On Thu, 17 Feb 2005, Vojtech Pavlik wrote: >If the raw resolution is larger than the cooked resolution, which is >usually the case, multiplication and division will be carried out to >transform in the controller.
Ok, but in this case the 3M raw resolution is 16k and calibrated resolution is 64k. >So we definitely want to do an one step calibration and want as much >resolution in the input data as the HW gives us. If the calibrated resolution is less than raw, you are right - but for the 3M hardware, the calibrated resolution is more (so nothing is lost). >> What about for a system where the hardware doesn't change but the software >> is replaced? Either the software calibration would have to be stored and >> re-used (for each system, if managing multiple systems), or software >> calibration would need to be re-run... > >In that case the best option is to store the calibration in the >touchscreen, but still use the raw coordinates and have the controlller >report the recommended calibration to the OS. Well since the hardware calibration doesn't lose any resoltion, I personally think just using the hardware calibration would be better and easier... >> Do you agree that for integrated touch panels, it is better to use >> hardware calibration, as no software calibration should be needed? > >"Should" is the correct word here. Often it is needed due to >nonlinearities where multi-point calibration is required, and moreso for >integrated touch panels, where you don't want to replace the whole thing >just because the panel is not so perfect anymore after long time use. But the hardware we're talking about, the EXII-5000UC 3M touch panel, does have 21-point linearization hardware calibration available. And since nothing is lost by using the hardware-calibrated coordinates, software calibration (21-point or 2-point) is still an option. >> Sure, sure, I'm just saying, why _force_ software calibration? If the 3M >> panel reports only raw coordinates, I _have_ to use software calibration. >> If it reports hardware-calibrated coordinates, I don't have to use >> software calibration... > >One reason is to make the behavior of all touchscreens as similar as >possible, so that applications don't have to care. But...you want to force software calibration just for consistency? Really? >Another is the information loss above. Not for this hardware... >Another is if you switch video modes (think CRT display), you have to do >software coordinate transforms, even with a perfectly calibrated >touchscreen. But I don't think that is true. The evdev interface provides the min and max range for X and Y. The X driver then calculates a (floating point) percentage for any given X/Y value and then aplies that to whatever screen resolution is in use. No re-calibration is needed when switching between resolutions, i.e. the min and max values are the same and linearization doesn't change based on screen resolution (besides hardware calibration fixes linearization problems). >If you can get the 3M panel to send you the calibration coefficients, >you'll still use software calibration, with all the benefits of the HW >one. But as far as I know, I can't, and it still would be easier to use hardware calibration, where it "just works" with no software calibration... -- Dan Streetman [EMAIL PROTECTED] --------------------- 186,272 miles per second: It isn't just a good idea, it's the law! ------------------------------------------------------- SF email is sponsored by - The IT Product Guide Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now. http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click _______________________________________________ linux-usb-devel@lists.sourceforge.net To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
