I had a chance to try Gear the other day. Seems pretty okay overall. Visuals seemed nice enough, at least in the Age of Ultron "trailer" I watched. Picture quality on Netflix was surprisingly poor despite the screen nearly filling your FOV. To be fair though, that may also have been due to rural Internet. The room itself looked pretty nice though, and dimming the lights when the show started was a nice touch. I wowed the crowd when I used StreetView to pull up my house. Standing in front of my house that was actually hundreds of miles away was pretty cool, but being 3m tall was kinda weird. lol
I didn't care for the touch-sensitive d-pad as I kept activating it when simply searching for it. I suppose I'd get used to it and it may last longer, but I think I might prefer a physical button. I also had a lot of trouble with focusing. I'd adjust the knob until focus seemed ideal, but within a few minutes it would feel like shit and I'd be compelled to refocus. Eventually I decided the problem was the "pitch" of the screen; it seemed like the bottom edge of the display was closer to my eyes than the top. Everything looked better when I pulled the bottom edge of the mask away from my face a bit. It's worth noting that I didn't try to adjust the straps or anything, as it was just being passed around at a party.
It was my first time using VR since Virtuality though, and it was pretty positive overall. It makes me more confident about my PSVR pre-order than I already was. Doubling the refresh rate will reduce latency and motion blur a lot, and having positional tracking will be a huge improvement, obviously. It also seems like it will be less reliant on the shape of my face to ensure proper alignment.
Given the power difference between a phone and a PS4, I'm not at all concerned with visual quality, not that I particularly was to begin with. I mostly judge games on gameplay, assuming playability/performance is up to snuff. Prettiness is pure bonus IMO; it's nice when it's there, but I don't miss it at all when it's not. I'd generally prefer readability to realism anyway, so abstract aesthetics typically appeal to me more in the first place. Really, I'd like to see everyone devs and gamers alike reset their expectations about per-pixel computation and focus on things that improve the experience itself, like refresh rates, lighting cues, etc. First work on solidifying presence,
then use any excess power on eye candy, since photorealism seems to be one of the least important metrics when it comes to establishing and maintaining presence. I'd rather live in a cartoon than stare at a photograph. Perhaps Im in the minority there though.
Here is a crappy analogy. You need to go to the store to get bread. It's 1:30 and you need to get it done by 2:00. The store is 5 miles away. You can either take your honda civic parked out front, or your enzo ferrari that's in a parking garage half a block away. Sure, the total time will be lower if you use the ferrari, but why bother? You'll make it back in plenty of time with the civic.
Personally, Id take the Ferrari because I count
any minutes saved as a clear win. By doing so regularly, I may even accumulate enough spare minutes to take a shower before my guests arrive to break bread. More likely, someone will call me up and ask for a favor, claiming my found minutes for themselves.
Then once Id gotten the bread, then much like AMD, Id park the Ferrari at my
own damned house, even if I needed to build another garage. Then I wouldnt need to spend the rest of my life wasting minutes walking back and forth to the fast car, on top of all of the minutes I waste trying to decide if its even worth the walk. I simply take the Civic when I need the extra seating, and the Ferrari when I dont.
reprojections takes 0.5ms running as an async compute job on the GPU. See here
https://youtu.be/3RNbZpcfAhE?t=653
Hey, nice find! They mention the 0.5 ms for reprojection also includes the distortion. It also says that the tracking routines take 0.8 ms, 60 times a second, and that's to process the feed from both cameras. Then the reprojection/distortion takes 0.5 ms, either 120 or 90 times per second.
also worth noting that as it is an async job, it is likely to run without removing 0.5ms of time from your own work - it should be able to slot in between during idle times.
Right. The graph shows tracking actually runs fully concurrently with rendering, and even reprojection overlaps a bit. I wouldnt be surprised if tracking was queued with a comparatively low priority. Occasionally failing to complete the camera math wouldnt be any more catastrophic than a single-frame occlusion. Is it possible for a job to get hotter the more times it fails to complete? Can tracking be queued with a priority equal to the number of frames since its last successful return?
Well, from what we know at this point Oculus decidedly doesn't have room scale coming. Every time someone asks them about it they reiterate that they are not targeting that.
I mean, of course all of that could change, but even if they do perform such a re-shift in focus (and personally I don't think this looks likely, although I'd be very happy if they did), the Rift HMD itself isn't designed for a room scale target.
The cables are too short, the FoV is angled upward and focused on the horizontal over the vertical, and there is no camera to orient yourself or grab a drink off the shelf.
Of course, none of those are absolute dealbreaker problems in and of themselves if they can get the tracking working at room scale sufficiently well, but it is rather obvious that one of these systems was designed for 360° / room scale and the other wasn't.
