Forum - View topicNEWS: Speech Synthesis Software for Anime Announced
Goto page 1, 2, 3 Next Note: this is the discussion thread for this article |
Author | Message | |||||||
---|---|---|---|---|---|---|---|---|
10円
Posts: 605 |
|
|||||||
Will ANN PLEASE start giving us names to go along with these 'news stories?' Seriously, I want to know who picked this would be PRESS RELEASE and tried to pass it off as real news. This thing apparently hasn't even been used by anyone yet. Talk about jumping the gun. How about some news we can actually use?
|
||||||||
Josh7289
Posts: 1252 |
|
|||||||
Not good. People losing jobs is not a good thing, especially when it's something most fans would never really want.
|
||||||||
aluria
Posts: 367 Location: New Westminster, B.C., Canada |
|
|||||||
I know, who would want to listen to computer generated voices expressing emotions? Somehow, I don't think I'd feel for the character. Its the voice actor/actress that makes the character real.
|
||||||||
Richard J.
Posts: 3367 Location: Sic Semper Tyrannis. |
|
|||||||
I don't know 10円, this story could potentially have a pretty nasty affect on the anime industry. It's not really that big of a deal now but it could have a major impact if it turned out the voices generated were at least comperable to minor VA's work. I'm not seeing this replacing major VAs any time soon, but low level, few line characters could potentially be voiced by machines if this software actually generates half-way decent voices. (Sort of like how CGI has largely replaced the practice of using mass numbers of extras for large battle scenes in movies.)
Potentially, that could mean a massive decline in entry-level VA work in Japan and, if it catches on over there, you know someone will start doing it over here. Anything to keep a few extra bucks right? At the very least, it could provide people who would otherwise have never been able to afford to make an anime the oportunity to do so, which could mean new creative forces entering the arena. I'm not saying you're wrong about this story, I'm just saying I can see how someone would think it's important enough to put up. Besides, it's better to have the opportunity to say "that's silly and pointless" rather than having to say at some point later on "why weren't we told about this?" It's always better to know and be able to judge for yourself, at least I think so. By the way, I do agree that ANN should include a name with their news articles. Like I said above, it's good to know stuff. |
||||||||
Egan Loo
Posts: 1362 |
|
|||||||
While I am not the writer of the article and thus cannot speak for the writer, here are some examples of FineSpeech's narration:
Sample Sample Sample There is a form to submit your own short message (registration required) and have it read back as seven kinds of male and female voices at three speed options and three pitch levels. There are more options, but not in the demo. Form |
||||||||
Splitter
Posts: 1276 Location: Knockin' on Heaven's Door |
|
|||||||
I pray to God this does not catch on.
|
||||||||
Deltakiral
Posts: 3338 Location: Glendora, CA (Avatar Hei from Darker than BLACK) |
|
|||||||
I'd don't have a problem with this software as along as it's used for Narratorion and such. Sorry but use real people for voice acting, and that includes extras who only have a few lines in the entire series.
|
||||||||
daedelus
Posts: 743 Location: Texas City, TX (ajd: 6/11/05) |
|
|||||||
While the two female samples sounded fairly realistic, the male sample totally sounded synthesized.
Perhaps it could be used for "train arrival announcements" or in a control room or the bridge of a ship for background chatter. I don't believe it will replace primary or secondary characters anytime soon. At least, I hope not. |
||||||||
BleuVII
Posts: 672 Location: Tokorozawa, Japan |
|
|||||||
I think this could be a good thing, as long as it is used SPARINGLY!
Imagine though, if every voice in a video game was spoken. Right now, budgets don't allow that. But with this, it could work. However, I see it working a lot better for Japanese than for English, purely because of phonics. Japanese sticks to phonetic rules a LOT better than English. |
||||||||
10円
Posts: 605 |
|
|||||||
I'm rather doubtful of that. If we're truly interested in sideline issues with the potential to do harm to those who work in the anime industry then why not show concern for more obvious factors such as rampant piracy? As it is, relatively few folks seem to think that's going to do any lasting harm to the global anime market. By comparison this little bit of currently untested synthesis code is probably going to have an extremely minor impact, even assuming it gets any use at all. But hey, maybe the Engrish dialog will actually improve a bit if this software solution becomes common place.
Absolutely, and this is one of the reasons I can't stand The Economist. Not all journalists are equal. Seeing who covers what over time allows us to get a feel for their abilities. Unlike the aforementioned periodical, I don't seriously think ANN is really trying to hide anything but at the same time I'm still at a loss to understand why even pennames are not given with each news story. |
||||||||
Egan Loo
Posts: 1362 |
|
|||||||
Anime News Network on piracy
Konami has been using it in Japanese arcades since 2005 for narration and speech that change each time a new game starts. That's probably why the female voices may sound familiar to someone who has been in a Japanese game center recently. What's new are the software hooks between FineSpeech and the applications developed by Celsys. RETAS! and other Celsys software have been used by Sunrise on anime such as Escaflowne, Cowboy Bebop, and Steamboy; Toei Animation since 1997 (Precure, One Piece), TMS (Detective Conan), Pierrot (Bleach, Naruto), BONES (Wolf's Rain, Eureka 7), and several other studios in Japan.
This software is designed for Japanese voice patterns—I haven't heard it used for English, if it can be at all. (I am not the original article writer, but I did a little exploring.) |
||||||||
enurtsol
Posts: 14893 |
|
|||||||
Heh, it's not like anime lips (and faces) are animated all that well anyways!
|
||||||||
KyuuA4
Posts: 1361 Location: America, where anime and manga can be made |
|
|||||||
If it can be done it Japanese, it can be done in English. Yet, unless the voice intended is to sound like Stephen Hawking, y'can't replace the quality of real voice acting. |
||||||||
oost_the_great
Posts: 6 |
|
|||||||
Maybe it's just me.. But the voice acting is a big deal!
On top of the fact, in Japan, (unless I'm mistaken) most fans follow and watch specific shows because of the voice actors involved in the shows. This just seems like yet another step to brining down the quality in order to increase the quantity. *Sigh* |
||||||||
Beatdigga
Posts: 4631 Location: New York |
|
|||||||
The only context I could see this working in is with one-line background characters and extras. That's it.
|
||||||||
All times are GMT - 5 Hours |
||
|
Powered by phpBB © 2001, 2005 phpBB Group