Bongofish

Screen Tablet malarky => Heyaaaalpppp => Topic started by: bumhee34 on October 31, 2013, 03:17:03 AM



Title: Help with calibration with waxbee
Post by: bumhee34 on October 31, 2013, 03:17:03 AM
am trying to find best mapping with su 1206e 01x

the real active area of my panel is 245.70 (H) X 184.275 (W)

and the hid listener is showing the following message when i upload debug template.
[attachment=1]

Please tell me what to modify in template. Thx in advance.


Title: Re: Help with mapping with waxbee
Post by: bernard on October 31, 2013, 03:32:44 AM
Hum, it seems the debug did not correctly found the max X and Y. the Pressure is also way off and the Tilt also.  All values are wrong. Do you always get those weird values?  try it again please. Power the Teensy and the board (i.e. connect the USB) without the pen touching it.



Title: Re: Help with mapping with waxbee
Post by: bumhee34 on October 31, 2013, 07:53:21 AM
Just like this without pen


Title: Re: Help with mapping with waxbee
Post by: bumhee34 on October 31, 2013, 07:54:44 AM
with pen in proximity, it shows different numbers


Title: Re: Help with calibration with waxbee
Post by: bernard on October 31, 2013, 01:59:07 PM
This is not normal. Without the pen it should have received the data. It seems only 3 bytes got in.

Next step is to do it manually:  (I should write a fixed page about this, I think it must be the 20th+ time I write this in this forum).

Flash the virtual serial port .hex file using the Teensy loader. This is found WaxBee download section. This will create a COM port on your system. Use a serial terminal tool like RealTerm that can display hex bytes.  Set it to the appropriate baud rate and enter the commands as per http://sourceforge.net/apps/mediawiki/linuxwacom/index.php?title=ISDV4_Protocol
Set the baud rate to the matching speed in the template.  (was that 38400?)  In RealTerm, put the display to "Hex".
Type 0 (zero) to stop sending pen coordinates, Type 1 to to start it back.  -- this is to check that the communication works.
Type * so it outputs the special packet that talks about the X/Y max, etc. It is all binary, but this packet has more bytes than the usual packet for pen coordinates.



Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 01, 2013, 02:10:39 AM
I gave up. Don't know why, but under windows 8, the imaginary "COM" port is not activated.
So i could not communicate with my tablet board through usb serial port. Tried with putty, teraterm, and realterm, none of them worked out.
Maybe I did some mistakes with installing usb serial driver. I uploaded usb serial hex on teensy anyway.
Maybe I just need to modify Slave X and Y size by manual (increasing numbers or decreasing..step by step until perfect match).


Title: Re: Help with calibration with waxbee
Post by: bernard on November 01, 2013, 02:17:24 AM
http://www.pjrc.com/teensy/usb_serial.html

Big sorry. Forgot to tell you about this. Follow the steps to enable on Windows.


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 01, 2013, 04:21:04 AM
I know about this one. But, actually i cannot install the driver, strange.. Error occurred while installing all the time eventhough with Admin privileges.


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 01, 2013, 04:35:17 AM
Solved, there is an issue on windows 8 with signatures. I will update the results later.


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 01, 2013, 04:41:55 AM
Just for testing, when pen is in proximity, realterm shows the following characters.

Actually I do not understand what to do with the link, "http://sourceforge.net/apps/mediawiki/linuxwacom/index.php?title=ISDV4_Protocol".

What to do next?


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 01, 2013, 11:48:15 AM
This is what i've gotten upto now.

with 38400, if i send *, no response.
after changing to 19200 when i type *, it shows C0 2F 7E 23 and so on.

Is this important information?


Title: Re: Help with calibration with waxbee
Post by: bernard on November 01, 2013, 02:18:45 PM
Yes!  After the * it is supposed to be 11 bytes.  Let me interpret those bytes... (this follows the following table: http://sourceforge.net/apps/mediawiki/linuxwacom/index.php?title=ISDV4_Protocol#Format_of_Stylus_Query_Response )


0  1  2  3  4  5  6  7  8  9  10
C0 2F 7E 23 7F 7F 41 00 00 00 05

                      #7654 3210
bits of byte 6: (41) = 0100 0001


(numbers are in hexadecimal)
Max X = 2F << 7 + 7E << 2 + 2  = 1780 + 1F8 + 2 = 197A  or decimal 6522
Max Y = 23 << 7 + 7F << 2 + 0  = 1180 + 1FC + 0 = 137C or decimal 4988
Max Pressure =  1 << 7 + 7F = FF or decimal 255
No tilt X nor Y
Version = 5

In short:

Max X = 6522
Max Y = 4988
Max Pressure = 255 (as expected)
No Tilt (as expected)

If we look at the physical active area that you told us : 245.70 (X) x 184.275 (Y) that makes about the same ratio and about ~0.0376 mm resolution.


