138 lines · plain
1.. include:: <isonum.txt>2 3=========================4N-Trig touchscreen Driver5=========================6 7:Copyright: |copy| 2008-2010 Rafi Rubin <rafi@seas.upenn.edu>8:Copyright: |copy| 2009-2010 Stephane Chatty9 10This driver provides support for N-Trig pen and multi-touch sensors. Single11and multi-touch events are translated to the appropriate protocols for12the hid and input systems. Pen events are sufficiently hid compliant and13are left to the hid core. The driver also provides additional filtering14and utility functions accessible with sysfs and module parameters.15 16This driver has been reported to work properly with multiple N-Trig devices17attached.18 19 20Parameters21----------22 23Note: values set at load time are global and will apply to all applicable24devices. Adjusting parameters with sysfs will override the load time values,25but only for that one device.26 27The following parameters are used to configure filters to reduce noise:28 29+-----------------------+-----------------------------------------------------+30|activate_slack |number of fingers to ignore before processing events |31+-----------------------+-----------------------------------------------------+32|activation_height, |size threshold to activate immediately |33|activation_width | |34+-----------------------+-----------------------------------------------------+35|min_height, |size threshold below which fingers are ignored |36|min_width |both to decide activation and during activity |37+-----------------------+-----------------------------------------------------+38|deactivate_slack |the number of "no contact" frames to ignore before |39| |propagating the end of activity events |40+-----------------------+-----------------------------------------------------+41 42When the last finger is removed from the device, it sends a number of empty43frames. By holding off on deactivation for a few frames we can tolerate false44erroneous disconnects, where the sensor may mistakenly not detect a finger that45is still present. Thus deactivate_slack addresses problems where a users might46see breaks in lines during drawing, or drop an object during a long drag.47 48 49Additional sysfs items50----------------------51 52These nodes just provide easy access to the ranges reported by the device.53 54+-----------------------+-----------------------------------------------------+55|sensor_logical_height, | the range for positions reported during activity |56|sensor_logical_width | |57+-----------------------+-----------------------------------------------------+58|sensor_physical_height,| internal ranges not used for normal events but |59|sensor_physical_width | useful for tuning |60+-----------------------+-----------------------------------------------------+61 62All N-Trig devices with product id of 1 report events in the ranges of63 64* X: 0-960065* Y: 0-720066 67However not all of these devices have the same physical dimensions. Most68seem to be 12" sensors (Dell Latitude XT and XT2 and the HP TX2), and69at least one model (Dell Studio 17) has a 17" sensor. The ratio of physical70to logical sizes is used to adjust the size based filter parameters.71 72 73Filtering74---------75 76With the release of the early multi-touch firmwares it became increasingly77obvious that these sensors were prone to erroneous events. Users reported78seeing both inappropriately dropped contact and ghosts, contacts reported79where no finger was actually touching the screen.80 81Deactivation slack helps prevent dropped contact for single touch use, but does82not address the problem of dropping one of more contacts while other contacts83are still active. Drops in the multi-touch context require additional84processing and should be handled in tandem with tacking.85 86As observed ghost contacts are similar to actual use of the sensor, but they87seem to have different profiles. Ghost activity typically shows up as small88short lived touches. As such, I assume that the longer the continuous stream89of events the more likely those events are from a real contact, and that the90larger the size of each contact the more likely it is real. Balancing the91goals of preventing ghosts and accepting real events quickly (to minimize92user observable latency), the filter accumulates confidence for incoming93events until it hits thresholds and begins propagating. In the interest in94minimizing stored state as well as the cost of operations to make a decision,95I've kept that decision simple.96 97Time is measured in terms of the number of fingers reported, not frames since98the probability of multiple simultaneous ghosts is expected to drop off99dramatically with increasing numbers. Rather than accumulate weight as a100function of size, I just use it as a binary threshold. A sufficiently large101contact immediately overrides the waiting period and leads to activation.102 103Setting the activation size thresholds to large values will result in deciding104primarily on activation slack. If you see longer lived ghosts, turning up the105activation slack while reducing the size thresholds may suffice to eliminate106the ghosts while keeping the screen quite responsive to firm taps.107 108Contacts continue to be filtered with min_height and min_width even after109the initial activation filter is satisfied. The intent is to provide110a mechanism for filtering out ghosts in the form of an extra finger while111you actually are using the screen. In practice this sort of ghost has112been far less problematic or relatively rare and I've left the defaults113set to 0 for both parameters, effectively turning off that filter.114 115I don't know what the optimal values are for these filters. If the defaults116don't work for you, please play with the parameters. If you do find other117values more comfortable, I would appreciate feedback.118 119The calibration of these devices does drift over time. If ghosts or contact120dropping worsen and interfere with the normal usage of your device, try121recalibrating it.122 123 124Calibration125-----------126 127The N-Trig windows tools provide calibration and testing routines. Also an128unofficial unsupported set of user space tools including a calibrator is129available at:130http://code.launchpad.net/~rafi-seas/+junk/ntrig_calib131 132 133Tracking134--------135 136As of yet, all tested N-Trig firmwares do not track fingers. When multiple137contacts are active they seem to be sorted primarily by Y position.138