like making new accounts or something)
unkempt coral
Yeah just check https://docs.aihub.wtf/ RVC > Cloud section, either Ilaria RVC o...
Thanks babe
normal karma
Yeah just check https://docs.aihub.wtf/ RVC > Cloud section, either Ilaria RVC o...
Can someone teach me how to make a rvc instead of just showing this
proper shale
also any free alternatives or any way to bypass the paywall? (any workarounds li...
Now that... is complicated
I don't really know much since I don't really use Suno
proper shale
Thanks babe
Yw :)
marble forge
Now that... is complicated
its really pricy so im just looking for a work around
proper shale
Can someone teach me how to make a rvc instead of just showing this
I mean... it shows you how though
normal karma
I mean... it shows you how though <:nails:1159569314848972891>
Ik but like
Can someone show me
How
It dosent tell me how
It just says what I have to do
No rvc links or nothing
Idk what to do
What website do I use
What links do I use?
slim geyser
It has links to everything you need
wait how do you have the model maker role while asking this
what shenanigans have occurred
queen kraken
brittle wing
check command prompt
so its still not responding after redownloading so im not sure whats happening
rare gobletBOT
Ayo? @cold comet level 1 !!!
cold comet
applio cant run cuz gradio is not installed
i ran the install program, so is there something els im suppose to do that i just wasint informed of
no model named gradio
proud elbow
redownload/unzip and make sure nothing is corrupted/missing
cold comet
the what version? wheres that at
proud elbow
the what version? wheres that at
proud elbow
who tf needs to pay for that
daring heath
wtf is rvc v2 disconnected
low shard
wtf is rvc v2 disconnected
rvc google colab to train models
slim hare
has anyone run into this issue before
this comes up only when i attempt to train 32k models
it works normally otherwise
low shard
Is there a disconnect preventer
wdym?
If u get disconnected for gou, you cant
if you get disconnected for disallowed code, you can try a guide
daring heath
this comes up only when i attempt to train 32k models
32khz is not a normal mf audio sample rate
48khz for video its a standard and cd is 44.1khz
just saying
not resolving ur issue
slim hare
32khz is not a normal mf audio sample rate
my dataset currently caps out at like 15-16khz so i cant use anything above 32k or else it increases the risk of artifacting greatly
daring heath
this shit slow
daring heath
my dataset currently caps out at like 15-16khz so i cant use anything above 32k ...
ok
is Google Colab slower than home pc with RTX 3060 Ti?
slim hare
is Google Colab slower than home pc with RTX 3060 Ti?
speed depends on a lot of stuff
whats your cpu
daring heath
Ryzen 5800x
rare gobletBOT
Ayo? @daring heath level 2 !!!
daring heath
16gb ram
i got a good home pc
i tested training on this laptop im using rn and it outputted faster than google colab
like it did the epoch shid and the saving different pth files faster
proud elbow
is Google Colab slower than home pc with RTX 3060 Ti?
it's fine already
slim hare
for me i have a 3060ti as well but its the 8GB version so i cant push my batch size past 8 (not that i would do that anyways, i prefer quality over speed)
mystic rune
Please help me how to solve this?
[Voice Changer] Pipeline is not initialized.
[Voice Changer] Waiting generate pipeline...
low shard
Please help me how to solve this?
[Voice Changer] Pipeline is not initialized.
[...
please for voice changer wokada ask in #🔍│help-w-okada or #1192011222023950368
mystic rune
thx
daring heath
for me i have a 3060ti as well but its the 8GB version so i cant push my batch s...
imma train the model home cuz google colab is shit slow
boreal sage
hey why does it do this
outdated trainer
rare gobletBOT
Ayo? @boreal sage level 3 !!!
brittle wing
outdated trainer
what does that mean?
boreal sage
what does that mean?
choose a newer one
brittle wing
like a newer model
brittle wing
use rvc disconnected
no im trying to use colab
rare gobletBOT
Ayo? @brittle wing level 2 !!!
boreal sage
no im trying to use colab
yeah rvc disconnected uses colab
on step 2 you will see blue text that leads u to the colab
brittle wing
yeah rvc disconnected uses colab
yh i figured
What would be great for Ai Training? RX 7900XTX 24gb or RTX 3090 ti 24gb or RTX Titan 24gb?
compact sinew
3090ti
wet harbor
yo guys im new, what do i do w the zip files i download in voice models to have them show up in rvc
rare gobletBOT
Ayo? @wet harbor level 1 !!!
cedar root
help
boreal sage
What would be great for Ai Training? RX 7900XTX 24gb or RTX 3090 ti 24gb or RTX ...
u cant train on amd gpus
proper shale
yo guys im new, what do i do w the zip files i download in voice models to have ...
unzip, load the .pth into the weights or models folder (depending on what you're using), put the index on logs
proud elbow
What would be great for Ai Training? RX 7900XTX 24gb or RTX 3090 ti 24gb or RTX ...
you need ROCm on linux for training with AMD GPU
gusty houndBOT
you know who else needs rocm on linux for training with AMD GPU? MY MOM!
low shard
https://github.com/facefusion/facefusion
i mean u can also send https://discord.com/channels/1159260121998827560/1163571683848900629 that has the colabs & kaggles as the official repo doesnt have them
daring heath
im currently making a ai model how do i continue training it in the future is there anything on this page I need to click btw its not google colab
gusty houndBOT
you know who else is currently making a ai model how do you cotiniue training it in the future is there anything on this page you need to click btw its not google colab? MY MOM!
rare gobletBOT
Ayo? @daring heath level 3 !!!
gusty houndBOT
you know who else * 3 * * ! ! ! < : lfg:1159355870119993496 >? MY MOM!
daring heath
HOW DO I STOP TRAINING, HOW DO I STOP EPOCH ITS AT 22 NOW I WANNA STOP TRAINING
EDIT ITS AT 23*
I WANNA CONTIUNE ON THE TRAINING SOME TIME LATER
compact sinew
just close the cmd
gusty houndBOT
you know who else donts wan na restart? MY MOM!
daring heath
and it will autodetect epoch 24?
compact sinew
and it will autodetect epoch 24?
depends on whats ur last saved epoch is
gusty houndBOT
you know who else hit the x? MY MOM!
daring heath
it took some time to close nice
gusty houndBOT
you know who else took some time to close nice? MY MOM!
brittle wing
anyone know why this could go wrong, i gave up already
gusty houndBOT
you know who else knows why this could go wrong , you gave up already? MY MOM!
queen flame
best way to filter background noises when recording?
proper shale
best way to filter background noises when recording?
Noise Gate in Audacity could probably handle the background noises, if not then use BSRoformer in UVR/MVSep
lunar quartz
If i have fake amd gpu (radeon rx 580) from aliexpress can i train with it?💀
proper shale
If i have fake amd gpu (radeon rx 580) from aliexpress can i train with it?💀
don't think so
proper shale
Noise Gate in Audacity could probably handle the background noises, if not then ...
Also other ones you could use:
Demucs DNR and Bandit Plus in MVSep
Demucs DNR could help with some forms of noise too
daring heath
using the 5.1 channels
daring heath
I haven’t ran it trough ai isolator yet
I did mid stereo at the stereo track at 5.1
So here’s where the vocal is
Most clean
proper shale
We've already gotten deleted once yknow
Don't want that to happen again
Also for our users too
ionic oak
Now you just read the Mainline guide, specifically the training part:
https://d...
But suddenly I only have an mp3/mp4 of a 5 minute vocal, what do I do?
proper shale
But suddenly I only have an mp3/mp4 of a 5 minute vocal, what do I do?
just insert the path to the folder it's at
and use mp3 instead, if you can
elder jolt
just insert the path to the folder it's at
You can help me too yk ;-;
ionic oak
just insert the path to the folder it's at
I'm lost on the RVC WebUI even with the site, can you help me? (I am French so the site appears to be in French)
proper shale
I'm lost on the RVC WebUI even with the site, can you help me? (I am French so t...
Just tell me what part you're confused with
ionic oak
Just tell me what part you're confused with
Literally everyone I don't understand I followed the steps we'll say but getting here and the rest I understand that dal and I don't know if I did the right thing
rare gobletBOT
Ayo? @ionic oak level 3 !!!
proper shale
Literally everyone I don't understand I followed the steps we'll say but getting...
Batch size is how many samples of your dataset it'll train at once - each sample is like 3 seconds of your dataset.
