Business Approach Comparison Sony PS4 and Microsoft Xbox
- Thread starter BRiT
- Start date
Sorry but I don't share your confidence.. Kinect was just little toy that could be used for simple games that were strictly full body dance or sport oriented. I have had Kinect1 since beta and while it was fine for Kinect sports it sucked for any broader scale. Voice (commands), which are still not supported here in Finland (not even english) suffered from low quality mics (normal speaking levels from 3.5-4m from kinect was no-no..)
Kinect2, weight distribution and muscle force are just cheap tricks that can be easily calculated from Kinect1 skeleton. Thump and forefinger seems to shake like some limb parts with Kinect1, so accuracy is quite weak for real games.
Overall, Kinect2 does not change foundation from how Kinect was/can be used in games.. so it will still used finely in similar games as before, nothing more.
I think most Kinects just gather dust at the moment, me and few of my friend that have Kinect, have not used it much if at all since it was launched. I was very disappointed when MS announced that Kinect2 is force fed with x1.
IMO, x1 would have been better off without Kinect2, lower price would have been positioned x1 little closer to ps4...
And one of the issues is that some people want to control their devices quietly. Especially during the night. In addition not everyone like using their vocals to do everything on a system.
It introduces some benefits but it doesnt get rid of the other methods of control. Its an addition not a replacement
My Kinect has also been collecting dust since launch. Waste of my money One of the worst purchases I have ever done. There was an initial excitement which got very fast worn off.
Despite that I think there will be potential to integrate more meaningful uses for K2 combined with the controller. But Kinect will only be a tiny companion. I dont think there will be many games that will use all its features as promised. Except some gimmicky casual games probably.
I find it sad that Sony wont pack their camera with the console though as I think THAT solution was superior for core games. The light bar on the controller and the gyroscopes, make the controller a powerful camera accessory. It elevates the functionality of the camera in a core game since it's intertwined with a method of control that has proven itself functional and useful for decades. Kinect 2 is more integral for browsing and nagivating at this point. Thats a huge missed opportunity in my eyes for Sony because their camera in combination with the controller would have been proven more useful for core games.
And now I wonder, what's the use for the light bar if not all developers will try to make best use of the camera? If this is the case, Sony invested money and added costs to the controller for a brilliant feature with lots of potential that wont be used
The reveal event - which appears intended to show "the XB1 is a great device for everything and everyone" instead showed "XB1/kinect as a TV interface".
To fix that mess, they went to E3 and basically banned TV/kinect from their presentation...
Given that kinect2 is their main advantage, they are going to have to get back on message and start talking about kinect in public... E3 was probably intended to build up enough goodwill to allow MS to 'balance' the PR message. And then the whole 'used games' mess...
Anyway, I would expect to see more focus on Kinect2 as we get closer to launch. The lack of confidence should be in the PR/strategy which has sucked.
You brought up good, valid points in the rest of your post, but in the true tradition of forums, I'm only going to answer one small part of it.
MS has two technologies running in the Kinect to deal with the scenarios you posit: echo cancellation and beamforming.
...
bkilian, many thanks for taking the time to explain. I did actually think about echo cancellation, but was unaware of the beamforming methods. Highly fascinating tech for sure.
Just to explain some of the reasons why I'm skeptical and my posts that come across as a tad too critical: The tech guy in me actually loves the stuff about Kinect. From a technology standpoint, I'm deeply impressed by how the tech has evolved and what potential it offers. As a personal thing, I also prefer it to the Move on PlayStation, since that is dependant on a hardware (the Move controller) helping out the camera. That's purely personal preference though - and I do admit, both directions feature different strengths and weaknesses.
The reason for why I'm a bit skeptical is more from a market perspective, as in Kinect and how it is included in the business approach of Xbox One. I do agree that voice-control offers some advantages over some inherent disadvantages of a remote (like switching from a game to a specific channel), but at the same time, I don't think that's such huge advantage - and that advantage (if you can call it that) still hinges on how well it actually works. Not every livingroom is the same - and I'm sure there are some livingrooms where Kinect will work more effective and some where it won't. The question is; at average, how well does it work? This is not something that people will accomondate their livingroom around Kinect - but the other way around. Most people have a livingroom that is already set-up in a certain way. The more casual you go, the less optimal it will probably be set-up. How well will it work?
