pimax did sent me a mail to participate in a survey about the eye tracking module
and also asked for comments, as this seems to be a one way ticked i thought we could just collect and discuss some comments and experiences here
-only works when foam it changed to a thicker non standard version (no vision kit yet received as backer)
-to get the calibration working the headset needs to be way to high on the face, resulting in blurry not sharp picture due to not looking into the sweetspot of the lens
-range of eye tracking does not cover the whole vertical and horizontal area
-seems not precise enough to be useful for input (tested with last stage of calibration tool)
-time to response to eye movement to high to be useful for DFR
-2nd side USB can’t be used with 5k+ when installed because USB port on the side is not prioritized (like when connected to USB 3 .0 instead of internal USB 2.0 from headset)
final comment:
not ready to sell as of lack/latency in DFR
if USB 2.0 connection is the reason (only 30Hz) then its a design problem and it should not be sold to backers or customers
I can’t get passed the calibration where the two green dots disappear and mentioned it in the survey. (AMCap gives an error that it can’t load a video filter)
I’ve been using an old 22mm PU leather face pad from my Vive since I got my 5K+ in early 2019, so I should be ok there.
Also, I asked in the survey for an extra pair of metal frames, I didn’t stick them on the lens yet, just in case I get my hands on a 8KX… The thick face cover holds them in pace without problems.
Very very beta stage… And I lack the time for lots of trials and errors at the moment.
make a request like this to the support and open a ticket, i dont think a comment i a survey will trigger anyrhing (at least not when its pimax)
the metal frames have a plastic edge on the upper and lower area so it can be hold in place by this without using the adhesive material on the back, i did not peel off the protection, just used it with the two plastic edges to keep it in place so i can use the mechanical guidance and magnets to hold the eye module in place, i guess the adhesive was added later as an addition
True! But I had a different reason: only when my eyes are far enough from the lenses, I don’t have a yellow glare/lightbleed. There is only a small sweetspot in distance of the lense that doesn’t ruin the experience for me…
i tried the eye tracking with usb 3.0 now, its way more responsive in the calibration, seems to go faster/better through calibration (in half the time), UserSDK.log shows 118 Hz instead of the 30 for the headsets internal usb2 port (5k+ owner)
EyeTracking camera = 0, rate = 118
slow eyeType = 1, idx = 1, rate = 118, average time = 1599, has fd = 0, no fd = 118
it also seemed to reach much further in horizontal range now but that could also be because of wearing the headset a little different (would need more testing) - at least the horizontal range is ok when it comes to the hardware capabilities
edit: also found out that if the usb side port of the eye tracking is used the headset gets a strong humming sound to the sound system of the headset - unusable that way for normal use (tried 6 different usb ports on the backside of the system board)
also PiTool does not recognize the missing eye tracking software (runtime and forwarder) when eye tracking is idependently connected to usb (3m usb-c cable to computer), needs to be started manually when connected this way
edit2: PiTool/headset does not properly work when eye tracking is connected independently, the whole picture is blocky and flickers sometimes as if the DFR is applied to the whole picture - atm thats not a option, would need changes to software/firmware
edit2.5: the independent connected eye tracking worked as it should be after rebooting, pitool found the missing aSeeVR components and offered to start them, and after a recalibration the steam home enviroment worked without the resolution problem i had before - so no changes in software or firmware needed to get it working
edit3: when connected to usb3 the calibration tool also works with the headset much lower on the face, it seems as if in usb2 the software not only reduces to 30Hz it also seems to reduce the area of the camera/picture used for eye tracking (lower in hor/vert area that is tracked)
The hum is probably ground loop, and it might hurt the electronics in the long run. You have both the top usb-c and side usb connected right? I had the same problem and had to pull it out of the top usb connector which also gave me the slack in the leads to properly position the lenses and pass the calibration.
More on groud loop on wikipedia: Ground loop (electricity) - Wikipedia.
Edit:
I can, for the life of me, not get it to calibrate with either upper or lower usb-c port (green dots disappear, have to move the 8K away 5 cm). If I connect it with an external cable to the pc usb 2 (30Hz) or 3 (120Hz track) it accepts the calibration.
no, i added a small spacer between hmd and eye tracking module to prevent the usb-c to connect to the headset
usb-c (side of eye tracking) and usb to hmd come from the same computer (ground) so if there is a ground loop its already there from the psu of the headset (and did not occur before), also psu and computer are connected to the same power socket on the wall
I am not sure if you just don’t know what a ground loop is.
In an electrical system, a ground loop or earth loop occurs when two points of a circuit are intended to have the same ground reference potential but instead have a different potential between them.
By giving a single device multiple connections to “ground” you are creating a ground loop.
Under ideal conditions the voltage/etc should be the same on both connections but, unfortunately, under real conditions there are often differences.
EDIT: Sources
myself (electrical engineer)
Henry W. Ott; Electromagnetic Compatibility Engineering-Wiley (2009)
Bruce R. Archambeault; PCB Design for Real-World EMI Control-Springer US (2002)
at least i get it now, by adding a 3rd connection (1st two are DP and USb of the hmd to the computer) with a shorter cable i get a shorter way and my already existing problem (two ways for ground) is now going though the hmd pcb (including sound part) to the usb of the eye tracking - so the source of the problem is that i have not disconnected the eye tracking entirely (the usb to the upper usb port must still have a connection to the eye tracking module) - i will try to use the plastic cap that came with the eye tracking to isolate the unwanted usb connection
edit: solved by using the plastic cap over the usb connector from the original packaging, the unit does not sit as tight now as i’d like it but that will be addressed later (with the KDMAS and comfort kit installed)
the additional usb cable would need to tight up by cable ties or double sided velco, i use a welding mask mod but i guess a vive DAS or the incoming pimax audio strap and comfort kit will have some space to tie it up
more of a problem might be my 3 meter passive extension - looks like there are some bundles with 8m (5+3 incl. a usb3 repeater), i ordered one for 25€
at least after a reboot and recalibration the eye tracking is working more in a way as expected (faster), still looks a little weird at the resolution edge as if things would pop into view in the steam home environment but i guess that can be fixed by software and might look better in other vr titels
? Thanks for the tip! Metal frames inserted/clicked now without the sticky part 