Game Commander and Peripheral Perfection by Thomas "AV8R" Spann
	Note too that you can import and export your saved GCA files so that you can share them with other squad members. While you can only have one FALCON4.GCA file, you create a sub directory where you save your GCA file. In doing so, you have now have the capability to have multiple versions of GCA files for the same game or application. (Yes you can have GC issue commands to other Windows based applications other than games). --- Example of a Falcon4.txt file --- The voice command name is listed first, followed by the associated keystroke(s) in braces. Gear {G} Wingman {W} Night vision goggles {N} Wheel brakes {K} Dogfight mode {D} Situational awareness {Shift+3} Change radar azimuth scans {F8} Change radar bar scans {Shift+F8} A C M boresight submode {Ctrl+F5} Request taxi {T}{4} Attack my target {W}{E}{1} Buddy spike {W}{E}{2} VOICE RECOGNITION TECHNIQUES: If I really have your attention, you have probably started to come up with the hard 'what if' questions. For example, how does GC recognize my voice versus who ever originally programmed it? Or how about what if a command that I mapped in the GC command interface isn't recognized in the game? Or what about if I speak with a thick accent or even in a different language like Chinese or Russian? These questions really are the same question and go to the heart of what makes voice recognition work. The short answer is that the user need only to use the TEACH interface to get GC to understand your language or accent on a command by command basis. This is done by repeating the command three times. Then the spoken word "Gear" is mapped to the written word "Gear," and indirectly to whatever game sequence you assign it (like the key "g"). Click to continue		Now for the long answer. I don't work for MindMaker so I can't give you their technoguy, but my swag at it is that this voice independant recognition program really doesn't care what language or accent you speak with. When the original GCA voice recognition files are scripted, the program converts the audible tones into electronic impressions or finger prints. What makes this technology an art as much as a science is how they write the software filters and digital signal processing algorithms. So when the spoken word "Gear" is taught to the program, they have balanced the recognition program so that it is smart enough to pick up that word from other words, yet not so sensitive that it has to be enunciated exactly like the way it was taught. This way the program will recognize the voice command within some reasonable tolerances from the way it originally mapped the word. So it turns out that voice recognition is phonetic in nature during the teaching process. Me262s in EAW When you issue the voice command "Gear", the program searches through its GCA voice mapping data to match it up with your programmed game commands. In turn, this game command is then mapped to the keyboard sequence which is then sent to the Windows keyboard buffer. The more similiar words sound, the more likely you are to get a confusion between commands. Thus it is to your advantage to make your commands as simple as possible, and as unique as possible. In that way you make the job easier for the program. For you technogeeks out there, you can also play with text to voice programs by clicking HERE. Go to Page Three
Copyright © 1997 - 2000 COMBATSIM.COM, INC. All Rights Reserved. Last Updated June 10th, 1999