I don’t think it’s something Pimax can solve by itself. From what I understand the issue is at Nvidia’s end. Hopefully driver updates will fix it. My budget has been blown on the portal and crystal, so I probably won’t be able to afford a new GPU 'til the 5090 series. Which will probably come with a new set of issues. Sigh.
Any new developments regarding this topic? I follow this thread as I experience the same issue with my 7900 gpu and 5K super hmd. I find it a bit difficult to believe the issue is soley Nvidias responsibilty since my gpu is Amd and the Pimax hmd is the only commonality.
The following is just a translation. Choose to read it or not, but I assure you that the conclusion is correct.
I am 100% certain that the SteamVR patch 1.25.4, which was supposed to fix the compatibility issues with the 40 series graphics cards, is ineffective for PiMax HMDs. It’s truly disappointing to see that PiMax couldn’t address the compatibility issues with the 40 series graphics cards even after two months of the release of these two SteamVR patches. Their inability to resolve this matter is truly disappointing.
As a user who upgraded from a 3090 to a 4090, I’ve noticed that this thread contains a mix of issues related to screen shaking, delay, and stuttering that are not necessarily caused by the 40 series graphics cards. In the past, NVIDIA addressed stuttering issues caused by monitoring software like MSI Afterburner in SteamVR. Additionally, the Pitool or Pimax client not being active in the foreground can cause stuttering or delay bugs. It’s worth noting that many casual VR users may not even notice screen shaking or delay with 40 series graphics cards.
In fact, SteamVR released two patches in March to fix the bugs related to the 40 series graphics cards. That’s why you can see that there are no more discussions about SteamVR and the 40 series graphics cards on other forums, and only Pimax users here are still searching for answers.
I’ve personally tried reinstalling a fresh operating system, Switching between the 3090 and 4090 graphics cards，switching between HDMI and DP connections for different HMDs such as Vive, Vive Pro, 8kx, and 5kxr。I’ve tested both the SteamVR version 1.25.3, which doesn’t fix timing issue for 40-series Nvidia cards, and the SteamVR beta version 1.26.2, which includes two patches for 40 series graphics card.
Based on my testing, The Vive Pro experiences the same screen shaking and delay issues as the Pimax with the 4090 graphics card when using SteamVR 1.25.3. However, there are no screen shaking and delay with the Vive Pro when using SteamVR beta 1.26.2. However, both the 8kx and 5kxr experience issues with the 4090 in different versions of SteamVR, while there are no problems when switching back to the 3090.
Combining this with Pimax inability to utilize SteamVR’s useful features such as Asyn Reprojection and OVR-SmoothTracking, as well as experiencing significant screen shaking in the SteamVR VR view compared to other HMDs, and spending six months trying to fix compatibility issues with Tundra Tracker pairing。
I would suggest potential buyers interested in Crystal HMDs to consider opting for Varjo Aero, which offers a similar screen but without the uncertainty of SteamVR functionality and patches that may be ineffective for Pimax.
Have you tried running @mbucchia Pimax XR runtime?
One big problem is that the lag is not a common issue across ALL 4090’s, some seem affected, others not.
I have a MSI 4090 Suprim LiquidX, and as another user with the same card here earlier stated, absolutely no problem with the 8KX. Dunno what magic MSI did on that specific card …
I second that. I have a gainward Phantom 4090 and no issues…
The 4090 graphics card experiences screen shaking and latency issues in Pimax Home, which is based on PVR (the native interface for Pimax) and does not require SteamVR.
I don’t have DCS and MFS games based on the native OpenXR environment.However, I believe that the issues present in Pimax Home would persist even with Pimax-OpenXR.
I Setting the SteamVR runtime to pimax-openxr did not make any difference.
Just like the DisplayPort interface of the 3090ti, which is a hardware design flaw by NVIDIA causing various HMD VR recognition issues, no one has released a patch to fix this problem for the 3090ti. The only solution is to use a DisplayPort 1.4 Repeater.
However, the situation is different for the 40 series graphics cards.
SteamVR has released patches for Vive Pro and Index, which do not require third-party software.
Posted Sat, March 4, 2023 SteamVR Beta Updated - 1.25.4
Compositor: Fixed timing issue for 40-series Nvidia cards.
