Voice Cloning with RVC - Step by Step

Voice Cloning with RVC - Step by Step - Easiest method

Рет қаралды 14,558

Күн бұрын

Пікірлер: 37

@BobDoyleMedia Ай бұрын

🌀✨Hey! Quick favor... ✨🌀 If you found value in this video, a simple LIKE 👍 and hitting SUBSCRIBE 🛎 helps more than you know. It lets KZbin know we’re onto something good here and helps others discover it too. ⚛ I’d love to hear what you think! 💬 Drop your thoughts, insights, or questions in the comments below. 👇 Your support means everything. Let’s keep exploring the creative uses of AI together! #BobDoyleMedia #LikeAndSubscribe #YourSupportMatters #ThankYou

@snaphaan5049 Ай бұрын

I just want to say Thank you. This has been the best thing I have come across for rvc training. I have been struggling with mangio-rvc and a 1060 GTX for months now. So went ahead and paid the $10 on mimicpc, uploaded my data (I've had months to perfect it) and within a hour (usually takes me 7-9 hours for 90 epochs) I had a incredible 150 epoch rvc model. F--- me. Subscribed.

@BobDoyleMedia Ай бұрын

I am SO happy to hear that. Sometimes, for whatever reason, people get mixed results and boy do I hear about it! :D Happy to hear it works as well for you as it does for me!

@johnzach2057 3 ай бұрын

Nice to see some progress in the field.

@SiliconSouthShow 3 ай бұрын

Crap in crap out! Love it! E5/E2-tts FTW, if you want to do you driver recording several times to get what your looking for.

@Mercurial_Phoenix 3 ай бұрын

What program can generate singing vocals with the style and characteristics of a specific singer? And is there anything better than RVC for singing?

@fulldivemedia 3 ай бұрын

Rvc is the one of best open source Ai out there for it, there is "ultimate vice removal" that is free and can do better job for separating vocals, you should watch some videos and learn to use it, you can even use the singers song and put your own voice on it, or mix two voices to make a new voice

@joseparedes380 3 ай бұрын

u da mad and u da man, that's it. Great content, unique, different.

@pinocho7746 15 күн бұрын

In the trainig process, the last field right bottom, give "Error", and i don´t know why.

@zhubarb 2 ай бұрын

Can the sample voice file be in a different language to the trained voice?

@troymcdonald 3 ай бұрын

Thanks for this. I’d love to be able to use the voice in TTS, along with an ability for emotions and extras like a laugh, giggle, etc. Think Google’s, Notebook LLM audio blog they can make for you.

@fulldivemedia 3 ай бұрын

There is a ree tts that is in beta but the beta version is amazing, F5-E2 tts is on Pinocchio is free and have emotional control, and doesn't need to train, just about 20 seconds of voice, and it is amazing

@TheJuice2024 3 ай бұрын

Big up Mimic PC, my favourite playground 🫶

@CaptainCarlossi 3 ай бұрын

Hi, thanks for your great work. I found your channel and your videos are exactly what I was looking for and what matches my interests. This video is also great, because it includes everything you need to know for training own models. One small question: I use Suno to create some songs, then I separate Vocals and Instrumental with UVR5. Afterwards I change the Voice of the Vocals to my Voice with RVC and finally put together all of this to "my Song" :D ... But I have two problems: A) The vocals are (intentionally) distorted because I mostly create rock songs. I know it's difficult, but is there a way to make the distorted vocals more or less undistorted again so that RVC can handle them better? My second problem is that the voice created with the my own model is usually too high. I then have to make it an octave lower (Transpose -12), but then it is either too low or sounds boring. With values between 0 and -12 I don't know how to adjust the music (e.g. with audition etc.). Can you please give me some hints? Thanx in advance ;)

@pdjay8912 3 ай бұрын

한번 Applio를 써보세요. 저두 한동안 RVC만 쓰다가 Applio로 바꾸었는데 결과가 상당히 놀랍고 프로세싱 속도도 상당히 빠른 편입니다. 그리고 Applio로 학습시키면 그 결과가 더 깔끔하고 좋습니다. 여러가지 기능들도 포함되어 있구요. 더이상 RVC를 쓸일이 없어요. 같은 pth학습파일을 사용하는데 폴더관리도 더 쉽습니다. 강력히 추천합니다.

