• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

HTC Vive Launch Thread -- Computer, activate holodeck

Status
Not open for further replies.
Man. I played like an hour of SuperHOT, and when I came out I caught myself moving as if time would only move when I did. Kind of a surreal moment lol.
 

Durante

Member
Did someone try the Vermintide VR thing? I forgot about it and it's 1am now, so I will give it a shot tomorrow. Reviews have not been that great.
The VR implementation is apparently really bad. Bad performance, bad IQ, inexplicable latency, and so on.
 

dsk1210

Member
I really dislike the optics in the vive, I hate the banding that the fresnel lenses cause.

Not only does the resolution need improved but the optics need a revision as well, black frame insertion while we are at it, I want perfect blacks.
 

Durante

Member
"Black frame insertion" doesn't provide perfect blacks. In fact, it doesn't change the contrast ratio at all.
It's a basic technique to implement a low-persistence display -- the Vive already uses low-persistence (with even less persistence actually than standard black frame insertion). IIRC the display has a 10% duty cycle, which is much less persistence than the 50% duty cycle of black frame insertion.

What you want is for OLED panel makers to improve the color switching performance when turning on a pixel that was completely off.
 

Haint

Member
I really dislike the optics in the vive, I hate the banding that the fresnel lenses cause.

Not only does the resolution need improved but the optics need a revision as well, black frame insertion while we are at it, I want perfect blacks.

Have you actually used a Rift or PSVR extensively and critically? Vive's rings are by far the least offensive of the 3 when compared to Rift's god awful god rays and PSVR's sub-millimeter sweet spot and significant CA.
 

Durante

Member
I'm still extremely curious to try the OSVR HMD's dual-lens optics. I'd love to know how they stack up, because clearly none of the other solutions we have currently are perfect.
 

dsk1210

Member
"Black frame insertion" doesn't provide perfect blacks. In fact, it doesn't change the contrast ratio at all.
It's a basic technique to implement a low-persistence display -- the Vive already uses low-persistence (with even less persistence actually than standard black frame insertion). IIRC the display has a 10% duty cycle, which is much less persistence than the 50% duty cycle of black frame insertion.

What you want is for OLED panel makers to improve the color switching performance when turning on a pixel that was completely off.

The DK2 did low persistence and maintained the blacks. If you ran without low persistence the blacks smeared just like the Vive did when the gamma had been adjusted in the beta.

I did mean low persistence but forgot the name for it.
 

dsk1210

Member
Have you actually used a Rift or PSVR extensively and critically? Vive's rings are by far the least offensive of the 3 when compared to Rift's god awful god rays and PSVR's sub-millimeter sweet spot and significant CA.

Used the DK2 quite a bit.

Some chromatic abrasion and a tighter sweet spot bit I preferred the look through the lenses.

Still love the vive though.
 

Durante

Member
I have a DK2 right next to me and used it from its launch until the Vive/CV1 launch. It had 2 modes of operation: true black smear or much more noticeable non-black than the current consumer HMDs.

Sadly this is an issue with the OLED panel tech used in all HMDs that needs to be fixed at the tech level.
 

Arulan

Member
I originally had a defect in mine where the right eye/screen wouldn't apply mura correction, thus giving me the black smear. Being able to compare the two as easily as closing one eye, mura correction is a necessary compromise in my opinion, but I'd love to see better panels address the issue in future iterations.
 

Zalusithix

Member
Cross posting this since seemingly nobody checks the regular VR thread, and it's prototyped on the Vive anyhow.

Road to VR has an article up on one of the upcoming wireless VR accessory contenders. Based on WiFi standards, proprietary compression scheme, less than 3ms total latency.

Mach-2K Specs:

  • Current resolution fully supported 2160 x 1200
  • Current frame-rate fully supported 90Hz
  • Planned resolution in the near future 4K per eye
  • Planned frame-rate in the near future 120Hz
  • Main CPU: FPGA
  • I/O: HDMI, USB 2.0, 12volts out.
  • Eye tracking input
  • Supply Power: 5 Volts DC
  • Current Frequencies: 802.11ac Wi-Fi 5Ghz
  • Future Supported Frequencies: Up to 60Ghz WiGig
  • On-board software: B.A.I.T “Biologically Augmented Image Transmission” Algorithm
  • OEM and SDK options available, allowing third parties to create application specific modules for the algorithm.
  • User select-able compression schemes
Not sure what the eye tracking is since there's no way they could achieve that with current headsets. Perhaps it's meant that they plan to hijack into future foveated headsets and use eye position to adaptively adjust compression across a frame with more compression used on the periphery.