Major Update Posted Fri, March 17 Introducing SteamVR 1.25
Full Update Notes SteamVR
Fixed hitching issue for 40-series Nvidia cards.
For HTC Pro 2 and Varjo Aero, which require third-party software, they have released patches specifically for the 4090.
I also conducted tests and found that when downgrading to SteamVR version 1.25.3, the 4090 and Vive Pro reproduced the same screen shaking and latency issues as Pimax 8kx and 5kxr .
Pimax Support replied to me only, indicating that they are planning to release an optimized version for the 4090 in June.
Well this is good news and will be welcomed by many Pimaxians should it resolve the issue for all. I certainly will be grateful to be rid of the tracking lag as it really does make the VR experience a bit nauseating particularly in faster paced VR games. After having dealt with it for months now I certainly will not miss it if they do indeed fix it.
For me this issue took away the joy of VR in the 8KX completely. I thought I was getting used to it, but I find myself rarely use the HMD\VR at all anymore due to this issue. The few times I’m in VR nowadays instantly reminds me of why. Hope this June update will fix it, but I wont get my hopes up too high. I for sure will not touch any other Pimax HMD in the future if this is not resolved. The uncertainty of getting this issue again is just not worth the risk. (Phantom 4090, now mostly used for watching Youtube videos and reading news … )
With all the excitement and even if it’s not nice. Are you aware that the problem only occurs with a proportion of users of the oldest version with the scaler from 2018? The following generation is not affected at all. Therefore it is difficult to find the cause.
So I’ve read, but unless I see that for myself I just don’t believe that. I’ve had friends trying my 8KX and not noticing anything strange at all, but for me it’s unplayable. Both me and my friends have VR experience from different HMDs as well, so this issues is indeed very individual, which makes it even harder to fix.
Not sure how you figure as it has been reported on 2076
However even if it was the older 8kX it needs to be fixed. Pimax used to demonstrate good support for earlier hmds. With now nit even supporting there newer models all that well since focusing on Reality & Portal.
Yes, they would need a hardware combination + HMD where the problem is proven to occur in order to analyze the data stream. They may have it locally now.
I’ve actually been following this thread from the start. The reports came, at least in the first few weeks. only by Gen1 users. With Gen2 it was more like “misunderstood”, “super sampling too high”, „false timing, “nocebo” etc.
Nah this was across all 8kX models. The problem now more seems like sensitivity to it. Much like how Seb is sensitive to a latency issue that most ppl might only see with the tap test. Still a big enough issue to fix.
It is now just less common an issue.
You and others lik @Raymondo may have this issue but are not sensitive to it. Much like many of us thought Seb was trying to create drama as we could not see the issue Sen described until another dev user explained they latency tap test.
@PimaxQuorra can you ask the team to further look into this issue? Maybe @CheezusCrisp might be able to help narrow down this issue.
If you have monitored the thread here and in Reddit tgen your aware it was both new sbd old 8kX models and some even reported on the 5kXR. So definitely also a sensitivity to the defect.
Some even noticed it when thet specifically looked for this defect. But was not a bother to them because they had to focus on noticing it.
Read back a few pages. There was also (at least) one user who switched to V2 and confirmed that the issue no longer occurs on V2.
As I said, I’m following this thread very closely. In the first weeks i bent my entire computer (bios, win settings, hardware) to get the problem also. No lags… Only on the Pimax moon could I force a small retightening. (8kx Gen1, 10850k, GB4090 first batch)
Normal hard overclocked standard system, the only hardware I’m missing against other users is the popular RealTek audio chip on the mainboard, which I deliberately avoided because it often causes problems.
Yes your looking for reasons to dismiss it. It is still on v2 8kXs as well. The same as some on original 8kX do not xp it. There is even a few ppl with other headsets that notice it and even noticed it after switching to a former gpu.
So it comes down to sensitivity. The person whom upgraded to a v2 simply maybe below his sensitivity and thus resolved or that 8kX v2 bom is different than those whom xp. Not that different frim v2 8kX that have compatibility issues with fw 2006.
Regardless of you opinion pimax needs to work on this. Considering the last fw upgrade for original 8kX is 299. Mean while pimax has released 2002, 2003 and experimental 2006.
Translation they have not tried to fix things in there end with 2703 to 2705 8kX models.
Overclocking to me these days is pointless as most things auto clock and overclocking often is minimal gains compared to wear on components.