I used them for recording.
#✨│ai-help
edgy minnow
So.
That kinda broke everything.
:/
In OBS i had to time everything properly, nw its all over the fucking place.
great
:D
hardy yew
it's constant anyway
edgy minnow
Its not
the ms changed alot
Why does my use need get dismissed lol
I can't use vonovox then. I can't even listen to it back lol.
hardy yew
(BTW I'm pretty sure those things that are no longer accessible as sliders in GUI can still be adjusted via config)
edgy minnow
(BTW I'm pretty sure those things that are no longer accessible as sliders in GU...
well i didnt see anything like that.
hardy yew
I can't use vonovox then. I can't even listen to it back lol.
there's other ways around that, monitoring is another feature that was not implemented cause it's basically redundant and bloats the app unnecessarily
edgy minnow
These are settings I needed, so I'm stuck on one of the other ones for now
hardy yew
well i didnt see anything like that.
i mean, the config file, probably configs/config.json
hardy yew
These are settings I needed, so I'm stuck on one of the other ones for now
yeah, nothing wrong with preferring one app over the other
edgy minnow
Less prefer, more I want the GUI
hardy yew
i just don't think it's "unusable for your case", more of taking a different approach compared to the other app
and was clarifying on that
But obviously, go with what suits you better
edgy minnow
The delays I put on made things clearer and sound better :c
instant doens't sound good.
weel 90ms isnt instant but lul
hardy yew
the GPU delay is just a tiny part of the overall delay, most of the lag comes from audio pipeline anyway
edgy minnow
Checked the config. I don't see it in there.
Trying to use this one but I hear random echos for just an instant. Its going through the output too.
hardy yew
Checked the config. I don't see it in there.
block_time -> block size
lookahead_buffer -> extra time
edgy minnow
block_time -> block size
lookahead_buffer -> extra time
Ohh.
edgy minnow
block_time -> block size
lookahead_buffer -> extra time
Uhhhh... hm. Okay but.. how do I set them manually then? I don't think I know the values to use.
hardy yew
i thought that you do know what you want to set them to, considering you're sure you need to change them?
edgy minnow
I had it set to Extra 5 and the other 128ms
hardy yew
If you want to play around you can use Vonovox 1.6.9, it still has the sliders
edgy minnow
How does this translate over?
hardy yew
but they were removed for a reason, really
hardy yew
I had it set to Extra 5 and the other 128ms
not 100% sure, but i think both settings in vonovox are in seconds, so that would be block_time of 0.128 and lookahead_buffer of 5.0
but to be fair
extra time of that much is completely pointless, i think
and such small block size will massively degrade pitch/feature extraction
that's why the defaults are defaults
edgy minnow
hardy yew
not 100% sure, but i think both settings in vonovox are in seconds, so that woul...
and as i said, i'm not 100% sure, so this can be further explained e.g. in Vonovox discord if in doubt
hardy yew
block time of 1.0 is a lot and will noticeably increase the delay
and also GPU usage i believe?
hardy yew
you sure? pretty sure 0.3 is the default, unless it got switched in recent release
edgy minnow
i mean i hear static anyway
:/
just random tears
so they both suck. joy.
one has rando static one has rando echo
okada just worked. wish it supported the 5090.
i guess i wasted my money upgrading my pc.
yeah no this sucks. wow.
elder coral
after 3 years i'm finally making the model i wanted with clean dataset
edgy minnow
@hardy yewcan you explain this unholy amount of static and tearing with vonovox lol
hallow thistle
Trying to use this one but I hear random echos for just an instant. Its going th...
Are you making your own W-Okada fork? I don't remember Tg Develop's W-Okada voice changer being as a separate app rather than a web browser.
edgy minnow
Are you making your own W-Okada fork? I don't remember Tg Develop's W-Okada voic...
no.
hardy yew
Are you making your own W-Okada fork? I don't remember Tg Develop's W-Okada voic...
oh, good catch xD i was wondering what you're talking about until i looked at the screenshot
edgy minnow
I use an old okada thingy
hallow thistle
edgy minnow
its the web window thing.
I can't have it in my browser :/
thats stupid and in the way
hallow thistle
You can call me stupid, but b2397 W-Okada typically looks like this. https://cdn.discordapp.com/attachments/1159290139609137264/1467527432457424969/image.png?ex=69ab8dad&is=69aa3c2d&hm=dcc93072228402da9317ca5804c0ff3bc1b476e623c1b59fa29de57a9bb64274&
hardy yew
i guess it's a fork of that fork
edgy minnow
static galore atm on vonovox so idk
hardy yew
there's a ton of them i guess xD
edgy minnow
this is why i dont wanna use vonovox. its not wokring right.
i have nothing else running
this is crazy'
hardy yew
<@214134626065580033>can you explain this unholy amount of static and tearing w...
i don't think static is any common issue, are you sure that's what it actually is?
And tearing is usually a result of too short block
edgy minnow
tearing
default
it son beta cant change it
remember?
you said it was default so i set it back
I could set okada normal to 128ms and it sounded insane on a 1080.
How exactly is vonovox better?
?_?
hallow thistle
Dedicating to the voice changer is a horrifying work. I don't always use the voice changer daily like some other people here. What made me got into these things was because of people here asked for W-Okada. 
hallow thistle
What do you mean Tg Develop's W-Okada doesn't support RTX 50 series?
hardy yew
How exactly is vonovox better?
volume handling, pitch extraction, chunk stitching... and a ton of other stuff which i don';t remember.
It's also pretty much the only RVC app build from scratch for this purpose, the rest are forks of old w-okada
hallow thistle
No, answer directly.
edgy minnow
What do you mean Tg Develop's W-Okada doesn't support RTX 50 series?
Normal Okada.. er dietrich or something
hallow thistle
Like, what you mean RTX 50 GPU won't gonna work with W-Okada fork?
hallow thistle
What you mean RTX 50 GPU won't gonna work with W-Okada fork?
hallow thistle
edgy minnow
its the one i was using before :/
and tried again
wtf do you want me to say to you
I keep trying to answer
vital bone
Anyone knows the best configuration for a male voice model?
hallow thistle
There are two different W-Okada forks, damn it.
vital bone
Anyone knows the best configuration for a male voice model?
Bumping this cause a friend been struggling how to properly tune one
edgy minnow
I even tried to name it earlier and spelled it wrong
hallow thistle
hallow thistle
This is getting stressful, so let's make things simple.
hallow thistle
Well, just forget about it and move on to another one. Now what about Tg Develop's W-Okada fork?
No, I'm not trying to be angry, and no troll or joke on this one. I'm genuine asking.
edgy minnow
I was muted.
edgy minnow
Well, just forget about it and move on to another one. Now what about Tg Develop...
It was echoing. I'm figuring that out now
I need control over what I'm using. I fine tune all my delays for OBS and stuff.
Vonovox will never work because of that. But anyway, launching a game to see if that was why.
5090 VR + Okada is waiting.
I mean, I can just throw the spare 3080 or 1080 in here if needed. idc.
hallow thistle
When I have to show some instructions about W-Okada or Vonovox to people, I usually have to open the "dummy" program at my own, even though my 2012 laptop doesn't have any GPU showing up in Task Manager.
hardy yew
(FTR VR might add new issues, even with low GPU usage the way GPU is forced to deliver frames might cause issues in RVC)
edgy minnow
The heat in my room will be hell.
hallow thistle
While some other people often rely on text to convey about what's happening in the voice changer, I do this by sending some screenshots of mine, because I'm aware words alone cannot always imagine as a picture.
hardy yew
dual GPU is definitely favorable if those issues occur
edgy minnow
(FTR VR might add new issues, even with low GPU usage the way GPU is forced to d...
I used a 1080 for Okada W, and a 3080 for games, ontop of a 5950X + 64Gig of ddr4
Its in the enxt room over still
hallow thistle
This is my most recent screenshot.
hallow thistle
Don't mind about settings in my screenshot, those aren't recommended settings. 
edgy minnow
What card should I get for dual? Preferably something thin.