The interesting thing to me is that it's a belt based unit with antennas that wrap around the body. Assuming this works, it rather neatly solves the high frequency problem. Rather than having the need for the transmitter high up in a central location to clearly see the head like the current Vive solution, this would allow it placed at desktop height and still have a LOS to the unit or its antennas regardless of body angle or position.
Product-Shot-02.jpg

Only time will show whether the compression is up to the task. They have these samples provided of a raw image compared to their old compression and new version, but it's easy enough to pull deceptive samples.
 
Cross posting this since seemingly nobody checks the regular VR thread, and it's prototyped on the Vive anyhow.

Road to VR has an article up on one of the upcoming wireless VR accessory contenders. Based on WiFi standards, proprietary compression scheme, less than 3ms total latency.

Mach-2K Specs:

  • Current resolution fully supported 2160 x 1200
  • Current frame-rate fully supported 90Hz
  • Planned resolution in the near future 4K per eye
  • Planned frame-rate in the near future 120Hz
  • Main CPU: FPGA
  • I/O: HDMI, USB 2.0, 12volts out.
  • Eye tracking input
  • Supply Power: 5 Volts DC
  • Current Frequencies: 802.11ac Wi-Fi 5Ghz
  • Future Supported Frequencies: Up to 60Ghz WiGig
  • On-board software: B.A.I.T “Biologically Augmented Image Transmission” Algorithm
  • OEM and SDK options available, allowing third parties to create application specific modules for the algorithm.
  • User select-able compression schemes
Not sure what the eye tracking is since there's no way they could achieve that with current headsets. Perhaps it's meant that they plan to hijack into future foveated headsets and use eye position to adaptively adjust compression across a frame with more compression used on the periphery.

The interesting thing to me is that it's a belt based unit with antennas that wrap around the body. Assuming this works, it rather neatly solves the high frequency problem. Rather than having the need for the transmitter high up in a central location to clearly see the head like the current Vive solution, this would allow it placed at desktop height and still have a LOS to the unit or its antennas regardless of body angle or position.


Only time will show whether the compression is up to the task. They have these samples provided of a raw image compared to their old compression and new version, but it's easy enough to pull deceptive samples.

I'm all over this as long as the price isn't too crazy. What are people thinking this is going to cost $300 or $400?
 

Durante

Member
The people making it seem to think $1200-1500 ;) (In the interview)

I don't think it's actually particularly expensive to build, from the description. The most expensive part is likely the FPGA (they might need a large/high-performance one) which could be replaced by a ASIC for a high-quantity consumer version.

I do wonder how their compression works, the given description is hardly more than "poof proprietary magic". It does sound like it might be non-temporal, and then it could just be a fast FPGA implementation of something like a variation of JPEG2000 or intra-frame x265 encoding.
 

kinggroin

Banned
Heads up to anyone who hasn't bought Superhot because they're worried about Oculus breaking revive in the future--the game is DRM free and will work even if you uninstall Oculus Home, provided you've copied the Oculus Runtime DLL's into the same directory as the game executable. You can backup that copy to a safe place, and the game is guaranteed to work forever.

WHAT?


Oh, this is incredible news!
 

kinggroin

Banned
I really dislike the optics in the vive, I hate the banding that the fresnel lenses cause.

Not only does the resolution need improved but the optics need a revision as well, black frame insertion while we are at it, I want perfect blacks.

We had them.


:-(

Of course BFI will help with the drawback of "all off" black levels with current displays (smearing, if I wasn't clear enough) - but according to Durante, the current panels already do a better job of this than BFI.

Is Superhot actually good in Vr?


It's not only good, it's probably the best VR experience I've had since first trying VR. Seriously.
 

Zalusithix

Member
The people making it seem to think $1200-1500 ;) (In the interview)

I don't think it's actually particularly expensive to build, from the description. The most expensive part is likely the FPGA (they might need a large/high-performance one) which could be replaced by a ASIC for a high-quantity consumer version.

