Spoken Briefings?

[PiHalbe]

Hi, I just wondered if there is any possibility to include spoken briefings in BfW. If it is possible, I guess Speex codec would be best, right?

And if it is possible, is there any interest to do it? Concerning users as well as speakers.

(And yes, I would be willing to work on this! At least for the German part.)

[zookeeper]

Technically possible, but it won't happen because we wouldn't get good enough speakers to do the actual voices and maintain them when the texts change or new bits are added.

This (or some variation) been discussed before several times, you should find some threads about it here in the music & sound forum. There's no need to repeat the same old discussions in this one.

[Jetrel]

Technically possible, but it won't happen because we wouldn't get good enough speakers to do the actual voices and maintain them when the texts change or new bits are added.

This (or some variation) been discussed before several times, you should find some threads about it here in the music & sound forum. There's no need to repeat the same old discussions in this one.

It's the underlined section that kills us.

Not only is it insanely hard to find voice-actors that don't suck; not only is it a herculean effort to make a separate recording for every single language the game is localized to (true whether different actors or the same actors do it), but to top it all off, we're quite human, and we have a habit of revising text, or even entire levels of campaigns, because, quite frankly, we've written quite a bit of crap. The thought of permanently 'freezing' any of our textual content just ain't gonna fly, because at best, we're inevitable gonna have a few small things to change about it.


Now, if someone had a good, open-source text-to-speech synthesizer, we'd be in candyland. That would be the perfect solution for us - I've seen games use the Mac OS's speech synthesizer before, and the idea works in practice. Someone should start a project to create such a thing, if it hasn't been done already.

[Viliam]

Just a sidenote: It is not necessary to make speech in all supported languages. As long as the text is displayed on screen together with speech, the English speech with localized written text would be acceptable -- it would be like movie with subtitles. Translated speech would be better, of course, but not necessary.

Speech synthesizer would be great, surely. Then there is a problem if it would work for all languages or only for English. Again, this could be solved by using English speech with localized texts when the target language synthesis does not work.

[vonHalenbach]

This seems to be a topic, which seems to come up again and again, because it seems so tempting for anybody to try voiceacting. It is no simple job. It is not done with talking a few sentences into a microphone and most of us, don't have a professional studio at home.

The idea with text-to-speech is certainly no option, because it produces not good enough output, to be usable for a game. I don't want spoken text without intonation in this game.

If someone really wants to do some sentences of his or her favourite campaign, please do it and upload it here as .ogg for our amusement.

We will have a good laugh at your expense.

EDIT:

Here is an example of text-to-speech done with a free tool. Rename the file to end with .ogg (it is not packed as .zip)

[hiro hito]

What about your idea of an original language you post here?

no accent/prononciation/translation problem.....

[scienceguy8]

You can make the speech from a text to speech generator sound human, but it will take a lot of work that most would be unwilling to devote. First, the text would have to be written in such a way that the computer would speak it correctly. For instance, if you type in my name into Microsoft Word and use its text to speech toolbar, it sounds like Anthne instead of Anthony. After the text is turned to speech and recorded as an audio file, the file would be uploaded into a sound editing program and intonation would be added by adjusting pitch and the space between words. After all this, the file would be added to Wesnoth, possibly with background music, to play at the appropriate time.

An open-source text to speech program could be modified to bring more life to spoken words by providing it with a pitch guide that runs alongside the text to be spoken, but even then it would mean more work because the intonation map would have to be changed whenever the text is also changed. Plus it would mean that one of Wesnoth's most endearing features, its ability to run on low-end machines, may be compromised.

[vonHalenbach]

What about your idea of an original language you post here?

no accent/prononciation/translation problem.....

Even more difficult for text-to-speech-systems, because they are trained to speak english texts. This was just a bad idea also, to make a completly new language. I won't build it further. You better forget, that it existed.

[PiHalbe]

and maintain them when the texts change or new bits are added.

OK, that is true. I have not been into campaign-design and had in mind that at least HttT, TtB and tSG were frozen already (or, sort of frozen at least, having only minor (gameplay-) changes). If this is not true, well ... than this is it, right.

Prove me wrong, but I think most of the campaigns do with four to six major characters and a dozen minor characters. It should not be that hard to find good speakers for them. You do not need a new speaker for every single horseman, at least that's my feeling.