Title: Re: Help with calibration with waxbee
Post by: bernard on November 01, 2013, 03:47:37 PM
I do not know why we are missing bytes when running in debug mode -- we only see the first 3-4 bytes and then nothing. Weird. Anyhow. Probably I should add some "wait" time in there to make it more robust.  It does not matter to you since this is only to know about what to put in the template.

Eventually, I might add a feature to auto-detect the size. There are some many variations of ISDV4 models, everybody has to go through a debug phase to find the max x and max y.


Title: Re: Help with calibration with waxbee
Post by: bernard on November 01, 2013, 07:10:15 PM
About windows 8 & signatures: How did you finally managed to install it?



Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 02, 2013, 02:09:00 AM
About windows 8 & signatures: How did you finally managed to install it?



I refer to this link http://www.pjrc.com/teensy/windows8.txt (http://www.pjrc.com/teensy/windows8.txt).

So finally, u interpret the bytes those are 6522 and 4988.
So, in the intuos template, what should I modify? Thx for your great help.


Title: Re: Help with calibration with waxbee
Post by: bernard on November 02, 2013, 03:29:47 AM
Modify the following:

SLAVE_X_MAX = 21240
SLAVE_Y_MAX = 15980

to be:

SLAVE_X_MAX = 6522
SLAVE_Y_MAX = 4988



Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 02, 2013, 04:18:58 AM
totally mismatching.. the cursor and the pen.

the resolution of the panel is 1440*1050.


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 02, 2013, 04:22:41 AM
When I use 24587, 18447 (no specific reasons for those number). the cursor and pen is somehow ok.
But in the case of 6522 and 4988, the discrepancy is really really big like the panel size itself..


Title: Re: Help with calibration with waxbee
Post by: bernard on November 02, 2013, 07:21:40 AM
I assume you kept the USB_X_MAX and USB_Y_MAX untouched from the original template, right?
Your mapping in the Wacom driver is set to fully to the screen and to not force proportion, right?

I personally find those numbers quite small -- just that it is really what the board reports.  Strange that is.

I sent you a special waxbee to try. Check your mail.



Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 02, 2013, 12:57:31 PM
I assume you kept the USB_X_MAX and USB_Y_MAX untouched from the original template, right?
Your mapping in the Wacom driver is set to fully to the screen and to not force proportion, right?
Both are right.
Still looks strange.


Title: Re: Help with calibration with waxbee
Post by: bernard on November 02, 2013, 02:53:22 PM
A0 00 2E 23 63 00 50 00 00
0  1  2  3  4  5  6  7  8

             7654 3210
byte 6: (50) 0101 0000


x = 00 << 7 + 2E << 2 + 2
y = 23 << 7 + 63 << 2 + 2


x = 0 + 184 + 2 = 186
y = 4480 + 396 + 2 = 4878



X was not at the right with just 186. Y was 4878 - was it close to the edge? Trying to find max values coming out of the tablet. Oh have you rotated the tablet 90 degrees?


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 03, 2013, 09:20:18 AM
nope, maybe I pointed to wrong edge point.

Still, the max y looks wierd, right? 4878 is still too small... strange.. no clue at all x.x


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 03, 2013, 09:24:07 AM
I test with 4 all edge points.

First : A0 00 38 23 7F 00 60 00
Second : A0 00 46 00 70 00 58 00
Third : A0 2F 0B 23 2F 00 68 00
Last : A0 2F 6D 00 4F 00 50 00


Title: Re: Help with calibration with waxbee
Post by: bernard on November 03, 2013, 06:40:33 PM
Hum, your packets have 8 bytes (I think it is missing the last one) -- I will assume 00 for the last one.


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 04, 2013, 12:09:56 AM
I checked again, sorry for my mistake.

Your assumption is right, the last one is 00 for all cases.