I do wonder how their compression works, the given description is hardly more than "poof proprietary magic". It does sound like it might be non-temporal, and then it could just be a fast FPGA implementation of something like a variation of JPEG2000 or intra-frame x265 encoding.

Yeah, I imagine the price point quoted is based on turning a profit while having a FPGA (expensive) and low estimated sales volume (VR arcades, theme parks, training etc.). Using an ASIC and going up to general consumer economies of scale would drastically change what they could charge.

No idea with the compression, and I doubt we'll ever find out. They obviously view that as a key linchpin to their product's viability.
 
Thumper was really cool before, but playing it in VR probably puts it really close to GOTY for me. The extra immersion almost feels like a vital component. Like it's not complete without it.

Is there any possible way we get Rez Infinite on PC? That's the only other game I can see coming close in a similar way.

Is Superhot actually good in Vr?

It's one of the best VR games currently available.
 

kinggroin

Banned
Thumper was really cool before, but playing it in VR probably puts it really close to GOTY for me. The extra immersion almost feels like a vital component. Like it's not complete without it.

Is there any possible way we get Rez Infinite on PC? That's the only other game I can see coming close in a similar way.
.

Maybe someone will mod a Dreamcast emulator and let us play Rez on the Vive that way.
 

ArtHands

Thinks buying more servers can fix a bad patch
Thumper was really cool before, but playing it in VR probably puts it really close to GOTY for me. The extra immersion almost feels like a vital component. Like it's not complete without it.

Is there any possible way we get Rez Infinite on PC? That's the only other game I can see coming close in a similar way.



It's one of the best VR games currently available.

yah, where the hell is Rez Infinite for PC?
 

Maiden Voyage

Gold™ Member
Anyone tried it yet?

I will likely pick it up at some point. The demo was good and I think a full game could be neat. At that price point it seems like an easy purchase for me. However, I'm holding off on buying anything until the Steam sale starts. I don't want to go too crazy.
 

Raticus79

Seek victory, not fairness
Yeah, SUPERHOT VR looks like a massive improvement in the gameplay, with tricks like bullet slices becoming relatively trivial.
jacksepticeye having some fun with it: https://www.youtube.com/watch?v=KuGz6IKldZI

Definitely agree with this comment: "Ok, someone needs to make a fucking pair of VR legs so you can fuckin Sparta kick the guy off of the helicopter"

Seriously, a top priority should be making a Lighthouse-based device you can strap around each shoe for foot tracking.
 

E-flux

Member
Is there any moving around in Superhot VR? Teleporting or anything like that? Or is it just a stationary game?

I tried out the Vermintide VR, it was super bizarre, i was in the way of the skaven, none of them paying any attention to me, just rushing past me and the stationary heroes who were just standing in the corner while i shot skaven that were trying to kill a door. It's also one of the worst looking vr games i have seen, everything was so blurry and the scale seemed to be way off, i'm pretty tall but while i was standing next to the witch hunter he was just towering over me, same with the other characters i saw around the place. It was pretty hilarious how when you hit Skaven with a melee weapon the flew to the walls at like 200 km/h even from the smallest swipe.
 

Zalusithix

Member
Seriously, a top priority should be making a Lighthouse-based device you can strap around each shoe for foot tracking.

Top priority is torso tracking. We need that just to get artificial locomotion working well and proper chest/belt interaction. Then feet. After that the elbows and knees.
 

E-flux

Member
Top priority is torso tracking. We need that just to get artificial locomotion working well and proper chest/belt interaction. Then feet. After that the elbows and knees.

Why would you need torso tracking if you have feet that are tracked? Wouldn't it be pretty "easy" to calculate which direction your torso is pointing based of your hands and feet?
 

Raticus79

Seek victory, not fairness
Top priority is torso tracking. We need that just to get artificial locomotion working well and proper chest/belt interaction. Then feet. After that the elbows and knees.

Hmm, I'm not seeing it (but maybe I'm missing something). If I could only pick one, I think the feet would make a much bigger impact just due to the sheer scope of the gameplay it opens up. *Godzilla-stomps some cars*

IK could do a better job of approximating torso position with the feet known at least. Maybe a belt as the next one? With positions captured for feet, pelvis, hands and skull, there's not much range of motion left. Elbows for sure, but I think knees might not be necessary.
 
