Strange wacom tablet behavior (clarification)

SDL Development
1

It’s been brought to my attention that my previous post was unclear.

I’m attempting to use a wacom tablet as a mouse.

In the test program graywin.c, the mouse functions normally in both
windowed and fullscreen modes. The tablet functions normally in
windowed mode, but in full screen mode exhibits some very unusual
behavior.

In full screen mode (640x480), cursor response to tablet movement is
extremely sensitive. Movement of a few millimeters on the tablet is
sufficient to move the cursor across the entire screen. Additionally,
the cursor only appears in specific positions on the screen, roughly
the points of a 100 pixel grid. It seems that movement of one unit on
the tablet elicits a change in posiition on the order of 100 pixels.

Any suggestions as to cause or further tests would be very helpful.

Thanks,

Theo Rhodes.__________________________________
Do you Yahoo!?
Check out the new Yahoo! Front Page.

2

I don’t profess to know anything about this, but to me it sounds like
your tablet, which is configured to run (I assume) at a higher
resolution than 640*480 – therefor stretching the movements out on
screen more pixels – is not re-configuring the movement to run at the
smaller resolution. I hope this makes sense.

What I’m getting at is a problem involving relative distances. The pen
moves 1 cm on the tablet, and to transpose that to screen coords, the
drivers move your mouse cursor so many pixels to make the movement seem
equivalent. When you switch resolutions, however, the "equivalent"
movemnt on screen SHOULD be much lower, but it’s not making the shift.
Thus it looks to move much farther.

The other thing I cam think of is that something with the pad is
doubling the mouse inputs, so that instead of moving 50 units, it gets
twice the input in the same amount of time, thinking you moved 100.

Again, I know nothing really about how the SDL code works, nor how the
drivers translate movement to screen coordinates; just some thoughts on
the matter.

–Scott

Skygod wrote:> It’s been brought to my attention that my previous post was unclear.

I’m attempting to use a wacom tablet as a mouse.

In the test program graywin.c, the mouse functions normally in both
windowed and fullscreen modes. The tablet functions normally in
windowed mode, but in full screen mode exhibits some very unusual
behavior.

In full screen mode (640x480), cursor response to tablet movement is
extremely sensitive. Movement of a few millimeters on the tablet is
sufficient to move the cursor across the entire screen. Additionally,
the cursor only appears in specific positions on the screen, roughly
the points of a 100 pixel grid. It seems that movement of one unit on
the tablet elicits a change in posiition on the order of 100 pixels.

Any suggestions as to cause or further tests would be very helpful.

Thanks,

Theo Rhodes.

3

I get similar behaviour in fullscreen mode for my tablet, an ACECAD Flair.

Anyone got some idea why?

/OlofOn Fri, 05 Nov 2004 11:11:26 -0700, Scott Harper wrote:

I don’t profess to know anything about this, but to me it sounds like
your tablet, which is configured to run (I assume) at a higher
resolution than 640*480 – therefor stretching the movements out on
screen more pixels – is not re-configuring the movement to run at the
smaller resolution. I hope this makes sense.

What I’m getting at is a problem involving relative distances. The pen
moves 1 cm on the tablet, and to transpose that to screen coords, the
drivers move your mouse cursor so many pixels to make the movement seem
equivalent. When you switch resolutions, however, the "equivalent"
movemnt on screen SHOULD be much lower, but it’s not making the shift.
Thus it looks to move much farther.

The other thing I cam think of is that something with the pad is
doubling the mouse inputs, so that instead of moving 50 units, it gets
twice the input in the same amount of time, thinking you moved 100.

Again, I know nothing really about how the SDL code works, nor how the
drivers translate movement to screen coordinates; just some thoughts on
the matter.

–Scott

Skygod wrote:

It’s been brought to my attention that my previous post was unclear.

I’m attempting to use a wacom tablet as a mouse.

In the test program graywin.c, the mouse functions normally in both
windowed and fullscreen modes. The tablet functions normally in
windowed mode, but in full screen mode exhibits some very unusual
behavior.

In full screen mode (640x480), cursor response to tablet movement is
extremely sensitive. Movement of a few millimeters on the tablet is
sufficient to move the cursor across the entire screen. Additionally,
the cursor only appears in specific positions on the screen, roughly
the points of a 100 pixel grid. It seems that movement of one unit on
the tablet elicits a change in posiition on the order of 100 pixels.

Any suggestions as to cause or further tests would be very helpful.

Thanks,

Theo Rhodes.

SDL mailing list
SDL at libsdl.org
http://www.libsdl.org/mailman/listinfo/sdl

4

It happens with many commercial games too.

DannyOn Sat, 6 Nov 2004 18:14:53 +0100, Olof Bjarnason <olof.bjarnason at gmail.com> wrote:

I get similar behaviour in fullscreen mode for my tablet, an ACECAD Flair.

Anyone got some idea why?

/Olof

On Fri, 05 Nov 2004 11:11:26 -0700, Scott Harper wrote:

I don’t profess to know anything about this, but to me it sounds like
your tablet, which is configured to run (I assume) at a higher
resolution than 640*480 – therefor stretching the movements out on
screen more pixels – is not re-configuring the movement to run at the
smaller resolution. I hope this makes sense.