Title: Re: Help with calibration with waxbee
Post by: bernard on November 04, 2013, 02:48:22 AM
Here's the interpretation of X and Y of those packets:



First :
0  1  2  3  4  5  6  7  8
A0 00 38 23 7F 00 60 00 --

                      #7654 3210
bits of byte 6: (60) = 0110 0000

X = 00 << 7 + 38 << 2 + 3 = 0 + 224 + 3    = 227
Y = 23 << 7 + 7F << 2 + 0 = 4480 + 508 + 0 = 4988

Second :
0  1  2  3  4  5  6  7  8
A0 00 46 00 70 00 58 00 --

                      #7654 3210
bits of byte 6: (58) = 0101 1000

X = 00 << 7 + 46 << 2 + 2 = 0 + 280 + 2 = 282
Y = 00 << 7 + 70 << 2 + 3 = 0 + 448 + 3 = 451


Third :
0  1  2  3  4  5  6  7  8
A0 2F 0B 23 2F 00 68 00 --

                      #7654 3210
bits of byte 6: (68) = 0110 1000

X = 2F << 7 + 0B << 2 + 3 = 6016 + 44 + 3  = 6063
Y = 23 << 7 + 2F << 2 + 1 = 4480 + 188 + 1 = 4669


Last :
0  1  2  3  4  5  6  7  8
A0 2F 6D 00 4F 00 50 00 --

                      #7654 3210
bits of byte 6: (50) = 0101 0000

X = 2F << 7 + 6D << 2 + 2 = 6016 + 436 + 2 = 6454
Y = 00 << 7 + 4F << 2 + 2 = 0 + 316 + 2    = 318



Results: Within these 4 samples, the biggest coordinates seen for X and Y are:

Highest X = 6454
Highest Y = 4988

Which fits **almost perfectly**  within the computed values seen earlier (is equals or less):

SLAVE_X_MAX = 6522  <<--- 6522-6454 =  68 * ~0.03mm = or about 2mm away from the edge!  
SLAVE_Y_MAX = 4988  <<--- dead on!

That confirms my theory: those SLAVE_MAX_x values are correct.

Re-reading your text, I think you might be doing a mistake (not sure): These numbers should be put within WaxBeeConfig and flashed to the Teensy. (NOT in the Wacom driver as the mapping "counts").  Is this what you did?

[attachment=1]


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 04, 2013, 04:55:13 AM
Yup, that is what i did exactly. Quite wierd... maybe we have incomplete knowledge about wacom sensor board.

Hope we can find something in future. At this moment, those numbers lead to "very" unreasonable result.


Title: Re: Help with calibration with waxbee
Post by: bernard on November 04, 2013, 06:57:09 AM
ISDV4 tablets are quite simple and have no configuration. In other words, the "tablet" is fine and behaves as expected. The problem either lies within the mapping logic & parameters in WaxBee/Teensy code or in the Wacom Windows driver mapping logic or parameters.

We'll dig more and see where it "breaks". There isn't much to it, we just have to be a little patient and we will find it, I am sure.  :)

In the meantime, make sure you also have the following in the WaxBee Config for the non-debug template that emulates an "Intuos2 XD-1218-U":

USB_X_MIN = 0           (left)
USB_Y_MIN = 1190      (top)        NOTE: this number is not 0 to avoid invoking the top menu stripe of the Intuos 2.
USB_X_MAX = 45720    (right)
USB_Y_MAX = 31680    (bottom)





Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 04, 2013, 07:20:27 AM
Yup, I made sure of it.


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 04, 2013, 08:07:15 AM
At this moment, I am using my theory(?).
With assumption that both intuos and wacom sensor board have same resolution, I am finding min x and max x of the slave.

Info : the panel is 12.1 inch. So, x is 9.68" and y is 7.26".
        intuos : x=45720 with 18", y=(31680 - 1190) with 12"

So, slave max x will be 45720/18*9.68= ~ 24587
And, slave max y will be (31680 - 1190)/12*7.26=~ 18447

And if you remind the active area of panel, it is amazing : 245.70 (H) X 184.275 (W).
Very close values!

Anyway, with above numbers, the mapping is quite good except along edge lines. Usually, tablet pc has some discrepancy between pen and cursor along edge lines, but in this DIY tablet, the difference looks larger than normal tablet PCs.

But anyway, with above numbers, it is ok in usual working area. Still wanna remove the discrepancy along edge lines.


