On Sat, Nov 12, 2011 at 2:37 AM, Cedric Sodhi <[email protected]> wrote:
> Hello everybody,
>
> are there any intentions or just the possibility to calibrate Wacoms
> with n-point (n+1-th order) maps? I assume there is a good reason why so
> far we only have 1st-order/2-point mappings? I think this is a point
> which matters for all those tablet users out there, like I am one.
>
> As I explained, the ISDv4 has notrious (in)accuracy problems, which is
> in particular annoying when the cursor derivates towards the edges,
> because then you have the coice between
>
> a) an inaccurate cursor
> b) a dead area at the borders
> c) a combination of a) and b)
>
> This problem can only be rectified with higher order mappings.
> Mathematically this is a trivial problem of replacing the linear map by
> an (arbitrary) higher order map. I assume there are other drawbacks,
> specfically performance?
>
> --
> Cedric

I don't believe anyone is working on calibration, though it's
something I too have often thought could use some improvement. There
is certain to be a performance penalty for improved accuracy, but it
really should be pretty trivial for low-order maps. The "tslib"
library, for instance, is able to provide three-point calibration for
touchscreens and is targeted at embedded system use. The two questions
I have are whether calibration should be handled in the driver (vs in
the server) and whether a transformation matrix or a mesh would be
better suited.

Jason

---
Day xee-nee-svsh duu-'ushtlh-ts'it;
nuu-wee-ya' duu-xan' 'vm-nvshtlh-ts'it.
Huu-chan xuu naa~-gha.

------------------------------------------------------------------------------
RSA(R) Conference 2012
Save $700 by Nov 18
Register now
http://p.sf.net/sfu/rsa-sfdev2dev1
_______________________________________________
Linuxwacom-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/linuxwacom-devel

Reply via email to