What I’m getting at is a problem involving relative distances. The pen
moves 1 cm on the tablet, and to transpose that to screen coords, the
drivers move your mouse cursor so many pixels to make the movement seem
equivalent. When you switch resolutions, however, the "equivalent"
movemnt on screen SHOULD be much lower, but it’s not making the shift.
Thus it looks to move much farther.

The other thing I cam think of is that something with the pad is
doubling the mouse inputs, so that instead of moving 50 units, it gets
twice the input in the same amount of time, thinking you moved 100.

Again, I know nothing really about how the SDL code works, nor how the
drivers translate movement to screen coordinates; just some thoughts on
the matter.

–Scott

Skygod wrote:

It’s been brought to my attention that my previous post was unclear.

I’m attempting to use a wacom tablet as a mouse.

In the test program graywin.c, the mouse functions normally in both
windowed and fullscreen modes. The tablet functions normally in
windowed mode, but in full screen mode exhibits some very unusual
behavior.

In full screen mode (640x480), cursor response to tablet movement is
extremely sensitive. Movement of a few millimeters on the tablet is
sufficient to move the cursor across the entire screen. Additionally,
the cursor only appears in specific positions on the screen, roughly
the points of a 100 pixel grid. It seems that movement of one unit on
the tablet elicits a change in posiition on the order of 100 pixels.

Any suggestions as to cause or further tests would be very helpful.

Thanks,

Theo Rhodes.

SDL mailing list
SDL at libsdl.org
http://www.libsdl.org/mailman/listinfo/sdl

SDL mailing list
SDL at libsdl.org
http://www.libsdl.org/mailman/listinfo/sdl

5

OK, this is a longshot I’ve come up with after trying out some tablet settings:

  1. In fullscreen mode,there is some kind of "virtual screen size"
    setup at some windows/directx/sdl level
  2. This size is much larger than the desktop resolution, say
    10.000x10.000 pixels or something
  3. If the tablet is in “absolute tracking mode”, each measurable
    coordinate of the tablet corresponds to lets say 100 pixels of the
    screen
  4. Hence we get the “grid behaviour”

I actually got it working in my full-screen drawing program after
changing the tracking mode to “relative”. Works great!

/OlofOn Sat, 6 Nov 2004 22:39:15 +0100, Danny van Bruggen wrote:

It happens with many commercial games too.

Danny

On Sat, 6 Nov 2004 18:14:53 +0100, Olof Bjarnason <@Olof_Bjarnason> wrote:

I get similar behaviour in fullscreen mode for my tablet, an ACECAD Flair.

Anyone got some idea why?

/Olof

On Fri, 05 Nov 2004 11:11:26 -0700, Scott Harper wrote:

I don’t profess to know anything about this, but to me it sounds like
your tablet, which is configured to run (I assume) at a higher
resolution than 640*480 – therefor stretching the movements out on
screen more pixels – is not re-configuring the movement to run at the
smaller resolution. I hope this makes sense.

What I’m getting at is a problem involving relative distances. The pen
moves 1 cm on the tablet, and to transpose that to screen coords, the
drivers move your mouse cursor so many pixels to make the movement seem
equivalent. When you switch resolutions, however, the "equivalent"
movemnt on screen SHOULD be much lower, but it’s not making the shift.
Thus it looks to move much farther.

The other thing I cam think of is that something with the pad is
doubling the mouse inputs, so that instead of moving 50 units, it gets
twice the input in the same amount of time, thinking you moved 100.

Again, I know nothing really about how the SDL code works, nor how the
drivers translate movement to screen coordinates; just some thoughts on
the matter.

–Scott

Skygod wrote:

It’s been brought to my attention that my previous post was unclear.

I’m attempting to use a wacom tablet as a mouse.

In the test program graywin.c, the mouse functions normally in both
windowed and fullscreen modes. The tablet functions normally in
windowed mode, but in full screen mode exhibits some very unusual
behavior.

In full screen mode (640x480), cursor response to tablet movement is
extremely sensitive. Movement of a few millimeters on the tablet is
sufficient to move the cursor across the entire screen. Additionally,
the cursor only appears in specific positions on the screen, roughly
the points of a 100 pixel grid. It seems that movement of one unit on
the tablet elicits a change in posiition on the order of 100 pixels.

Any suggestions as to cause or further tests would be very helpful.

Thanks,

Theo Rhodes.

SDL mailing list
SDL at libsdl.org
http://www.libsdl.org/mailman/listinfo/sdl

SDL mailing list
SDL at libsdl.org
http://www.libsdl.org/mailman/listinfo/sdl

6

In full screen mode (640x480), cursor response to tablet movement is
extremely sensitive. Movement of a few millimeters on the tablet is
sufficient to move the cursor across the entire screen. Additionally,
the cursor only appears in specific positions on the screen, roughly
the points of a 100 pixel grid. It seems that movement of one unit on
the tablet elicits a change in posiition on the order of 100 pixels.

If each of your tablet’s corners correspond to the corners of your
screen, then your tablet most likely uses a protocol designed for
absolute positioning (many cheap ones have no sense of absolute
positioning and blow to use for art.)

At any rate, what you’re experiencing is the fault of the driver for
the tablet, or perhaps, your OS vendor (gasp.) See if there is a newer
or different driver. I bought a tablet that only worked with its
specific drawing program a long time ago, big mistake. Someone was nice
enough to write a driver for it though so it would behave as an
ordinary pressure sensitive pointer device.

Good luck.On Nov 5, 2004, at 12:51 PM, Skygod wrote: