any freeware that helps change recorded voice?

[kenny1999]

I want to record some voice but I don't want to be recognized. Any freeware that can help alter the voice a little bit ? But, would it be reversible?

Thanks for help.

[redwudz]

Try Audacity

[pippas]

As redwudz suggests, Audacity should let you do want you need.... If you simply slow down or speed up though, then yes, it's reversible.... Try creating a Vocoder track with a an unknown and complex sound.

I would think trying to reverse that would be pretty difficult.....

There's a tutorial here that may help: https://www.youtube.com/watch?v=dXNrWrxIaRE

[kenny1999]

As redwudz suggests, Audacity should let you do want you need.... If you simply slow down or speed up though, then yes, it's reversible.... Try creating a Vocoder track with a an unknown and complex sound.

I would think trying to reverse that would be pretty difficult.....

There's a tutorial here that may help: https://www.youtube.com/watch?v=dXNrWrxIaRE

thank you so much for the information. I am learning now

[hech54]

By "reversible" do you mean by someone else so they can identify you?.....or do you mean that you want to basically destroy the original recording by altering it in Audacity?

[netmask56]

I've done commercial work to reverse a so called disguised voice recording that had been "processed" - wasn't too difficult to unpick especially if it is just raising or lowering the pitch, however the best way to disguise a voice is to actually 're-voice' it with another person who is far removed from the original personalities even to the extent of picking someone at random. This by far the most secure way. Most Vocoder work the end result often sounds comical or SciFi ish and can be distracting. It depends a lot on the "why" you are doing this. Remember in most security services they have access to individuals who posses perfect pitch and are skilled at "listening in" to a recording like listening to a Beethoven symphony and concentrating on and literal hearing in your minds ear just the cello part. Sometimes called "Listening with the 3rd ear"

[lomitus]

Not to be rude here, but free software and wanting to change one's voice so as not to be recognized...is it just me or does this sound like some rather questionable motives here?

I must be getting old....

[pepegot1]

Wave Pad sound editor can do it also. Get an older free version 4.5.2, as newer version work only for 30 days.

[pepegot1]

void

[EddyH]

I wouldn't bother with anything as simple as a simple pitch up or down, as that can be very easily undone if there's someone who has a vested interest in finding out who's being interviewed. A bit like the people who try to disguise a face or car license plate with a seemingly complex but actually very easy to undo "swirl" filter. You need to do something that leaves it understandable but altered in a way that's not easy to reverse.

Possibly a good place to start, unintuitively, would be with those child-focussed voice-changer phone apps that have a range of different options. It means having to send it through the phone mic of course (might be able to use 3.5mm cable?), and back out through the speaker (or same jack) if there's no file save option, but, hey, it's not like we're going for CD quality here. And if you do something like converting it to a robot voice then feeding that back in as the source to turn it into a duck or alien or something, that'll be quite hard to reverse engineer. Just have to make sure it's still understandable.

There might be PC based versions of them, or some that can take pre-recorded files?

But anyway, as otherwise said above, actual TV producers tend not to bother with face-blurring/pixelating/mangling and voice-processing any more. It's too jarring, and always carries the risk of being reverse engineered SOMEHOW. They just film the person in sufficient silhouette or with other physical obstruction that you get the impression of there being a talking head (which makes it a bit less weird) without being able to recognise them even with extreme video enhancement, and revoice them with an actor. It's not exactly expensive or longwinded, and in fact takes less time in postproduction than mucking about with filters. It also much reduces the risk of even a filtered voice being recognised by accent, delivery style, dialect etc, as the original dialogue can be rendered to plain text and given slight changes to make the language more generic before being handed over to the actor in an ADR booth (...and maybe someone specialised in doing such replacements without any original audio as guide, too, so they're not influenced and simply have to go by timings plus the visible movements of mouth and body)

A fun thing to do would be to feed the voice into a gender changing filter, then pitch up or down as appropriate to make it sound like the interviewee is actually not their stated sex and just doing a bad vocal impression of it (as male and female voices actually differ rather more by harmonics than dominant pitch, and a high pitched man is unlikely to be mistaken for a low pitched woman)... would take a smart investigator to spot that, and to be able to accurately change it back. Problem is, you still have the issue of speech patterns etc, and any filter of that type which is in any way effective (which requires a whole lot of harmonic overtone analysis/isolation and individual re-pitching of each one before recombining) is an expensive pro audio workstation plugin rather than a freebie standalone tool. (Believe me, I've looked before, on behalf of someone after recording a pretend radio drama for a university project who was having trouble finding any female volunteers to do those voices, or indeed anyone who'd do it for a fee that a student could afford...)

[JVRaines]

not to be rude here, but free software and wanting to change one's voice so as not to be recognized...is it just me or does this sound like some rather questionable motives here?

n a r c !

[EddyH]

Oh, I forgot before - there are also enough options out there for just revoicing using a synthetic text-to-speech engine. I don't know if there's anything that would do it directly (and certainly the chance of it taking what you've said, recognising it and then replacing the soundtrack with a synthesised version would be slim), but if you're happy enough to transcribe what's said into one of them, record the output and paste it into the video, it shouldn't be a particularly complex operation.

Heck, even Google Translate could be a suitable source for it. You can make it speak both the input and the output sides. For added amusement or obfuscation factor you can type in e.g. English text on the input side of a, say, Japanese to <whatever> translation job, or any other language with a reasonable quality TTS attached to it (some are REALLY good, some mediocre, some shockingly bad, and others nonexistent) and get a quite heavily accented but still recognisable attempt at rendering what you typed, literally like a native speaker of that language having a shot at saying the English words without quite knowing what they mean. (It's also the only option for changing gender or output speed with GTrans, unlike other speech engines - each language has either a male or female voice and a particular word rate)

[netmask56]

- Link Deleted -

[redwudz]

netmask56, I deleted the link in your post as a political reference. Please read our rules before posting.

Moderator redwudz

[netmask56]

Sorry didn't realise it had any party political content. I read it as the ability to soon be able to imitate anyone's voice.

The main body of the article was about a new piece of software called 'Lyrebird' maybe I should have dived down into the article for their web page. Hope this one is safe - they have many examples of what can be done.... https://lyrebird.ai/ Clearly in the political dimension could be dangerous but I'm more interested in the technology.