8 should be fine for most sets, 4 for smaller ones
ionic oak
Batch size is how many samples of your dataset it'll train at once - each sample...
hmmm ok then what should I do for this? I don't know any TB?
rich shell
hmmm ok then what should I do for this? I don't know any TB?
hey
in my voice changer the beatrice is working but i cant hear my rvc
please help me
proper shale
hmmm ok then what should I do for this? I don't know any TB?
ya need to download TensorBoard, check https://docs.aihub.wtf/rvc/resources/epochs--tensorboard/#tensorboard
proper shale
in my voice changer the beatrice is working but i cant hear my rvc
check your command prompt
rare gobletBOT
Ayo? @rich shell level 1 !!!
rich shell
nothing
proper shale
send screenshots too, that'll help
proper shale
alr, now just open it
rich shell
ionic oak
alr, now just open it
how ?
proper shale
input should be your mic
proper shale
send a screenshot, of what it's saying
rich shell
@proper shale when i click passthrough its working but my voice not rvc
can you come vc
proper shale
yes, I'm aware
proper shale
can you come vc
can't
I asked you to check command prompt to see if something went wrong with the initial install and it screwed up RVC stuff
rich shell
proper shale
alrighty, seems like we found the issue
delete the pretrain and model_dir folders and restart W-Okada
should do the initial install again and should work
rich shell
cant understand
proper shale
delete the two folders (pretrain and model_dir) and restart the program?
rich shell
rare gobletBOT
Ayo? @rich shell level 2 !!!
proper shale
close off cmd first tho
rich shell
done
ITS
OFF
BUT CANT UNNINSTALL
@proper shale i deleted whats inside
and
runned
the http one
ionic oak
gave you a link at the end there, now just open that
Is there a video to explain everything?
In fact I would like a voice type that is perfectly the same as the model I don't know if it is possible with "free" software but if there is paid software or the voice modifier is perfect without delay bug I would like it 'use even if it means paying and there on RVC I don't understand how it works at all plus with the voice message that I have I don't know how to do it even with the guide
gilded mica
how do you make this work in discord?
proper shale
train first, it should put all your training data afterward
proper shale
how do you make this work in discord?
need a virtual audio cable
proper shale
yw! :)
proper shale
so either its screwing up on installation or your gpu just cant run RVC
rich shell
.
proper shale
In fact I would like a voice type that is perfectly the same as the model I don'...
Sadly there is no such thing as a no-delay voice changer rn
rare gobletBOT
Ayo? @proper shale level 105 !!!
proper shale
rvc vol is 0 but ping and stuffy i can see
does it give you the same old "pipeline is not generated" thing?
rich shell
yeah
nvm
i downloaded another
@proper shale what to do in ffmpeg path
enviormental?
gilded mica
Ok i set up vb audio cable but i dont know if it works
rare gobletBOT
Ayo? @gilded mica level 1 !!!
gilded mica
Yippe it works
daring heath
CAN UVR OUTPUT 48KHZ??
regal bison
How do i make my voice less raspy? im using the e girl voice and my friends say it sounds abit too raspy. Whats the best setting for it?
crude flame
How do i make my voice less raspy? im using the e girl voice and my friends say ...
Mess with the tune
brittle wing
how can i make it better quality and sound better
crystal gull
Just checking in again, what should someone use for batch size if they are running an rtx 4070 (12gb vram)
I've heard some say this description isn't accurate and some say it is
"It's advisable to align it with the available VRAM of your GPU. A setting of 4 offers improved accuracy but slower processing, while 8 provides faster and standard results."
peak tusk
I've heard some say this description isn't accurate and some say it is
"It's ad...
less batch more accuracy but slower
if you want to put higher values, align it with your gpu vram
faster, but less accurate
crystal gull
Ah I've been training at 12gb which is my max vram
gritty pelican
Who can help my with a extraction from a MP3 file to PTH?
odd shale
Who can help my with a extraction from a MP3 file to PTH?
Don't use MP3s files.
Use wav or flac instead
odd shale
CAN UVR OUTPUT 48KHZ??
Nope, with the current UVR models the cutoff will be always 17k or 20k
odd shale
how can i make it better quality and sound better
Training with clean datasets and using the correct batch size along testing.
abstract kestrel
Applio no ui tutorial please 😭
rare gobletBOT
Ayo? @abstract kestrel level 4 !!!
odd shale
Applio no ui tutorial please 😭
I think we don't have a tutorial for that.
Use the current Applio Colab.
Daselos al man porfis.
scenic citrus
Where do I mix
gritty pelican
ok
upload your dataset to your Google Drive storage
I need to create a folder with name "dataset" in my google drive?
dataset audio can be a .wav file, right?
ok
thank you
i will try
mortal jasper
cosmic tide
i cant open the rvc
rare gobletBOT
Ayo? @cosmic tide level 6 !!!
digital otter
Hello, I'm new in here which videos would y'all recommend me to start training voices?
proper shale
Hello, I'm new in here which videos would y'all recommend me to start training v...