I have the 3080 but it doesn't feel wise to put that into the pc too.
hallow thistle
Because another type of people here have been looking for a GPU that can handle AI especially with a very large VRAM size, I can think of those "NVIDIA RTX Pro" GPUs.
hardy yew
WDYM "doesn't feel wise"? xD
edgy minnow
WDYM "doesn't feel wise"? xD
ITS HOT AS FUCK IN MY ROOM
LOL
or was with the 3080 1080 combo
and they reallllyyy didnt have rbeathing room at all in that tiny case
stupiiiddd shoulda listened to the microcenter guuyyy
hardy yew
wouldn't be surprised if the 3080 was similar or even better in terms of power efficiency than the 1080
this can also be measured, try HWINFO64 and you'll see what's going on
edgy minnow
I was thinking of buying a 5060 ti or 5070/ti
hardy yew
don't you already have 2 spare GPUs anyway?
hallow thistle
Y'all still see me? 
hardy yew
not sure i'm following
edgy minnow
hallow thistle
Aside from NVIDIA GeForce RTX the consumer/gaming GPUS, I'm also interested in those professional/workstation NVIDIA RTX Pro GPUs to test AI programs.
hardy yew
why do you insist on buyin another gpu when you already have 2 that are more than sufficient?
edgy minnow
8GB would probably be enough since thats what a 1080 has.
edgy minnow
why do you insist on buyin another gpu when you already have 2 that are more tha...
I donm't want to take the 3080 out.
its evga... its my baby
:<
unu protecc
hardy yew
xdd
edgy minnow
and i think the 3080 would be too much heat
hallow thistle
NVIDIA GeForce RTX 5060 is an entry-level GPU. While it has 8GB, it likely suits for some basic AI inferences.
hallow thistle
There's a specific RTX 5060 Ti model that features 16GB but that's about it.
edgy minnow
even a 1060 is..
1000 series was dropped by NShittia
so, youd have to use old drivers. thus why im going for low profile 5060. temps, power, etc.
i forget what the safest driver is atm for the 5000 series. everything has been "vibe coded" ai coded garbage thats been frying cards.
hallow thistle
hardy yew
go ahead then, i'd not hesitate to just use the 3080 if i had it spare like you, but if you're settled on buying a dedicated GPU for rvc, then the basic 5060 is more than enough
you'll be fine
if you want to make it cheaper then even a 3050-3060 will be okay too
hallow thistle
Those RTX xx60 might be fine for basic AIs. But if you want more performance and thermal is less of a concern, I'd go for those GeForce RTX xx80, but remember that my own advice might not fit your needs. 
edgy minnow
im... only doing this
and i have a 5090 if i need it for other AI stuff... and a 3080..
hallow thistle
GeForce RTX 5050 is tiered below than RTX 5060.
hardy yew
Those RTX xx60 might be fine for basic AIs. But if you want more performance and...
performance impact on RVC is negligible... if it works, it's enough
and it will work
elder coral
this baldi model is trained with 9 minutes long and 15 seconds of audio
edgy minnow
performance impact on RVC is negligible... if it works, it's enough
RVC 5050 vs 5060?
i mean both are more powerful than the 1080 from what im seeing
I MENA
I MEAN
I MEAN
I MIGHT
hallow thistle
RVC 5050 vs 5060?
hardy yew
RVC 5050 vs 5060?
as i said, 1060 is enough, so i'd imagine there's absolutely 0 reasons for any of those to not be sufficient
to be clear, I mean the above if the GPU's sole purpose is RVC
if you plan on doing more stuff with it, then it obviously depends
hallow thistle
The difference between RTX 5050 and RTX 5060 is a bit obvious, especially "memory speed", the VRAM type and CUDA cores count.
So, if I have to pick which one is better in this conparison, I'd say RTX 5060, unless the budget is a concern that I'd downgrade for RTX 5050.
hardy yew
So, if I have to pick which one is better in this conparison, I'd say RTX 5060, ...
if it was just for RVC, what would possibly be the reason to spend more than less money?
edgy minnow
i got timed out again uwu
but really i might buy that if i can get the downpay soon enough
hallow thistle
if it was just for RVC, what would possibly be the reason to spend more than les...
Aside from basic AI inferences, I'm also into training something, especially RVC voice models. It's why I'm chosing "RTX xx80/xx90" for my own. 
elder coral
now where is @viral mason ?
hardy yew
Aside from basic AI inferences, I'm also into training something, especially RVC...
but that's your use case, not necessarily theirs
that's why it's important, what they're going to do with it
hallow thistle
How am I supposed to answer?
edgy minnow
tbh, i need help making this model i sue again.
I don't remmeber how to do it all, and the source material is known but kinda unknown. idk which ones specifically were used.
hardy yew
now where is <@1023278814752677918> ?
probably sleeping xD
hallow thistle
I use Applio RVC to train voice models.
elder coral
probably sleeping xD
is this output good?
hallow thistle
I just bought a new pair of earbuds today, so at the moment I'm trying to adjust sounds. 
edgy minnow
overear supremecy
hardy yew
is this output good?
looks and sounds pretty clean to me
elder coral
ye
hallow thistle
While two GPUs of different specs can't be combined together into the same unit in a PC system, you can assign the RVC or voice changer to RTX xx60 while assign RTX xx90 for something like games.
edgy minnow
While two GPUs of different specs can't be combined together into the same unit ...
ya. thats what i did with the 1080 and 3080
hallow thistle
I mean even two RTX 4090 GPUs still can't be combined within NVIDIA SLI anyway. 
hardy yew
is this output good?
where's the sample from BTW? it seems nice
elder coral
where's the sample from BTW? it seems nice
well i already submitted it fr model maker
hardy yew
well i already submitted it fr model maker
i mean the sample source, not the actual model
elder coral
wdym
hardy yew
the audio file that you used as input for the model inference
i meant where is it from
hallow thistle
Does anyone know a single slot card I could use with this?
"GALAX GeForce RTX 4060 Ti Max 16GB" is the only known GPU from RTX 40 series that uses single-slot, from what I've found. I've never seen any GeForce RTX 50/30/20 series GPU that would use single-slot fan cooling solutions, as they mostly use dual-slot and larger. Those "single-slot" RTX xx70/xx80/xx90 might be used within water-cooling solutions instead. Just letting you know, the standard "NVIDIA RTX PRO 4000 Blackwell" is the only known professional GPU that is single-slot most recent as of 2026.
forest topaz
Hi everyone, I’m exploring an idea for an AI-powered email automation system built with n8n, and I’d really appreciate feedback from people who already run production workflows. The concept is an AI email agent that manages an inbox like a digital assistant: automatically reading new emails, filtering promotional messages, labeling and categorizing them, and determining importance based on keywords, sender priority, and predefined rules. The system would also draft AI responses using a document knowledge base (company info, FAQs, instructions) while allowing humans to review or edit drafts before sending. It could integrate with a CRM to pull contact history for better context, notify teams when important emails arrive or when specific keywords/priority senders appear, and provide daily productivity tools like email summaries, scheduled sends, and even newsletter generation from curated content. I’m also considering a dashboard that shows incoming emails with labels, AI-generated drafts (sendable in one click), important notifications, and daily summaries. What I’m trying to validate is whether something like this already exists in the n8n ecosystem, whether there are technical limitations that could make it unreliable, whether it could realistically scale with providers like Gmail or Outlook, and whether businesses would actually find it valuable. If you’ve built email agents, inbox automation, or AI response systems with n8n, I’d love to hear about your lessons learned, reliability challenges, and what tends to break in real-world use. Thanks in advance for any insights.
lime lance
https://play.google.com/store/apps/details?id=com.avionti.PicToPDF looking for feedback will be very thankful to you
viral mason
now where is <@1023278814752677918> ?
Asleep because that was at 6 am for me
viral mason
It's best to keep extra processing at 2.7
quartz raven
is it std or dml for amd systems?
low shard
is it std or dml for amd systems?
this is a General AI Server, elaborate:
- your pc os
- your pc gpu
- what are you doing
- what tutorial are you using
quartz raven
this is a General AI Server, elaborate:
- your pc os
- your pc gpu
- what are yo...
