That's a very good point, I see it now. However, I am not sure if such technology could be implemented. So basically ...


That's my opinion exactly.

The way I see it. You're saying "Why should we hire chefs to make us pizzas, when we can create a device that makes pizza by itself?". An interesting idea, but I am completely clueless on whether we do have the technology needed to make the pizza making device...AHEM! Voice synthesis, or not.

I've actually been keeping tabs on TTS technology aka voice synthesis. It's not perfect by any means, but it's come a long way, especially in the hands of such companies as http://www2.research.att.com/~ttsweb/tts/demo.php and Lernout & Hauspie. The demo page I linked to does text-to-speech on the quality level of the dialogue of M.A.R.Go.T. from Fallout 3, and I think L&H has even more advanced TTS technology.

When we have developed greatest automatic hamburger maker, i.e. graphic engines and the hardware that followed eventually, we can have high hope that if enough developers get interested in the concept, it would be achieved, and the specialized hardware might follow as well.

In the OP I have offered a solution that although might not be a slick and fast solution, but it can be a safe solution that would eventually reach the desired quality.

And also in the OP in the middle of the post, I reached the conclusion that the best solution would be a middleware, and reflected that at the end of OP, so you see, I'm agree with you that Bethesda should not waste their time with the new technology, and let's hope a group of young hot heads snatch the idea and make a middleware of it.

Edit:

Let's hope somehow it finds its way into future TES games, even if with the help of special hardware.

http://www.oddcast.com/home/demos/tts/tts_example.php?sitepal is a really good text to speech software.

exactly my thoughts.

Vision is our most important sense. Next to that lies hearing. Graphics have reached a high level and in the coming years they will become photorealistic. That happened because they severely limited our creativity. Just plain linear interactive movies could be done just with pictures, no real time interactivity could be achieved. On the other hand sound doesn't limit us that much. We can still make linear fps with voice acted dialogue. But with RPGs, with freedom of choice and a huge amount of content, voice acting reached its limits. So we can either limit content to fit the old technology or advance the technology itself. Voice synthesis is the future, as Sphagne suggested.

http://www.youtube.com/watch?v=bgxjq3UPDK0&feature=related
http://en.wikipedia.org/wiki/Speech_Synthesis_Markup_Language



That is pretty good

I must agree; far better than the demo I posted.

But both are far away from realistic

I do not see this happening within the current generation of console's time lines. Nothing to do with the technology in the consoles. Just a time frame.

I don't think voice synthesizers have come nearly far enough to be actually taken serous in a video game. If I heard any of those synthesizers that were posted in TESV I would not be able to help but fall on the ground laughing.

I think give it maybe 15-20 years, and it could be a possibility. But for now leave the synthesizers for GPS and phone use.

I think your estimate is way off base. I also think that it's not as bad as you make it out to be. Sure, each line would need to be listened to and tweaked (with an invisible "phonetic equivalent" which provides pronunciation for those lines that aren't pronounced right by the TTS engine), but then that line can easily be transferred to every voice in the game; this opens up the possibility of a multitude of different voices, not just one per race, or one per several races, but potentially hundreds of voices, enough for each NPC to have a distinct voice. This also opens up the possibility of using the player character's name in voiced dialogue.

I don't think it's necessary to wait 1 1/2 to 2 decades for this when the technology is good enough, now, to do it with a little effort - much less effort, and much less cost, than voice acting.

Those responses are what I hope not to hear from the ones who's opinion matters, just like what I posted earlier:

:glare:


Exactly!

As I have suggested in OP, each in-game line of text can be made of different threads, and one of those threads could contain the phonetics, emotion codes, lip movements, and the like.

By the way, I know that my long OP is hard to read and decipher, but I think that I have offered a solution in there that could eventually reach some passable results.

Edit:

Glad to know that I had a similarly minded friend out there. :wave:

Lol, you can say that all you want, but it doesn't change a thing. "Bare with initial responses"? You really expect a big name company to want to go along with something like that? Besides I said I wouldn't be opposed to it when voice synthesis is developed better. But right now, it does nothing to add immersion to the game, and I think people would honestly choose text over that (if voice acting was out of the picture). So I think just let the people who are working on it to keep doing it until it gets there.



You might be right, but I haven't been exactly presented with this "current technology" of voice synthesis. All I've seen so far are these things where you type in text and it gives a very artificial voice saying the line. I can't even comprehend how somebody would prefer that over a real human voice with emotion and everything. If they had to spend that much time to tweak it how are you sure that it would be worth their time more than actual voice acting? I realize the point that it would offer us a lot more lines, quests, and spoken lore (books are always there too), but would we really want to sacrifice quality for quantity? I mean that's really what it comes down to for me.

Also, my "estimate" probably was way off base. It wasn't really based off of anything more than pure guesswork, but to be honest I haven't really seen any amazing new developments with voice synthesizing.

I did not suggest that the current games which are in development should switch to voice synthesis, (I wish TESV would), but I hoped some middleware developers would work on that so that the technology was available for the next wave of the games, like TESVI, for instance.

As for currently developed deep-style RPG games, I think Morrowind's method is way better for them.

:facepalm: Sorry, I now recall you suggesting that middleware developers do that.

I still am a bit weary of it though. It still seems to me no matter what it will feel artificial compared to actual human voices.

No more so than the graphics side of things.

But I can live with that. Video games just happen to look better with artificial creations, rather than pictures of actual things. And the voice sounds better from a human, rather than artificial. Which is my opinion.

That is because we are we are spoiled, but if we had switched to synthesized voices in parallel with synthesized images, we would have bad voices to start with bad graphics and then at this point we would have acceptable to great voice synthesis along side acceptable to great 3d graphics.

