You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi @ all,
did you think about adding a further optional parameter to ts_calibrate / function getxy() ?
I thought about something like "min_samples", so that a minimal count of samples needs to be recognized within calibration (and maybe validation as well?).
By testing v1.22 of tslib, sometimes there are faulty touch-events on the touchscreen (could be because of old resistiv touchscreen). This could generate a faulty calibration and let validations fail more often.
If there is a minimal count of 'samples' needed for calibration / validation, these faulty-events could be ignored and would not take big effect to the calculated, touched point.
These 'min_samples' could have a default value given and overridden by another optional parameter within terminal calls.
What do you think about that?
The text was updated successfully, but these errors were encountered:
that's a good idea in general, but not so easy.
1st: some touchscreens (firmware) doesn't send multiple (or many) sample when "tapping".
2nd: I think ts_calibrate has to rely on the input to be correct. It creates it's filtering from the raw input. Try not to let endusers to the calibration, but create a known-good one and resue it, maybe. Verify what's going on by watching evtest maybe, while doing the calibration.
Hi @ all,
did you think about adding a further optional parameter to ts_calibrate / function getxy() ?
I thought about something like "min_samples", so that a minimal count of samples needs to be recognized within calibration (and maybe validation as well?).
By testing v1.22 of tslib, sometimes there are faulty touch-events on the touchscreen (could be because of old resistiv touchscreen). This could generate a faulty calibration and let validations fail more often.
If there is a minimal count of 'samples' needed for calibration / validation, these faulty-events could be ignored and would not take big effect to the calculated, touched point.
These 'min_samples' could have a default value given and overridden by another optional parameter within terminal calls.
What do you think about that?
The text was updated successfully, but these errors were encountered: