After some great impressions from around the internet, I thought that maybe, just maybe the Kinect 2.0 would be a major improvement over the original. To my disappointment, it hasn't been. I hooked up my TV to the Xbox One without issue, however, the Kinect consistently misinterpreted my speech when trying to change channels and watch shows; even with another couple calibrations, I still couldn't get the thing to accurately interpret my speech more than 60% of the time. General other functionality has been problematic as well. I had my wife attempt to use it with no luck either. I use google now constantly which has never been a problem of me but it looks like the Kinect just doesn't work. I've seen several live streams with similar issues. I think MS over promised again and despite some people claiming it works almost every time, I don't buy it. Give it a few weeks and we'll see what people think.
Already Disconnected the Kinect
Hearing this sort of stuff in regards to the Kinect, I'm left wondering why Microsoft thought it would be a good idea to produce a second one in the first place. While I hope some of these issues can be possibly fixed in the coming months, I keep wondering why?
@machofantastico: We can only speculate, but I think it's the same reason Sony reused the playstation eye for playstation move. They've spent all this money into developing the tech, so they can't bring themselves to completely cut the cord and abandon it.
I assume the mindset of the xbox one developers were along the lines of: "The Kinect was a great success, we already have the architecture established, just refine it and put it out there again!" It was likely one of the cheapest parts when they were developing the console, and some suits up in the corporate chain figured it would be great for profit margins if it was a mandatory purchase, to make sure it's put in as many homes as possible.
If it can't deal with different accents perhaps they shouldn't force it onto people and instead make it optional.
@cabbages: That's a shame, I'm Northern Irish and I've had very little problems with it picking up my accent, it seems to work most of the time when its just me in the room talking to it.
I had issues with changing to a couple specific channels, but after adding those channels to the favorites category as Jeff suggested on the Bombcast, it has been smooth sailing. You do need to learn the proper syntax for the voice commands, as well as a particular rhythm. Once I got the cadence down, the voice commands work really well.
@barrock: I read somewhere (Kotaku?) that if you're already watching tv and then say that, it'll turn off your tv and cable box because its just hitting the power button for them? That seems like a huge bummer, but I guess you can get around it by just turning on the Xbox with a controller? Or does it still do the same thing?
There are definitely issues with it taking the odd second try to get it to recognize perfectly, but I'm actually amused by the things I can accomplish while doing anything which requires me to not be holding a controller (specifically, cooking). It still might not be living up to its full potential, but man, now that the hardware is there all it means is that the software needs to catch up.
I will 100% agree with the sentiment of it not necessarily working with all languages and accents, however. I'm lucky enough in that department to be the vanilla, American-English speaking person which Redmond obviously put the most resources into.
huh have had no issues with the voice commands probably cause i sit a couple of feet away from it.
I'm the same, you have to speak a little louder the farther you move back but I think the biggest problem people have is forgetting the Kinect uses specific commands and is still limited in some regards. Regardless, controlling my Xbox through Kinect is FAR more convenient than without and I find it a chore now to cycle through menu's with my controller alone.
Well if it's anything like Siri, then it's got no chance. Last time I used Siri to record a note it started calling up my cousin, thing is useless.
Even if it doesn't work the greatest I don't really see a reason to disconnect it. It's worth being on for the face recognition sign-in alone.
It takes all of 5 seconds to sight in manually and no one else in the house will use it except for me. Other than the voice control issues, we also watch DVR the majority of the time and the Xbox One is worthless for that.
@machofantastico: We can only speculate, but I think it's the same reason Sony reused the playstation eye for playstation move. They've spent all this money into developing the tech, so they can't bring themselves to completely cut the cord and abandon it.
I assume the mindset of the xbox one developers were along the lines of: "The Kinect was a great success, we already have the architecture established, just refine it and put it out there again!" It was likely one of the cheapest parts when they were developing the console, and some suits up in the corporate chain figured it would be great for profit margins if it was a mandatory purchase, to make sure it's put in as many homes as possible.
Wow, cynical. Maybe Microsoft just legitimately thought it was a good idea and a cool thing in general. Which hey, when it works, it totally is both of those things.
Some people seem to be having a lot more luck than others. I'd guess a lot of that is issues with accents, as some in this thread have already guessed.
This is why I want a kinect-less version. Watching the live stream showed me all I needed to see. It doesn't look intuitive or fun to use. And it's intrusive to boot.
@horseman6: Yah but why disconnect it? No offense, maybe it's not your intent, but it just seems like trying to sensensualize a non-issue.
Does it matter if Microsoft even had a 100% reliable spoken command system?
The irony is that commanding inanimate objects or people in you home is not something we are used to. Its a fun diversion at first, but after awhile people become self conscious and bored by having to speak at all. We don't realize it but most people don't talk to issue most commands. We nod, we quirk and eyebrow, we move five parts of teh face to say yes or no. So much of what we do in non-verbal, very subtle, and a matter of context. We issue commands in microseconds, but often those commands come after minutes of other non-verbal actions. And it is teh fact that commands are contextual over time that make ist impossible for Kinect to even be serviceable for gesture or even micro facial commands. Everything about New Kinect is still MACRO; a strong clear voice is needed for voice commands and strong very defined hand, arm, leg, and head gestures are needed for visual commands.