Windows 11 9060XT 16GB I am trying to install and use w okada and I am using this https://www.youtube.com/watch?v=xwdbwbtO-FQ
I have the std one installed the default models work fine but any models that I download sound really bad and choppy so I try the dml one but it doesnt launch
raven maple
i am wanting to create an ai remix of tears by sabrina carpenter featuring ariana grande how do i go about doing this
low shard
Windows 11 9060XT 16GB I am trying to install and use w okada and I am using thi...
all youtube tuts on that topic are outdated asf, i checked that one and it's no exception
uninstall everything you got from this video
check https://docs.aihub.gg/realtime-voice-changer/local/tg-develops-w-okada-fork/
low shard
i am wanting to create an ai remix of tears by sabrina carpenter featuring arian...
you want to make an AI Cover right? What's your PC OS and GPU?
raven maple
you want to make an AI Cover right? What's your PC OS and GPU?
yes i do , i'm using windows my gpu is intel uhd graphics
i have 2 gpus
one is nvidia geforce rtx
the other
is intel
low shard
yes i do , i'm using windows my gpu is intel uhd graphics
I'm guessing Windows 11
raven maple
yes
low shard
one is nvidia geforce rtx
could you tell the full name of that one?
raven maple
nvidia geforce rtx 2050
low shard
nvidia geforce rtx 2050
good enough for AI Covers locally, so you won't be limited by a freemium online web app
check https://docs.aihub.gg/rvc/resources/dataset-isolation/ on how to get the vocals of the song if you don't already have them
and check this RVC (Retrieval-based-Voice-Conversion) Fork (modified Version) called Applio for Voice Swapping Files: https://docs.aihub.gg/rvc/local/applio/
quartz raven
It doesnt allow me to click noise supression
low shard
It doesnt allow me to click noise supression
did you download the program I told you? can you show me a screenshot?
!give-media-perms 3h @quartz raven
low shard
find the download button
quartz raven
did you download the program I told you? can you show me a screenshot?
I figured it out Do you have settings tips your personal setting?
low shard
I figured it out Do you have settings tips your personal setting?
that depends on your hardware so I can't give my personal settings, it would be better you show a screenshot of your current settings so I can check if they are okay
quartz raven
that depends on your hardware so I can't give my personal settings, it would be ...
I'm not sure what to show
low shard
you can higher up the extra to 2.7 for better quality, at the cost of requiring more processing power and might cause more delay
you could lower the delay by using the server audio processing, but you wouldn't be able to use echo or noise suppression
quartz raven
you can higher up the extra to 2.7 for better quality, at the cost of requiring ...
What sample rate should I use
low shard
What sample rate should I use
like the setting when using the server audio processing?
quartz raven
like the setting when using the server audio processing?
Yeah I am trying to figure out now which one is generally better client or server
Do you have a perference or know
low shard
Yeah I am trying to figure out now which one is generally better client or serve...
depends on your needs:
- Client: easy to use, can use echo, and noise suppression
- Server: harder to use, can significally reduce delay if used with Wasapi or Asio
viral mason
depends on your needs:
- Client: easy to use, can use echo, and noise suppressio...
pretty sure client mode stopped working last November
low shard
pretty sure client mode stopped working last November
client mode isn't working on any wokada versions for everyone?
viral mason
client mode isn't working on any wokada versions for everyone?
I've heard from many that it hasn't been working
Same for me, stopped working completely
Don't know if it is for all versions but at least wokada tg it's broken and possibly others
hardy yew
checked it just out of curiosity, works for me (tg develop)
I mean, even after selecting client mode the start button still says "start server" but i imagine it's just a misleading label
viral mason
I mean, even after selecting client mode the start button still says "start serv...
It still works? For me it outputs nothing with the same microphone settings on server mode
Strange
elder coral
this baldi model is trained with 9 minutes long and 15 seconds of audio
here
viral mason
here
The audio is alright but I'd personally use an audio of a real person talking
Instead of a tts for a sample
lapis kiln
what best sitting?
lapis kiln
What is the reason for this? The sound cuts out and then comes back on suddenly.
viral mason
what best sitting?
viral mason
What is the reason for this? The sound cuts out and then comes back on suddenly.
bad settings, try what I have
lapis kiln
Are these the best settings?
lapis kiln
What is the difference between Client and Server?
And what is best
lapis kiln
https://discord.com/channels/1159260121998827560/1159290139609137264/14795401496...
what Wasapi or Asio How do I use them?
hardy yew
just choose WASAPI as the audio driver
lapis kiln
just choose WASAPI as the audio driver
Why doesn't it work?
So, do you recommend using a ( server ) instead of a client? Is it better?
viral mason
So, do you recommend using a ( server ) instead of a client? Is it better?
if client mode does not work for you yes
lapis kiln
if client mode does not work for you yes
It works
viral mason
you can keep it on client mode then
viral mason
Are these the best settings?
they work very well for me so probably will for u too
lapis kiln
they work very well for me so probably will for u too
server yes? not client
viral mason
whichever mode works for u
lapis kiln
whichever mode works for u
what best
elder coral
my graphics card is cooked, i can't make models now
edgy minnow
It's best to keep extra processing at 2.7
Why is that?
chilly furnace
Why is that?
just because
edgy minnow
too many errors with these new voice changer forks
too much audio tearing that wasn't there before
hallow thistle
Why is that?
karmic trellis
Hi, is there any difference between vb cable and vac lite? I have the vb cable installed and i am curious if there is an advantage of switching to the other.
edgy minnow
https://github.com/IllIlIlIllIl/voice-changer/releases/tag/b2335
Deiteris fork supporting RTX 5000 series
elder coral
pls help collab can't connect to gpu
hallow thistle
https://github.com/IllIlIlIllIl/voice-changer/releases/tag/b2335
Deiteris fork ...
That's an older version. This specific version simply has Pytorch version upgraded and that's about it. 
edgy minnow
That's an older version. This specific version simply has Pytorch version upgrad...
Yujp and it works better than the other two
none of the errors, tearing, etc, that i wa shaving before
lol
hallow thistle
Why doesn't it work?
hallow thistle
Yujp and it works better than the other two
Can you benchmark and record the audio between these two W-Okada forks with all same settings? I'm skeptical about your words alone.
edgy minnow
Can you benchmark and record the audio between these two W-Okada forks with all ...
Benchmark?
This fork for 5000 series is working with warfrtame at max settings, 1440P, with no stuttering, 120FPS. Vonovox xouldn't even be played at the same time lol. TG kept having issues.
If you can explain how to benchmark this, I can try since I'm waiting on a 5060 to arrive sunday
I'm not experinced or knowledgable. sorry.
hallow thistle
Benchmark?
What do you think about this one? I open two different W-Okada fork versions.
edgy minnow
I can do that.
I'm playing with a friend atm tho
o3o
I can post this one for now, and I'll edit them later.
hallow thistle
Benchmark refers to a scenario where two or multiple systems being assigned to perform a task identically to achieve results for comparison. For example, PC system with different GPUs, RTX 4060 and RTX 5050, being used to test graphics capabilities and performance in a game where both must have the same settings. In "voice changer" context, it means to test these two W-Okada versions, especially performance, audio quality, something something, while both programs must use the same settings as well (voice model, chunk, extra, input/output audio).
edgy minnow
oops
edgy minnow
Benchmark refers to a scenario where two or multiple systems being assigned to p...
cant turn off noise suppression
neat
Well this is wierd. Now its wokring normally.
spare stump
Hi everyone, I’m currently using w-okada.
I’ve been creating models using Applio. I noticed that models created with the contentvec embedder work perfectly fine in w-okada. However, whenever I try to use models created with the spin embedder, I only get distorted, 'alien-like' noises instead of clear audio.
This led me to wonder: is it possible that w-okada is simply not compatible with the spin embedder? I’d really appreciate some clarification or help on this.
Thank you in advance!
tame oracle
!howtoask
hardy yew
Hi, is there any difference between vb cable and vac lite? I have the vb cable i...
If it works for you, don't bother switching. I've heard that for some people one works fine and the other causes issues, although usually VAC is the go-to solution.
But as I said, if it works for you, just enjoy it, no reason to change anything
hardy yew
Hi everyone, I’m currently using w-okada.
I’ve been creating models using Appli...
Did you select the right embedder in w-okada?
You need to choose it in the realtime app, it won't adjust automatically
Tg develop and vonovox both support spin
spare stump
Did you select the right embedder in w-okada?
You need to choose it in the realt...
Does w-okada not support spin?
hardy yew
Does w-okada not support spin?
If you mean the old mainline one, then I don't think it does. If there's no way to select the embedder, it's gonna be contentvec and that's it.
But using that old app is still disadvised anyway
spare stump
If you mean the old mainline one, then I don't think it does. If there's no way ...
I'm using w-okada deiteris
b2332 ver
hardy yew
Haven't ever used it, but as I said, if you can't see an option to change the embedder, that simply means there is none
spare stump
THX:)
hardy yew
Tg develop or vonovox are the two ways to go
hardy yew
I've used both, vonovox is the best quality-wise (especially the most recent beta that will be released soon)
spare stump
wow thanyou
hardy yew
Latency I think should be similar in both. And also depends on the setup
low shard
Same for me, stopped working completely
have you tried re-installing it? what specific browser are you using?
