Freigeben über


Touch and pen input issues

Today in my RSS reader, I found a review from PC Magazine dated 1/7/07 on the new HP Pavilion TX1000 Tablet PC.

In the review, Cisco Cheng writes "But the OS did not recognize many of my pen strokes. One of the big reasons is that HP elected to go with a touch screen instead of a digitized screen from Wacom. Users often have a tendency to lay the palm of their hand against the screen when they write, thereby causing touch-screen functions to interfere with their pen strokes. This is a major problem: During testing, the system often didn't recognize the last stroke of my letters. The glass on the screen seems a little thicker, too, so you don't get that pen-on-paper feel when you write. Overall, the writing experience did not meet my expectations. By contrast, the Wacom-enabled X60 tablet and the Fujitsu LifeBook T4210 , both of which I upgraded to Vista Build 6000, offer impressive accuracy in handwriting recognition. "

This is one of the challenges we have tackled while adding touch input to Windows Vista. How do you determine what is a benign touch that can be ignored, and what is a "real" touch, where the user intends for input to occur?

In general, the digitizer's driver measures the size of the touch contact (2d area) and passes the area information along with a calculated confidence value, to Vista's Tablet components. Vista then determines whether or not to ignore the touch contact, or give it control of the cursor.

You also need to be aware of two different types of digitizers, capacitive/combination ("digitized screen") and resistive ("touch screen").

Specifically addressing Mr Cheng's issue, when a combination digitizer detects that pen and touch are both in range/use, Vista prefers the pen and suspends touch for as long as the pen is in range. This is a simple way to remove contention between pen and touch. It gives a great user experience because it is very much like using a writing instrument on real paper. But because the HP is using a resistive digitizer, it simply sees multiple touches, and cannot tell which is the pen and which is touch.

Vista does not currently have a method to pick one valid touch out of multiple valid touches. How would we determine which one was "most valid"? If you put two (or more) fingers on a Tablet with a touch input screen at the same time, either you get total rejection (no touch input), or the cursor gets trapped between the valid touches. The cursor is released when the touches are released, but it's still not an optimal solution.

So what Mr Cheng is probably seeing is, that a) the driver that HP used on his review machine probably didn't implement area/confidence; and b) resistive digitizers ("does not have electronics") do not behave the same way as active ones do when both pen and touch are in use simultaneously.

I hope you all had nice holiday seasons. I will be posting more Vista Tablet features soon.

MSDN reference: https://msdn2.microsoft.com/en-us/library/ms698092.aspx