Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Multitouch on the MacBook Air and Beyond

TBi said:
You are looking at this the wrong way. The computer sees the first press and moves the mouse to their. Now it sees the second press as if your hand had moved very quickly to another part of the touchpad and it moves the cursor accordingly. That's all, no more no less and no averaging.
Click to expand...
First of all, touchpads don't move the mouse on first contact. Try putting your finger down anywhere on your touchpad and tell me how far the cursor moves. However, I appreciate that isn't your key point.

If I've understood correctly then I argue that if I put a finger down in the corner and then put another down somewhere else then the computer sees movement because the average of all the pressed locations has moved. You argue that the touchpad forgets about the first touch because it is mostly only interested in movement.

I don't see that the results of "two fingers in two places" would be different according to either interpretation. What we probably need is for someone to switch off cursor acceleration and do a test. But we'd also need them to be able to guarantee that the averaging isn't being done in firmware or the OS, so that seems unlikely.
TBi said:
The capacitance type pads also use a matrix system (except for the ipod scroll wheel but that's a different beast).

...

Resistive and capacitive pads both use a matrix to read the values, they just measure them a different way.
Click to expand...
No. Using a matrix system is one option. Capacitance shunt is another. Analog Devices, who happen to manufacture one of the latter types provide a nice datasheet explaining it.

There's also a brief bit on it in the wikipedia article on touchpads, but obviously that isn't going to be as accurate as the original source.
 
The current laptops can detect three fingers, i read somewhere that this is how you right click in ubuntu. When MBA is realeased someone will hack it so atleast it will be there for the techies. But i do doubt apple would release this feature for free.
 
Thomas Harte said:
First of all, touchpads don't move the mouse on first contact. Try putting your finger down anywhere on your touchpad and tell me how far the cursor moves. However, I appreciate that isn't your key point.
Click to expand...

No they don't, to be correct i should have said point of last contact.

Thomas Harte said:
If I've understood correctly then I argue that if I put a finger down in the corner and then put another down somewhere else then the computer sees movement because the average of all the pressed locations has moved. You argue that the touchpad forgets about the first touch because it is mostly only interested in movement.
Click to expand...

The trackpad doesn't forget anything. It just reports coordinates to the software, the software dictates whether it is registered as a cursor movement or not.

Thomas Harte said:
No. Using a matrix system is one option. Capacitance shunt is another. Analog Devices, who happen to manufacture one of the latter types provide a nice datasheet explaining it.
Click to expand...

EDIT 2: Gosh darn it i'm wrong again!

Just reading through the design brief for this part again! It uses a grid of different sized capacitors across the surface. So if you touch the right side of the pad it see's a higher (or lower) capacitance based on the layout of the pad.

Unfortunately this means that it won't support multitouch!

EDIT: I'm a bit wrong here. It can be used in an XY trackpad. But it also uses a matrix of shunt sensors. Similar to the standard method.

That's actually the different beast i mentioned for iPods. From what i remember that chip was designed specifically for iPods. To my knowledge t only works in one direction, i.e. you'll only get the X reading and no Y data. It works on the iPod because the touch wheel only measures the the X data, and doesn't need Y.

It wouldn't work for a normal touch pad, and if it did the method used would be a lot more difficult and expensive to implement than the standard matrix.

Also it doesn't show it very well but the Rx and Tx parts of the "shunt" are actually milimeters apart. (I've seen it in person)
 
MacBook pro revisions

When are these impending MacBook pro revisions? I really want to buy one, and if it is going to be June, I will just get one now. If it is feb...
 
Manic Mouse said:
It would be very nice if remote disc and multi-touch gestures were in 10.5.2 :D
Click to expand...

Yes, and it would help sell the current line of laptops (ie., MacBooks for those who don't want the expense of an Air, and MacBook Pros for those who need more power than the Air.

Any new feature they add to the current hardware set only makes them more attractive to new buyers and switchers.