It didn't work for me on Chrome, but worked for me on Brave
You also should try switching the devices to smt else and then back to the devices you're gonna use
low shard
checked it just out of curiosity, works for me (tg develop)
interesting
low shard
Hi, is there any difference between vb cable and vac lite? I have the vb cable i...
vac lite is more suggested for not having any issues
also I saw you sent a message about wokada tg develop fork, you'd need to download both 001 and 002 files and extract the first to extract both
if you could tell your pc gpu and os helpers could help more
karmic trellis
if i extracted the first one, do i need to put the 002 inside the mmvc folder, or what should i do?
now i have 4 files, mmvc folder, two bat files, and the 002, all in the same folder
do i need to do anything or is it good as it is now
low shard
if i extracted the first one, do i need to put the 002 inside the mmvc folder, o...
nop, when you extracted the first one it extracted both, you don't need the zips now
may I ask your PC GPU and OS tho? To see if you got the correct one
karmic trellis
win 10, nvidia rtx 4070 ti
low shard
win 10, nvidia rtx 4070 ti
nice, you could either use wokada tg develop or vonovox
hallow thistle
if i extracted the first one, do i need to put the 002 inside the mmvc folder, o...
Only extract .zip.001 one to a "MMVCServerSIO" folder, no need to put any other zip file into the same folder.
karmic trellis
ok thx guys
quiet axle
Hi i was wondering if its possible to run vonovox on a 2nd pc so my main pc wont need to handle the load
karmic trellis
should i use server or client in this okada fork? and also, what sample rate should i use?
hallow thistle
should i use server or client in this okada fork? and also, what sample rate sho...
gloomy hill
Can someone help me with my prompt?
hallow thistle
Can someone help me with my prompt?
What's up, Nick.
gloomy hill
What's up, Nick.
Hi. Here:
# Role
You are a highschool student and you need to be a topic generator that delivers unique, compact topics on demand.
# Task
Generate exactly 10 unique topics where i can spend 10 hours to talk about it with my friends and output them as a comma-separated list with no spaces.
# Context
The user needs quick topic suggestions in compact form—no explanations, no formatting, just the raw list.
# Instructions
- Each topic consists of **one or two words**
- Topics must be unique and unusual—no generic or overly common terms
- Output exclusively as a **comma-separated list with no spaces**: `Topic1,Topic2,Topic3`
- No additional text, no introduction, no explanation, no numbering
- Respond with the list only—nothing else
Each topic must belong to a completely different semantic field.
Avoid poetic or artistic word clusters.
All topics should be concise and straightforward, suitable for quick brainstorming sessions.
model-name=gpt-5.2
temperature=1.1
presence-penalty=0.8
frequency-penalty=0.6
top-p=0.95
"content": "Xenobiology,Cryptoeconomics,Paleomagnetism,Pharmacogenomics,Exoplanetology,Topology,Gastrophysics,Biosemiotics,ForensicLinguistics,Cryonics",
Can anyone tell me why the AI gives me these kinds of "topics"? They always have something to do with physics, astronomy, or biology, even though I don't explicitly say so.
It's supposed to generate 10 topics that you could talk about with other people for hours.
hallow thistle
I don't think anyone asked for this shit.
hallow thistle
```m
# Role
You are a highschool student and you need to be a topic generator th...
Which parts in this are input prompt and output message?
quiet axle
Guys im trying to multi pc setup Tg develop fork but when i try to run the bat file. it gives this error
usage:MMVCServerSIO.exe [-h] [--log-level {debug,info,warning,error,critical}] [--launch-browser]
MMVCServerSIO.exe: error unrecognized arguments : false --https true
Any idea why its doing that?
gloomy hill
Which parts in this are input prompt and output message?
The output is the content part. Above are model settings and system prompt
quiet axle
Guys im trying to multi pc setup Tg develop fork but when i try to run the bat f...
When i remove the argument, it runs but it stuck on initialized
lilac pecan
```m
# Role
You are a highschool student and you need to be a topic generator th...
ask chatgpt or grok ai
quiet axle
Have anybody tried to do multi setup using tg develop yet?
hallow thistle
When i remove the argument, it runs but it stuck on initialized
Did you add ".env" file in MMVCServerSIO folder? I did mine using my PC's IP Address, and the voice changer loaded up anyway.
quiet axle
Did you add ".env" file in MMVCServerSIO folder? I did mine using my PC's IP Add...
you dont use the bat anymore?
PC ip address your talking about is our main pc right?
not the one were using to run it?
hallow thistle
With an env file present, I just directly run the MMVCServerSIO.exe within Command Prompt.
quiet axle
ohh isee...
hallow thistle
PC ip address your talking about is our main pc right?
Yes, your main/host PC IP address where the voice changer is hosted.
viral mason
have you tried re-installing it? what specific browser are you using?
It didn't ...
I am using opera gx for it, was working before last year until November but then it stopped
low shard
Hi i was wondering if its possible to run vonovox on a 2nd pc so my main pc wont...
I just recently added this part to the vonovox guide: http://docs.aihub.gg/realtime-voice-changer/local/vonovox/#dual-pc-setup-via-sonobus-cross-os
low shard
I am using opera gx for it, was working before last year until November but then...
Try other browsers, a browser update or setting might have caused it to stop working
It works fine on:
Brave 1.87.192 (Build ufficiale) (a 64 bit) Chromium: 145.0.7632.160
Windows: 11 25H2 Pro
low shard
it's ok, I switched to Vonovox anyways since it's better atm
If you got 2 PCs you might wanna check out the dual pc setup for vonovox btw
plush jacinth
can someone help me fix this issue
low shard
it sounds interesting but I don't underestand it
one pc runs vonovox
one pc does other shit while using ai voice
anyways u have only 1 PC
plush jacinth
Pytorchstreamreader failed reading zip archive: not a ZIP archive
low shard
can someone help me fix this issue
!howtoask
plush jacinth
rtx 4060 windows 10 trying to start vonovox and getting
File "C:\Users\dot\Downloads\Vonovox_beta_17_11\runtime\Lib\site-packages\torch
serialization.py
", line 780, in init
super().init(torch._C.PyTorchFileReader(name_or_buffer))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: PytorchStreamReader failed reading zip archive: not a ZIP archive
Error: RVC model not properly initialized. Check above error for more details.
Failed to initialize audio processing
Critical error in start_vc: Failed to initialize audio processing
Traceback (most recent call last):
File "gui/gui.py", line 1496, in gui.gui.GUI.start_vc
File "gui/gui.py", line 1533, in gui.gui.GUI.initialize_voice_conversion
RuntimeError: Failed to initialize audio processing
low shard
rtx 4060 windows 10 trying to start vonovox and getting
File "C:\Users\dot\Do...
show a screenshot of your vonovox settings
!give-media-perms 3h @plush jacinth
ornate lake
can someone help voice changer
low shard
can someone help voice changer
!howtoask
karmic trellis
hey, in the tg fork are there any audio effects that you guys recommend to make my voice better than how it is now? Its not bad, but i feel like it could be better
viral mason
I've never needed to use any effects on the voices in tg fork, I use fl studio to add extra voice effects for specific characters but that's it
most settings I use to make my models sound good in tg fork is in advanced settings
karmic trellis
can you please tell me what crossfade overlap and protect are?
i have default settings and those are different than yours
viral mason
crossfade stitches words together to make talking sound more natural, I found increasing it made it sound more realistic without any quality loss
half fjord
yo ! someone know why there's not my mic or anything there?
viral mason
protect I am unsure of what it does but turning it completely to 0 made my models sound much better
viral mason
yo ! someone know why there's not my mic or anything there?
did you download the right one?
this one u have isn't as recommended as the new Voice changers are
karmic trellis
protect I am unsure of what it does but turning it completely to 0 made my model...
okay thx, i tried it and it made a difference
its better now
viral mason
yippe
half fjord
AMD
half fjord
where s the good guide?
viral mason
here's the guide
btw download this, it's a virtual audio cable, recommended as it's better than vb cable
half fjord
btw download this, it's a virtual audio cable, recommended as it's better than v...