I'm not sure if this is an issue in the states, but in Europe (or at least my country) most modern appartments/houses have bigger livingrooms with open kitchens. That means, the noise level has generally increased and because the room is bigger, the distance to where ever the mic is located, is probably further away. Add to the fact, that as people buy bigger screens, they are more likely to move further away from the screen than closer. I'm sure this is to some degree solvable (or the effect minimized) by the techniques you mentioned, but the challenge doesn't stop there:
There's probably a large distinction between how well a command is understood to how well basic strings/words are translated into correct letters and spelling on the screen. That's why I'm assuming that the basic interface will work well, because from the 20 to 100 odd voice commands, they are predictable to some degree and the error-margin is smaller. It gets interesting as soon as we are talking about strings that are not basic commands. Channel X is already a higher challenge (where channel is a pre-fixed command and X a string) - or program Y, where program again is a pre-fixed hard embedded command and Y is a string like "Dr. House" where the program itself is perhaps only named "house", which might apply to other programs running at the same time on multiple channels. These are just some examples of the challenges awaiting. I'm sure I could think of more complex ones. Then there is the question - if these programs are running through some Xbox TV app or something - and how voice-commands are linked with dynamic content? If I am to take a guess, it's all down to how well the technology will translate strings through voice-command into correct spelled words which are then matched up by the TV guide to see what fits and the result with the highest matchup will then display (like a modern movie scraper for instance). Sounds great in theory, but how well will it work, especially if there are multiple hits etc - or more crucial, if the strings are misunderstood, miss-spelt and not found?
Also, surely, there must be some rules. I doubt you can simply just talk to your Xbox and it will sort out the rest. I don't think this will be possible because Xbox is a machine without any form of intelligence - thus, context is difficult to put into commands. So you might have rules or bounderies where a command will need to be spoken in a specific order if you want it to work, at least for a higher accuracy.
As I already mentioned - I find this tech extremely fascinating from a technical point of view - but more to the point, if this doesn't work well on a daily basis to the casual consumer, people will stop using it. *If* the gamepad will have to be used (or will be, purely by convinience) because voice-controls isn't that accurate in real-world usage, then surely it defeats the whole point of using it in the first place, which is when people will instinctively grab the gamepad instead of fumbling with voice-commands that on average don't get the desired results. It's only a good replacement if it's a full replacement.
I actually think the prospect of being able to install a Xbox-Controller-App for your smartphone or tablet (any OS mind you) to be far more exciting than any promising tech via voice-recognition. But has Microsoft even proposed something like this is in the works? I haven't heard anything, although IMO - this would be a brilliant move, probably easy to realize technically - and a whole lot more resource friendly too. Although not as exclusive as voice-commands.
I see why Microsoft is betting on it. It's their ace-card, their joker card so to speak, offering something that is not yet on the market to this extent. But going by how they demonstrated it, it's also in their interest to make you think it works flawless. And to some degree and judging by the euphoric-like responses by a few in this topic, I somehow think that a lot are thinking especially about the "potential", the endless posibilities, looking through rose-tinted glasses where the belief is strong that somehow it'll just work like portrayed in so many science-fiction movies. I think once it comes out, some people will be perhaps very disappointed simply because the expectations on what it can do, will far exceed what it actually will. And reality will set in, on how limited it will end up being. Simply because the complexity to voice-recognition is huge and a whole lot less predictable, unless you set certain bounderies or rules on how to use it. Which at that point will make it a glorified super remote controller with some inherent flaws along the way. Or I might be all wrong and this will turn out to be biggest thing since the invention of the IR remote. But somehow, when thinking of the technical challenges, I somehow doubt it.
Of course, having said that, even if it fails on expectations, whatever the results, the tech and research done will still be hugely fascinating and impressive.
Now, to another point I wanted to point out - Kinect in games. Are there some that are really excited about this? Core-gamers?
I bought PS EyeToy back on the PS2 and was fascinated by the minigames and challenges to moving your hands in front of a screen. That was when I still played my games in front of a small TV in a small room where I could set it up. Then, I got bored with it. I got bored with moving my hands, my head to actually play a game. It was a good excercice, but it wasn't what I bought a console for, not "gaming" in the sense how I am used to and want to play. It's a fun toy for when you have people over and want to play some party games, but on the whole - I just prefer to sit or lie on my couch and play games comfortably.
Now, I know Wii is a huge a success, but despite this huge success, I really wonder how often these games get played. Sure, they had a lot of sales, but all this points to, is that everyone bought something that was considered cool and new. Just like I bought EyeToy and count as a +1 sale, I'm sure so did many people with the Wii. It's cheap, small and refreshing new. But does it actually get used a lot? I'm willing to bet that games get used far more on X360 and PS3 - and without looking, I would also think that the software sales to userbase ratio is higher on these platforms than on Wii.
So, while the world (well Microsoft and Nintendo) is betting on these new casual games, I'm just not too sure. Sure, they get sold, but do people really want it as much as people buy consoles to play core-orientated games? I guess that could be irrelevant for Microsoft, because a sale is a sale, regardless if it's used a lot or not. As a gamer though, I just don't want my games to move in that direction. I think it's a short lived prospect, one that could do more damage than good. I'd rather we split them up: if people want to buy these kinds of games, go ahead, get a Wii or get the add-on. But let us play games with (and only) a controller who want to and prefer it that way. To the same point, even if Kinect has like 20 million sales - how many were bundled? And how many bought it simply because it was a fascinating accessory and portal to new and refreshing games? And how many still use it at all?
I'm not sure, but I do think Tablets and smartphones are where this casual market is heading. I'm not sure Kinect or Move is the future. It may introduce some exciting features in core-games, but on the whole, I don't think the tech (specific to games) is that a revolutionary concept. It's refreshing and new, but it gets old and tiresome quickly.