Title: Re: Help with calibration with waxbee
Post by: bernard on November 04, 2013, 03:04:42 PM
Mapping issue on edges:

If the mapping is non-linear OR LCD edges is missing sensor active region underneath (i.e. sensor smaller than LCD or sensor misaligned), then we can always use WaxBee "anchor" feature to tweak the mapping to "approach" a little bit more to your taste. Anchors are used to "break" the mapping into 3 segments per axis instead of one, so you can align differently in the middle vs. the edges.

But before playing with anchors, we should make sure we did the maximum on "normal" mapping.


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 05, 2013, 12:08:55 AM
Yes u r right, and in my opinon the mapping is non-linear based on my experience.

The clue is that the cursor is following somehow linear but when approching the edges the moving speed is weird. looks like slow to fast. (Maybe I am not correct, just guess)

Still we have no idea about maximum values.

Anyway, to play with anchor, what number should i type on anchor box?


Title: Re: Help with calibration with waxbee
Post by: bernard on November 05, 2013, 07:00:23 AM
Anchors:  Not sure how to explain. You deal with one axis at a time.  For example, if we take the left side and right side (X axis), you have to put a numbers in the slave anchor x and numbers in the USB anchor x.

The following example is "normal linear" mapping from MIN to MAX.  SLAVE (Serial) MIN must be converted to the value "USB MIN"
        MIN                      MAX
Serial:  |--------------------------|
                               
linear mapping to:
USB:     MIN                                        MAX
         |-------------------------------------------|



 

The following example is has "anchored" mapping from MIN to Anchor points to MAX.  SLAVE (Serial) MIN must be converted to the value "USB MIN", the Slave Left Anchor must be converted to the USB Left Anchor (same thing to the right anchor).  This creates 3 segments instead of one. We just this, we can "simulate" simple non-linear "acceleration" near the edges.

        MIN   Left         Right   MAX
Serial:  |-----|-------------|------|
             Anchor        Anchor           
        
"Anchored" mapping to:

USB:    MIN  Left                       Right        MAX
         |----|---------------------------|------------|
            Anchor                      Anchor         


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 05, 2013, 08:50:44 AM
This is very good idea. I will try with some numbers by manual.

So the point is that we can convert non-linear mapping to linear-like mapping or vice versa.


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 05, 2013, 09:15:32 AM
One simple question, is the anchor point value from the nearest surface? or from left or down?


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 05, 2013, 09:38:50 AM
No need to play with anchor. My sensor board was not flatten due to my case.

With 24587 & 18447, the mapping accuracy is acceptable.

But still have question,, why does the set, 6522 & 4988, not work?


Title: Re: Help with calibration with waxbee
Post by: bernard on November 05, 2013, 02:30:47 PM
We'll try to discover that tomorrow over our live exchange.


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 06, 2013, 01:27:13 AM
Found one interesting problem.

There is no jitter.

BUT! When I draw, then the drawing looks have jitter!

I compared this new tablet to my DIY intuos, this has too many jitter-like line.

I upload the result on MS OneNote. The lines are not smooth, and i can tell that is not smooth cause i have many experience with tablets.

Maybe the emulation is not perfect? I am not sure. And I do not know whether this kind of problem is unique for me or others also have.


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 06, 2013, 01:33:47 AM
And this is not the problem of LCD. I drew something with LCD turned off, and still there are lots of jitter-like lines.


Title: Re: Help with calibration with waxbee
Post by: bernard on November 06, 2013, 04:09:48 AM
This digitizer has the lowest resolution I have ever seen. I do not know if it is just that or there is really a problem. I computed around 0.03mm resolution.  I have an idea on how to test that. Remap the screen to a 5 cm square on the tablet. Configure a paint software to only paint 1 pixel dots (not lines) so we can see the "steps". Then we see if stuff are happening normally or not. Use a ruler to draw a diagonal (30 degrees roughly) line.


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 06, 2013, 05:25:09 AM
Ok, let's figure it out later.

And I also tested with windows journal(?) (the default tablet pc note).

I upload for the cases windows journal and MS OneNote.

Looks onenote is worse.

And actually, I do not understand your suggestion with MS painter. Should I enlarge the diagonal drawing result?


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 06, 2013, 05:28:11 AM
I don't know whether this is what you want.

Anyway, I draw a line using a ruler on normal screen and I enlarged the drawing to see the pixels.

Second drawing is that I draw with a ruler, but on the enlarged screen.


