Arch Linux

kplaceholder50

Member

Registered: 2021-10-24

Posts: 5

Help with setting up a Wacom Cintiq 16 with KDE Plasma

Hello. I recently acquired a Wacom Cintiq 16 that I am attempting to connect to my computer. So far, it's honestly being a pain in the a** to set up, mostly because of unexpected interactions with other software. I'll also state that I have done my research prior to writing this post, but I couldn't really find anything useful. I am running Arch Linux with KDE Plasma on X11 on a desktop computer with two monitors (besides the Cintiq). I owned a Wacom Intuos prior to this, and it worked right away. This post will be about multiple problems I have been encountering, some of which I have been able to mitigate but not solve, others not at all. I would like to share my frustrations with the community to see if some brighter minds can help me deal with this, since I feel like I'm going insane just trying to get this to work.

Admitedly I'm not sure where to make this post. It's kinda about hardware, but since this deals mostly with Plasma-related issues, I'm posting it on this category.

Hardware specs:
- CPU: Intel i7-8700 / GPU: NVIDIA GeForce GTX 1060 / Motherboard: MSI Z370 / Memory: Kingston HyperX Fury, 8GB running at 2400 MHz
- Storage: An NVMe SSD with the Linux filesystem, a separate SSD with the Windows partitions, and an additional HDD without OS that is permanently mounted.
- The device has two 1920x1080p monitors connected to it, one via HDMI-0 and the other via DP-0. The Wacom Cintiq 16 is connected into the system via an HDMI-to-DP adapter, in turn connected into DP-3.

Software specs:
- OS: Arch Linux x86_64, currently running with the Linux Zen kernel version 6.5.8
- Desktop environment: KDE Plasma 5.27.9 with Xorg.
- The GPU is running under propietary NVIDIA drivers.

Issue #0
I'd like to preface this with a preexisting issue which I have with Kscreen: One of the monitors is physically damaged, and to compensate for this, I used the NVIDIA settings application to introduce a 40px-wide underscan to the monitor. This means that, while originally the BenQ was 1920px wide, the underscan makes it go down to 1880px wide, which is not officially supported, but works as expected. I have been using that monitor this way for more than a year, and it's good enough. The main issue being that Kscreen does not acknowledge that resoultion. Instead, it now treats the monitor as having a 1880px wide viewport but a 1920px wide screen, which has a consequence that I really hate: it adds "panning" to the entire screen layout, so that if the mouse touches the right border of the monitor on the right, all screens unalign themselves from the monitors and pan to the left, to reveal an empty 40px wide black gap. If I want to undo this panning, I have to move the mouse all the way to the left of the setup to "push" in the other direction.

Let me be clear: The objective of this post is not really solving this particular issue. While this undesired behavior cannot be disabled from Kscreen, I can outright disable Kscreen, and Plasma still works properly as long as I don't disconnect either of the monitors, which I never need to do. Under regular circumstances, this issue would be solved. However, now that I'm trying to add a third monitor, this issue has joined the tangled web of problems that I have with my new Cintiq.

Issue #1
With Kscreen disabled, the Wacom Cintiq 16's screen is disabled every time Plasma boots. The system does recognize the Cintiq's screen properly as the boot up text shows in it, and it goes as far as SDDM correctly detecting it as a third screen. As soon as Plasma starts though, if Kscreen is not running, the Cintiq just turns off. I can go into Plasma's screen settings or NVIDIA settings and see that the Cintiq's screen is properly detected, but disabled. I can re-enable it and use it normally... for the remainder of the session. The next time Plasma starts, this happens again. No reboot needed — Logging out and back in already disables this screen. If I restart the SDDM service, SDDM again keeps the tablet's screen on, but Plasma does not.

Since this seemed to be an issue with graphical settings not persisting into xorg.conf, the first thing I tried was to save them with NVIDIA settings. Without success, that is. In fact, as far as my knowledge of xorg settings go, my xorg.conf properly recognizes and enables the display on DP-3 –which would be the Cintiq–, and the tablet's screen working properly outside of the desktop environment should be proof of that, but Plasma seems to ignore it.

So far, the only reliable way that I could find to avoid Plasma turning off the Cintiq's screen, is to re-enable Kscreen. With Kscreen enabled, this issue does not apply at all. But this unsolves issue #0, which makes my setup barely usable. I tried to reboot with Kscreen enabled, then disabling it to see if Plasma's screen layout is somehow persisted this way, to no avail. The only constant is that if Kscreen is on, the Cintiq remains on; but if Kscreen is off, the Cintiq turns off. This does not happen with the other two monitors.

With this post, I am hoping that somebody helps me get out of this problem. So far, my solution has been leaving Kscreen enabled, and then create a Pyautogui-based Python script that runs on startup, which is perpetually checking if the mouse is approaching the far right end of the display, in which case moves the mouse to the far left end, countering the undesired screen panning and leaving some sort of useable "wrap around" effect. So far, the script is working with mixed success. I'd rather not depend on this hacky approach to deal with Plasma/Kscreen messing with my screens, but it's the best approach I have right now.

Issue #2
I'm guessing that this one issue could be in a different post altoghether, but since it's happening with my Cintiq alongside the aforementioned issues, might as well add it here. The correspondece between the touchscreen region and the visual region of the screen, as configured in Tablet settings within Plasma, does not persist across sessions. The Cintiq's touchscreen region defaults to the entire display across all three screens. This is obviously not what I want, so I go into Tablet settings, set the correspondence to DP-3 screen only, and it works... for the current session only. Rebooting, or restarting the SDDM service, resets the correspondence back to the entire display. This time, this seems to be the case even before Plasma starts, since it also happens in the login screen. With the Intuos tablet that I already had, the default was also the entire display, but Tablet settings would be properly persisted.

Admitedly, I haven't had much time to research or test this out though. I don't really know how I'm supposed to deal with this. I guess I could make another script that runs on startup which sets the Cintiq's touchscreen correctly, but how many more scripts am I going to have to make in order to work around all of these Plasma shortcomings?

---

So, to use my system right now, I either have to:
- Disable Kscreen so that the undesired panning won't plague me. I let the tablet be off by default, then whenever I want to use it, go into NVIDIA settings to enable it, then go into Tablet settings and set the correct touchscreen correspondence. Not great to avoid procrastination honestly.
- Enable Kscreen so that the tablet is on from the get go, but still have to go into Tablet settings to fix the touchscreen correspondence, and even when I don't want to use the tablet, trust that my Python script will mitigate the panning, but then avoiding other stuff that I know will break the script, such as logging out and back in via the GUI.
- Disable Kscreen and forget about the Cintiq entirely. Remove the Python script from startup as it's no longer needed, keep using my system with only two monitors, and keep using my Intuos. Then, cry about the money I spent on this thing

I would be really thankful if someone could lend me a hand in getting out of this mess.