|
On the original topic, is there any speculation as to why there is a calibration issue? I've looked over gtktscal, and it appears to store linear mapping parameters (from Wacom sensor data space to screen resolution space), based on a 5 point calibration system.
Increasing the number of points, IMHO, should not change things much. It looks like there are strange non-linearities when you draw on the screen, I've got the screen edges and the interior to match perfectly while .5-1cm away from the edge there is a mismatch. Anyone else observe this? If this is true, then the actual mapping of Wacom data to the screen is non-linear.
I wonder why this is (and find this a little hard to believe), I've used the Intuos line and I've never had calibration problems. Unfortunately, gtktscal dumps the linear parameters (X/Y shift and X/Y scale, along with X/Y swap for rotation) directly using low level hardware calls, so I don't think we can change this. More calibration points while sticking to a linear map will not do the trick, if the Wacom spits out non-linear data.
Fortunately, one should be able to write a software map on top of the hardware calibration, say, by providing multiple points to calibrate and then using separate linear maps for small parts of the screen (approximating any non-linearities with piecewise linear maps). Just thinking aloud.
|