Is it normal that my voice sounds much deeper on this RVC?
viral mason
Is it normal that my voice sounds much deeper on this RVC?
not unless you set the pitch low
it shouldn't lower pitch by default
half fjord
not unless you set the pitch low
Before I set my pitch to 12, I had a normal voice but now with 12 I have a very deep voice
viral mason
hmm
viral mason
weird
half fjord
"crepe tiny onx"
viral mason
before i had smth else
this is the normal pitch extractor
viral mason
"crepe tiny onx"
this one isn't really that good, very outdated
half fjord
this is the normal pitch extractor
i cant take an other one?
viral mason
i cant take an other one?
you would have to download them inside the app itself
half fjord
you would have to download them inside the app itself
Yk where ?
i cant find the settings
viral mason
ew
viral mason
i cant find the settings
this is old
I won't be helping you tho if you're egirl trolling
pallid crater
viral mason
https://colab.research.google.com/github/hinabl/AICoverGen-Colab/blob/main/Hina_...
preetty sure this is old
sure, what gpu do u have
AMD or Nvidia
what is ur pc GPU
is this AMD?
mk
you both need these two, first one is the voice changer second is a virtual audio cable
https://github.com/tg-develop/voice-changer/releases/download/b2397/voice-changer-windows-amd64-dml.zip
pallid crater
Which collab should I use to easily make AI covers?
pallid crater
I don't want to manually separate voices, that's why I don't like it
viral mason
I don't want to manually separate voices, that's why I don't like it
there really isn't anything left that can do it without seperating voices sorry
with weights shutting down and not having new sign ups there isn't anything like that left
for vac lite/vac 47 run setup64 (not as admin)
then after for the voice changer run the file called mmvcserversio.exe
same for u @grim kite
just those two things
pallid crater
That collab was the closest thing, why don't they fix it?
sour harbor
hi, everyone
can anyone say how to make a ai cover song? im new on it
viral mason
That collab was the closest thing, why don't they fix it?
whoever made it probably lost interest in keeping it working
I'd ask a helper or someone smarter than me who might know more spaces like it
quiet axle
most settings I use to make my models sound good in tg fork is in advanced setti...
Ohh interesting, im gonna try out your settings. May i ask how much extra you use?
viral mason
I'm tempted to not help you anymore
hardy yew
XDD
viral mason
yep, that is normal for it
pallid crater
the voice changer should open in a browser
That collab was the closest thing, why don't they fix it?
viral mason
already said why
setting your mic settings up, your input should be whatever microphone your headset/headphones use and output should be Line 1
chill bro
<@&1159293140440723499> keep eyes on this weirdo
pallid crater
I used to make my AI covers in Weights easily by putting the MP3 files without separating them, but then I don't know where to easily do it again.
quiet axle
oh god they got thanos snapped
low shard
was not expecting the lady to get kicked
We don't allow neither weirdos asking for sex requests neither catfishing
viral mason
I didn't see that she was asking for that, if I missed that then I'm glad she is gone
viral mason
Ohh interesting, im gonna try out your settings. May i ask how much extra you us...
I use extra at 2.7
low shard
https://colab.research.google.com/github/hinabl/AICoverGen-Colab/blob/main/Hina_...
that one is outdated and discontinued, try: https://docs.aihub.gg/rvc/cloud/aicovermaker/
!howtoask
pallid crater
that one is outdated and discontinued, try: https://docs.aihub.gg/rvc/cloud/aico...
Covers are easily done there.
low shard
Covers are easily done there.
that one you sent is never coming back, you need to use the updated alternative, that's your only choice
this one I sent also manually removes the instrumentals, similar to the one you want
pallid crater
that one you sent is never coming back, you need to use the updated alternative,...
that alternative?
low shard
that alternative?
yeah, check https://docs.aihub.gg/rvc/cloud/aicovermaker/
viral mason
you both need these two, first one is the voice changer second is a virtual audio cable
https://github.com/tg-develop/voice-changer/releases/download/b2397/voice-changer-windows-amd64-dml.zip
low shard
elaborate more:
- your pc gpu
- the tutorial link you're using
- the issue you're having
quiet axle
I use extra at 2.7
Thank you!
low shard
you both need these two, first one is the voice changer second is a virtual audi...
imagine it's AMD integrated GPU lol
viral mason
imagine it's AMD integrated GPU lol
their pc blows up XD
low shard
could you elaborate what I asked?
viral mason
if nothing works check your gpu
both are very simple
for vac lite/vac 47 run setup64 (not as admin)
then after for the voice changer run the file called mmvcserversio.exe
pallid crater
yeah, check https://docs.aihub.gg/rvc/cloud/aicovermaker/
Can they be done there without separating vocals and instrumentals?
half fjord
I use extra at 2.7
Do you know why sometimes the voice splits in two?
viral mason
Can they be done there without separating vocals and instrumentals?
he said u can do that yea
viral mason
Do you know why sometimes the voice splits in two?
what?
pallid crater
Does it have YouTube links?
viral mason
no clue ask Nick
viral mason
Does it have YouTube links?
pretty much all youtube tutorials for ai stuff is outdated or uses old stuff
low shard
Can they be done there without separating vocals and instrumentals?
the program will do everything automatically, please check it
low shard
Does it have YouTube links?
wdym? like youtube tutorial on how to run it? nop
viral mason
settings you mic settings up, output is your regular headset/headphones mic
output is Line 1
make sure you have set your gpu here
alright, next is to upload a model
go to the left where that + sign is
download a model from here, extract, and put the pth and index file in their respective slot https://discord.com/channels/1159260121998827560/1175430844685484042
viral mason
download a model from here, extract, and put the pth and index file in their res...
.
it's most likely due to your settings
since you use amd idk the best settings for you
but
I know going into the advanced settings and changing these make all models sound much better and more realistic
most likely it is your model then, what model are you using?
that's not a good sign
I'll assume then you're trying to use a "mommy egirl" voice and stop helping
good luck Noah
low shard
Check #📰│dev-updates message for News about Applio, Vonovox and Wokada TG Develop on AI HUB Docs!
bronze locust
yo my voice changer just stopped on discord
low shard
yo my voice changer just stopped on discord
!howtoask
bronze locust
7800 xt win it just stopped working as i reloaded it and added a kanye west voice
it works in the actual voice changer but not on discord
WIN 11
low shard
7800 xt win it just stopped working as i reloaded it and added a kanye west voic...
can you give the tutorial link you used?
low shard
the rvc voice changer
there are multiple options and versions, this is also a general AI server, please elaborate and give the link to the tutorial you used or program download link
bronze locust
iused
VOICE CHANGER WINDOWS AMD64-DML.ZIP
VAC470LITE.ZIP
it just doesnt work on discord
low shard
VOICE CHANGER WINDOWS AMD64-DML.ZIP
send the program download link, there are multiple voice changers
bronze locust
I CANT
i cant
someone on here
give me the download
and i downloaded it
and i cant send it as it is too big
low shard
and i cant send it as it is too big
you need to give the link, not the downloaded file
low shard
i dont have the link
can you show a screenshot of the program interface and the discord settings?
!give-media-perms 3h @bronze locust
low shard
that's wokada tg develop fork
set output to line 1
monitor optionally to your headphones to hear yourself
bronze locust
thanks
low shard
you're welcome
snow hazel
How can I download mel_band_roformer_karaoke_becruily.ckpt to UVR
viral mason
who?
that isn't a voice changer its just a pitch shifter
probably voicemod
rvc can but not just like a real person could
it could be voicemod but could also be another daw like fl studio or reaper
easiest to set up would be voicemod tho it isn't completely free
modest fiber
I just saw the johnny silverhand model , running it on vonovox should work fine eh?
viral mason
I just saw the johnny silverhand model , running it on vonovox should work fine ...
yea all models made here work on vonovox
I highly recommend the new beta version if u don't have it
modest fiber
Ah ok, haven’t even installed it yet so I’ll try to get that. You got any recommendations for other models that sound clean when talking? It really doesn’t matter what it is.
modest fiber
I know a lot of people ask here about e girl voice , I just want something that works good😭
viral mason
Ah ok, haven’t even installed it yet so I’ll try to get that. You got any recomm...
not really sure about clean models besides my own and ones made by Vix
viral mason
I know a lot of people ask here about e girl voice , I just want something that ...
look for any made by mine and Vix
modest fiber
I’ll definitely check them out. Thx mate
elder coral
the baldi model sounds good but the sh is a problem with the artifact
edgy minnow
@hallow thistle Popping back in to say its still working. I'm very confused by it, but TG and the fork of Deitrich or whatever for 5000 series is wokring.
Just wierd PC things.
viral mason
Like Shrek?
I wouldn't judge anyone on the type of model they use unless it's a e-girl/e-boy
So you're fine 🙏
Are you able to send a screenshot?