Not a video, but good enough and very informative:
digital otter
Not a video, but good enough and very informative:
https://docs.aihub.wtf/
Okay thank youuu
proper shale
You're welcome :)
proper shale
move it to C: drive, downloads folder
hearty barn
@brittle wing hello my g
cosmic tide
move it to C: drive, downloads folder <:nails:1159569314848972891>
I fixed it, i forgot to tell you
brittle wing
you need ROCm on linux for training with AMD GPU
I’ll look into that, is there a guide you recommend?
haughty shadow
brittle wing
https://rentry.co/rvcrealtimeguide
Sorry for the ping but I did a bit of testing with my GTX 1660 super and it works perfectly with (0.40-0.45 SL) with 5.00 Extra
violet heron
i cant open the rvc
Which file are you opening?
hot lagoon
synthv Help me fix this FL Studio (21.1.0.3713) [64Bit] r37347 has encountered an error and needs to restart.
Crash log saved to C:\Users\user\Documents\Image-Line\FL Studio\Support\Logs\Crash\ folder.
Please use Ctrl+C to copy this message and report to support.
OS Version:
Windows 10 (Version 22H2, OS Build 19045.4412, 64-bit Edition)
Exception:
Access violation at address 00007FF81449E22C in module 'synthv-studio-araplugin-x64.vst3'. Read of address 0000000000000138
Callstack:
00007FF8140B0000:00007FF81449E22C: synthv-studio-araplugin-x64.vst3
00007FF8140B0000:00007FF814189620: synthv-studio-araplugin-x64.vst3
00007FF8140B0000:00007FF81418AC98: synthv-studio-araplugin-x64.vst3
00007FF8140B0000:00007FF81418A808: synthv-studio-araplugin-x64.vst3
00007FF8BCCD0000:00007FF8BCCDEF75: USER32.dll
00007FF8BCCD0000:00007FF8BCCDE69D: USER32.dll
0000000073A70000:0000000073FB2EFA: FLEngine_x64.dll
0000000073A70000:0000000073FB2F68: FLEngine_x64.dll
0000000073A70000:0000000073FB343A: FLEngine_x64.dll
0000000073A70000:000000007474B56D: FLEngine_x64.dll
00007FF684D80000:00007FF68505A2DD: FL64.exe
00007FF684D80000:00007FF68505A4FD: FL64.exe
00007FF684D80000:00007FF684DB0976: FL64.exe
00007FF8BC7F0000:00007FF8BC807344: KERNEL32.DLL
00007FF8BDEF0000:00007FF8BDF426B1: ntdll.dll
violet heron
my model doesnt have an index, how do i fix that
If you made the model, you have to train a index, if not then 
Some models have the accent baked in so there’s no need for a index
amber orbit
Some models have the accent baked in so there’s no need for a index
ahhh
dusty mural
what do people use to make ai covers now?
i need help making ai covers cause idk what link to use anymore
can anyone give me a link?
hollow hull
why cloning is so complex
can anyone help me with cloning I really try all things and it doesn't work
azure marshBOT
Suggestion for @dusty mural
hollow hull
thanks broooo
dusty mural
thank uu
white bone
How do I prevent my RVC from crashing every time I try to boot it up after I load up a voice model?
ok nvm now it isn't opening at all
EDIT: NVM I just had to wait
young finch
so my rvc comes out as static. i was wandering how to fix this. can someone help me please?
rich shell
how to create rvc own voice model
proud elbow
how to create rvc own voice model
mossy sable
@proud elbowBut this is just how to make model generally
How to make model fit my voice? I can find 100 different gawr gura models but rarely one that fits my voice specifically even tho it perfectly fit some other person from a video I watched
can I train it to fit my voice?
low shard
How to make model fit my voice? I can find 100 different gawr gura models but ra...
Wdym fit ur voice
young finch
Go advanced settings and put the trancate up to 300
It did not work.
halcyon cliff
using the RVCdisconnected to make a model and upon trying to train it said "FileNotFoundError: [Errno 2] No such file or directory: '/content/drive/MyDrive/rvcDisconnected/SmallV1/%s/SmallV1/filelist.txt' " and completed the cell. i followed the guide. whats going on?
its been a while since ive been on here and im not rly used to all the updated colabs and sections so sorry if i broke or missed something. i tried to use easygui but that had no guide, im kinda used to doing work on making a model anyway
rare gobletBOT
Ayo? @halcyon cliff level 1 !!!
halcyon cliff
nvm found it in the forums. weird bec i used original instead of TITAN like that other user was using and kept my model in alpharthmic as well (there sonly one number in it) ill re connect and try again
giving the same error
brittle wing
so its still not responding after redownloading so im not sure whats happening
any ideas?
opal kelp
-colab
slate snow
does anyone know how to fix this? i using Hina_Modified_Realtime_Voice_Changer_on_Colab.ipynb but the model voice didnt show up and i cant event upload
wispy lodge
does anyone know how to fix this? i using Hina_Modified_Realtime_Voice_Changer_o...
