how to convert male voice to natural female

If you are new here check this forum first, your question may have been answered.
RELATED
PRODUCTS

Post

UltimateOutsider wrote:
whyterabbyt wrote:I already got that from your comments on Vocaloid supposedly being realistic, yes. The fact remains that the first set of examples were deemed 'believable' enough for film, and this software been being used for that purpose for several years now.
Well, Birdemic is a feature film, so forgive me if that credential doesn't impress. And any film with a budget higher than Birdemic's can employ actual women for speaking parts.

If you think the Ircam thing sounds like a real woman, well... you might need to meet more women.
Perhaps you can't read. The fact that its been used for that purpose is proof that its been deemed believable enough for that purpose.
That's all that needs said. Like anyone who's used a lot of audio software, I'm pretty familiar with the kind of artefacts etc to listen out for, which is why I didn't say whether it sounded real to me, or not. Nice try, though.
For someone who's not so familiar, eg an 'audience' I think it's likely to be believable enough[\i].

There are many reasons why a properly-programmed Vocaloid can pass better than the real-time speech processors or TTS can, not the least of which is the fact that it employs actual speech samples of actual females (or males, depending on the voice package)


Im quite aware of how vocaloid works; im also quite aware that the use of samples is in itself insufficient to provide realism, as evinced by every vocaloid example ive ever heard. it certainly isn't up to the vast range of levels of nuanced expression in a genuine voice, something which is quite important to speech. Voice processing preserves those nuances, and the level of expression far more effectively.

And of course, feel free to gloss over how much time and effort it might take to get vocaloid to get near it.

But you're the guy who recommended Antares Throat a few posts back, so I see no point in defending it further. ;)


Yes, I the guy who recommended a tool that will do what the OP requested; transform his male vocals into something sounding female. It's not great at it but it can be picked up cheap fairly frequently, and I think it'll probably do better than some of the previous suggestions.
But please, enlighten me how vocaloid, or any other speech synthesis, can perform that specific task, I'm dying to hear that.
my other modular synth is a bugbrand

Post

Imho, "Best tools" requested in original post are (1) your ears, to meticulously study a female vocal slowed down by 5-6 semitones (2) your vocal tract to reproduce this (3) any old DAW that lets you transpose it back up 5-6 semitones & run it through an equalizer.

In case the moderator considers the thrust of the original post to be strictly software-oriented (as all other responses have) I shall not detail the astonishing strategy of trading surplus male resonance for female pitch any further here (but do check Dalal's thread at http://www.kvraudio.com/forum/viewtopic ... 2&t=365086
...if interested).

All I can say for software is that I struggled for an insanely long time manipulating formants using CM magazine's MHarmonizer (setting=vocoder, high quality), & Elastique in Reaper, and reluctantly reached the same conclusion as ntom, tradivoro, UltimateOutsider and Timman (posts 2, 9, 10, 11)... Quality conversion to female is simply not possible.

We may blame our mothers for the core 'problem' (?!), establishing in infancy our ultra-sensitivity to the tiniest nuances / inflections / textures which constitute vocal authenticity... The tension between Whyterabbyt & UltimateOutsider (post12 onwards) is typical of how quickly & deeply this operates, & shows why software which wanders into the fray faces such crushingly strict judgement.

Post

Zoned wrote:We may blame our mothers for the core 'problem' (?!), establishing in infancy our ultra-sensitivity to the tiniest nuances / inflections / textures which constitute vocal authenticity... The tension between Whyterabbyt & UltimateOutsider (post12 onwards) is typical of how quickly & deeply this operates, & shows why software which wanders into the fray faces such crushingly strict judgement.
Erm, no. Its as shallow as the fact that the suggestion of voice synthesis software that generates vocals from scratch is not the answer to a question about voice processing software that can transform one vocal timbre into another. Nothing more, nothing less.
my other modular synth is a bugbrand

Post

I felt the exchange was highly informed and heartfelt on a very difficult subject, whyterabbyt, that's all. Peace.

But back to the matter in hand: we all need a transcendent female vocal to crown / center / validate our various compositions... Since men are sweaty, megalomaniacal, emotionless brutes - there's little doubt about that - the transgender problem remains. Here are my most recent attempts using the aforementioned FRET system (Female-Retardation Emulation + Transposition).

http://k003.kiwi6.com/hotlink/bfbk838u5 ... emale_.mp3

http://k003.kiwi6.com/hotlink/7lubd5j35 ... emale_.mp3

Original post asks for before / after examples - so for a rough "before" (ignoring EQ) just transpose down 5 semitones.

Are they any good? They sound promising to me but I just don't know. Feel free to criticize. I won't cry (not openly anyway - I'm a MAN remember !!!)

Post

Indeed... I always suspected Rick Ashley & Kylie Minogue being one and the same session singer; only difference being Kylie is pitched up and Rick is pitched down.
We are the KVR collective. Resistance is futile. You will be assimilated. Image
My MusicCalc is served over https!!

Post

I could argue that the real difference is that Kylie rocks and Rick rolls.

But I can't be bothered.

Post Reply

Return to “Getting Started (AKA What is the best...?)”