Damn, hope this is out by Friday as I'm having a VR party and this looks fun to play.

Also I should probably get Blarp!

If you don't mind me asking, what games do you play for VR parties? I might get a headset in the near future and it'd be awesome to know what games are good for something like that.
 

Zalusithix

Member
Why would you need torso tracking if you have feet that are tracked? Wouldn't it be pretty "easy" to calculate which direction your torso is pointing based of your hands and feet?
Feet would narrow in the position and direction of the torso, but you'd still be guessing. With movement, you don't want to be guessing. You can stand with one foot facing forward and another angled. Is your torso aimed at the straight forward foot or the angled foot? IK won't tell you accurately; it'll just make up a possible solution. This affects natural interaction with smaller regions of the torso as well. We're quite adapt at using our body (torso) as inventory. The concept of various pockets and belt pouches is ancient and we can blindly reach to these areas accurately IRL even if they're rather small. This is impossible right now as the torso doesn't rotate like it does in real life and doesn't accurately reflect our bending over at all.

Hmm, I'm not seeing it (but maybe I'm missing something). If I could only pick one, I think the feet would make a much bigger impact just due to the sheer scope of the gameplay it opens up. *Godzilla-stomps some cars*

IK could do a better job of approximating torso position with the feet known at least. Maybe a belt as the next one? With positions captured for feet, pelvis, hands and skull, there's not much range of motion left. Elbows for sure, but I think knees might not be necessary.

The amount of interactions we do with our feet to the world is secondary to the possibilities the torso tracking would provide to self interactions. Accurate tracking of the feet also provides other problems to deal with. Right now you can be on uneven terrain and since you don't see your feet, nothing is a big deal. Once you track your feet, they'll either be clipping into the environment or floating over it. If you match the feet to the environment, then you get a mismatch between the VR and IRL positions. Better to tackle the easier to solve solution first. A belt for torso tracking can also conveniently carry a battery pack for wireless. Two birds with one stone.
 

Vanmunt

Banned
Got a million views on a VR video after one day - it works. I just assume it's a persona being put on for the show and don't sweat it.

1 Million views.... Christ on a bike, I was definitely born 25 years too late as I can be annoying as hell with out even trying.
 

E-flux

Member
Feet would narrow in the position and direction of the torso, but you'd still be guessing. With movement, you don't want to be guessing. You can stand with one foot facing forward and another angled. Is your torso aimed at the straight forward foot or the angled foot? IK won't tell you accurately; it'll just make up a possible solution. This affects natural interaction with smaller regions of the torso as well. We're quite adapt at using our body (torso) as inventory. The concept of various pockets and belt pouches is ancient and we can blindly reach to these areas accurately IRL even if they're rather small. This is impossible right now as the torso doesn't rotate like it does in real life and doesn't accurately reflect our bending over at all.

You would be guessing but given how vive already tracks your head and hands accurately i think that the torso tracking would be good enough if feet would be added in.Onward looks derpy as hell but even in that your body is where you think it is 90% of the time when you are reaching for your holsters or any other item on your body.
 

Zalusithix

Member
Entertaining is not a description I would ascribe to this guy. Horrifying and annoying, perhaps.

Welcome to the Youtube generation. =P

You would be guessing but given how vive already tracks your head and hands accurately i think that the torso tracking would be good enough if feet would be added in.Onward looks derpy as hell but even in that your body is where you think it is 90% of the time when you are reaching for your holsters or any other item on your body.

90% of the time isn't good enough. If your controllers had a 10% error margin where they didn't do what you expected them to, they'd be in the trash.
 

E-flux

Member
Welcome to the Youtube generation. =P



90% of the time isn't good enough. If your controllers had a 10% error margin where they didn't do what you expected them to, they'd be in the trash.

That is why i think that with feet tracking that last 10% could be eliminated, and with the virtual body the problem is not as severe as controller not functioning properly, since it's not that off putting when your body is in a 20 to 40 degree angle instead of facing forward, which it will do as soon as you take a step.
 
Status
Not open for further replies.
Top Bottom