What do you mean you cant upload? You should click edit button and there will be menu with model slots where you can upload
slate snow
yes. i've tried that but nothing happend
i saw someone tutorial and after he create the server, he got 4 voice models
and i dont get it even though i follow the instructions
wispy lodge
Check colab notebook for errors. The cell where you run the server shows server logs, it may show the reason why your model does not appear and you can post it here
wispy lodge
i saw someone tutorial and after he create the server, he got 4 voice models
This colab version does not download default 4 models on startup and this is expected
slate snow
i dont understand about this part "notebook for errors"
orchid pine
i need help 
i cant hear myself when im using a voice and i cant find out why
cosmic tide
Which file are you opening?
I fixed it
wispy lodge
i dont understand about this part "notebook for errors"
When you run a cell in google colab, it shows server logs under it. It will be something similar to this
Even though this is Kaggle, the same should be on Google Colab
slate snow
ohh okay, i will see it
wispy lodge
If there's a problem with your model, it will show log entries that have "Exception" or "Error" or something similar
slate snow
oh man
rare gobletBOT
Ayo? @slate snow level 2 !!!
slate snow
i dont know what happend, but thank you very much!
one last thing, am i need to restart the server when its 2 hours?
wispy lodge
That's the amount of time you will be able to use for the current day, so restart won't help. Google Colab provides around 4 hours per day for free for instances with GPU
I think it will shut down automatically when the time is over
wispy lodge
Yes, you either buy compute units or the other way is to create multiple accounts
slate snow
That's the amount of time you will be able to use for the current day, so restar...
that means i cannot using the app when the duration is over right?
wispy lodge
that means i cannot using the app when the duration is over right?
Basically, yeah
slate snow
Yes, you either buy compute units or the other way is to create multiple account...
i see! Thank you Kage! I appreciate it
proper shale
I fixed it, i forgot to tell you
Nice! What fixed it though?
stiff goblet
Hello. Does anyone here use the X-minus pro site? I need to use BVE V2 model for a 4 minute song. I would be very grateful if you can help
cosmic tide
Nice! What fixed it though?
I put the rvc and python on c: drive
proper shale
I put the rvc and python on c: drive
ah makes sense
proper shale
https://erdem.is-a-virg.in/6C6Hc7lVA.png
outdated colab, use newer ones instead:
Ilaria RVC - https://docs.aihub.wtf/rvc/cloud/ilaria-rvc/
Applio - https://docs.aihub.wtf/rvc/cloud/applio-colab/
see which one you prefer
unreal oak
oh thank you
sharp hazel
guys dose ContentVec better than HuBERT?
proper shale
guys dose ContentVec better than HuBERT?
Supposedly yes, but test out and see.
ContentVEC and Hubert help on getting the pronunciation of stuff correctly, btw.
glad zealot
outdated colab, use newer ones instead:
Ilaria RVC - https://docs.aihub.wtf/rvc...
no mainline sadge
warm peak
i downloaded a rvc v2 voice but its .pth , should i convert this to the wav?
odd shale
Well, results also will depend on how clean is your dataset.
odd shale
i downloaded a rvc v2 voice but its .pth , should i convert this to the wav?
No, if you wanna use a model use the Applio Colab and paste the Huggingface link to the model on the Download tab.
It's kinda hard for me to explain that. But in fact 32k is more stable, even more if your dataset is clean and consistent enough and reaches at least 16khz
warm peak
No, if you wanna use a model use the Applio Colab and paste the Huggingface link...
this is my first time actually i only know RVC v2 Disconnected colab site and this site wants me to zip the wav voice into drive so what im gonna do with this .pth file
odd shale
this is my first time actually i only know RVC v2 Disconnected colab site and th...
You know you need also the .index file too, right?
warm peak
You know you need also the .index file too, right?
yea i have that file too but i dont know how to use two of them lol
odd shale
yea i have that file too but i dont know how to use two of them lol
put them in a folder
then name that folder to your liking (a legible name)
and then zip up the folder
warm peak
and then zip up the folder
done
rare gobletBOT
Ayo? @warm peak level 1 !!!
odd shale
done
then upload the zip file to your huggingface account (you'll need obviously to create a huggingface account)
warm peak
on it
odd shale
on it
Scroll down to the "Uploading to Huggingface" part
after you succesfully uploaded your model to HF, use the Applio colab.