Title: Re: Help with calibration with waxbee
Post by: bernard on November 06, 2013, 06:09:47 AM
The idea is not to see "pixels zoomed in".  The idea is to change the mapping in the Wacom driver to map a tiny area on the tablet to "huge area" on the screen. (like the whole screen).  That way we can see the resolution of the tablet on the screen. We can analyse the positions visually and interactively.





Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 06, 2013, 06:16:46 AM
I see. Let's do some experiments when you have time.
Today, I need to focus on my own research (a graduate student).
Just let me know what to do then I will update this post again.
Thx, bernard, you are really helpful.

*and sorry for my poor Engish :(


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 07, 2013, 01:39:40 AM
And I wanna know that I can only use intuos2 emulation for this tablet.

Is it impossible to use ISDV tablet driver? or Graphire something like that.

I tried with Graphire 3, but failed. Maybe trying with other drivers help to fix the jitter-like drawing issue.

I cannot imagine this sensor board have low resolution.


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 07, 2013, 02:16:45 AM
Found little bug in waxbee template.

If you look at the template PEN to Graphire3, the slave protocol is set to wacom adb.

If I modify to ISDV4 protocol, it works. But still jitter like inking..:(


Title: Re: Help with calibration with waxbee
Post by: bernard on November 07, 2013, 03:44:49 AM
Thanks for pointing out the bug about the Penenabled-Graphire3 template, indeed using the ADB protocol is totally wrong!

For your "jitter", we can investigate to see what is the root cause of the issue and if there is something that can be done about it.

Emulating a USB penenabled would be doable, but it is not implemented. Takes time to do these things and beyond having a 1:1 match with the original hardware there isn't much concrete benefit to it other than benefiting from a screen calibration utility (which is a nice feature actually).

Like I explained earlier, the next step is to analyze the "jitter" closely.  For that, map a tiny area of your board to a huge area on the screen (like the whole screen).  Close enough to see the resolution of the tablet -- if you draw on the screen with just 1 pixel dots (no lines), you would be supposed to see a spaced "imaginary grid" of dots. Like one dot per grid intersection. Then once we have that setup, we clear the canvas and draw a single diagonal line following a ruler:  The dots that are picked should follow a "regular stairs" pattern. There shouldn't be dots appearing outside of that stair pattern, if there are then it is jitter if not, then you are seeing the resolution of the tablet.  There is also a possibility that the scaling logic introduces issues with "aliasing". (before getting on the screen, there are 2 scaling that are happening:  coordinates of the original tablet scaled to the emulated USB tablet, then it is re-scaled again to the screen coordinates.

One way to reduce the chance of aliasing is to use an "emulated" tablet with the highest resolution. Here I am thinking the Intuos2 12x18 -- there is well enough precision in there o avoid any effect from the double re-scaling.


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 07, 2013, 05:16:15 AM
Cause I am a newbie, I have one stupid question.

"aliasing" algorithms are different from OS? I am using Win 8 64bit now. Not sure whether it is related.

And I will try your suggestion later.


Title: Re: Help with calibration with waxbee
Post by: bernard on November 07, 2013, 05:32:50 AM
I cannot answer your question because I do not understand it.  I think I'll just say: Forget about the word "aliasing (http://en.wikipedia.org/wiki/Aliasing)" -- I wasn't sure how to describe the type of "number rounding-errors" that can arise when performing 2 "scaling" operations in a row. (under certain conditions).


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 07, 2013, 06:39:29 AM
Using tablet setting, I can remap small portion of sensor board to the whole screen.

In this way, I do not need to modify waxbee template, am i right?

So, now I need to test with the small area of sensor board.


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 07, 2013, 06:48:16 AM
And this is the capture from the drawing with a ruler.

If I understood wrong again, I would feel shy, haha.

Anyway, is this correct way to check?

Anyway, I can see jitter-like drawing cleary.


Title: Re: Help with calibration with waxbee
Post by: bernard on November 07, 2013, 07:39:57 AM
I think you got the point, just that we need to draw dots -- not lines -- I mean -- if you draw with your mouse and move very fast you should see many pixels that are not touching each other. Because of the lines we cannot analyse the data. 

Whatever software you use to draw the dots does not matter.



Title: Re: Help with calibration with waxbee
Post by: bernard on November 07, 2013, 08:36:38 AM
Ok, I tried to find a free software that would do this but no luck. They all "connect the dots"!

