Syntrillium Backup Forums :: View topic

Iscander · Posts: 4

I've recorded speech of a man with a neutral voice, without
emotions.
How can I color his speech with some emotions (particularly,
I need to make his voice merry and sonorous)?

Iscander

Gulliver · Location: Estonia Posts: 442

If you want emotions - get a new, 'emotional' voice!
Seriously, you can't simulate them with electronics (be it hardware or software). Sonorous? Maybe, if you'd try to EQ and/or compress it somehow. But merry?... Forget it and try to get a new voice, really. (Or make him read his speech again - emotionally!)
Good Luck! ;-)

Edited by - Gulliver on 07/10/2002 01:03:49 AM

Craig Jackman · Location: Canada Posts: 909

Digital Audio Workstations are wonderful things. they can fix all kinds of little mistakes and let you create what you want. That being said, the original phrase of computer programming still applies to digital audio .... "Garbage in = Garbage out".

Graeme · Member Location: Spain Posts: 4663

This is about the best example of fixing a problem at the source, on the forum at this time.

You have just discovered why v/o specialists are paid so much money to read a few words into the microphone.

VoodooRadio · Location: USA Posts: 3971

I am not a Bass player per se.... but, I managed to get a track recorded using my bass. Now, how do I get it to sound like Jaco Pastorius? Wink

Good Luck Voodoo

oretez · Posts: 181

word on the street was that Bach was one hell of an organ player. once his compositions were published than just anybody could do what he did.

the debates raised by these type of questions (how do I use technology to control and manipulate expression) probably predate but are certainly exacerbated by technology that reproduces performance.

the questions are legitimate even if motivations are not always pure. Not infrequently its exploring the boundaries of what's fashion and what's physics that both get undone and new approaches to old art and new art forms emerge . . . while I have no intention of getting rid of the 50's fender tweed, 60's HiWatt twin, some obscure formulation of a baby Princeton I'll also throw a Line6 Pod into the arsenal . . . and the Line6 Pod produces results I would not have thought possible 10 yrs ago . . . and while my guess is that emotion in human speech is a bit more difficult to model than even an unmodified mid 60's AC30 (whose primary purpose in existence was to mock and betray you) I've also seen modeling synthesis and voice generation software come a long way in the last 10 yrs. (and original poster was asking about emulating merriment not how to write jokes that made him as funny as Robin Williams)

with that said do I think that with Cool Edit's tool box synthesizing emotion from flat affect is easy or even possible . . . nope I don't . . . personally I think it would be easier to hire Robin Williams to reread the script & then convince the client that you'd studio magic'd it (and I've done a fair amount of additive synth, a bunch of editing and fleshing out of live performance for eventual CD release)

even if currently possible it would still be quicker and cheaper to hire someone to reread (though next year their might be a Line6 Robin Williams module out . . . and what with Botox pretty soon the only expression will be electronic expression.