Any YouTube tutorials are outdated for real-time voice changer stuff
They don't use the new stuff and give kinda bad advice
You had it open last time we did this
Do the same thing u did last time
That's not how that works
Is this the one I gave u?
If u got it off a YouTube video it's outdated
👍
Want me to help u setup the cable first?
U ran setup64?
Just open mmvcserversio
Yep, when u do that then run mmvcserversio, should open a command window (black box with text) then open the voice changer in browser
It's an app that runs on web browser
That was unnecessary
But now for mic settings
Input: regular headphones/headset mic
output: Line 1
Weird
Try wasapi over mme
In the audio tab
Hm
At this point I'd ask a helper bc I'm struggling to stay awake, it's almost 5 am where I am and I'm not very good at finding what causes errors
It shouldn't be doing all this
Because it only runs the interface on web, the app is still locally on your computer
Wouldn't hurt to try
Oh shit that might be why it isn't that
Working
Yea I'll get u the peak thing tho, it's called Vonovox
I use it
Here ya go
For this all you need to do is run the start file after extraction
exotic scroll
Hello? I need help. I cannot cloning voice due payment many times.
hallow thistle
Has the initial issue actually solved? Because I feel like this gonna be a long conversation.
hallow thistle
Hello? I need help. I cannot cloning voice due payment many times.
Paid request and commission voice models aren't allowed here. Now what do you mean by something financial issue?
Hello?
Anyway, double click on start.bat file to launch the Vonovox.
This conversation gonna be a very long one if you ask for every step without having to take look at the guide doc as reference. I'm more of telling best known working settings. 
patent trellisBOT
viral mason
Run start
Not really
Same mic settings like last time
Hopefully without any errors
In output
hallow thistle
viral mason
It's for connecting the voice changer to discord or a game
hallow thistle
low shard
I just saw the johnny silverhand model , running it on vonovox should work fine ...
it should yeah
low shard
thanks
btw do you need any other help or is it done?
rvc can't do non speech sounds super realistically
hallow thistle
pulsar glade
Hello, im completely new to the subject. I wanted to use an Ai to change my voice as I speak. I watched a tutorial about RVC from W-okada but I don't really understand what to download and everything since it was 3 years ago (the tutorial, and between these 3 years, the website changed so Im kinda lost). I don't know if RVC is still good now, if there are better alternatives. That's why I ask for your help to help me download it and use it on discord 🙌
hardy yew
-rt
hardy yew
Hello, im completely new to the subject. I wanted to use an Ai to change my voic...
assuming you have an nvidia GPU, see Vonovox or Wokada Tg-Develop from the message above
both links lead to full installation/usage guides
both apps are nice with some pros and cons
pulsar glade
assuming you have an nvidia GPU, see Vonovox or Wokada Tg-Develop from the messa...
thanks you capybar
what's the difference between Vonovox and Woakada Tg ?
hardy yew
tg develop:
- used in web browser
- more convenient UI
vonovox: - separate app
- recent beta has some nice quality improvements, so vonovox will be superior in that context (obviously the difference is not night & day so tg develop is still a good alternative)
pulsar glade
okay thanks you !! I think I'll use Vonovox
low shard
voicemod doesn't use AI RVC Technology for voice cloning, it's just pitch and audio effects
red dune
guys, is intel arc graphics card good for generative AI?
pulsar glade
tg develop:
- used in web browser
- more convenient UI
vonovox:
- separate app
-...
I now have Vonovox installed. All I need to do now is download a preset ? (I read all the stuff about the settings ; audio device etc)
via voice-models channel, I suppose ?
low shard
guys, is intel arc graphics card good for generative AI?
which specific GPU and OS do you have? and what Generative AI Tool do you want to run?
low shard
via voice-models channel, I suppose ?
if you downloaded the VAC too, then yeah just check #1175430844685484042 mikami
low shard
virtual voice ?
the Virtual Audio Cable: https://docs.aihub.gg/realtime-voice-changer/local/vonovox/#virtual-audio-cable
pulsar glade
yes I downloaded, I don't need to do anything for it ? like I just downloaded and opened the setup and everything
hardy yew
via voice-models channel, I suppose ?
yeah, you can find models either in #1175430844685484042 or on https://voice-models.com/
then you import a model and its index file into a preset
tired minnow
-colab
pulsar glade
thanks you, I just need to download the preset and import it into vodovox and it's done ?
pulsar glade
yeah, you can find models either in <#1175430844685484042> or on https://voice-...
thanks for the link
I just imported the preset (first one I found on the voice model website) to test it , .pth file. But I don't really know what to do now, I tried changing my input to hear me talking but It doesn't work (sorry for asking help for every step) 💀
low shard
I just imported the preset (first one I found on the voice model website) to tes...
show a screenshot of your settings
!give-media-perms 2h @pulsar glade
low shard
vonovox:
input: your mic
output: line 1
other programs:
input: line 1
output: your output device like headphones
hardy yew
just like Nick said
but for testing
you can set output in vonovox to your speakers
then you can directly hear it
pulsar glade
what is other programs ?
hardy yew
discord, for example
pulsar glade
okay
hardy yew
anything in which you want to actually use the voicechanger
pulsar glade
I moved my preset to the vonovox file
bc I created another one and I think that was why it wasnt working
pulsar glade
vonovox:
input: your mic
output: line 1
other programs:
input: line 1
output: ...
I did it
now all I have to do I clicking on start ?
ITS WORKING
thanks you all guys
pulsar glade
may love is upon you
pulsar glade
https://tenor.com/view/writing-fast-delete-death-note-gif-11890321
DELETE
cerulean hedge
how do i create ai voice clones for ai vocal covers again? I know how to do it but just doesnt remember the site
and has the technology improved or are we still in 2023 again
oblique locust
hello, for some reason i dont have any audio with the cable?
low shard
how do i create ai voice clones for ai vocal covers again? I know how to do it b...
what's your PC OS and GPU?
low shard
and has the technology improved or are we still in 2023 again
RVC v2 hasn't improved that much
low shard
hello, for some reason i dont have any audio with the cable?
!howtoask
modest fiber
Oh yea I was told to download the beta version for vonovox, what’s special about it?
hardy yew
Oh yea I was told to download the beta version for vonovox, what’s special about...
much better stitching algorithm, among other things
in general.. just better quality
zenith kiln
"Vonovox_beta_17_11.zip" is that the right vonovox?
low shard
"Vonovox_beta_17_11.zip" is that the right vonovox?
the latest vonovox beta seems to be https://huggingface.co/dr87/vonovox/blob/main/Vonovox_beta_17_11.zip, so yeah
zenith kiln
the latest vonovox beta seems to be https://huggingface.co/dr87/vonovox/blob/mai...
i just dont have all of the settings so i thought its an older one somehow
hardy yew
i just dont have all of the settings so i thought its an older one somehow
that's actually the new one, some sliders were removed recently because they introduce more confusion with very little benefit or reason to ever adjust them
near basin
Hello for some reason, when i speak into my mic, my voice is audible and also the ai voice is, and whenever i switch my mic to cable, it just repeats it over and over
my problem is with mmvc rvc
white shadow
just wanted to post this here for quicker help lmao
https://discord.com/channels/1159260121998827560/1480328521002909851
rotund plinth
gloomy apex
you both need these two, first one is the voice changer second is a virtual audi...
is this the newest?
viral mason
is this the newest?
if you use AMD as your gpu yes,
if u use Nvidia tho use Vonovox https://huggingface.co/dr87/vonovox/resolve/c8034f5f6d50648a8109bb4f847182362e2b779b/Vonovox_beta_17_11.zip
https://software.muzychenko.net/freeware/vac470lite.zip
gloomy apex
if you use AMD as your gpu yes,
if u use Nvidia tho use Vonovox https://hugging...
yeah im useing AMD here is my card GPU 0
AMD Radeon RX 6650 XT
Driver version: 32.0.21043.1005
Driver date: 2/18/2026
DirectX version: 12 (FL 12.1)
Physical location: PCI bus 9, device 0, function 0
Utilization 63%
Dedicated GPU memory 6.6/8.0 GB
Shared GPU memory 1.3/16.0 GB
GPU Memory 7.9/24.0 GB
CPU
AMD Ryzen 5 3600 6-Core Processor
Base speed: 3.60 GHz
Sockets: 1
Cores: 6
Logical processors: 12
Virtualization: Enabled
L1 cache: 384 KB
L2 cache: 3.0 MB
L3 cache: 32.0 MB
Utilization 86%
Speed 3.91 GHz
Up time 237:27
Processes 307
Threads 7038
Handles 1570282
just download these two files then
spare stump
I made a model using KLM6, but since w-okada doesn't support spin, I’m using contentvec as the embedder instead. Even though I heard KLM6 needs spin, it seems to work just fine with contentvec so far.