Plus it would mean that one of Wesnoth's most endearing features, its ability to run on low-end machines, may be compromised.

I guess it would be possible to include the audio-files in an extra package (has been discussed earlier for unit-responses). So it would be up to the gamers choice to have them or not.

[aManFromMars]

There is BASICally a confusion here between free as in speech, and FLOSS as in software/BEer. Speech is not Free, and actors in vocal virtual SYNthesis cost $$$ Micromanaging the Macro of Human Global Perception. And IT does this QuITe Simply by forwarding ESPecially crafted IDeas/Zero dDay Trades, Openly [hiding in full sight, as here, in an very public Internet message] which can be seen as a precursor to target/energise ITs Utilities and those who sit on/at their Controls.

And it is pretty obvious why all of that is happening. Vocal Virtualisation is the Intellectual Server Space for Hypervizars.... AI Programmers Heaven. Campaign freeze is backtracked by pulling leading-edge features in Heir to the thrONE, an Enigmatic Prize Submission in Finely Honed, Machine Intelligence. And an InterNetworking Global/Universal PlaySpace for the ARGonaut.

Voice playback allures/allows and it for nothing less than the MetaMorphosis of that other Alternate Reality, the Underground, to exercise an Overarching IT Control of Reality Plays, Virtually.

[Darth Jordius]

Technically possible, but it won't happen because we wouldn't get good enough speakers to do the actual voices and maintain them when the texts change or new bits are added.

This (or some variation) been discussed before several times, you should find some threads about it here in the music & sound forum. There's no need to repeat the same old discussions in this one.

It's the underlined section that kills us.

Not only is it insanely hard to find voice-actors that don't suck; not only is it a herculean effort to make a separate recording for every single language the game is localized to (true whether different actors or the same actors do it), but to top it all off, we're quite human, and we have a habit of revising text, or even entire levels of campaigns, because, quite frankly, we've written quite a bit of crap. The thought of permanently 'freezing' any of our textual content just ain't gonna fly, because at best, we're inevitable gonna have a few small things to change about it.


Now, if someone had a good, open-source text-to-speech synthesizer, we'd be in candyland. That would be the perfect solution for us - I've seen games use the Mac OS's speech synthesizer before, and the idea works in practice. Someone should start a project to create such a thing, if it hasn't been done already.

You've forgotten all about Microsoft Sam!

[PiHalbe]

Whatever that should mean concerning spoken briefings. At least for my part, I have not been talking about synthesis but recording.

hiding in full sight, as here, in an very public Internet message

I guess there is a lot hidden in my sight here ... I just don't get your point. Or anything at all.

[West]

There is BASICally a confusion here between free as in speech, and FLOSS as in software/BEer. Speech is not Free, and actors in vocal virtual SYNthesis cost $$$ Micromanaging the Macro of Human Global Perception. And IT does this QuITe Simply by forwarding ESPecially crafted IDeas/Zero dDay Trades, Openly [hiding in full sight, as here, in an very public Internet message] which can be seen as a precursor to target/energise ITs Utilities and those who sit on/at their Controls.

And it is pretty obvious why all of that is happening. Vocal Virtualisation is the Intellectual Server Space for Hypervizars.... AI Programmers Heaven. Campaign freeze is backtracked by pulling leading-edge features in Heir to the thrONE, an Enigmatic Prize Submission in Finely Honed, Machine Intelligence. And an InterNetworking Global/Universal PlaySpace for the ARGonaut.

Voice playback allures/allows and it for nothing less than the MetaMorphosis of that other Alternate Reality, the Underground, to exercise an Overarching IT Control of Reality Plays, Virtually.

Are you on drugs?

[turin]

I would say "spambot", but what is it spamming?

[PiHalbe]

So what is it about the spoken briefings now? I think at least campaigns like Tutorial, HttT and TtB are quite final and will face only minor changes (at least as far as story is concerned). So it should be possible to do.

I tried some exemplary recordings yesterday and I think it is possible to find people willing and able to give their voice to our heroes and villains from Wesnoth. I already managed to put the recordings inside the scenario to get some enhanced audio-experience.

I have a Samson C01U microphone and feel it is quite suited (and affordable).