What do you mean ?
The non-standard game interfaces discussion thread (move, voice, vitality, etc.)
- Thread starter patsu
- Start date
What do you mean ?
specwarGP2
Regular
Well MS charges $60 for a 16gb USB thumbdrive so no one should be surprised that kinect costs $150 with no game.
(((interference)))
Veteran
Well yes, but apparently even the actual manufacturing cost for MS is $150
Bleh, the working while sitting thing is still a rumor, so I'm unconvinced. There's one fact: we're 5 months from launch and nobody knows shit till then. How's that?
Tommy McClain
The speakers placed across the kinect body
They can be used to focus on one player tracking where the voice is coming from through out the room. Or they can be used normaly and pick up as many voices as you want. You can even have each mic pick a person and track for that voice only.
You mean the mic (microphone) !
There is no speaker in the Kinect accessory.
Yes, that's the 4 mic array we referred to above. But Kinect only supports 1 (human) speaker voice recognition according to Microsoft's store specs.
They have not demoed simultaneous/multiple speaker voice recognition, have they ? 1 speaker should be sufficient for device control (and of course SP games and turn-base MP games).
The speakers placed across the kinect body
They can be used to focus on one player tracking where the voice is coming from through out the room. Or they can be used normaly and pick up as many voices as you want. You can even have each mic pick a person and track for that voice only.
You are just making things up. You cannot use each mic to track separate voices. The point of having an array of mics is to have a direction for a sound source. Each mic receives every sound (with slightly different delay).
You guys are miss-reading what is there.
You can do anything you want with those mic's if you write proper software for it. And yes I know the point for the 4 mics.
How ?
It says, "Array of 4 microphones supporting single speaker voice recognition"
Would you be able to elaborate on the mechanics ?
Sounds like Ken Kutaragi all over again... Promises, promises...
What would you want me to elaborate on ? If mics can pick up sounds and the software is working to isolate a single voice and track it through those 4 mics you can easily replicate that with multiple voices. The only problem will become processing over head. In a game like scene it , there should be processing power to spare.
We will know in nov who is right tho I guess.
Stop it already. Even if you can track multiple voices (which may be theoretically possible, but not practical at this point), you can not make each mic track separate voice with software. It's an hardware issue.
(((interference)))
Veteran
But, it's not a rumour:
http://kotaku.com/5565777/xbox-kinect-does-not-play-well-with-couch-potatoes
Device is not out yet. He seems confident that it will be working at launch. Perhaps its already working but the software isn't completely tested or ready for the show floor.
Let's be clear here - what you said with each microphone tracking one person is decidedly incorrect. That makes it sound like each is a motorised, highy directional mic! All four microphones pick up all the sounds, but with slightly different timings, and using these four different flavours fo the same sound, you can process them to isolate a particular sound. I haven't done this myself, but I believe the basic idea is you can time-shift two samples so the voice you want exactly matches which will anplify that voice, while the alternative sounds that aren't a perfect match will muddle to become noise.
For this to work, you need the sounds you're picking up to be reasonably isolated. In a room full of reverb for example, the end result would be very muddy. And if two people are sat next to each other shouting answers, that's going to pose a problem. It appears my suspicions that the size of the device are for the spacing of the mics was correct, as a greater distance between mics means great sound separation. I expect PSEye's array is too narrow to be much use, hence the fact it's not doing any sound isolation in my experience. There are still going to be major limits though, and you cannot just "do anything you want" with them, unless someone has invented a brand-new audio detection engine as sophisticated as human hearing! Voice recognition is ropey at best, and with four voices cross-talking, it'll be struggling. hence it's not surprising that the specifications state 1 voice recognition. One at a time is good enough for a quiz, but we're a long ways still from Star Trek style talking to computers.
That only confirms the gesture controls for movies will work, period. That suggests it'll work only while standing. Whether Kinect works while sitting on a couch is still unconfirmed, according to that spokesperson.
It doesn't even confirm that. MS is certain that it will, but unless they are infallible, the best laid scheme o' mice an' men gang aft agley and it may not work as they intend.
Dunno why this thread got closed sp I've opened it up again pending some other mod saying, "how dare you!," and putting an end to it.
Perhaps I was wrong with each microphone tracking a voice. But software should be able to do it
I believe they will be able to track multiple voices in quiz show type games. For each team they can simply have the team pick a person to be team leader and only that person is needed to be followed. Mabye for quick games as a way to buzz in it might not work .But the 360 itself should be more than capable of picking out multiple voices from the audio stream of 4 diffrent mics
