Mandatory Voice Disguiser

This is not exactly a feature suggestion for Linden Lab to implement. More of thinking out loud.

Supposing you had a voice chat system in which the use of a voice disguiser system was mandatory. No one would get to use their real voice. If this idea was applied in the SL voice system, the mandatory nature of it would be determined by a parcel owner, not something that would apply universally. This idea could be used in voice chat systems outside of and having nothing to do with SL.

Does that seem like it might be fun? You'd get the fun of being able to talk but still not be identifiable on the basis of sex and age and other factors that bear on one's voice characteristics.

One could make it sillier still by having times when only one disguised voice pattern was allowed, so everyone would sound like a male, or everyone would sound like an female, - or all Darth Vaders, all squirrel voices, etc.


---- second thought ---

So far as I am aware voice modifiers operate by altering the frequency balance of the voice signal, adding echo, and doing other modifcations to the original voice signal.

Voice to text systems operate by doing a phoneme recognition first, which produces ambiguous results. Then various forms of anaylsis such as grammar checking , being part of a phrase that exists in a sample document, etc are applied to the candidate words to help reduce the ambiguity.

Suppose you had a voice conversion system with a somewhat different purpose, a phoneme recognition program. This sort of program wouldn't be concerned with figuring out the correct words a person was saying but rather the correct sounds.

If a phoneme recognition system could be made to work in realtime then you could have it output a phoneme using a stored voice such as Neospeech Kate or one of the AT&T voices. This could be used as a voice disguising system in which the output voice had no resemblence to the input voice.

Well, my first question is if SL voice will actually be usable with a voice disguiser? That is, will there be a way to impose the disguiser software between the sound card and the SL client?

I've long thought that SL AV's should have both a voice disguiser and speak in an accent and/or manner consistant with their character. Part of the SL experience would be staying in voice character.

Interesting idea.

Personally, I'd rather have a voice *transcriber*.

But that seems to be a but further off, let's set so double the killer delete all

That does sound the fun way to do it, that way less people are disadvantageds, but the deaf and speech impaired still would be

I've used the text-to-speech option on a mac in another world...still get to type (think before sending) and sounded pretty darn reasonable.

So the reason of this idea, you still can pretend you're a lady?

Honestly, it doesn't seem worth the trouble for LL to jump over while they are at it.

However, doesn't mean you can't get a 3rd party software to do that for you.

http://www.audio4fun.com/voice-over.htm
http://www.screamingbee.com/product/MorphVOXJunior.aspx
http://www.blazeaudio.com/categories/vchanger.html

That's a lot of good links you've got there, Vincent. Is there something you want to tell us eh?

I know the two first and they are sadly very deceiving.

A voice to text system is a phoneme parser. Morpheme analysis is used to enhance the accuracy of the phoneme parser. You could just use the parser by itself, but why? It'd be slightly faster, but far more inaccurate.

Voice on SL is the wrong way to go. Just forget it. Its basically Second Life defeated by half. The next half being a webcam! Its all part of the so called onward march of development and improvement which will bring Real Life into Second Life and destroy it.
Sure it is said you do not HAVE to use the voice inclusion (when it is released) but not using it clearly demonstrates you have something to hide.
Voice is a totally different medium for communicating with. It changes everything, believe me I've used it. Most importantly it gives greater recognition. For a start you cant think before you speak, with any success, like when you type text when you basically have all the time in the world to reply. If you use voice once you've spoken the words, thats it really, to backtrack and say you didnt mean what you said causes all kinds of confusion and distrust. Its the wrong way to take. The logical step beyond voice is your webcam and THAT totally defeats any second life whatsover. I'll bet they are even looking at that feature already.

Your response doesn't indicate that you read my post. Please try to post comments that are relevant to the thread topic. If you can't find a thread where your comment is appropriate, all you have to do is start a thread of your own.

There are a number of threads discussing the merits and demerits of adding voice to SL. This is not one of them.