I'm not going to piss and moan if they don't offer it on current lines, even though I agree with previous posts that it appears the current hardware could probably do it, but I think it would be short sighted of Apple to not offer functionality, even for a small upgrade price, to current line of machines.
 
Current gestures

It seems I can not get the 'Click and Drag' working on my MBP? Any tricks with this?

I find 'Screen Zoom' quite useless. I would have liked it if one could zoom pictures in iPhoto, but simply magnifying the screen is not something that I need that often.
 
And I want finally the MacBook Pro update! I desperately need a new Laptop for business and I don't want to shell out that kind of money for a machine that will be probably outdated in a matter of weeks.

The new touchpad excludes the possibility now to buy the current one, since the gestures would be definetely a big help with my daily work...

:apple:!!! Be a little more open about updates to your computer line-up!!! What is the point about being so secretive about product updates?

groovebuster
 
techgeek said:
I second that!
Steve if you're reading this (and we all know you do) release a USB multitouch trackpad. You'll sell at least 2 of them :D

It would be awsome to use a device like that on my iMac. Much better than a mouse.
Click to expand...

Make it wireless... No wires please.:p
 
Easy Features to Add

I would love to have simple gestures for Expose, Dashboard, etc on my macbook...

For example, for expose, rap three fingers left to right. For dashboard, right to left, etc. it seems like even if the current MB or MBP trackpads don't register three fingers, something like this would be easy to add.
 
mrmma said:
I would love to have simple gestures for Expose, Dashboard, etc on my macbook...

For example, for expose, rap three fingers left to right. For dashboard, right to left, etc. it seems like even if the current MB or MBP trackpads don't register three fingers, something like this would be easy to add.
Click to expand...

Use hot corners :)
 
So much for touchscreen. This will last 5 years at least. First was the mouse. Then the touchpad. We would have affordable fourth, fifth gen touchscreen monitors by now.
 
TBi said:
Use hot corners :)
Click to expand...
I've tried that...I seem to accidentally activate them frequently, much to my frustration. Plus it seems like a three finger tap would be much much faster.

One could do a double rap to minimize, etc...
 
Thomas Harte said:
I seriously doubt that they could. I think the current trackpad can tell the location of a touch and the size of the object touching — enough information for the scroll and right click taps on current MacBooks and MacBook Pros but not enough for MacBook Air style gestures where you need to be able to tell two independent finger locations.
Click to expand...

So basically you are saying only one of the fingers movements can be tracked. The location and therefore the change in location, or movement, of onl one point can be analyzed. Thus two touch scorll works by observing one large point and its movement, fiuring that the large area means its more than one finger.

That is not true though. If you split your fingers into a peace sign and place ach finger ont he side of the trackpad, far away and not in alignment and move them. The page scrolls. If it could only track one objects movements, then splitting up the finger would only move the mouse since whichever object is picked, its one finger.

But no it sees both and even moving them out of synch produces a scroll. The trackpad is tracking both objects movement and size. BOTH! Try to put two fingers one one side and a third apart from it. And it wont scroll. Thus it knows the size of both objects on the trackpad and their movements.

But wait, something might disprove my line of thought. Hold one finger steady and move the other, and it still scrolls. So I guess it can see both objects but can only track one of them's movements. Unless they programmed it to scroll when there is two.

One more test. 2d scrollings. Enabble 360 degrees on my tracpad goin into preview. I start my two finger sperate. My left one is going to move to the right. My right one is going to move down. Thus if it can only track one movement, the area will only scroll in one direction down or right. But wait, its scrolls in both directions. Two fingers. Each with different movements, and it understands both independently.

So for sure we know that it can understand multiple objects on the tracpad. It can also determine their sizes. So mulitple objects and their sizes can be observed. Also the final test shows that it can track movements of independent objects, non-touching. So it can understand everything that seems neccesary to use these gestures. Recognizing multiple objects, recognizing their size, and their movement.

