Filthyfrankster Posted February 16, 2016 Posted February 16, 2016 I'm tired of the old fashion robotic women having few selected lines, and not being able to create more Aurora friendly voiced announcements. My suggestion? New announcer, Something like the XCOM bald fellow with "Hello Commander." I can do said voice, (Because I have the voice of a sex god.) and can throw in a pitched announcer reading if anyone wants. That way we can have more variant types of announcements, AKA, CCIA coming to visit and such.
Guest Marlon Phoenix Posted February 16, 2016 Posted February 16, 2016 Can you provide an example sample? I like the lady.
Skull132 Posted February 16, 2016 Posted February 16, 2016 Yesss, but we want an example of what Frank can pull off. I actually wouldn't mind getting new soundbites, if they're good. Gives us freedom to do more shit.
Guest Menown Posted February 16, 2016 Posted February 16, 2016 If Frank can at least make the fucking voice match the words, I'll be happy. I get all OCD when she starts talking and it doesn't match the announcement.
Conservatron Posted February 17, 2016 Posted February 17, 2016 you could do some quick samples via http://vocaroo.com/
Killerhurtz Posted February 17, 2016 Posted February 17, 2016 As a side note; I'll have to find it again, but I remember playing with a voice synthesizer software that was holy-shit good. Like, if I remember correctly it was ACTUALLY good enough to create a synthetic radio host with. But yeah, the bottom line is: I'm confident in the new TTS capabilities that I say we should consider it. I'll be back with samples.
LordFowl Posted February 17, 2016 Posted February 17, 2016 The voice shouldn't be like ol' Baldie, as cool as he is. It should be neutral, emotionless, and without intonation. Think like the black mesa Vox system in half life 1. This makes it more dynamic, as a plus.
Carver Posted February 18, 2016 Posted February 18, 2016 The voice shouldn't be like ol' Baldie, as cool as he is. It should be neutral, emotionless, and without intonation. Think like the black mesa Vox system in half life 1. This makes it more dynamic, as a plus. We already have the Vox sound files in the old code(Because Bay removed them for some unfun reason I'd imagine, wouldn't be a difficult port though). They've been there for years, and if an admin felt like it they could still declare "Johnson Inspection Day".
Killerhurtz Posted February 18, 2016 Posted February 18, 2016 OK SO. I've scoured the internet and by far the most relevant TTS voice to us is the Ivona family. We can either use it online via http://www.textaloud.com/ and recording it via Audacity, or if we want to be straight up, the license for both the software and the voice comes under a hundred dollar (permanently). Here's my samples so far (not actual lines, just stupid shit to showcase the voices): Brian (Normal, Delta ) Eric (Normal, Delta (I can probably make it better)) Marlene (German Accent)(Normal, Delta) Raveena (Indian accent)(Normal, Delta) Salli (Normal, Delta(I can also probably make it better/less noisy)) Personally I really like Salli and Eric. But let's see what Frankieboy has to give us - it would be a lot more dynamic and better if he's good at it.
XdwbX Posted February 20, 2016 Posted February 20, 2016 I like what you have, Killerhurtz. Good stuff. I'm still undecided if I prefer a natural voice that's been heavily edited, or a tts with only light editing. Any opinions on this? Here's a couple I slapped together: Departure Shuttle Called: https://clyp.it/bg2restl Departure Shuttle Docked: https://clyp.it/521wrxmv
Recommended Posts