There you got a guide too.
warm peak
sorry for my dumb brain, where should i upload
odd shale
sorry for my dumb brain, where should i upload
I've told you to scroll down to the Uploading to Huggingface part on this doc.
warm peak
oh okay
warm peak
sooooooo whats next
warm peak
yes sir
odd shale
yes sir
Then go to the Download tab
You'll need the huggingface link to your model zip.
For this, go to your model's repository you created
and right click on your model's zip file and click copy link
warm peak
bruh i waited so long i think its says connection error
should i restart applio?
No interface is running right now
odd shale
should i restart applio?
Try clicking again on the Public url
warm peak
Try clicking again on the Public url
No interface is running right now
odd shale
No interface is running right now
Restart Applio then.
rare gobletBOT
Ayo? @warm peak level 2 !!!
odd shale
this app hates me
Including the fact you put the wrong link.
warm peak
Including the fact you put the wrong link.
software community hates me
odd shale
software community hates me
?
warm peak
?
just a joke
i right clicked and copied link let me check it
got it waiting for procces
odd shale
got it waiting for procces
I'm not sure what you're doing wrong.
But let me teach you myself how to upload models to HF
warm peak
i think im doing well rn
odd shale
i think im doing well rn
Change "blob" for "resolve" on the link you just put on the Download model tab
warm peak
Model downloaded successfully
odd shale
Model downloaded successfully
Okie, then go to the Inference tab and click "refresh"
odd shale
can i use this voice model for text to speech?
I think you can.
You gotta click on the TTS tab ig.
warm peak
makes sense
odd shale
yeaaaa
But for RVC inferencing, drag the audio file you wanna use to where it tells you.
Wait till the audio finishes downloading
and then just click on convert
that's all
warm peak
text is gonna be turkish so i choosed a turkish tts voice named tr ahmet neural. is this okay?
odd shale
text is gonna be turkish so i choosed a turkish tts voice named tr ahmet neural....
It should be okay.
I can't help you further from this point since i haven't ever used Applio TTS
sharp hazel
Is the least noisy and most stable one.
could you explain why?
warm peak
I can't help you further from this point since i haven't ever used Applio TTS
you helped a lot bro thank you very much
odd shale
you helped a lot bro thank you very much
Ur welcome.
zinc cloud
it keeps saying link invalid when i input huggingface link?
is there a max file size or length of audio clips u put in???
it works now tho
finite galleon
brittle wing
what's difficult one click trainning & train model
rare gobletBOT
Ayo? @brittle wing level 1 !!!
brittle wing
Then, can I press the training model to learn at the beginning and press the training model to learn additionally?
Yes, I have to press the train model button when I finish 300 times at first and do an additional 600 times, right?
which file do i open to run the program
in the folder
brittle wing
in the folder
You can open the go-web.bat file for RVC Webui
oh i was opening rvg-gui bat it just opened the ui as u pinged me
did i open the wrong thing
it was taking really long on the cmd window i thought it was broken
is there a way to send files
rare gobletBOT
Ayo? @brittle wing level 1 !!!
brittle wing
oh thats how
cinder oar
which version of okada should i get for rtx 3080?
brittle wing
if i record the vocals myself instead of using an ai to remove the vocals will it sound better
stuck anchor
Hi everyone ! i need help for rvc i cant start it they said moduleNotFound error: no module named _distutils_hack and pywin32_bootstrap and tkinter 
obtuse rapids
how to make ai voice model? 
keen tartan
can someone help me, on the pitch detection algorithm its written in chinese but everything else is english and i cant select "start audio conversation" because it says it cant be written in chinese
proper shale
if i record the vocals myself instead of using an ai to remove the vocals will i...
probably
had to delete cuz... copyright
brittle wing
its not even a copyrighted song its a song my friend made lol
proper shale
can someone help me, on the pitch detection algorithm its written in chinese but...
what are you using?
keen tartan
what are you using?