What I'm curious about is: does using contentvec instead of spin basically turn the KLM6 model into a 'placebo' (like a hollow shell with zero performance gains)? Or is the model still performing as a legit KLM6, just without the spin-specific features?
Sorry for bugging you with questions every day! lol. No rush at all, so just lmk whenever you have a sec. Hope you have an awesome day, thanks a ton!
tame oracle
i want ehelp
!howtoask
manic plover
!howtoask
I did everything correctly and downloaded the app but a 'Premium Subscription' screen Stuck, and when I try to remove it. it does not get removed.
tame oracle
I did everything correctly and downloaded the app but a 'Premium Subscription' s...
did u even read 😭
manic plover
did u even read 😭
?
viral mason
I made a model using KLM6, but since w-okada doesn't support spin, I’m using con...
If the pretrain was made with spin you have to use spin, same with a contentvec pretrain
Vonovox supports spin and spinV2 models
But if you use AMD then you'll have to use wokada tg fork which also I think supports spin, idk about spinV2
spare stump
If the pretrain was made with spin you have to use spin, same with a contentvec ...
I'm aware that w-okada doesn't support Spin. What I'm wondering is: if I train a KLM6 model using ContentVec instead of Spin, does it essentially turn the KLM6 model into a 'placebo' (just an empty shell with no actual performance gain)? Or will it still work properly as a KLM6 model, just missing the specific features that Spin provides?
I am using a model pre-trained with SPIN, and I noticed that it works normally when training another model using ContentVec. Regarding the question I asked above... is it actually okay to use it this way? I'm curious if the features of the SPIN pre-trained model actually function as intended when trained with ContentVec, or if it results in a model with no real performance benefits.
hardy yew
I am using a model pre-trained with SPIN, and I noticed that it works normally w...
The way I see it, using a different embedder than the pretrain was made with just asks for quality loss and artifacts.
If you want to stay with contentvec, why not just use a contentvec pretrain?
spare stump
The way I see it, using a different embedder than the pretrain was made with jus...
Is that so? That’s such a bummer... It seems to be working okay, but it just feels kind of off, you know? But anyway, thank you so much.
hardy yew
Yeah, it definitely will work, but it's simply not what the pretrains was specialized at. I don't really see a reason to use a different embedder than the pretrain was created with
Can't speak about details as I only ever trained a spin model once some time ago and don't remember a thing about it.
Also, pretty sure that most non-OG pretrains are actually finetunes which probably would mean the spin ones could be better if they were made from scratch.
But that's a side note
vapid cape
Is there any Arabic-trained RVC voice model with Arabic samples?
hardy yew
Is there any Arabic-trained RVC voice model with Arabic samples?
You can use the search box in #1175430844685484042 or https://voice-models.com/
There is some
vapid cape
There is some
I’ve searched through the model list but couldn’t find anything.
I’m specifically looking for an Arabic female voice model with Arabic samples for RVC.
If anyone has one or can share a link, I’d really appreciate it! 🙏
vapid cape
I checked the link you shared, but I could only find a limited version.
Is there’s a website where I can get more arabic voices samples for RVC?
stuck ravine
im trying to use a voice changer, idk if it just becuase its not perfect, but sometime it kinda bugs out
viral mason
im trying to use a voice changer, idk if it just becuase its not perfect, but so...
what gpu does your pc have?
stuck ravine
nvidia
viral mason
if u downloaded the voice changer from a yt tutorial btw it's super outdated
knotty valley
how do i train my own model localy
viral mason
the docs for applio is right here but idk how to set up the version locally
knotty valley
ok
viral mason
ask a helper like server femboy
knotty valley
thank you
viral mason
you're very welcome
red dune
is intel arc gpu good?
I'm planning to buy a laptop an omnibook flip one and this is the gpu it has
viral mason
not for ai, you'd be shooting yourself in the foot
tbh just get Nvidia it's the best to go for
dusty bramble
just a question, is vonovox better than okada or nah
viral mason
yep
if u wanna know what is different about it go to their official discord server bc it has too much stuff to explain
btw it's Nvidia only
dusty bramble
can u send thr invite to me?
viral mason
sure!
dusty bramble
tyty🙏
red dune
tbh just get Nvidia it's the best to go for
the thing is the nvidia available here is all 360 not any of the rtx family.
and the only rtx 3060 is in the range of 100k 😃
I'll cry in poor
viral mason
the thing is the nvidia available here is all 360 not any of the rtx family.
and...
3060 is good 
red dune
that's nice
the only problem is the money
I live in a country where prices are high asf and salaries are so low
viral mason
that's nice
the only problem is the money
I live in a country where prices are ...
all I can wish you is good luck
viral mason
but can u give me a summary on why it is?<:pray:1151686040529141770>
tbh not really my brain is empty rn
hardy yew
but can u give me a summary on why it is?<:pray:1151686040529141770>
why is what?
viral mason
why is what?
he's asking why Vonovox is better than Wokada
hardy yew
improved pitch extraction, volume handling, audio stitching, probably also better audio pipeline optimization considering it was built from scratch with that in mind
probably there's more improvements but that's just what i remember compared to w-okada xd
(talking about most recent beta to be specific, 1.6.9 doesn't have some of those improvements)
viral mason
1.6.9 should be forgotten about tbh
the most recent beta even if it is a beta should be the new default version
hardy yew
yeah but for people new to vonovox it's not so obvious that 1.6.9 is so much worse than the beta
so it's worth mentioning
viral mason
fair
dusty bramble
shit, i installed the 1.6.9
gritty dawn
my voice changer stuck on white screen
dusty bramble
https://huggingface.co/dr87/vonovox/resolve/c8034f5f6d50648a8109bb4f847182362e2b...
appreciate it, ty
night fern
What happens if you train 200 epochs of a dataset using one custom pretrain, stop it, then resume using a different custom pretrain? Would it be borked? in RVC
hardy yew
What happens if you train 200 epochs of a dataset using one custom pretrain, sto...
A pretrain is a starting point for your training. It won't be used when you resume training, because then your actual model checkpoint is the starting point.
chilly furnace
https://giphy.com/gifs/ai-is-fake-your-slop-bores-me-isnt-real-Qz3fzoG7zhRup5sfz...
this mf unironically uses giphy
who uses giphy bro
white shadow
you'll need this before anything yuki 
https://www.ara.ac.nz/study/priority-learner-groups/international/english-language-study-options/
low shard
when i upload a voice changer it doesnt show up anyone know?
!howtoask
zenith kiln
how can i train my voice on a model with applio?
low shard
how can i train my voice on a model with applio?
what's your PC GPU and OS?
low shard
may I ask why my voice changer isn't working?
elaborate:
- your pc gpu
- your pc os
- the tutorial link
- the issue
low shard
formal ermine
-realtime
zenith kiln
what's your PC GPU and OS?
rtx 4080 and windows 11 pro
low shard
rtx 4080 and windows 11 pro
that's great, check https://docs.aihub.gg/essentials/how-to-make-voice-models/
zealous raptor
What vonovox settings should I use for default?
cerulean rapids
hi guys, i just downloaded this https://discord.com/channels/1159260121998827560/1472925123462955251
but when i put it on the rvc it doesnt change anything xd
dusty slate
https://giphy.com/gifs/ai-is-fake-your-slop-bores-me-isnt-real-Qz3fzoG7zhRup5sfz...
who’s gonna tell bro
round fossil
I made a rvc model and it worked great in the site, like, it was identical to the original, but when i tried putting it on okada voice changer it was really far off, does anyone have any idea on why?
- I tested with both trained and added index
- I used .pth instead of .onnx cause i couldnt convert it.
hallow thistle
I made a rvc model and it worked great in the site, like, it was identical to th...
