VaM 1.x Create Voice Audio - Changing the Voice Male to Female

[Hipper]

I'm looking to record my voice (a male) for both male and female characters in a scene.

The stages as I see it are:

1. Prepare a written dialogue. I can do this.
2. Record that dialogue to my PC. I can do this.
3. Convert that dialogue to other voices, notably female. This is my question.
4. Once 1 to 3 are completed it seems that this audio file can be added to VaM and hence a scene, and can be Lip-synced if necessary. I will research these two matters in more detail later.

So, how do I convert my recorded voice, stored on my PC, to other voices, in particular female?

I've seen on here some ideas and searched online, but it seems a major battlefield to find one that will suit me. Also with AI there seems to be a plethora of products claiming AI usage. There are also reports of software, once downloaded, leaving active bits behind once its uninstalled. It seems not only a battlefield but a minefield!

I want software (I'm happy with payware as I assume that is safer and more reliable) to put on my PC. I don't need real-time conversion etc.. My only purpose is to make speech for VaM.

Any recommendations appreciated please.

(I don't want to use Voice.ai. I went on their website and I had no intention of downloading anything, yet somehow voice.ai-Downloader.exe arrived in my Download file. Not only that, but it keeps entering my Download file everytime I open my Browser (Microsoft Edge). I contacted Voice.ai and they offered a couple of solutions which had no effect. My only option it appears is to change my browser. I'm not happy!)

 

[SlimerJSpud]

You might try the app Audacity, which is free. It can do various sound effects.

If Edge has the ability to flush cache, cookies and history, try that. Sounds like malware to me.

 

[Clownmug]

Have you tried RVC? It's free, open source, and you can run it locally. The setup is pretty easy with the guide below. I prefer RVC-TTS personally, it's probably better to go from a female TTS (text-to-speech) voice rather than your own in my opinion.

RVC:

GitHub - RVC-Project/Retrieval-based-Voice-Conversion-WebUI: Easily train a good VC model with voice data <= 10 mins!

Easily train a good VC model with voice data <= 10 mins! - RVC-Project/Retrieval-based-Voice-Conversion-WebUI

github.com

RVC-TTS:

GitHub - litagin02/rvc-tts-webui: Text-to-Speech Gradio webui using RVC and edge-tts

Text-to-Speech Gradio webui using RVC and edge-tts - litagin02/rvc-tts-webui

github.com

RVC Setup Guide:

 

[murn]

You can use VoiceModVam to get a bit of change in voice. I'm not saying from male to female outright. You'll have to put in the voice acting work and then mess around in VMV. If you're looking for something paid I will suggest VoiceMod itself. You'll have less effort in making your voice sound female and the AI voices works really great but it's all about acting the part to make anything work anyway. You may have to do some recording magic to record from the pc but it works well. If I was in a place that was not having dogs barking all the freaking time or some other crap going on I would have been doing my own recordings for voices.
I use to use voice director from replica studios to convert voices but they had shut down a month ago so I can't recommend where you can change male to female audio files... yet.

 

[Hipper]

Thanks everyone for your replies.

SlimerJSpud - I already looked at Audacity and have heard of it before in the Hi-Fi world but it just seems to do things I can do already (using Filmora Wondershare for example) like using an Equaliser, which from my brief efforts doesn't work out very well.

Regarding Edge, I did the cookies, cache and checked extensions. I even searched my three internal hard drives. In the end I've replaced Edge with Firefox and made that my default browser (probably something I should have done anyway) and since then Edge itself no longer downloads the Voice.ai-Downloader.exe. I wonder if there was a three or four day time limit on its activity? Anyway, the bloody thing has gone and I'm not going there again!

Clownmug - RVC looks way more complicated then a simpleton like me wants to deal with just to get it on my PC. Thanks anyway.

murn - VoiceMod looks interesting. Unfortunately I couldn't find any pricing on the site for their Pro version. Fortutanately I found a promotor of VoiceMod on YouTube and he quotes £4 per month or £50 Lifetime licence.

I'll check it out.

 

[Hipper]

I tried VoiceMod but it seems complicated to work with pre-recorded material. I'll keep looking!

 

[murn]

To get voicemod to work with pre-recorded audio you will have some work rerouting inputs and outputs of audio sources. Voicemod is for real time effects not pre recorded but when there's a will there's a way. I think Elevenlabs has both tts and voice changer so you can use that.

Yep ElevenLabs has it.

 

[Banana2025]

I use https://app.kits.ai I love it .

 

[Hipper]

At the moment I've come to the conclusion that SlimerJSpud, the first responder, called it best!

I tried an online voice changer which had some 200 possibilities but there were only three British female voices. Whilst on the surface it appeared quite good in fact it seemed to struggled with some words. Many of these so called free trials still involve signing up for something.

And be warned, even the free Audacity is a bit tricky to download without getting other software you don't want. The button to 'Download Audacity 3.7.4' is a trick. Underneath it says 'Download without Muse Hub' and you should click on that if you only want Audacity (the button should of course clearly say 'Download Audacity with Muse Hub'. I fell for it and had to uninstall Muse. Not only that but it places a folder (an empty one) in your 'Users/Your Name' folder that you have to delete yourself. The internet pisses me off more and more as in many places there are a series of traps you have to avoid to get what you want and they are deliberately designed that way).

Audacity ® | Free Audio editor, recorder, music making and more!

Audacity is the world's most popular audio editing and recording app. Edit, mix, and enhance your audio tracks with the power of Audacity. Download now!

www.audacityteam.org

Anyway, I got Audacity and did the following:

1. I tried to speak in a more feminine voice. It's fun to try and also to try to speak slower, faster or in other different ways. Another skill to develop.
2. Use Audacity to raise the pitch. I found 1.5 to 2 semi-tones were about right. Any higher and you are in 'chipmunk' land!
3. Use Audacity Equaliser and roughly followed the guidance I'd found on the internet which said (I can no longer find the link):

1. Make a High Pass filter from 100Hz (in other words lower to 0dB from 100Hz downwards in a smooth curve).
2. Boost 1-2kHz slightly.
3. Subtle boost 3-5kHz.
4. Cut slightly 400-600Hz
5. Reduce 5kHz slightly to reduce sibilance.
6. Seemingly contradicting (2), Cut 1-2kHz slightly to reduce nasal sound - presumably if there is one.

Reducing pitch has the most effect but the EQ does have some. Obviously there is room here for experimentation.

I also found to act a male but make the voice slightly different from yours, perhaps 0.5 to 1 semi-tone change might help.

The results are OKish but not what I had intended and could certainly be considerably better. Perhaps that will come with practice. Maybe also the AI type voice changers will improve and expand.

Thanks for all the advice.

 

[SlimerJSpud]

"Alvin!"
(chipmunk voice) "OK!"

It always helps to read the fine print to know what button to click on...

 

[Hipper]

The thing is the right place to click doesn't look like a button. It can't be an accident.

 

[murn]

The thing is the right place to click doesn't look like a button. It can't be an accident.

Not accident. It's like that on purpose. Welcome to the internet.