RVC1006
proper shale
its not even a copyrighted song its a song my friend made lol
didnt wanna risk anything tho sorry
brittle wing
na i get it its fine
keen tartan
sure
proper shale
how to make ai voice model? <:boohooh:1176674698629750975>
https://docs.aihub.wtf/ has everything you need to know
keen tartan
rare gobletBOT
Ayo? @keen tartan level 2 !!!
keen tartan
some of this is in chinese
im turkish so it does translate some
but like its still chinese so when i click start it just crashes saying "phyton cant run chinese language" or whatever
keen tartan
what should i use?
low shard
check our docs instead https://docs.aihub.wtf
proper shale
you know the solution i assume
low shard
THAT IS NOT RVC GUI
alr im js not gonna talk abt local shit sorry😭
proper shale
yeah that's the thing
proper shale
alr im js not gonna talk abt local shit sorry😭
NOOO KEEP TALKIN
proper shale
yeah that's the thing
did you get a turkish version of RVC or it just automatically translated?
low shard
NOOO KEEP TALKIN
i mean as i cant use local so idk much of it, i js wanted to help 😭
ngl first time i see someone use rvc realtime, usualyl everyone uses wokada lol
proper shale
might've found the problem.. use the same type of audio inputs
one is using mme and the other is using windows directsound, use mme for all
keen tartan
oh alright
proper shale
ngl first time i see someone use rvc realtime, usualyl everyone uses wokada lol
RVC realtime is pretty decent, only lacks decent noise suppression
keen tartan
THANK YOU MJ
YOU ARE MY SAVIOR
MY SUNSHINE
I LOVE YOU
oh wait
sooo
when i use dio sound model
it says "cant use chinese characters"
keen tartan
alright
elfin bay
How long should my dataset be for a 250 epoch voice model
thin stump
How long should my dataset be for a 250 epoch voice model
There's no set amount. You're going to be observing kl, d/total, mel on the tensorboard. G/total is the average of that https://docs.aihub.wtf/rvc/resources/epochs--tensorboard/#how-many-epochs-should-i-use-for-my-dataset
hmm think of it as a game with a horrible UI
calm moat
Hello everyone, I don't really understand neural networks, I downloaded a model of the Russian singer Lena Katina, added it to the rvc GUI
Then I added the a cappella of the song and clicked convert
At the moments when backing vocals sound, the neural network seems to suffocate and begins to wheeze. what should I do?
I am from Russia, I write through a translator
proper shale
mj have you figured it out
give me the error message real quick
keen tartan
give me the error message real quick
sure
.pth file directory doesnt use chinese characters
proper shale
Hello everyone, I don't really understand neural networks, I downloaded a model ...
You do need to isolate the lead vocals - use https://mvsep.com for that - either MDX B karaoke model or Ultimate Vocal Remover HQ > BVE
keen tartan
My directory is like this
C:/Users/Tarık/Desktop/AIVOICECHANGER/RVC1006Nvidia/assets/weights/dio.pth
this is my directory
proper shale
.pth file directory doesnt use chinese characters
this what it says on cmd? send a screenshot
keen tartan
this what it says on cmd? send a screenshot <:nails:1159569314848972891>
its in turkish
calm moat
You do need to isolate the lead vocals - use https://mvsep.com for that - either...
thank you!
keen tartan
oh u mean cmd
rare gobletBOT
Ayo? @calm moat level 1 !!!
keen tartan
its
proper shale
thank you!
You're welcome :)
keen tartan
nothing on the cmd
proper shale
nothing on the cmd
send a screenshot regardless
keen tartan
rare gobletBOT
Ayo? @keen tartan level 3 !!!
keen tartan
Oh my god.
I FIGURED IT OUT
mj
C:\Users\Tarık\Desktop\AIVOICECHANGER\RVC1006Nvidia
since ther is
small letter ı
which is only in turkish
it crashed
but then i did like
calm moat
uvr🥹
keen tartan
but then i did
assets/weights/dio.pth
only
and it worked
so if anyone else comes up with this issue
u know the fix
normal ruin
Guys, it's a know issue that RVC is tagged as a trojan ?
calm moat
uvr🥹
I think it's because of the Cyrillic alphabet in the file path, I'll try to reinstall
proper shale
Guys, it's a know issue that RVC is tagged as a trojan ?
that's... probably a false positive
where did you get rvc from
proper shale
yeah dw, it's safe
calm moat
What can I do?
keen tartan
do you guys know how i can stop the real time voice thing playing back to me whenever i speak?
proper shale
What can I do?
atp i'd probably use mvsep
as i linked earlier
wait
wait
WAIT
move that out of program files if it's there
go to a downloads folder instead
keen tartan
mj
can u really quickly
tell me what i need to do for it to stop playing back my voice to me
if you know ofc
proper shale
tell me what i need to do for it to stop playing back my voice to me
there's like
2 chinese options
right?
calm moat
transfer from program files to downloads?
proper shale
in the bottom
proper shale
transfer from program files to downloads?
