I see slight delay in motion when I turn my head very quickly (5k+)

issue

#21

Well I finally had a teamviewer session with Doman. As best as I could tell from watching, he ran some diagnostic routines but did not “calibrate” anything or make adjustments. He pretty much just read and analyzed my FW, apparently doing some static checks. I say static because the HMD sat stationary on a stool the whole time. No problems were detected. I was hoping/expecting there would be some motion tests and measurements. After he was done with his checkout he asked me if I wanted to update the firmware to the latest. I said why not. Might as well try something. As expected it made no difference.

I must say though that it is nice to have Pimax tech support be so responsive. Though it is in their interest as much as the customer’s to explore possible anomalies so they can improve production quality.

Edit: And for completeness I went ahead reconnected my Vive and must correct myself and say that it is not “rock solid”. That would be near impossible. My mental comparison of the two would be that the Pimax does have a bit more lag. (Maybe due to the type of panels being used?) And maybe that “bit” added to what I saw in the Vive is enough to make the lag jump out at me. So I guess I’ll accept that my HMD is not defective and just get used to the lag that I’m experiencing as a new norm.


#22

Just checking if anyone has found a solution for this question? I received my 8K a few days ago and noticed this straight away. last night I got one lighthouse up and running, but that didn’t make any difference. It’s a bit confusing, my girlfriend would definitely get sick if I let her try it!

Doesn’t seem to be latency as such, my first tests were with nothing running and last night I had exactly the same effect with 100%CPU load (the PC is running a render at the moment)

It was almost as if as my head was rotating, my eyes were moving around the circumference of a circle that was bigger than my head.

My Rift definitely didn’t have this problem.


#23

You may be experiencing the poor distortion profile issue that I mentioned yesterday. With one eye closed I get warping on the nose side of the lenses when I rotate my head. Obviously this will only effect the geometry in one eye at a time as the warped geometry is not rendered at the edge of the lens for the other eye. With both eyes open the result is a weird disconnect between what both eyes see. It could be perceived as latency and I think it’s causing my eye strain.


#24

I don’t think it is the same thing, to be honest. If I am in the grey room and look ahead at the lines going to the horizon, I can see them following me as I move, then settling back into position when I stop. It isn’t at the edges, it is the whole display.


#25

Hi all. New 5K+ owner (finally) - my first impression is that I noticed this rotational latency immediately. I made a video to demonstrate it. Coming from rift and vive - this sticks out like a sore thumb to me, and gives me a headache.
Pimax 5K+ rotational latency demo

I’m very interested to see how many other people notice this, and if not - what are they doing that we are not. I will submit a support ticket with pimax soon.

Edit: couple of pokes as suggested by commenters: @Heliosurge @Doman.Chen


#26

i did notice it straight away, reported it t Pimax but they claimed that there was no issue and that this was normal.
I use motion rig when racing and sometimes depending on track or motion settings it becomes unplayable.
Did open a ticket and sent videos but like i said they claim that it is ‘normal’…


#27

i think you should poke all Pimax team in your video, they need to see this.
@piwok


#28

@piwok , great job documenting the problem in your video. They had asked me to do so when I first contacted them about the problem but I couldn’t figure out how to video it with only my cell phone camera. I’m sure by now Pimax knows it’s an issue (I’m sure they’ve known all along from internal testing) and if they haven’t fixed it (or even acknowledged it) by now I’m doubtful they can or will.

Fortunately for me, I only do sim racing (not in a motion rig) and I don’t snatch my head around too much so I can tolerate it as is. But yeah Pimax needs to not ignore this if there is a possible fix.


#29

I have to admit that Pimax did surprise me a lot with the “IPD offset” setting.
Now that we have it documented and we can clearly see it they will make it a priority (sure they knew about it).
But i find surprising not seeing more report of it (we might need another poll :joy:)…
Might be faulty batch…??


#30

Yes, the motion to photon latency is significantly higher on my Pimax 5K+ compared to my OG Vive. However sometimes I didn’t notice it, so I wasn’t sure if it’s different on some settings or I just don’t notice sometimes.
Edit: Can you maybe make the same video with the Vive, just as a comparison? Just out of curiosity.


#31

Hi @Willyfisch - I’ll do one on rift if I get a chance, but I can tell you with confidence the rift and vive are very similar and have zero perceivable rotational latency, which is why they feel locked in and more like real life compared to the Pimax. This is the reason Oculus specified 90FPS minimum for the achievement of “presence” - to get the motion to photon latency well under 20ms.


#32

Thanks for the kind words @ricknau . You dont need to snatch your head around fast to observe this rotational lag - once you know what you’re looking for you’ll be able to see it slightly even with slow smooth movements. Its absolutely unacceptable for a 90fps headset.

I’m hoping its not a faulty batch @Jean-Mehdi but rather a tweakable firmware parameter. I suspect they need to reduce number of samples in an IMU smoothing/averaging buffer, or increase the IMU sample rate significantly. Both suggestions are free of charge pimax :stuck_out_tongue_closed_eyes:


#33

Valve Did need to do adjustments to Lighthouse tracking as Beatsaber players demonstrated could move faster than originally thought.

It’s interesting though. @mixedrealityTV noticed a latancy wobble in the 8k that required a special tap test for majority to even notice. Since he has a sensitivity in theory this does not affect all 5k models out there but hard to say for sure.

Since pimax now only seems to have 3 shared beta testers with Xtal. Maybe @SweViver & @VoodooDE could also use your vid sample to test & give pimax feedback as well on the subject. Great job on producing a vid; as this is often an owners frustration when asked to produce pictures and/or vids.


#34

Exactly what i wanted to say but couldn’t find the words
thank you :joy::joy:


#35

I have the same problem with my 8K. It’s unusable. Instant nauseating. As a side note, the lighthouses are not recognized by Pimax.


#36

Thanks @Heliosurge. Pimax support ticket lodged - SUPEN-4487 for reference. I’ll keep you all updated if we make any progress on that front.


#37

As promised. Comparison of latency between rift and Pimax 5K+

Read youtube video description, i had trouble finding the exact rotation point on the rift side. The companion cube also moves “globally” because my rotation point is not quite correct, but you need to compare latency only - ie lag as the movement starts. I included a clear example at the end of the video.

I’m not going to test Vive unless someone really badly wants it. I can already tell you the answer.


#38

I can confirm that Pimax are aware of the issue and engineers are looking into it :slight_smile:


Why is the VR 'Media' Blackballing Pimax?
#39

Thanks @Jean-Mehdi and @Doman.Chen . This news is most welcome i’m sure for many people. Here’s hoping the Pimax engineers agree that this fix is top priority over new features.


#40

I agree with that, big big priority for me and many others backers/preorders.
I’ve been helping few preorders here in London with the setup and all of them ask ed me about this issue.
My guess is that Pimax was well aware of it and now thanks to your amazing video we are finally able to prove our point, big thanks for that !!! :+1:t4:

Let’s hope it gets fixed sooner rather than later…

Edit : did you run the test on 90 hertz mode ?