I need some help pls

[hushed bison]

I'm trying to download the voice changer but I couldn't find the "MMVCServerSIO" thing, I don't know if I downloaded the wrong folder...

[stoic zodiac]

This is a General AI Server, we won't be focused on voices anymore

Elaborate:

• your PC GPU
• your operative system
• what you want to do
• what tutorial link are you using
• a screenshot of the program

[final trail]

The folder or the .exe file? If the exe program file went missing after extraction, it could be one of your antivirus programs removed it as a false positive. W-Okada is a free and open source program. Can you tell me which tutorial video or website are you following and what is your PC GPU?

[hushed bison]

my pc has a Nvidia RTX 3060 Ti, window 11, and ryzen 7 uhm, I was trying to install the W-okada voice changer with the guide but I couldn't find the application that was mentioned, also I tried to go to other section with all the versions and when I tried downloading them the folder was corrupted or something cuz WinRar can't open or extract the folder...

[stoic zodiac]

-realtime

[river flicker]

💻 Local Realtime RVC

Guides for Programs that use RVC Models in Realtime for Calls/Games

• Wokada Deiteris Fork

Most suggested. GUIDE

• Original Wokada

ONLY the latest alpha comes close to the Deiteris Fork performance, older versions in youtube tuts are way worse. GUIDE

• **RVC GUI Mainline Realtime**

Unavailable, the guide is outdated and the program is worse compared to the ones above, and much less updated

[stoic zodiac]

1st link

Wokada deiteris fork

Don't use the GitHub download, it's splitted, just follow the guide

[hushed bison]

I downloaded 2 folder, the first in the W-okada fork guide and the in the releases link

[hushed bison]

but is it a link to a direct download? cuz that's what I was looking for

[stoic zodiac]

You need to read the guide

It's not easy to download the program and use it

You must read all of it, it has a link to download the program ofc

If you don't read the guide, you won't be able to use it basically

It's a must

Also because you need 2 programs not just 1

[hushed bison]

the RCV right?

@stoic zodiac look

[hushed bison]

uhm @final trail sorry, can you help me with this, I downloaded it but when I use it sounds terrible, cuts my voice a lot and sounds too robotic

[stoic zodiac]

gpu: rtx 3060 ti
f0: rmvpe without onnx
extra: 2.7
chunk: 120
input: microphone
output: line 1
monitor: headphones optionally to hear urself

[hushed bison]

k thx

[hushed bison]

uhm btw, is there any like mod that I can use to improve the model? bc I was using it and it's worse than the W-okada app version, the one that runs on my pc, not in the browser

[final trail]

That's because you just run W-Okada anyway out of the box without ever setting anything on W-Okada, hence the robotic sound.

The lower "extra" number (0.5 s) indeed makes audio low quality, so 2.7 s extra is always recommended.

[hushed bison]

no, I did edit some stuff but still making the robotic sound, let me send you a ss

[hushed bison]

that's why I was asking if there is anything I can add to the client to make it sound better, or if it is just because of the model

in that case, can you recommend me a certain type of model that can sound more "natural", I'd tried with Japanese models, English models and Spanish models...

[final trail]

Wow, ok. If chunk and extra settings still don't make audio to sound better, it can be the voice model itself. Try a better one from #1175430844685484042

[hushed bison]

so what should I look for bc I saw that not all the model have like the same numbers and idk what does that mean

I downloaded one that says 400 epochs

and I saw another one with 250 I think

[final trail]

[hushed bison]

I'm sorry I'm new with all this kind of stuff

[final trail]

The epoch number found in RVC voice model name doesn't mean anything; it basically tells how many epoch used during training of a voice model in an RVC program. I forgot how to explain about this.

[hushed bison]

alr so the numbers doesn't matter at all

[stoic zodiac]

on wokada, you can optionally use more advanced settings for benefits:

• Force FP32 mode: on (THIS IS OFF BY DEFAULT! Turning this on improves stability. Increases VRAM usage by 200 MB)
• Disable JIT compilation: off for faster loading speed of the program, on for slightly better performance (10-15 ms) for Nvidia only)

[stoic zodiac]

no, all the info you see doesn't matter in the quality, it's just training info, the only way to check the quality is via listening to it or using it

get a better model

[hushed bison]

How can I turn FP32 on and disable the JIT?

[hushed bison]

k, thx

[stoic zodiac]

Check the advanced settings, it's in there