Or you can point with an onscreen cursor, swipe/wave your hand to change screens...
When there were just keyboards people decried mice. Once there were kb/Ms people decried touch. Things are progressing as they should. Voice and motion are next because they are the most natural interfaces possible. We just have to work on precision and command recognition.
That is interesting. Wonder if they will capture your motion and render it on screen for a game like FIFA.
Now can it render your thumb motions as discrete button presses, like the way you'd use your right thumb on different face buttons of the controller?
That would kind of be a requirement for replacing the controller. Or maybe different gestures to represent button 1, button 2, etc. and different forefinger gestures to represent shoulder button 1, trigger, etc.
Essentially, games would have to replace all the physical controls with a gesture-equivalent but will people prefer that? Would they bother to learn the gestures for the different buttons?
On the Wii, all you had to do was wave the wand thing and it could render a tennis racquet swing, a baseball bat swing. You didn't have to have proper form actually and it seemed all it cared was the timing of the gesture and the acceleration of the motion. So I could wave it in front of me and it would do either a forehand or a backhand.
For motion control to really take the next step, it should penalize you if you use the wrong motion or if your motion and mechanics don't pass a minimum threshold of accuracy. So if you do a forehand swing but your player is only in position to do a backhand, you should completely whiff on the ball. Or if your grip is wrong so that the racquet face would be pointing up, it should hit a lob instead of a nice flat stroke. Or if you torque your wrist as you swing, it should render a topspin or a backspin, with the appropriate ball physics when it hits the surface of the court (which would also influence ball physics, depending on what kind of surface it is).
But I'm not sure that game designers want to necessarily encourage you to develop the right mechanics or technique, to require that kind of precision in a video game. Players could get frustrated if their learning curve isn't fast enough and may wonder if they really wanted to be good at a sport, they'd be practicing the sport instead of playing a video game representation of that sport.
That has nothing to do with the quality of the mics. It is a function of how good your calibration is. You need to calibrate the Kinect at the loudest you will ever play a game, and if you move the Kinect or your furniture, you need to calibrate it again. I have tested the Kinect understanding whispers at 3.5 meters, while the console was outputting 5.1 audio, the destruction scene in "Independence day".
Most people get the calibration wrong, they're actively researching ways to do a better job with a crappy room impulse, or ways to generate the room impulse during gameplay.
Thats fairly amazing. If you can recall, what was the decibel level of the scene?
Orion
Regular
interesting.
I don't see how it'd be pleasant to change volume by voice, with the remote you get a feeling of the loudness of the show and can stop when it feels right. Also there are shows that have explosions and loud noises that demand active lowering and raising of the volume(unless the tv has an automatic volume adjustment feature).
How would that work with voice, saying "higher" "higher" or "up" "up" for each volume number to be raised, or would you simply say raise the volume, and then give a stop command? either way doesn't sound satisfactory.
the tv controller, only requires a few button presses. One to switch to cable/satellite, and another few for the numbers of the channel.
Just want to say I appreciate your post, you are always calm and reasoned and add something to the conversation.
I have done that calibration about 20-30 times and I just tried it again few times. No matter how loud or silently I run the setup it still can't hear me when talking normally. (while sitting on couch 3.5-4m from tv/kinect) "You're speaking too softly"
If I speak lot louder than normal conversation would be it says "ok" but then I have to speak voice command loudly too.. would irritate me and others at house.
My audio/tv setup should be ok, 5.1 sound and 64" plasma, kinect 10cm from top of the tv and (about 160cm from floor)
It also requires you to have a TV controller in your hand. Voice doesn't take anything away, it adds convenience and a more intuitive UI.
For volume, you could have presets to activate with voice. Or say "volume" and a number. I don't know how it will actually work, but it doesn't have to be "higher, higher, higher, lower..."
I think it was 85-90Db at the Kinect, it's been a while. We were standing 10-11 feet from the sensor, and were quite surprised when it started picking us up at below normal conversation level. It was a controlled test, but in a pretty bad environment for Kinect (One of our conference rooms, hard walls everywhere, very echoey)
Back to the topic at hand: I'm curious as to why Sony did not include the PSEye in their box considering how much attention they lavished on it at their reveal. I'm wondering if they originally were going to include it, but then pulled it to hit the 399 price point that they were afraid MS was going for.
Hmm, what's your room like? This could actually be bad mics in your case, although that's very rare, or it might be that your room is so big that your DB level at the kinect is below our threshhold for the "say the numbers" test. Note that the test doesn't actually do any voice recognition, you can say anything there, all it does is measure the Db level of you speaking and report OK if it is over a certain level. If I was still there I would encourage you to do voice collection and send the session ID to the team, they can normally work out if it's a bad mic. We only found maybe ten of those in total in the millions of devices sold.
If you have any friends with a Kinect, have them bring it over and see if it works better than yours, if so, then yours is bad.
I'll stop now, this is OT