So I wrote one in javascript. ->  http://jsfiddle.net/SAEfU/   

Hit "Run" and draw in the black rectangle on the right. (tried in Google Chrome)


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 07, 2013, 08:58:31 AM
Here, I tried to draw rectangle.


Title: Re: Help with calibration with waxbee
Post by: bernard on November 07, 2013, 09:17:26 AM
I mean, with the special mapping described above?


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 07, 2013, 09:36:10 AM
Yes, I track the cursor and try to draw rectangle cause the sensor board area is remapped into small area.

Is not it what I should do?


Title: Re: Help with calibration with waxbee
Post by: bernard on November 07, 2013, 02:18:09 PM
yeah, but it seems it is not "powerful" enough to see some space between the pixels.  Try to map a much smaller area on your tablet (like 1 cm or even less) -- put tape or something to remember where it is physically on the tablet.  What you can do is to modify the "counts" (section number 3. in the device driver window) instead of playing with the little square (for which there is not much precision).   Make the count difference in each axis to be around 500. For example,  lefT:10000 and right:10500  and top:10000 bottom:10500.   Hopefully 500 is significantly less than your screen resolution. (like at least half less).



Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 08, 2013, 12:09:04 AM
What you can do is to modify the "counts" (section number 3. in the device driver window) instead of playing with the little square (for which there is not much precision).   Make the count difference in each axis to be around 500. For example,  lefT:10000 and right:10500  and top:10000 bottom:10500.   Hopefully 500 is significantly less than your screen resolution. (like at least half less).
I made around 2.5cm square because it is minimum requirement in the tablet setting.
Anyway, I do not get it. Modify "counts"? I don't know what device driver window you mean.
Sorry for my poor understanding.


Title: Re: Help with calibration with waxbee
Post by: bernard on November 08, 2013, 12:59:18 AM
Counts is the english label of the item 3 in the dialog woth the 4 numbers - I mean your screenshot in this thread. I cannot reproduce the labels because it uses an alphabet I cannot know how to reproduce here.


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 08, 2013, 01:23:57 AM
Ok, but as the setting says I cannot make it 500 cause it is too small.

Anyway, I set like as follows, as small as possible i can achieve.


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 08, 2013, 01:24:48 AM
If this is not enough, I need to modify waxbee template. In that case, should i only modify slave size? or USB also?


Title: Re: Help with calibration with waxbee
Post by: bernard on November 08, 2013, 03:05:18 AM
What is the smallest "Count" you can achieve (do not drag the handles using the mouse, try to type a few numbers until you find the smallest)?

I have issues here and can't try it on my machine at the moment.



Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 08, 2013, 05:16:35 AM
These are minimum differences I can get. They are around 3000.


Title: Re: Help with calibration with waxbee
Post by: bernard on November 08, 2013, 04:40:05 PM
ok, now if you draw on the jsfiddler page, what do you see?

I might be forced to build an app to analyse this further. But time is on essence here, I'll try to think what can be done.


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 09, 2013, 05:22:02 AM
Same as before.


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 16, 2013, 02:30:28 AM
Update.

I am pretty sure that it is not the problem of wacom board.

I am working with another wacom board from TC4200, cause one of my friends asked me to mod it.

After soldering and emulating, this different board from what i have, also shows jitter-like lines when drawing.

Maybe is this problem from OS? I am using win 8 64 bit, but no clue that it is coming from OS. Maybe I need to test on another OS.

Before testing on another OS, I tested with another VGA cable (I've used DVI cable) still no luck.


Title: Re: Help with calibration with waxbee
Post by: bernard on November 18, 2013, 08:47:34 AM
jitter: If you didn't already, you can try with the digitizer alone (no LCD in the way). 

Can you explain a little bit better what is the jitter you are seeing exactly. It is not clear which part is jitter and which is not. (make an overlay of what is supposed to be the expected output).

Does the jitter appear to always be from one axis only? (X or Y)


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 21, 2013, 01:40:55 AM
Tried with digitizer alone, but same.

Looks both axis have jitter lines.

I upload the capture, and circled with red line that i think jitter-like lines are. (the circle is done not by pen. don't misunderstand)


Title: Re: Help with calibration with waxbee
Post by: bernard on November 21, 2013, 04:00:26 AM
ok, I trust you when you say you have jitter, but I cannot appreciate/understand the effect from your analysis.   