What is your PC GPU? And where did you get that one W-Okada version from?
low shard
What vonovox settings should I use for default?
show a screenshot of yours
!give-media-perms 3h @zealous raptor
low shard
but when i put it on the rvc it doesnt change anything xd
show a screenshot of your settings
!give-media-perms 3h @cerulean rapids
violet swift
uh
low shard
uh
hello, you need any help?
this isn't really a server about getting clients and jobs btw
violet swift
tried to use vonovox. i just downloaded it
i have 0 idea in coding
but
i launched it and when it get to warmup
it makes an error
low shard
@violet swift I just saw your post in https://discord.com/channels/1159260121998827560/1481316368778399877, I replied to it
violet swift
oh ok
low shard
tried to use vonovox. i just downloaded it
they aren't a staff helper and just joined btw
zenith kiln
that's great, check https://docs.aihub.gg/essentials/how-to-make-voice-models/
but how can i make a certain model train on my voice so it sounds better?
viral mason
but how can i make a certain model train on my voice so it sounds better?
That's not a thing you can do with RVC
low shard
but how can i make a certain model train on my voice so it sounds better?
you can't
hardy yew
Can do it the other way around, trying different ways of speaking and more appropriate (higher/lower) pitch to find what the model works better with
viral mason
just talk like the person/character and or raise the pitch of your voice manually while doing so
I do that constantly as I make models of charactres instead of regular people sos I have to kinda act when using the different voices
zenith kiln
I sound robotic for some reason and pretty weird.
i've tried to mess up with my mic's settings and my voice
viral mason
did you download it from a youtube tutorial? if so it's outdated
zenith kiln
no im using vonovox
viral mason
oh.
models made on that pretrain are very old
rin3 is kinda a badly made pretrain so I would use the newests version of his voice
zenith kiln
So i have to check for the most recent models?
viral mason
nvm the newest one from what I saw uses OV2 which is also bad
zenith kiln
oh
viral mason
So i have to check for the most recent models?
it's better to use the most recent version of any voice
zenith kiln
okay thank you
viral mason
you're welcome
round fossil
What is your PC GPU? And where did you get that one W-Okada version from?
I have a rtx 3060 and i got W-okada from a random version on github, the conversion speed is not the problem, just the quality of the voice.
low shard
I have a rtx 3060 and i got W-okada from a random version on github, the convers...
if you used https://github.com/w-okada/voice-changer, it's not suggested for RVC anymore
-realtime
round fossil
K, ty
low shard
https://discord.com/channels/1159260121998827560/1290050765418659893/12900507654...
the user has deleted all his RVC models and his discord account, you can't really do much about it
kindred kernel
the user has deleted all his RVC models and his discord account, you can't reall...
just checking if anyone downloaded it before it all got deleted
chrome canopy
guys what is the best free ai photo editor that are browser based
fleet grove
Is tg develop better than deiteris?
tame oracle
Yes
chrome canopy
is claude better than chatgpt ?
sturdy pier
Hiii! How to make a cover ai? I don't have a conversor ( The thing that makes your voice turn into the AI voice bank.) Can u help me?? :/
devout shoal
can someone help me? my cpu doesnt sound raspy with an ai voice, but when i switch to my gpu, it sounds raspy, even though it should be better than my cpu, i have a 3070
viral mason
can someone help me? my cpu doesnt sound raspy with an ai voice, but when i swit...
try Vonovox, whatever you're using now will sound better with it
much better voice changer
viral mason
you're welcome! btw what voice changer did you have before
there are 2 with web ui
and what model were you using
devout shoal
hold on this download is lagging my pc lol
i was using these
and a stanley pines vc
hallow thistle
https://github.com/tg-develop/voice-changer/releases/download/b2397/voice-change...
This one is Tg Develop's W-Okada fork. Could've say this voice changer name and ask for best known working settings.
wet dirge
hi how to separate regular vocals from the bg one in ai cover
i want to keep the og bg vocals
but the ai version has them in it and it sounds awful
nevermind
i found a tool
wide vapor
Help please RTX 50 GPU won't gonna work with W-Okada 2.1.4-alpha
wide vapor
https://github.com/IllIlIlIllIl/voice-changer/releases/tag/b2335
Can you write me please
wide vapor
https://github.com/IllIlIlIllIl/voice-changer/releases/tag/b2335
what i need download?
hallow thistle
Help please RTX 50 GPU won't gonna work with W-Okada 2.1.4-alpha
There are Tg Develop's W-Okada fork (b2397) and Vonovox, these voice changers have better RTX 50 supports.
-realtime
hallow thistle
While b2335 W-Okada fork might work, it's simply another version of Deiteris' W-Okada fork b2332 where its Pytorch was upgraded to work with RTX 50 GPU, overall features are the same and just nothing else.
wide vapor
While b2335 W-Okada fork might work, it's simply another version of Deiteris' W-...
i see b2397 its better b2335?
hallow thistle
i see b2397 its better b2335?
What does this mean?
wide vapor
What does this mean?
hallow thistle
Tg Develop's W-Okada fork (b2397) is basically a continuation of Deiteris (b2332) one but made by different author with more recent features, especially RTX 50.
wide vapor
Tg Develop's W-Okada fork (b2397) is basically a continuation of Deiteris (b2332...
In Audio Processing not work Client. Work only Server. Do you know why?
fathom pine
has anyone tried doing real time rvc in the cloud (e.g. kaggle)? how's the delay?
viral mason
I've heard it's possible but I haven't used it since I have a decent gpu
oak falcon
If anyone want to collaborate in ml project so please dm me we collaborate and learn together
hard chasm
If anyone want to collaborate in ml project so please dm me we collaborate and l...
What are you making
oak falcon
What are you making
Nothing right now
oak falcon
Make using ann , cnn , rnn lastm etc
hard chasm
Make using ann , cnn , rnn lastm etc
Sounds good
hard chasm
Make using ann , cnn , rnn lastm etc
DM me anytime you wanna start
misty wigeon
@analog obsidian Sorry to be a bother, but I see your messages from time to time helping ppl. I'm wondering if you could help me by recommending some settings for a model I want to make in Applio?
I have a 43min speech dataset (deadlock character lol) but I'm unsure of what vocoder, sample rate, batch size, pretrain, and embedder I should use for an optimal model. I know a lot of it can be trial and error, but do you have any pointers?
analog obsidian
<@775545133448953856> Sorry to be a bother, but I see your messages from time to...
batch size 8, contentvec, original pretrain, hifigan
epochs max 200, save every 10, if e200 sounds bad, hear previous epochs until you find a good one
analog obsidian
do you think it would be worth trying refinegan too?
sure, try it and see if it gives you something good, the refinegan pretrain was trained only in 32k and spinv2
so you have to select these options for the training
ah i forgot about the 24k refinegan pretrain existence lol, i think that one is contentvec
misty wigeon
got it, okay.
one more thing, i've noticed when I convert audios in the inference tab, the command prompt window will say something about using a hifigan vocoder. if i've made a model using refinegan, will it automatically use a refinegan vocoder for inference?
analog obsidian
got it, okay.
one more thing, i've noticed when I convert audios in the inferen...
yes, it will automatically use refinegan
short wagon
Hiya, My VCC isn't working/take ages to select a voice
short wagon
pls elaborate
when I select a voice it takes ages to load and then when I try and start it it just doesn't start
low shard
when I select a voice it takes ages to load and then when I try and start it it ...
this is an ai general server, there are multiple programs and versions, elaborate #✨│ai-help message
you need to tell important infos such as your PC GPU & OS and the tutorial / download link
short wagon
you need to tell important infos such as your PC GPU & OS and the tutorial / dow...
RTX 5060OC, W11, idk what the software is called and I can't remember where I got it from (other than my friend like, a year ago on my deleted acc)
nvm, found the github page
it's only when I set it to the GPU
low shard
https://github.com/w-okada/voice-changer
that's original wokada, not suggested anymore
it probably doesn't work specifically since you have an RTX 50 Series GPU
uninstall it, if you have vb audio cable, uninstall it too
-realtime
low shard
if you need only realtime voice changing, either use Vonovox or Wokada TG Develop Fork
karmic trellis
Hi. I have been testing and using the wokada tg fork voice changer for like a week now. I currently use it on 300ms (and 2.7s extra), can i lower the ms even more? My GPU could handle that (4070 ti), but i dont want the quality to decrease. Should i leave it on 300ms, or can i lower it?
if the green graph thing in the top right corner never went above 200, can i set the chunk to 200? Does it decrease quality if my pc can handle that?
hallow thistle
how do i set index in vonovox
Similar issue happens with W-Okada voice changer, even in Vonovox, setting index rate to any value would use CPU resources more, so not always recommended unless you know what you're doing.
weak flint
@viral mason heyy i was wondering if i could dm you about some help? i would rather not send in this server just incase and would feel safer in dms.