It is not even a case of learning to speak clearly or guesture a certain way, it that doing so is tedious to humans when so much we do with other people, pets, or even devices like keyboards, mice, or touch screens can be more subtle. It is also a issue of social grace, there is little polite interactivity in talking in a loud clear voice to nobody just to turn down you TV's volume or to change between TV and a game. The people in you life will eventually tired of you saying "Xbox do this" or "Xbox do that" when you could just press three fucking buttons and STFU. If your not a dick, you will probably tire of your own voice as well.
@monkeyking1969: If this were a MSFT-only thing, I could understand your rant. But, like it or not, voice control is the new "it" thing, and I don't think it's going anywhere any time soon. All of the major TV providers (at least in my are..Comcast, DirectTV, etc) are advertising voice-controlled boxes...HARD. Apple and Google are clearly on-board. We probably all need to get used to human beings talking to machines..
@oginor: Seems like a forced "it" thing. Much like 3D movies.
More like diskinected, am I right?
Hey, guys?
WINNER!
I am from Kentucky and have a little Southern drawl. Works perfectly fine for me. Speaking slow and casually seeme to deliver 95% accuracy for me. I love it.
The voice control stuff is working at about 90% accuracy for me. Being able to fast forward, play, and pause my TiVo from the kitchen(small apartment) with my voice is pretty helpful. Have tried out the motion controls with Kinect Sports Rivals Preseason and Fighter Within so far. The first requires finesse and small movements, and the second is a fun flail fest. I'm keeping mine hooked up.
Some people are blissfully unaware of how poorly they enunciate their words. Unfortunately, trying out the new X1 and Kinect is a very poor time to learn about this, especially since the Kinect may to be blame for much of the recognition problems anyway.
I would agree with you, except that's not the issue. Google Now works very well but this just doesn't. Anyways, I plugged it in again because my wife wanted to Skype her family. That worked pretty well. I also tried to use it again, just not the TV stuff. I'm getting maybe 80-90% accuracy now but you really really have to slow your speech for it to understand anything. This is far from the Star Trek computer that we all should desire, this is more like a Star Trek Computer with processor damage and bad programming. It's worse than Siri, and that's saying something.
Wow, cynical. Maybe Microsoft just legitimately thought it was a good idea and a cool thing in general. Which hey, when it works, it totally is both of those things.
Some people seem to be having a lot more luck than others. I'd guess a lot of that is issues with accents, as some in this thread have already guessed.
I consider my stance more along the lines of a realist than a cynic, I never downplayed the Kinect or Microsoft, just stated the most likely reason why it's still around.
The point still stands that the Kinect's a mandatory buy-in, and people couldn't tell their respective retailers "No thank you! I'll buy one when I have incentive to later!"
Kinect is hit and miss for me.
When I first set up the system, I was honestly amazed that I ran through all of the app install/update process almost exclusively with voice while I was working around the house on other tasks. The only exceptions where when I had to put in a password or activation code to a website. That was a "next gen" type of experience that I was stoked to see how it would work in games and with general media use.
However, when I actually started using it with games and media, it's accuracy fell apart and I found myself constantly misunderstood or unheard at all. I have figured out a couple of things that has brought the accuracy and utility way up.
First, I had set up the Kinect right in front of my TV on top of the base for the TV itself. I have the speakers enabled on the TV so the audio coming through the TV is causing some vibration or noise issue in the TV base that was breaking the Kinect noise filter. As soon as I moved the Kinect to a shelf above the TV, the recognition rate went way up.
Secondly, the typical unnatural voice pitch that I've adapted to for other speech recognition systems actually performs worse with Kinect. I have to remind myself to speak in a normal, conversational tone, but with some slightly longer breaks between words. I like that it can be more conversational, but it's a reflex for me to respond to a failed voice recognition with the stilted voice that has worked with other voice recognition software. The slightly of typical conversation cadence pauses are getting easier after a couple days using the system, but they are at weird spots in the overall control flow so it is definitely something that requires more getting used to than I prrefer.
There are some pretty major gotchas that I've found. Kinect does not do well picking out commands from an active conversation. That's not too bad unless you have a couple friends over or someone is on a phone conversation or something of that sort. It does really badly when a podcast is playing on my tablet next to me while I play.
The new Kinect still feels like a huge improvement over the previous one and I really like how well integrated it is with nearly all of the system, but they still have a ways to go to be up to par with Google Now which for me right now is the gold standard for voice recognition.
Being able to say "Xbox On" and the console/tv/cable box all turning on at once in response is reason enough to keep it plugged in.
Or, you know, you could just press the Xbox guide button? Same level of effort with less frustration.
Does that work as well? Neat. I haven't been frustrated with the voice stuff when it comes to turning it on and off. I walk in the room, say "Xbox On" and by the time I sit at my desk where my controller usually is everything is turned on and waiting.
Being able to say "Xbox On" and the console/tv/cable box all turning on at once in response is reason enough to keep it plugged in.
Or, you know, you could just press the Xbox guide button? Same level of effort with less frustration.
Does that work as well? Neat. I haven't been frustrated with the voice stuff when it comes to turning it on and off. I walk in the room, say "Xbox On" and by the time I sit at my desk where my controller usually is everything is turned on and waiting.
If it worked on the 360, I see no reason why using the guide button wouldn't power on the Xbox One. Microsoft has to know at this point that not everyone is willing or able to use voice commands.
Please Log In to post.
Log in to comment