Seeing all the options players are being given in some of todays video games in customizing the appearence of their avatar, very few if any ineractive media appear to give the ability to customize the voice of your protaginist.
From a production standpoint, it is obsviously cost prohibitive to record a dozen different voice overs. But shouldn't it be possible, with current generation hardware, to give the ability to change some fairly basic qualities of a pre recorded voice over? i.e. pitch, timbre, tonal quality etc.?
Something as complecated as modifiying the enuenciation of a prerecorded voice (like changing an accent) is obviously still pretty far out and would take a crazy amount of computing power, but to have a couple pre recorded voices (one male and one female) and change some simple qualities during playback seems totally doable. Even older model Creative Labs offerings may be able to do some of it in hardware, and if not in software. (Though I'm not sure about the current generataion of consoles).
Hell, it would be pretty cool to hear how your character changes during character creation as the post processing takes in account the change in timre to your characters voice as you modify the facial structure and change the structure of the skull and nasel canals. And though not applicible to ME3, even adding a reverb or slight echo for a non-human or mechanically assisted voice on the fly.
Yeah, it could get kinda silly with people making Femshep sound like she's huffing helium, but it would give a level of customization that I have not seen er...heard before in a video game and completely doable from a technical standpoint as far as I can tell.
Querry to Audio Devs-Voice Customization in Video Games/ME
Débuté par
streamlock
, juin 08 2011 05:46
Retour en haut
#2
Posté 08 juin 2011 - 05:55
Walker White
-
- Members
- 933 messages
Sound rendering is in its infancy. Right now, we can render things that you would normally foley, like items breaking or falling. And this itself is groundbreaking research (it is one of the NSF projects that a right-wing senator railed against as a waste of money).
As for speech generation, we have can do some minor audio filtering/post processing. But it still has to start out as recorded, or at least as a bank of phonemes. And our ability to modify it is very restricted.
As for speech generation, we have can do some minor audio filtering/post processing. But it still has to start out as recorded, or at least as a bank of phonemes. And our ability to modify it is very restricted.
#3
Posté 08 juin 2011 - 06:05
streamlock
-
- Members
- 668 messages
Walker White wrote...
As for speech generation, we have can do some minor audio filtering/post processing. But it still has to start out as recorded, or at least as a bank of phonemes. And our ability to modify it is very restricted.
Correct, I'm not talking about actual speech generation, but in game post processing of pre-recorded voices only.
#4
Posté 08 juin 2011 - 06:39
Walker White
-
- Members
- 933 messages
streamlock wrote...
Correct, I'm not talking about actual speech generation, but in game post processing of pre-recorded voices only.
I think it would sound very artificial, and you are likely to lose some emotional inflection (cue Mark Meer jokes here).