One way to test is to not move the pen (put it in a plastic container or a in a glass or in ceramic cup or anything else that can make it stands relatively vertically with no influence from your hand) and see if the mouse on screen never seem to stop moving/jumping erratically around the pen.  Another way to test is to make straight lines with a ruler. 

Another thing: -- do you have more jitter when applying more pressure?

I did a fix recently but I did not publish that version. I can send you that version privately for you to try.  If we are loosing bytes on the serial port, this is (supposed to) fix it. This version has not been extensively tested and may not work very well, but, on the other hand, it might actually make a difference!  Who knows.


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 21, 2013, 04:50:54 AM
I can say that there is no jitter at all. That is why I am using the term "jitter-like".
There is no movement of mouse points when the pen is stagnant, but it only appears when I draw the line.
I checked the new waxbee, but still no luck.
The pressure effect is negligible.
And for the help, I attached the drawing on photoshop.
In this case, the lines are better maybe due to algorithm, but still u can see jitter-like lines compared to other wacom tablets.


Title: Re: Help with calibration with waxbee
Post by: bernard on November 21, 2013, 02:29:22 PM
Sorry, until I see a more clear evidence of what is "wrong", I won't be able to help you. Something I can verify like "making" a straight lines with a ruler for instance (something I can compare) -- here you are just highlighting piece of letters. Like I said: I cannot make the difference between your intention and the "jitter"!  These look like normal hand-drawn letters.

Reduce your issue to the smallest and simplest stroke possible.  A stroke when I can clearly see the issue.

Have you tried lazy nezumi ?  Some people appreciate what it does.

If your line is "too long" then have you played with the Wacom pressure settings to cutoff at the "good moment"?

If you have an algorithm in your head that could help you out and cannot be done with other means, then I might be able to pull it off within WaxBee.



Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 22, 2013, 04:53:54 AM
Here, you can see the problem clear.

The lines are drawn using ruler, and as you can see the lines are now smooth.

I mean there are something like step-jump during drawing (red circles).

This must be the problem of low resolution of the wacom sensor board. And as we discussed before, the debug mode gives us wrong resolution information.

I think waxbee or something else might interpret the signals from the wacom board "wrong". This is not my filed, so i cannot modify something to cure this.


Title: Re: Help with calibration with waxbee
Post by: Aerendraca on November 22, 2013, 11:33:17 AM
Hmm, this is interesting. Do these little ramps appear in the same place every time?

You use PhotoShop right? So, I would try sticking a ruler to the screen with some tape (so you can keep the path the same), draw a line on a blank canvas zoomed to 100%, then where the ramps appear use the guides to mark their position. Once done erase the line and draw again, see if it happens in the same place. Alternatively, draw lines of 1px width changing the colour after each stroke, then see if you can notice any color other than the last colour you used; make sure you use colours that have good contrast against each other and against the background.


Title: Re: Help with calibration with waxbee
Post by: bernard on November 22, 2013, 05:28:08 PM
Ok -- let's try again using the same jsfiddle . I did this to show "single pixel dots" only -- no lines. The  "lines" "fudges" the reading. We cannot make a good analysis of what really is happening.

http://jsfiddle.net/SAEfU/

Please do the following. It should be very easy to do:

#1- Do many lines using a ruler -- do the lines very slowly to avoid skipping pixels, and be extremely careful to stick to the ruler.

#2- Attempt to "fill" a small region with pixels to make it a solid color -- just by drawing pixels. If we are "skipping" pixels, that means some pixels you cannot access and you won't be able to fill a region.


Title: Re: Help with calibration with waxbee
Post by: bernard on November 22, 2013, 05:46:38 PM
That jsfiddle will only test at the "pixel" accuracy. To see if we are "skipping pixels".

I believe in Photoshop you can get "sub-pixel accuracy" which is something I am not sure how to test (yet).  And that could be related to your issue.  Once we are clear that we are -- or are not -- skipping pixels, then we can look further. 


Title: Re: Help with calibration with waxbee
Post by: bumhee34 on November 23, 2013, 01:07:53 AM
Quest #1

I was not extremely careful T^T, anyway looks like there is no skipping pixels with this test.


Title: Re: Help with calibration with waxbee
Post by: bernard on November 23, 2013, 03:15:35 AM
yes, you are right, no skipping of pixels indeed. 

Scratching my head on how to make a test that would reveal with the actual accuracy (or jitter) we are getting at the "sub-pixel" level...