A lcd panel don’t have any electronic part on it*, it’s a slave component to the part of the monitor called a lcd controler Shop by Category | eBay
the Gsync protocol is on the controller part
- it’s a simplification I know
1 Like
For the first year and a half my Vive was solid wrt tracking/image stability. Then, coincident with last spring’s Windows Update, something changed and I would get the same kind of slight shift you describe when I moved my head left and right in SteamVR Home (oddly, if I started apps directly rather than from Home, this wouldn’t happen). Also interesting is that the controllers were properly tracked while this was happening. At some point in late July, I believe, there were updates to SteamVR that, possibly in combination with further updates to Windows and/or new nVidia driver updates, made the problem disappear and now everything is solid again.
What this tells me is that this kind of glitch is breakable/fixable through software updates and is unlikely due to the refresh rate (which didn’t change on my Vive while this was happening).
Distortion doesn’t change, but if it’s there it means that the lateral position of an object is incorrect. This in turn means a linear motion is no longer rendered as linear. Finally this means that when you move your head, objects that are static will appear to change position.
I update gpu driver and steamvr to be steamvr beta as we talk in other topic, those issues are solved (except controller still have some shift).
@mixedrealityTV Try to check this, I think you may found the same issue.
I can play the apex construct although I closed the reprojection right now, I think it is about reprojection bug (closed the reprojection has more issues in apex constructor, open have a little issue). Although reprojection not effect to pitool (off/on), but may it still have some problem (I think “on” is the default value).
controller still have some shift although steamvr alert to update the firmware.
I test by return to use “steamVR” from “SteamVR Beta”, I found the issue again, but less than before gpu update. When try to close the reprojection, the frame drop run so fast again.
Honestly have avoided topic as it sounds like it is very elaborately describing the effect of a lower frame rate.
This would not only be wrong but also very apparent and easily fixed. They’re not just going to copy their profiles without adjustment. The lens will act upon a portion of the panel, regeardless of panel size or position even. So at first run you check the alignment and adjust the image sent to the panel accordingly. And done. Maybe that’s not done to perfection and is that the reason it looks off. It certainly has nothing to do with panel Hz as Sebastian said imo
2 Likes
I agree, I think distortion is the wrong term for what he (believes he) observed. The objects would NOT be distorted, they would simply be out of position because while you move your head and the objects should remain in their position relative to your head/eyes, they move along a little bit until the next frame with updated positioning of the objects comes along.
That is how I understand MRTV’s explanation.
However, if this is truly a low persistance display - that should not be possible because instead of seeing the last frame until the next one is delivered to the display, it should black out, shouldn’t it ?
So what he say only makes sense if for some odd reason you have the same old frame being displayed for a second time. That could be the case if the display worked at 90 Hz and the renderer only produces 80 FPS, so 10 frames per second are recycled that way and it leads to such out of position objects. But these display only manage 80 Hz, but for every of the 80 cycles they should be receiving a new frame, so there would be no repetition of any frame.
I don’t understand how MRTV’s explanation could make sense, to be honest. And we have not heard about that particular issue from any of the other three testers (inluding Yangfeng) nor from the backers so far.
So I personally will not base my Decision (Trademark claim… - ooops, LeBron already took that one) on this alleged distortion/persistance issue because I am not yet convinced that we have an issue here. Perhaps there is something for people specifically sensitive to something but I didn’t notice it at the time with th V2 and hardly anybody seems to notice it today, so I will leave this one at the general level that having 90 Hz is generally nice, but not crucial by any means so will not weigh much for me.
3 Likes
this indeed literally sounds like an exact description of brainwarp. really do hope those patents are filed.
2 Likes
Yes, I pointed that out about the persistance but for some reasons everyone is talking about distortion…
This can come from, either the panels, too slow, or the upscaling chip that is too slow.
The 4k had serious pixel persistance, the second annoying thing on it (first was fov)
1 Like
Actually, the moving along effect sounds like bad motion prediction. That can be adjusted in software. When rendering a frame, the software does so using a predicted pose where the headset is expected to end up when the frame is shown. This prediction naturally isn’t quite prescient; sudden movement will throw it off, but typically just for a single frame. That single frame represents a longer time in a lower frame rate, however. Since the time the prediction is for is dependent on how often frames are shown, the prediction must be adjusted to match; this would be hard to tell but would make the whole world lag or lead just a little bit (proportional to the mismatch and movement) even for smooth moves. So if it predicts for 90Hz and displays at 80Hz, you’ve moved just a little further by the time it shows up, and the world is sliding along with your movement - but realigns when you stop. Does that resemble the behaviour you see?
1 Like
Head movement + rendering are part of the same loop and calculated constantly, if you move your head, that new position is sent back from the HMD to the HMD driver to the engine so a new image can be calculated correctly.
erm, this explains quite well how that works with Oculus ATW
Maybe Pimax have implemented something like this as they are working on it.
Holy SHit. Would this be the reason PImax went silent on Brain Warp? MS beat them to the patent office and now they can’t use it anymore without paying M$?
1 Like
So MS have patented Left / Right eye alternate rendering that has been in discussion for ages?. How did that get approved? The only conclusion I can think of is that Pimax were not the first to show this working.
Same as how did Oculus manage to slap a patent on curved VR displays. Seems the patent office don’t do very good research these days, as long as you have big pockets.
1 Like
This is normal, US patent offices don’t do a check for prior work.
1 Like
I don’t know how patents work really. A curved screen or alternate eye rendering hardly seem innovative and worthy of intellectual property rights indeed.
Can I patent chopsticks if nobody did it before me yet?
1 Like
Assuming MS only applied for a US patent.
Software patents are not recognized in most countries, however it is in the US, it’s extremely dumb. But I don’t think it will matter much, Pimax should just be able to point to prior work is it brought up.
It is very possible that Pimax already have a patent and the press didn’t pick it up, but did so when MS patented the same thing
1 Like
@SweViver @mixedrealityTV what steam version are you using? normal or beta.
And sebastian, what game do you found this issue most?
I don’t found problem like this in fallout4, but found in apex construct, rick & morty, beat saber on title logo. Found little on steamvr home (difficult to notice).
I wonder if this technique isn’t useful for every headset out there. Even phones. Basically this should be free to use as it’s a VR booster and will help the industry forward
Artifact (error)
In natural science and signal processing, an artifact is any error in the perception or representation of any information, introduced by the involved equipment or technique(s).[1]