Have you gave up on me? :)
How tablet is initialized? is there any way to calibrate it?
On Sun, Jan 23, 2011 at 20:52, AndrewP <[email protected]> wrote:
> I refreshed my memories with Windows7 and I must say, distortions I've said
> about earlier might be a software issue, because I didn't see them on plain
> desktop. So everything feels fine in Windows7 I sketched several pictures
> there with my natural style and handling, but didn't notice anything
> outrageously wrong. I must admit cursor don't coincide exactly with the tip
> as in Linux, even after a so called "calibration", but still everything
> feels normal. I can't get any logs from there so you just have my word. I
> have only one pen, and it is the same in either case.
>
>
> On Sun, Jan 23, 2011 at 21:41, AndrewP <[email protected]> wrote:
>
>> Chris, I managed to try your suggestions:
>> this "if (!coord.tip && !coord.eraser) ds->pressure = 0; " didn't discard
>> "false" clicks. As there is Tip pressure reported by isdv4-debugger.
>> 2. and "if(ds->pressure > 128) ds->pressure = 0;" possibly shoved off
>> more than we needed, because I lost click instantly, and couldn't achieve
>> getting a pressure level. I mean I press on the screen clicking on
>> something, but even I still press click is gone.
>>
>> I did have "false" clicks in Windows 7, but far less frequently, also had
>> there a weird distortions on the edge of the screen, really bad ones (cursor
>> will move away from the edge and click somewhere), they wasn't constant, but
>> sometimes, anyway never had such thing in Linux. Only when I tilt my pen
>> steeply enough cursor can get away from the edge by some 25mm, then when I
>> straighten pen up a little it goes back to the tip, but never had a click
>> when that happened.
>>
>> Do you really think it could be a pen issue? Frankly, I'm willing to
>> believe that, because it will be lesser evil for me.
>> But I don't know where to find a compatible pen, could you tell me which
>> one I can try to find? if getting another one I wish to get thicker one more
>> like Intuos Grip Pen.
>>
>>
>>
>> On Fri, Jan 21, 2011 at 14:51, Chris Bagwell <[email protected]>wrote:
>>
>>> Wow, I do not know where i got line 171 from. Its around line 721
>>> that it needs to be done. To give you some context, it would be added
>>> after this to reset pressure values:
>>>
>>> /* pressure */
>>> ds->pressure = coord.pressure;
>>>
>>> BTW, Nicolas posted agreeing with Ping that tilt issue sounds very
>>> much like a stylus problem.
>>>
>>> Can you confirm one more time that you do not see similar behavior
>>> under Windows with same stylus? I think I recall you saying behavior
>>> is different. Is it different after you made MaxZ=255 change?
>>>
>>> Chris
>>>
>>> On Fri, Jan 21, 2011 at 12:16 AM, AndrewP <[email protected]> wrote:
>>> > Are you sure it should go there? I couldn't compile with this
>>> correction.
>>> > http://dpaste.org/dF7J/
>>> > On Thu, Jan 20, 2011 at 14:21, Chris Bagwell <[email protected]>
>>> wrote:
>>> >>
>>> >> So, it sounds like its basically working for you now?
>>> >>
>>> >> For tilt reporting pressure, I suspect this is just way inductive
>>> >> technology works with limitations touchscreens impose. Maybe Ping can
>>> >> confirm.
>>> >>
>>> >> The only additional tweaks I can think of:
>>> >>
>>> >> 1) Maybe pressure values from hardware should be ignored unless either
>>> >> tip or eraser press is reported by hardware. Does tip/eraser map to a
>>> >> real internal button indicating physical touch? I suspect so since
>>> >> there is a delay in ability to detect pin vs. eraser.
>>> >>
>>> >> To test this idea, edit wcmISDV4.c around line 171 and add "if
>>> >> (!coord.tip && !coord.eraser) ds->pressure = 0;"
>>> >>
>>> >> 2) Maybe hardware is reporting maximum of 127 values for a reason.
