wacom's newest tabletPC driver issues!

Discussion in 'Fujitsu' started by stoneseeker, Oct 12, 2012.

Thread Status:
Not open for further replies.
  1. Plognark

    Plognark Scribbler - Standard Member

    Messages:
    176
    Likes Received:
    25
    Trophy Points:
    41
    I've been considering it too :D

    I'll have to give this a shot tonight. It's consistent with what we know so far. Losing pressure sensitivity in PS is a serious downer; I use a lot of layer tricks and brush settings that aren't available in SBP.

    Not knowing exactly how the firmware communicates, I wonder if it's getting double requests from two different drivers that's causing the glitch? I've seen some threads on other forums talking about SBP where they had to disable the ISD drivers to get proper pen performance back, but nothing like what we're seeing here. If the glitch does still occur in SBP without the wacom ISD drivers installed then it's probably not a driver conflict, although that certainly could be making the issue worse.

    SBP glitches a lot less for me than Photoshop. If it's related to how the drivers communicate back to the firmware, I'd imagine that the stock windows pen drivers are a lot less intense.

    ...

    You know... as I'm thinking about it, a lot of the time when the glitch triggers for me, it's when I'm opening windows or menus up, or when I close a window and jump back to the drawing canvas in PS.

    Do you think it's possible that photoshop is actually using both drivers? When moving over menus it would use the basic windows pen drivers, but then when it goes back to the canvas it fires up wintab again?

    That might explain the doubled data rate we see when the glitch hits... maybe each set of drivers is concurrently filling up a buffer of pen cursor data, and somehow there's a collision, or overlap, where they're both trying to update the pen cursor at the same time?

    I'm not sure why the glitch would still hit in SBP which doesn't use wintab... I guess I can test it out with wintab completely removed. I could test it in photoshop too, albeit with no pressure sensitivity, but that's fine. I'll use a 20% opacity brush, so when the glitch hits I'll be able to see when it strikes. And I can always run a USBpcap log along with it.
     
  2. stoneseeker

    stoneseeker Animator and Art Director Senior Member

    Messages:
    3,606
    Likes Received:
    1,077
    Trophy Points:
    231
    Hold on here, there's a few things not lining up.

    1. I get the issue just as much if not more in Manga Studio 5 and Sketchbook Pro as I do Photoshop. In fact I was working on a monstrous Photoshop file last night for 4 hours, (500mb file!!) and it maybe glitched twice in that time period. Using Flash is almost guaranteeing a glitch every 5 minutes though. So my experience is inconsistent with yours in that photoshop is better for me. Cs6 on win 7.

    2. Yes, Photoshop is most certainly using both drivers. If you look closely, everytime the cursor switches from the wacom white arrow (or tool cursor) to the windows tiny white star cursor (hover over your desktop to see the cursor I am talking about) you are witnessing the other driver being used. This is true of all windows tabletPC's I believe and is not unique to our systems.

    3. Pretty sure someone early on in this thread established that the glitch happened even when the wacom drivers are not installed, just windows default drivers. They did a fresh install and confirmed that after a lot of drawing on a fresh install with only the bare bones windows drivers installed (no Fujitsu drivers) the glitch still happened.

    Wintab and the wacom drivers are certainly not causing the problem, they just make it worse somehow. I think your firmware hunch is still quite likely though. Plus, uninstalling wacom's ISD drivers is hardly a work around. The whole point of this machine is it's Wacom inside, and unlike Ntrig, has wintab support. Certainly, son enough there will be no need for wintab anymore as all the software makes the switch to windows API. uninstalling integral and useful drivers is hardly a work around and doesn't solve our problem at all, just introduces new ones.
     
  3. Tab Rocker

    Tab Rocker Pen Pal - Newbie

    Messages:
    52
    Likes Received:
    6
    Trophy Points:
    15
    Hi Jan,

    Just checked the second USB and there is nothing saying Wacom. It looks similar to the first USB. I will try to get another screen shot later tonight....still at work.
     
  4. Plognark

    Plognark Scribbler - Standard Member

    Messages:
    176
    Likes Received:
    25
    Trophy Points:
    41
    Yeah, all valid points.
     
  5. Tab Rocker

    Tab Rocker Pen Pal - Newbie

    Messages:
    52
    Likes Received:
    6
    Trophy Points:
    15

    I do recall someone testing the glitch in XP(I think) and in Linux and still getting the problem. That said, I cannot remember if they installed a driver or not.

    I definitely agree this is not a good work around in general, but for me it works OK until the problem is solved. For my work, I only need pressure in SBP. That said, I find it strange that without the Wacom driver installed I can go for hours without any issues. With the driver installed I start getting issues within 10 minutes. Again, this tells me that the Wacom drivers are making the issue much worse at the very least. It would be interesting to see the USB packet activity without the Wacom driver installed...