yea
keen tartan
its these i think
calm moat
ok
proper shale
disable the first one
keen tartan
i cant
proper shale
basically
keen tartan
like i said
calm moat
proper shale
output, change that to the virtual audio cable
keen tartan
oh
proper shale
well shit
calm moat
well shit <:nails:1159569314848972891>
Cyrillic alphabet?
proper shale
Cyrillic alphabet?
idk
calm moat
🥳 🥳 🥳
rare gobletBOT
Ayo? @calm moat level 2 !!!
proper shale
ah nice
proper shale
what am i supposed to do
being honest w/ you, im... not sure 😭
what's your settings rn tho, send a screenshot
proper shale
can you select only one of the bottom ones or can you unselect both?
keen tartan
I can select only one
and one displays my real voice
while the other one is the ai one
and i cant unselect
i have to select between the two
proper shale
yeah i think you're gonna have to listen to your own ai voice
proper shale
@pastel oak correct me if im wrong
keen tartan
but then no one would hear me
proper shale
cant i just mute the output
you could probably... mute the window?
keen tartan
OH!
good idea
i give up
it did work but it was
coming off ragged
like
e-m w-t sigma
instead of
erm what the sigma
obtuse rapids
https://docs.aihub.wtf/ has everything you need to know
how do i get RVC
proper shale
e-m w-t sigma
increase sample length, decrease extra to like 0.9 or something like that
proper shale
how do i get RVC
just check the RVC > Local section
obtuse rapids
just check the RVC > Local section
ok
proper shale
which pretrain to use for a voice that's robotic
just go for original, if it doesn't do well maybe try titan?
boreal sage
just go for original, if it doesn't do well maybe try titan?
this is original
and its robotic
how do i fix it
proper shale
this is like... poorly trained, or it screwed up on feature extraction
how many epochs & steps?
boreal sage
450/25650
keen tartan
no no i meant
its not the smaple lenght
it cuts off mid sentence into nothingness
then goes back
well its okay
thank you for trying to help
proper shale
its not the smaple lenght
hm
obtuse rapids
just check the RVC > Local section
is this it?
boreal sage
just a video of him talking
proper shale
yeah it's the echo
boreal sage
is this it?
boreal sage
yeah I'll get something else
proper shale
https://docs.aihub.wtf/rvc/local/mainline/
^
brittle wing
any ideas?
hi im still having trouble, my rvc isnt responding when i try to activate any voices. id really love some help simply because i dont want to share my irl voice with anyone
obtuse rapids
https://docs.aihub.wtf/rvc/local/mainline/
ok
proper shale
hi im still having trouble, my rvc isnt responding when i try to activate any vo...
give a command prompt screenshot
brittle wing
give a command prompt screenshot <:nails:1159569314848972891>
like i mentioned, it doesnt seem like anythings happening
the top line says "is half true" i didnt mean to cover it that much
and the whole thing doesnt respond even if i dont use the virtual cable
proper shale
like i mentioned, it doesnt seem like anythings happening
MME/DirectSound duo might not work, use CABLE Input with MME instead
That might be it
brittle wing
MME/DirectSound duo might not work, use CABLE Input with MME instead
oh huh coz im tryna get the output to be the cable since i was following a tutorial that said it would need to be the output if i wanna use it via discord n such
proper shale
oh huh coz im tryna get the output to be the cable since i was following a tutor...
yeah i'm aware, it's just that there's these types of audio "things" (honestly dont know what to call them) and they don't go well together
see if there's MME cable input tho
brittle wing
see if there's MME cable input tho
ohh okok i did find a way to make the output also mme and it seems to have worked now!
rare gobletBOT
Ayo? @brittle wing level 2 !!!
proper shale
ohh okok i did find a way to make the output also mme and it seems to have worke...
Niiiice
Thought it was another Okada issue lel
brittle wing
haha thanks for the help! 
is there any way to set up my sound so I can also hear it or is it just not a thing LOL 💀
proper shale
haha thanks for the help! <:SkyDeepBow:649389650519261205> is there any way to ...
the second option in the bottom, "output converted voice", should do that
brittle wing
the second option in the bottom, "output converted voice", should do that
hmmm i have that turned on 🤔
proper shale
#✨│ai-help message maybe use this thing?
brittle wing
https://discord.com/channels/1159260121998827560/1159290139609137264/12380935617...
that looks complicated. ill look into it!!
tysm!
slim geyser
<:blobsweats:585893206441590831> that looks complicated. ill look into it!!
Or you could just go to your audio devices in your control panel, and select listen to the output on your virtual audio cable.
obtuse rapids
what does this mean?