But as recorded sounds was a lot easier to use than recorded images, they became the norm without thinking about the future limits of the method, just like the Year 2000 problem that caused a lot of problems and re-designs as the computer reached their original designed limits because of nearsightedness.

Now the computer generated graphic side of the games has become real efficient, we are reaching the limits of the easier compromise of the sound side of the games, (the recorded voices), and this is the time to focus on this problem, because those shackles are starting to become too tight as our prisoners are growing steadily.

We have to act before our prisoners suffocate, (did I mention that those shackles are around the neck of our quests and AI?) :whistling:

I want to emphasize this.

Yes! People are spoiled!

Remember back when people accepted the fact that a video game is just animations on a screen and didn't expect developers to fool them into believing they're inside a movie? Back then, the only thing expected of a video game was to be FUN.

Now, it's just a showcase for pretty technology.

Wouldn't you prefer a game with more stuff to do than to just look at and listen to? The voice synthesizer is a way to accomplish this, potentially.

Aside from that, it's never gonna start sounding better until people start practicing with it.

All that being said, however... since the voice synthesizer doesn't seem to add that much more realism than dialogue in text-form. Why don't we just keep game dialogue mostly pure text until the technology catches up?

I think making an RPG all voiced was jumping the gun. And now, people who are used to Oblivion and Fallout 3 are uneasy about the idea of dialogue without voice-overs, because they're not used to it. In other words, they're spoiled, as you said.

They're afraid it'll make the game less fun, but if they could just experience an RPG with text-based dialogue and see for themselves how much more there is to do, and how much more freedom they get in the dialogue, and how much more immersive the culture and lore of the world becomes (as opposed to just the novel sights and sounds), there would be no question that fully-voiced dialogue ruins RPGs.

Indeed that's what I would go with. Just let the technology catch up, and until then I would say don't try to do full voice acting. I mean when you have a game like WoW that doesn't really have any voice acting, and is still so popular I don't get why people are thinking it's impossible these days to make a game without full voice acting. I mean the quests do tend to svck in WoW, but it's still along the same lines of an open world with lots of quests. The way I see it is games that have a linear, movie type story would probably need it, and it wouldn't effect any of the development of the game, but with games like the elder scrolls which shouldn't play out in a single plot-line like a movie should be able to have more options, and I think full voice acting really effects that in a negative way.

Yes, I say that if they can not use synthesized voices yet, they are better off with Morrowind style voice acting.

Back then I did not think that Morrowind lacked anything in that regard, and even now when I play it, I dont miss anything. It seems so natural, or maybe because I expect such a thing from Morrowind, that it seems natural, so it all goes with expectation, and as I have said earlier, we are spoiled and expect more and more from games.

But if they revert to Morrowind style of Voice acting to wait for the voice synthesis to mature, they would gain the time and resources that would go into the voice acting process and they would remove the barrier that voice acting forces on imaginative new ideas for quest developments, AI, and prettily any other system that has to deal with dialogs, or voiced texts in any way.

In Morrowind method, as the pressure on each voice actor was minimal, they could find more voice actors that could do the job that was needed from them to be done, so they had more different voices available to chose from, and the end result was more varied in-game voices than oblivion.

But in voice synthesis method, or my suggested voice building method, we can potentially have different voice for each freaking npc out there, and as said before we can finally have npc'c that call our in-game characters by their names, and hear them do it, and more important than all, we can start to hear what our character says to others, with our selected voice, from a sea of available voices to choose, plus the tweaked pitch and effects to make that voice unique.

And if that barrier is removes, I can guarantee that you would see a revolution in what quests could do and what is expected from AI, believe me, and I say that playing without voice or with sub-par voice synthesis with new revolutionary styles of quests and AI, would be so much better that current, dumbed down style of quests and limited AI, that they could not even be compared with each other.

Edit:

Yes, yes, yes, yes, and exactly!


But as you have just wrote before: it's never gonna start sounding better until people start practicing with it.
So let's hope some developers jump the gun in this direction, and start practicing with it.


My exact sentiment about the matters.

Now, I'm going to write about the ideas that I have been working on, about new styles of quest development, and after that about new character AI, as I fill in the details, as if the barrier is removed, so that jumping the gun in this situation looks even more attractive when people see the true potential of RPGs regarding quests and AI, although the quests could be done for adventures as well, and the AI part is universal.

Something they did in Bloodmoon, to respond to the blandness of Morrowind. And it was also done in Baldur's Gate, I remember.

Most of the characters say the first line of their dialogue, which sort of pulls you into it. Of course, the extra time it takes to lip synch is something that didn't need to be considered before.

It's good to see I'm not the only one with these sentiments towards game content these days, by the way.

I wouldn't mind an optional TTS with some voice acting in the next Elder Scrolls; you'd be able to have just the TTS, just the text, or both; I'd use both, personally, since I use subtitles for TES IV and FO3 as it is.

I'd love to see TTS developed to a point where it could be used in the ES series. That said, I wouldn't want it added to ES unless it sounded natural, otherwise I would just.... cry. The only problem is, I can't really foresee any application for TTS in games outside of Bethesda-esque RPGs, as I think most other titles are pretty content using VA. Until then, I think our best hope for the immediate future is better data/audio compression, and, for the love of the divines, a second disc dedicated to audio files.

Agreed with the OP, especially the RPG-Cancer part. Oblivion's system didn't make my game more immersive AT ALL. Games which are fully voice acted I just click before they can even begin the sentence. I read far faster than their talking speed. IMO, fully voiced dialogue to important NPCs, text-only for random NPCs. Like Fallout 1 and 2, only with the ability to hear muttering and short phrases when you go near NPCs.

Also, don't hire people who charge $10,000 for 1 line of text please. :rolleyes:

"Speak quickly, outlander!" English, Catherine UK. It's the closest from the list in terms of fluid realism.