    On another note, the Wacom branded notebook is getting announced in two weeks. That may be my replacement if this never gets resolved. Love this notebook except for this issue.
     
  6. Plognark

    Plognark Scribbler - Standard Member

    Messages:
    176
    Likes Received:
    25
    Trophy Points:
    41
    Oh cool, I've been waiting for Wacom to get that thing done.

    I agree that the T902 is awesome except for this one stupid issue. I think that's a big part of why it annoys me so much. That and my chronic OCD :D

    The real proof for the packets are whether they're being duplicated at the firmware level or by the drivers. That's part of the limitation of software based packet analyzers. The communication does go both ways; there are handshake packets sent back by the drivers whenever a chunk of data gets sent (and vice versa), so if the drivers are sending back malformed packets and the firmware isn't coded with proper error checking things could go horribly wrong. We can't see those with a software packet analyzer, we just see the data packets that get through to the operating system. I don't think we can capture any token or handshake packets at all.

    It could just be a data volume issue too if the firmware buffer isn't large enough (or is too small), or if the driver buffer has issues as well. Bad code could be sending interrupts at the wrong time,

    So really what I'm saying is there are ten thousand possible things that could be happening, but we know where to look at least :p

    I'm reading up on this: USB in a NutShell - Chapter 3 - USB Protocols

    It's a lot to take in.
     
    Last edited: Aug 9, 2013
  7. jan_c

    jan_c Pen Pal - Newbie

    Messages:
    48
    Likes Received:
    4
    Trophy Points:
    16
    I'm going to uninstall the Wacom driver and just use the default Microsoft driver while trying out SBP this weekend. If I can replicate the improvement Tab Rocker noted, that will be an interesting result.

    Also, while fiddling with Wacom drivers yesterday, I downloaded the Cintiq 24HD driver and noticed that it actually includes the ISD driver as well! It's a newer version than the Wacom/Fujitsu drivers we've been using. I forced it to install via Device Manager. Unfortunately, the glitch still happens. Right now, I'm thinking of modifying the driver INF files so that the T902 can use the Cintiq DLLs and control panel as well -- the Wintab32.dll is definitely newer. (Reminds me of the old days when I had to modify INFs to make an old ATi video card compatible with drivers not designed for it... hopefully some of that knowledge remains...)
     
    Last edited: Aug 9, 2013
  8. jan_c

    jan_c Pen Pal - Newbie

    Messages:
    48
    Likes Received:
    4
    Trophy Points:
    16
    I uninstalled the Wacom driver and played around with SBP this morning, so it's using wisptip without any wintab interference. There was no difference on my machine -- the glitch appeared just as often. I made a video recording of it, but the camera ended up having bad focus again, so it's hard to see the cursor.

    I captured the traffic using USBlyzer and got some logs. The repeated data is consistent with USBPcap. There is this weird pattern I noticed -- the glitch triggers when 4 interrupt packets containing the exact same data are sent in sequence. I have no idea what that means; I'll forward these logs to the Fujitsu rep too.

    BTW, Wacom released a new ISD driver, version 7.1.1-16, which I just tested -- the ISD_TouchUser.exe memory leak has been fixed! http://cdn.wacom.com/U/drivers/TabletPC/ISD_DualTouch_711-16.exe
     
  9. stoneseeker

    stoneseeker Animator and Art Director Senior Member

    Messages:
    3,606
    Likes Received:
    1,077
    Trophy Points:
    231
    Wacom released an actual fix for their driver??

    Just when I thought I had them all figured.

    thanks for letting us know, now I can actually try wacom touch drivers in art programs for gestures... windows sucked for touch support.
     
  10. Tab Rocker

    Tab Rocker Pen Pal - Newbie

    Messages:
    52
    Likes Received:
    6
    Trophy Points:
    15
    Very interesting. I wonder why I barely get the glitch without the Wacom drivers installed. Perhaps it's happening but I'm just noticing it less? I do have other services disabled in combination with the Wacom drivers removed so perhaps it is a combination of things I stripped away. I posted steps about removing these services a while back in this thread and that seems to keep the glitch to a minimum.

    OK, I installed the new Wacom drivers and restarted. The glitch came back in SBP fairly quick. Bummer. On the plus side, I noticed that my calibration is WAY better in all corners after doing a normal calibration. They must have improved this area with the update. I will have to try a long drawing session again to see if the glitch is reduced with the memory leak being fixed.
     
    Last edited: Aug 12, 2013
Thread Status:
Not open for further replies.

Share This Page