07-06-2011 10:11 PM - edited 07-06-2011 10:13 PM
I hope the engineers at Lenovo also look at my third video too.
It describes the same problem, and compare with my X61T.
My tablet is returned, and I'm just going to stick with my 4 year old X61T.
I also wonder if X201T works okay on the edges while X200T has similar edge problems, because the X201T bezel design is the same as X61T. Is there a reason why Lenovo switched bezel design back and forth?
X61T (frame type) -> X200T (flat type) -> X201T (frame type) -> X220T (flat type)
(edit) Sorry, I see andrer's reply about X201T having same issue in previous page.
This might belong in another topic, but another problem with the X220T touchscreen is that it doesn't do palm rejection. In other words, when writing with a pen, and you lift the pen enough for a moment, the screen recognizes the palm and thinks you are trying to do a multitouch gesture or do something with your finger. This is a big problem for example in Onenote2010 because it zooms in/out or moves around the paper while writing when the pen tip is in the air. Of course, this is not severe as the edge problem, but I think Lenovo should address this problem.
07-06-2011 10:47 PM - last edited on 07-06-2011 11:02 PM by andyP
please have 'em look at the images we posted as well. The best way to see the distortions accurately is to go and draw angled lines over the edges using a ruler as shown here many times.
Using this very same method it would be possible to come up with an easy, user friendly calibration method to de-distort the 2d plane.
- Distortion: By drawing arbitrary angled (!) lines with a ruler across the screen-edges
- Translation (= shift up/down, left/right) by 4 points the user has to click on (just like now)
Now you go and recognize each line drawn and fit a "true line" to it, starting from the center. The "difference" then will be mapped and finally interpolated using a 2d spline function for instance, this should suffice, it's no rocket science!
One can either use the resulting vectorfield as a "lookup-table" or come up with a transformation function. Whatever pleases the software engineer most.
Seriously, if the engineers struggle to implement this, I am willing to provide the algorithm for it!
Moderator edit; large blank space below post adjusted
07-07-2011 01:14 AM
At least that opened my eyes. The X61T is worlds apart from even my X201T which I thought performed well.
UPDATE: Recalibrated it (once more) and it's better, but.. there's still some of this "pillow effect" visible.
07-08-2011 11:35 AM
anddddddddddddddddddddddddddddddddddddd Thanks for the quick response Lenovo!
Tablet pc is now in the hands of UPS. Enjoy refurbishing it. Good luck everyone else.
07-08-2011 09:46 PM
07-09-2011 02:10 AM - edited 07-09-2011 02:11 AM
It seems that the problem we are talking about here is the pen orientation.
@ egolfb: I'm afraid we're mixing the terms here In the thread above the pen stayed in landscape mode even if the user turned the screen resolution to portrait. This led to the cursor having its x and y axis flipped 90 degrees in relation to the screen. However, in this thread people are talking about the angle they're holding their styli in their hand as 'pen orientation'.
07-09-2011 09:43 AM
we're not talking about "angles" here either.
The problem at hand in this thread is the nonlinearity of the digitizer in close proximity to the borders of the screen.
See these lines drawn with a RULER:
07-09-2011 09:47 AM - edited 07-09-2011 09:56 AM
we're not talking about "angles" here either.
You have been actually. I was referring to when people do talk about pen orientation in this thread, they're referring to the way they're holding their stylus and wheter or not it affects the wiggliness (or nonlinearity as you put it) of the lines (such as in your picture). Just tried to explain the difference in meaning in regards to the other thread egolfb was referring to.
For example Sciurine wrote :
...is if I hold the pen at anangle less than rougly 30 degrees (relative to the screen, more than 60 degrees away from perpandicular to the screen) and moving very quickly around the edges of the screen.
The other thread also refers to pen orientation, but there it has to do with switching between lanscape and portrait orientation, not to issues with the digitizer.
But yeah, let's move along here
07-09-2011 06:13 PM
I had my x220t for more than 21 days now, if they dont fix this issue Im then screwed and bought a laptop for 1.3k for nothing?
If ever they do not fix this because of a hardware issue, what will happen?
This issue is very annoying, you cannot write in landscape mode and then in portrait mode because the calibration will be way off. Basicly, all the good features that made me buy this tablet are not working right.
07-09-2011 08:36 PM
whenever you work in landscape, calibrate and then switch to portrait, it's obvious, that the calibration will be off, because you'll look at the screen from a different angle! .. thus you'll have to re-calibraten AGAIN.
Obviously, if Lenovo would do a proper job, they'd store calibrations for each orientation... but apparently nobody thought that far!
don't want to sound like a smart-alec but: This is clearly something to add to the list of a proper digitizer-driver/tools.