@theheroes-nn7gn 2 ай бұрын

Where can I get it and is it free?

@amkire65 3 ай бұрын

If I use three four or five-minute audio of three different voices, will it merge them into something totally new? Also, what about "borrowing" voices from audio books to train?

@fulldivemedia 3 ай бұрын

You can train two separate voices and mix them later, watch more videos about rvc

@fulldivemedia 3 ай бұрын

I think you should talk about the "e2-e5-tts" it is beta but it can do a perfect clone from 20-30 second audio, no training needed, and have emotional control, and all open source and free, it is also on Pinocchio

@BobDoyleMedia 3 ай бұрын

@@fulldivemedia I’m literally playing with it right now. It is amazing and I will definitely do a video on it. And I used Pinokio.

@fulldivemedia 3 ай бұрын

@@BobDoyleMedia I always enjoy your videos, not just for the news, but also your spirit, I miss your funny music you make with Ai, and I would like you to make a video about prompting, specially in text to image, some KZbin rs have free pdf about it, I think it can help a lot of your audience, maybe a video about making faceless KZbin channels using free Ai, I wish you long and happy life

@snaphaan5049 Ай бұрын

@@BobDoyleMedia Only issue is it can (at the moment) only accurately clone chinese and english language voices. Anythign else sounds... weird. I'm Afrikaans so it's propably gonna take years before it's included.

@tuckerbugeater 3 ай бұрын

Have you tried F5-TTS?

@BobDoyleMedia 3 ай бұрын

@@tuckerbugeater I have not!

@john_blues 3 ай бұрын

@@BobDoyleMedia I'd be interested to hear your take on it. It has a lot of hype. I wasn't overly impressed. Perhaps I'm missing something.

@pdjay8912 3 ай бұрын

F5-TTS 이거 예술입니다. 15초정도의 음성데이타만 있으면 말하는 느낌이나 발음,억양등도 비슷하게 보이스모델로 학습시킨뒤 텍스트를 음성으로 읽어줍니다.

@fulldivemedia 3 ай бұрын

@@pdjay8912this can really help, some times I put my voice training for more than 24 hours on 4-5 hour of audio to get good outputs, in this way I can use several 15 seconds voice frome the same person, one yelling, one happy, one sad, one laughing..... And create my own emotions, but they have emotional options too, it is beta but it showe lot of promises

@Nafas2021 3 ай бұрын

When I'm doing a one-hour conversion to another voice, there are some artificial and interrupted parts. Is there a way to convert one hour to another voice with very fluent and non interrupted parts?

@geo4design 3 ай бұрын

I recently installed RVC on my old MacBook Pro after switching to Linux, and I must say it was a bit of a process, but it's free. After experimenting and creating some KZbin videos, I found that it takes several attempts to get things right. Given the effort and the number of failed attempts, I don't see the point in paying for it. Instead, I recommend getting a Linux PC and doing it for free. It's a more cost-effective and reliable option. Are you being sponsor dont seems to help us as consumers.

@wheresmyvibe. 13 күн бұрын

help i’m trying on macbook

@collinbloo 28 күн бұрын

Is there a way I could do this for free?

@hihosh1 Ай бұрын

But this still doesn't explain how to easily set up pytorch and everything you need to even make the program run

@BobDoyleMedia Ай бұрын

@@hihosh1 the point of this video is to show why you don’t need to do all of that if you use mimicPC as your interface. I do provide links for downloading to your own computer, but that’s not part of this video.