Norman said in his review that to him it seemed that games would render 110 degrees and that pimax would then ‘stretch’ this to 200 degrees. @PimaxVR said this was incorrect and that the game actually really rendered 200 degrees. Yet I find it weird that Norman though otherwise. He must have had reason to believe this.
I wonder why Norman thought this. Is the render distortion not good enough so that it just looks ‘stretched’ at the sides ? Or is there some kind of ‘trick’ involved ? Would it be possible to show us a 200 degrees screenshot, dumped from the SteamVR compositor maybe ? That would help a lot.
I asked them the same question and they confirmed that it was the render distortion they needed to work on. They said don’t worry about it, they will definitely fix it by release. We shall see…
Thanks, interesting thread. I find this post very interesting. It seems it actually might be quite difficult to do it right, which is worrisome:
[quote]Anything over 180° (and up to that, with massive increasing inefficiency) means you can no longer render a rectilinear viewport. You need to start rendering multiple actual views per eye with their own viewports (not just shifted frustums like with Lens Matched Shading). This is the sort of thing where you start having to customise the game engine to get it to do what you need rather than shifting around the existing view.
Also, that documentation is just for the application-side API to talk to SteamVR. For the drivers to talk to SteamVR there is a different API.
However, there does not appear to be a way to pass multiple projection matrices per eye available, so OpenvR would have to be ‘hacked’ to add the functions needed (like with the Kinect body tracking demo where 3 points on the skeleton were turned into virtual Vive trackers because OpenVR has no way to pass a skeleton post). One way would possible to present the HMD as two HMS that happen to be in the same physical location but with different optical geometries. A huge pain to keep things synced up though.
In order for things to be done ‘properly’, like any extension to OpenVR it requires you to wait for Valve to implement what you want then update SteamVR to support that extension.[/quote]
@PimaxVR Will you be able to improve the FOV rendering and the lense distortion problems by Janurary? Is it only a software problem. Please confirm thanks
Do you have any idea why Norman felt it was 110 degrees stretched to 200 ? And can you post some screenshots please, from the SteamVR compositor screenshot system ? THanks !
I must say I’m still worried about this. Norman felt it was 110 degrees to 200 and @PimaxVR hasn’t given a possible explanation for this. Can you dump some screenshots from within the SteamVR compositor please ? Thanks
Jeremy is the one to listen to on the stretching front in that video, as he doesn’t wear a pair of thick rimmed glasses with lenses like bottle bottoms unlike poor Norm.
Tested just did a follow-up on their Pimax 8k impressions and the FOV warping issue is discussed in great detail again as a really big negative for anything except virtual desktop or movie watching use. @PimaxVR if you do in fact have a solution already and there really is no warping going on as described then it is crucial that you get evidence out to support that from a 3rd party. It would be best if you could somehow get Tested to try it again with something that you feel shows that the warping is not an issue. A lot of people follow Tested and this latest video from them is much more negative than their overall initial impressions video and basically all of the negativity is coming from the warping affect that they experienced.
Perhaps any warping can be corrected in the drivers or maybe it might be the source game engine or even DirectX that’s the problem and needs a rendering correction. All 3D graphics are essentially a projection during rendering to a 2D flat plane to “take a snapshot” for each frame sent to your monitor/HMD. To be fair to Tested they point out triple screen setups also suffer from this (I’ve had one of those for over a decade and its not that much of an issue).
I did warn about Tested’s independence and I would wait for other impressions before basing a purchasing judgement on this, especially with their coziness to Oculus in the past (up until they found out about Palmer Luckey’s political persuasions).
I don’t understand why everyone expects this Pimax 200 FOV 8K HMD to be done correctly anyway. It’s just a race to be the first and trying to be as technically ‘ok’ as possible. I personally trust Pimax to do the best they can and from what I see happening with the 4K model and software, they are working on it with great enthusiasm and I find their results good enough. Too bad I was too late to see the kickstarter go live and missed the early bird reward because I’m sure Pimax will keep working on improvements and we’ll see the HMD get better and better over time regardless of what we know today
I don’t expect the 8k to be perfect because no HMD ever will be, but I do want to be fully aware of it’s shortcomings so I can make as informed a decision as possible on whether or not to get it. I am sure that Pimax will do the best that they can, but if that best truly does result in warping sever enough to distract from immersion in the VR space then I want to know that up front.
@PimaxVR I follow Tested’s reviews and this Fov distortion is an absolute crucial point as the most important thing in VR is to keep the 1:1 mapping ratio all across the Fov else it will look tricked un-natural and disturbing. If you do have a solution already adressing this point and there really is no more warping going on, please follow acegamer’s suggestion above and let us know.