>>> >> Even though its returning 0-255 we might need to ignore all pressure
>>> >> values below 128.
>>> >>
>>> >> To test this idea, edit wcmISDV4.c around line 171 and add if
>>> >> (ds->pressure > 128) ds->pressure = 0;"
>>> >>
>>> >> Chris
>>> >>
>>> >> On Thu, Jan 20, 2011 at 4:30 AM, AndrewP <[email protected]> wrote:
>>> >> > Finally I saw maxZ=255 in Xorg.0.log when set maxZ to 255 in git
>>> drivers
>>> >> > also xinput shows level 2048 when I press much stronger. But still
>>> it
>>> >> > catches
>>> >> > pressure very willingly:
>>> >> > I get values about 800-1000 with light press.
>>> >> > 0-168 when not even touching it. Don't see anything wrong when
>>> holding
>>> >> > pen
>>> >> > exactly straight up. But when holding in my natural way it's even up
>>> to
>>> >> > 168.
>>> >> > If I tilt it
>>> >> > even more I get values around 600-800 easily (no touching).
>>> >> > After values of 1600 it feels pretty linear.
>>> >> > Tried several Threshold values as it should be it affect only when
>>> >> > "click"
>>> >> > is actually
>>> >> > recognized, thus limiting a pressure range.
>>> >> > On Wed, Jan 19, 2011 at 15:02, Chris Bagwell <[email protected]>
>>> >> > wrote:
>>> >> >>
>>> >> >> On Tue, Jan 18, 2011 at 4:18 PM, AndrewP <[email protected]>
>>> wrote:
>>> >> >> > How do I tell if some other driver is using the device?
>>> >> >> > I removed tablet device from xorg.conf
>>> >> >> > and I guess inputattch won't keep working if serial port is
>>> already
>>> >> >> > used.
>>> >> >> > I get pressure in gimp but it feels way less then even 127
>>> levels. of
>>> >> >> > course
>>> >> >> > it is very
>>> >> >> > subjective but I get around 30-40 levels and I need to press
>>> really
>>> >> >> > light to
>>> >> >> > be able
>>> >> >> > to use it.
>>> >> >> >
>>> >> >>
>>> >> >> After you upgraded to latest git, you did not re-patch wcmISDV4.cc
>>> to
>>> >> >> force MaxZ to 254. Can you do that and try gimp again?
>>> >> >>
>>> >> >> Please realize that without this change that pressure of 127 to 254
>>> >> >> will all equal 2048. Also, what did you end up setting your
>>> Threshold
>>> >> >> to? Somewhere around 800?
>>> >> >>
>>> >> >> Divide what ever that Threshold value is by 16. Example:
>>> 800/16=50.
>>> >> >> So this example means that only pressures of 50-127 are useful for
>>> >> >> drawing. A very small range indeed and very much shifted to light
>>> >> >> pressure side.
>>> >> >>
>>> >> >> If you change MaxZ=254 then it becomes 800/8=100 so 100-254 are
>>> valid
>>> >> >> pressure ranges. A much greater range. You'll want to also
>>> adjust
>>> >> >> your Threshold value down by 1/2 since Thresholds are different
>>> units
>>> >> >> once the MaxZ=127 bug is fixed (the 16 vs 8 I mentioned). So
>>> halving
>>> >> >> to Threshold=400 would give you pressure ranges of 50-254.
>>> >> >>
>>> >> >> Chris
>>> >> >
>>> >> >
>>> >
>>> >
>>>
>>
>>
>
------------------------------------------------------------------------------
Special Offer-- Download ArcSight Logger for FREE (a $49 USD value)!
Finally, a world-class log management solution at an even better price-free!
Download using promo code Free_Logger_4_Dev2Dev. Offer expires
February 28th, so secure your free ArcSight Logger TODAY!
http://p.sf.net/sfu/arcsight-sfd2d
_______________________________________________
Linuxwacom-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/linuxwacom-devel
