I feel the same, for the next one we could print some ID’s with our nicknames and avatars hehe
2 Likes
The desktop PC had a 2080Ti, so I apologize I said 2080 (I was just confused). Although, the heating problems in that case caused a lot of throttling despite no overclocking, so I wouldn’t say it was running on maximum performance. The laptops are both 2080.
7 Likes
Do you know what the 8KX ED demo settings were? Eg was it native res with FFR and Motion Smoothing on? 75 FPS or half that with MS on?
2 Likes
Hi Martin, question about that PC. What settings did you use for the terrain in the flight sim? I felt it had no detail at all, not bad pixels but just terrain and trees but not much else. I wonder what the possibilities were and if they were turned down to make it all buttery smooth.
And nice meeting you in person, it was good to see you really are exactly as in your videos. A very nice person.
4 Likes
I agree, it was very nice to meet you in person, and Sebastian, Ard, Cas and Chary as well, I didn’t see Voodoo there though
2 Likes
Saw Nathie walk around too
1 Like
Weird that they use PiTool at 1 when it has been confirmed both by PimaxUSA and users that higher PiTool settings coupled with lower SteamVR settings give better image through the headsets.
I notice myself on the 8K that at PiTool 1.75, there is much better colour and clarity on the headset but hey, maybe the 4K native 8K-X doesn’t require such tweaking and can make PiTool 1 look good.
Might though give users better experiences with panel contrasts and colour. I also notice much more gradation detail in artwork at higher PiTool settings especially in Sims with upper atmosphere effects and landscape details which are just not displayed at lower Pitool settings. Even if the Render Target Resolution in SteamVR are similar.
2 Likes
Thanks for coming to this forum while you are so busy . I hope you are getting enough sleep . You looked buggered on the last video I watched .
3 Likes
PiTool at 1 is actually already super sampling a bit
1 Like
Not just about super sampling.
Kevin is on record in these forums stating there are other benefits running PiTool at higher than 1.
The SS overhead can be reduced at SteamVR’s end and yet there is still benefits in image quality and lens distortion profiles even at the similar target render resolutions as stated by SteamVR.
If only he wouldn’t clam up when asked what benefits.
Try as I might, I fail so see it, I’m afraid. My eyes have slowly been going the past few years, and continue to do so, but…
2 Likes
PiTool at 1.0 is the default value (which many ppl wont even change, maybe because they arent well-informed), and there is no reason to go higher or tweaking for a VR Days demo. I dont know how many hundreds of people we had in our booth at VR Days, but every single person was super-impressed with the 8KX quality. Most people were also very surprised how good the 8K+ handles the upscaling.
There are cases where PiTool 1.5 mixed with SteamVR SS at 50% or even lower looks great. But for some reason, I think DCS, X Plane 11, Aerofly FS2. Elite Dangerous, Assetto Corsa and similar sims loses some clarity when moving down the SteamVR SS slider already below 80% (no matter what PiTool value you use). For a more consistently good image quality, I prefer PiTool 1.0 all day.
Of course you can probably increase the sharpness and level of detail by doing some tweaks, but I wouldnt say it will do any good for the framerate.
The flight simulator we did on VR Days was Aerofly FS2, and I thought a smooth experience is more important that maximizing the supersampling to crazy levels.
Oh and by the way, currently with a new in-house version of PiTool, using render quality at 1.0 & SteamVR SS 100% with 8KX (Normal FOV) gives you 3840x3160 per panel. Yes its 1000px more on the vertical side (real native 4K is 3840x2160), but those extra pixels are necessary for the distortion correction. I had an 1 hour lesson with the Pimax engineers who explained to me why they need more vertical resolution. The distortion is handled in a different way on HMDs with canted panels, compared to Cosmos, Vive Pro or whatever. Hard to explain, but yeah, I hope you get my point.
20 Likes
Waiiiit… You once said that it was about 3800 * 2400 right?
Then, did Pimax implemented better distortion correction algorithm that causes performance decrease, and we get better image?
Anyway, Thanks for your information Martin! 
Hang on a second - are you saying this is when not using parallel projections mode (EDIT: …for which several people have already demonstrated the need, here on the forum)?
I guess there was misunderstanding somewhere, but the resolution @SweViver noted (3840 x 3160) has an aspect ratio:
3840 / 3160 = 1.215
While my Pimax 5k+ gives in the native mode, PiTool RQ=1.0 and the Normal FOV the recommended render target resolution as 3200 * 2632, which gives the aspect ratio:
3200 / 2632 = 1.215
If we were talking about the res in Normal FOV, then the new headset has exactly same aspect ratio as the old one, which, technically, it should, if the lenses are same.
EDIT: Do not get too hung on the 3840 horizontal res. In Normal FOV only about 2/3 of the panel is used so these pixels are getting squished by 1/3 anyway.
4 Likes
Ok, so just the regular base sampling for 1:1 render_resolution:resolution_in_the_centre_of_the_barrel_distortion ratio, then. He made such a big deal out of the vertical there, that I had a kneejerk reaction. I guess he indeed got hung up on the 3840 thing (EDIT: …and I fell right into it, without a second’s thought). :7 (EDIT2: …should also have nothing whatsoever to do with canting the screens, unlike the parallel-to-native reprojection.)
Well, technically if you want to compare the sampling rate on 8k-X to 5k+, then it is definitely worse (relatively speaking) on 8K-X. 5k+ native horizontal res is 2560, while the recommended render target res is 3200, which is a factor 1.25.
While in 8k-X case the same factor is 1.0. But how exactly it translates to actual sampling in the optical center I have no idea. But given the lenses are the same, it should be the same on both.
The point is that 8k-X probably takes an advantage of the higher res in general, so even the sampling factor is slightly lower, the overall image is better, because less is lost in the pixel conversion.
2 Likes
Yep, yep, I follow you (…or so I like to think :P).
…and at some point, somebody spears a stick into the ground, at a more or less arbitrarily chosen spot. 
1 Like
The lcd-panels are smaller so they had to redo the hole distortion profile and they manage better.
1 Like