Ok - update. My Reverb G2 lasted all of an hour. I called HP tech support and they balked, claiming I needed to hook up with Connection, since they sold it to me. Nevermind that all of the 800#'s on the email sent me to HP Support.
I’m disappointed, and now that I’m back to being self-employed, I’m frustrated as hell at the expense and fight over a warranty that’s just over a week old. I’ll send it in, since I still have all the equipment, but damnitall, I just wish they’d done a bit more robust construction or testing before release.
I’m guessing shipping is part of the problem…
Seems like all retailers ship the G2 in the black box. The hardware is snug inside, but any acceleration forces from dropping the box, being hit by other boxes, etc, will be translated straight into the hardware, with very little damping. I wish they had packed it in a larger box, with some cushioning around it.
When I picked up my Odyssey Plus headset, the reprojection was definitely inferior to that of Oculus. It was one of the reasons I went back to my OG Rift, and then on to a Rift S about 6 months later. It was a very pleasant surprise to find that the WMR/SteamVR reprojection has not only caught up, but actually surpassed that of Oculus.
I’ve seen how @Troll packages things. He is a master of impact dampening engineering. I want to send him $20 just to ship me an egg someday…just to see if he can engineer something that will get it all the way here…
As mentioned above in that mass of words above , WMR is doing the reprojection and SteamVR is just being used to turn the feature on or off, as WMR doesn’t have a ‘per Steam library’ entry to use.
(this image isn’t ideal, but I found it and in a rush but you can imagine it right to left as how a game is drawn, distorted and then viewed in the HMD)
So the lens is on the left with a physical shape to give a wide field of view, the ‘barrel distortion composition’ redraw is in the middle, and then the nice flat grid that you can imagine a 2D game is drawn in on the right. The reason for the 140% or 100% (or 100% and 50% on the G2) is that you’ll notice the grid pattern in the barrel shape has more area in it in in the centre. Picture a canvas drapped over a barrel I guess.
So what 140% resolution (or 100% on G2, they effectively ‘build in’ the 1.4 ratio this time to confuse everyone) does is ensure the edge of the grid is also getting a 1:1 game pixel to HMD pixel ratios. That means it’s clear in the center and the edges, otherwise the game pixels look stretched as you get further away from the center.
For flight sims we tend to care about the center area more, so we can read everything clearly, so getting the 1:1 for the middle group of squares means 50% (G2) is fine. For people playing room-scale then sometimes the peripheral being more clear as well is good. Of course it’s also just a balance to get a decent framerate vs clarity as well.
SteamVR measures how often your video card doesn’t make the native refresh rate and provides an ‘automatic’ resolution out of the box. It’s not great, as you could play Beat Sabre one day, and it’ll recommend 250% for you on the next time you play DCS, and your PC will turn into an abstract art installation where pictures of planes are drawn once every 2 seconds. ;). It’s better to set the SteamVR resolution either to 100% globally, and then adjust it as a ratio per game, i.e. 100% in ‘General’ setting and then ‘50%’ in the DCS entry, so you’ll get 100% x 50% = 50%. You can also use the Motion Smoothing setting per app that way as well, so handy.
One thing also worth pointing out with anything VR ish is that it’s eyes dependent, so if it works for one person it might not work for another, we all see differently in fovea details and flicker etc. The general tip is ‘get as many pixels in the center you can without destroying framerate’.
My notes on the video (remember the t-shirt), fwiw:
Using SSAA x 2 (!) and SSLR on in DCS VR seems like crazy-talk and burning that poor GPU for no good reason (doubling the resolution in DCS only to then halve it in VR etc). I think he said he’s running the resolution at 50% on a G2 and then locking down to 60 Hz, so an odd set-up or maybe I don’t get it. Perhaps there is something in it, but definitely unusual advice I think? It’s like we have fifteen different ways to upscale resolution and no-one knows what they all do so it’s like witchcraft “If you try SS-Eye-Of-Newt at 2.3 then it looks better, but only combined with Bats-Wing-Ansiotropic - give it two clockwise stirs as well!”
Someone called ‘theVRPilot’ should sort out showing fpsVR in OBS for performance videos - we need charts not feels!
He showed a lot of AI flights that is really on CPU bound, so could be running a Voodoo GFX and DCS would do what it does.
Flat shadows and x2 MSAA seems to be the way to go in DCS until performance is improved. It’s a shame as it does look flatter compared to a nice 4K 2D display.
So much this. I’m often guilty of just saying “meh…it feels OK” but there is no accounting for what I’ve been drinking or how hypoxic I am (not at the same time of course unless I’m on a bender in Vail). It is always nice to see side by side comparisons with real numbers.
I mean, it’s good to set stuff how you like it and staring at FPS meters in VR is diminishing returns which could be time spent flying, but in a video called ‘RTX 3090 + Reverb G2 DCS - Virtual Reality benchmark’ I want to see some numbers.
Mia culpa. I began watching the video at work and realized that if I didn’t shut the lid on my laptop soon, a YouTube shame spiral was close at hand. It began promising enough given the hardware, sim of choice, and “50 fighters” mission.