Okay I had this idea this morning so dont flame me for not having thought it all out. There have been several posts and topics with the idea of using voice recognition in spellcasting. Obviously very cool idea in MMORPGs, the problem always being that it takes WAY too much bandwidth. I''m gonna try to explain my idea. The microphone would still be used and you would still say the keywords or whatever for the spell. But instead of transmitting your voice over the line you could have the client determine if what you spoke correctly activates what you have access to. If it does it tells the game server that you casted a spell. It tells the server what spell, and the server tells those around or in the battle you casted that spell. Now on everyone''s client machine would be voice/audio files that came with the game. These files would would have a prerecorded spell chants. So when the client is told you or someone else casted a spell, its simply told what spell and the client plays the prerecorded audio file. So even though you don''t actually send speech over the wire, you get to hear it being chanted anyways. Did I explain that clearly? Hope I did. I know it wouldn''t be as great as actual speech but it does add something to the game. The downside I see immediatly is that if theres a lot of spells then the players might have to instal large amounts of data. Comments, ideas, questions, suggestions? I want to hear what you think.

Actually, you have an even better idea tucked neatly away in the magic spiel. Voice recognition can be used to convert speech to text, and that text can be sent to other uses in real-time (ala, chats in ICQ). So basically you get real-time speech without any bandwidth concerns since the clients are doing all the work. You can also limit the through-put so someone doesn''t spam the server by rambling nonsense as fast as his lips can move. Maybe speak the phrase and it appears on-screen, and you hit a key to send it. I think that would super-enhance existing games, and you wouldn''t need to send wave data.

Great idea, Hydra.

This sounds like a super-cool ideea. I''m already loughing my ass thinking about what would a player of your game look like: shouting fourious orcish curses at the screen. The only thing missing are some special position sensitive gloves that would also force the player to make certain gestures to enable a spell.

The idea could very well be used in WWII games: make the player shout orders in German or Russian

Have any of you tried speech recognition lately? Office XP for instance... you should

I''m afraid I''ll burst your bubble to some extent, but it''s not a viable solution for some time...

Michael Heilemann
---------------------------
Designer on Singularity - Sysop at Nerve Impulse
Let us never allow ourselves the sin to forget our dreams!

quote: Original post by TechnoHydra
I know it wouldn''t be as great as actual speech but it does add something to the game.

Actually I think it would be better as with using actual speech there''s potentials for spoiling the atmosphere of the game - You are in the middle of a dangerous battle and then all of a sudden you hear a loud "screeeech" sound as one of the players has trouble with their mics. One sure way to bring you crashing back to reality.

Also, most peoples voices would sound stupid when they are chanting due to the lack of the authentic medieval accent, which again would spoil the atmosphere of the game. And what if you''ve got a guy playing a female wizard...

In regard to Tom:
I had a similar idea a few days ago. I thought that you could uses speech recognition to convert the speech to text. This would be sent to the server which would relay it to the clients. The client apps would then convert the text back to speech using a speech synthesis.

Players would be able to chose their voices so this would prevent squeaky barbarians and suspiciously husky barmaids from spoiling the atmosphere The synthesised speech would be sound authentic to the time period.

It could even go as far as converting the text to authentic medieval style old English before it''s sent. It could be used to get around the language barrier too as the text could be converted to what ever language the client users desire.

I know that this is a little out of reach at the moment due to the current state of speech synthesis and language converters but it''s something to think about for the future.

- Kaijin

"If you find a job that you love you''ll never have to work again."