Of course I stress the seems, because I am not knowledged in the mechanics of the trackpad or the software. So as for sensing, it seems to be bale to. Software wise or some other hardware limitations, who knows.....
 
sjamaan said:
Here's an intresting experiment: hold one finger still on the touchpad and move the other one up or down. You'll notice that this gesture is recognised as two-finger scrolling. It's not just the average that is reported, because scrolling would then occur at half speed, which is not the case. However, it's still possible the touchpad 'feels' something moving and detects two fingers without registering the relative and absolute positions of the fingers. Position detection is a requirement for gestures like pinch and rotate.
I think the question is "can our current hardware detect the position of a finger on the trackpad, or does it just register movement?"
Three finger-swiping must be possible using our current hardware because this gesture requires only movement detection, and we know our trackpads can detect the difference between one, two or three fingers.
Click to expand...
Why would we need location. Pinch could work by two objects moving in opposite directions. It may allow for some little quirky waus to pinch, if u pass ur fingers. But te importnt part is it would work for pinch.

I dont see why location would be neccessary for pinch and rotate. Rotate you just see on object stationary ad another moving in a circular directions. If you try to scroll in 2d, it can do it currently. Hold one finger stationary, and rotate the other one in a clockwise direction. It scrolls down and to the right. If it can observe that, why not just program that ggesture to rotate a picture?
 
samh004 said:
I should think even if Apple doesn't introduce this in 10.5.2 for the rest of us, it should be supported if someone knew which files were necessary to copy over.

Maybe, just maybe, the three-fingered gesture wont work, but there's no reason the pinch/expand and rotate shouldn't work.



This will be answered as soon as someone pulls apart the MBA and compares the trackpad part number to the existing MB/P's part numbers.
Click to expand...
You might be right ont he 3 finger. Might be to much information for it. Like the bandwith of the sensors is so small, that detecting the existence as well of the movement of three fingers might be too much. That would be understandable but I see nothing holding us back from the others.

Not neccessairly. While same part numbers would prove it could work. Different ones don't prove that it won't since different part number can still have similar functions.
 
TBi said:
If two fingers are on the same horizontal then it can tell the difference by looking at the vertical scale. Remember you have both X and Y coordinates. The X might be the same but as long as the Y is different the trackpad knows where it is.

The limitation isn't really in the trackpad itself, but more in the interface between the trackpad and the mainboard. If the interface only sends the co-ordinates it's current/newest touch then multitouch won't work. If it sends the co-ordinates of every single touch then multi touch will work.

I'd say the macbook trackpad could almost do 4 finger touching.



Every trackpad only registers position the position of your finger. It's the cpu which reads the changing positions and translates this into movement on the screen.

Basically as you move your finger across the pad it gives off X,Y coordinates like this

(1,0) ... (1,1) ... (1,2) ... (1,3)

and the cpu moves the mouse cursor appropriately
Click to expand...
It can send the movement of two fingers independently.
 
Chef Medeski said:
<SNIP>
Of course I stress the seems, because I am not knowledged in the mechanics of the trackpad or the software. So as for sensing, it seems to be bale to. Software wise or some other hardware limitations, who knows.....
Click to expand...

Everything you said is right except for one thing. The trackpad doesn't see movement. All it sees is where something is on it.

Basically the computer asks the trackpad for info a few thousand times a second. The trackpad responds by saying

"I've something at this point (X1,Y1)"
then says
"I've something at this point (X2,Y2)"
and then
"I've something at this point (X3,Y3)"
and on and on

Or for two fingers

"I've something at this point (X1,Y1) AND (W1,Z1)"
then says
"I've something at this point (X2,Y2) AND (W2,Z2)"
and then
"I've something at this point (X3,Y3)"**
and on and on

** assuming here you lifted the second finger.

It's the software which translates this into motion.
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back