Additionally, the limited coverage of the Constellations versus the Lighthouses means a less efficient use of whatever space you
do have. Yeah, while Rift will clearly be
capable of room scale support whether its supported officially or not, its really hard to make a strong argument for Rift over the Vive if its something youre interested in.
I wrote something about my direct comparison of the "seated experience" with both HMDs:
Comparing the seated Vive and Rift CV1 experience in Radial-G
The most surprising thing to me was how differently the optical artifacts manifested, and how preferable one was over the other.
Nice writeup, and yeah, while the lens artifacts on Vive seem more unnatural and out of place than the spooky glow of the Rift, it does seem less offensive overall. I was surprised you didnt notice any difference in apparent resolution, but as you say, the difference in FOV isnt gigantic, and I suppose Wipeout(?) isnt the best game for noticing fine detail. lol
It sounds like you felt the field of view was more or less the same in most directions, but extended in the inferior direction? So maybe theyre simply applying less distortion to that area of the screen? Dig my shitty visualization:
So it it possible both teams squeezed 40% of their available pixels in to the sweet spot, and 15% in to each of the remaining quadrants? Then Vives apparent resolution would only be reduced in the inferior quadrant where theres less squeezing going on, but still leaving it with similar resolution and FOV across the rest of the visual field.
So are you able to confirm or deny any of this? Have you done any comparisons of text readability? That seems like it might be the best way to expose any differences in apparent resolution. I was thinking that despite the increased focusing, Rift may have a hard time competing on detail simply because of the smearing produced by the lens; its pixels are effectively less distinct. But as I was typing this, I realized than in a readability test, while the Rift would look like someone came by and ran their thumb across the wet ink, the Vive will basically have faded copies of the original text copied and transposed all over the place, which may actually reduce readability if you have multiple lines of text.
Incidentally, your opening line
Because my body was complaining a bit after 2 days of room-scale madness, I decided to test the seated experience in the Vive today.
is why I think room scale will ultimately be thought of as something you go and do at Dave & Busters or Six Flags rather than in your home. It doesnt really matter how much your brain is still having after two days, because eventually your
body will say, Enough. Then, as you say, you will happily trade fun-and-exhausting for fun-and-not-exhausting. Which isnt to say youll never return to room scale, but by definition, physical exertion can only be sustained for a limited time, and the more you exert yourself, the less time youre going to have before your body fails. This will be true regardless of your fitness level. So while appeal may be broad and sustained, participation will be self-limiting.
And by the same token, I think that strong and enduring appeal and the enthusiasm it created will actually help to drive room scale enthusiasts even further out of their home and in to the arcades. It seems like a lot of room scale aficionados just want more, more, more, and arcades will easily be able to shame anything youre realistically able to set up at your house, not just in terms of scale, but in terms of physical interactivity in terms of obstacles, stairs and ladders, doors, switches, etc. Sure, I suppose you could build something just as good at your house, but at that point, you may as well start renting it out to people.
head to head comparison on tested
https://www.youtube.com/watch?v=EBieKwa2ID0
Oculus seems the better headset in many areas - fit, adjustability, comfort, optics. But motion and tracking are so important they may just make up for it.
Makes me wonder what might have happened if Oculus and Valve didn't fall out. A rift using lighthouse tech would have been pretty much perfect..
From a business perspective, I dont really understand whats going on with Oculus. When Facebook took over and Zuckerberg said they werent a hardware company, I assumed their strategy would be much like Valves; design something nice, then start trying to persuade hardware companies they should be building it. Thats why I wasnt at all surprised when the first Oculus headset turned out to be a Samsung headset instead. But not only have they come out with their own headset, instead of making it a simple peripheral to help grow the tech, their primary interest seems to be control. If they want to control the platform end-to-end, thats all well and good, but then it seems to me they shouldve simply created their own platform which they controlled. Some sort of dedicated VR platform or whatever.
The most interesting quotes from the video:
Norm: "Every time I play a Vive game, 10 minutes in, I think to myself, boy I wish I could play this exact same with the tracked controllers wearing an Oculus Rift
Jeremy: "I can say the exact same sentence, in fact when yesterday I was playing on the Vive, I had to take it off and say, UGH, I really miss my Oculus Rift, because it is just so much more comfortable"
Jeremy: If Touch was out now, there would be a lot less favourability with the Vive
Comfort is king. Once the wow wears off, the hardware and software that people stick with will be the stuff they find most comfortable.
3. The headset is more comfortable than the Vive, and I especially like being able to adjust the vertical tilt.
Oh, that's adjustable on Rift? Nice. Like I said, I think that was my biggest problem with Gear.
