• 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.

Vive was delayed because of "a very, very big technological breakthrough"

tuxfool

Banned
Wireless headset without compromising latency would be big, hope it's that, but I doubt it.

You're correct to doubt.

Wireless really is out of the door for quite a while, though I can think of one possible candidate, it don't think it is foolproof. Take it from somebody that works with wireless devices all the time.
 

Plum

Member
Advancements in A.I. that means you can interact with your waifu as if it were real life.

They delayed it to mitigate the catastrophic events it will have on the birth rate.
 

Speely

Banned
You're correct to doubt.

Wireless really is out of the door for quite a while, though I can think of one possible candidate, it don't think it is foolproof. Take it from somebody that works with wireless devices all the time.

I would be interested in any elaboration upon this candidate that you might be willing to share. :)
 

Bsigg12

Member
Better ergonomics? What could they possibly design, prototype and implement in a timely manner to both get into the consumer headset and if it requires any sort of developer attention, get out to devs in time for them to actually use said breakthrough? It seems more like positioning from the marketing team rather than just saying it's not ready yet and the late 2015 date was a mistake.
 

tuxfool

Banned
I would be interested in any elaboration upon this candidate that you might be willing to share. :)

This is just speculation on my part. But I imagine 60Ghz UWB radios could potentially yield some results, maybe.

For a current headset, you would need a transfer rate of 2560 x 1440 x 90fps x 24bpp. This yields a data rate of approximately 8Gbps just for image data. In the next Wifi standard they're going to be integrating such radios but those only have a maximum throughput of 4.6Gbps.

Now we are also ignoring the fact that these tests are done in laboratory conditions, we are ignoring sound data (and positional data though that is extremely low bandwidth). Performing any compression or encapsulation of packet data will inevitably yield more latency and there can be no retransmissions or dropped frames.

So as I said, very doubtful, but this is the only possible candidate. It should be noted that my previous assessment was based on the fact that I was under the impression that current standards had a higher throughput. Calculating the numbers makes one considerably less optimistic. This isn't even considering that future headsets will increase in resolution. Who knows somebody may still think of something?

e: looking up the bandwidth provided by display port at 1440p @60Hz is approximately 5.8Gbps.
 

kabel

Member
I hope this has anything to do with the size of it.

I don't want monstrosities like this:
HTC_Vive_(16).jpg

This is something I wold pay good money for:
 

t hicks

Banned
I hope this has anything to do with the size of it.

I don't want monstrosities like this:


This is something I wold pay good money for:

see you in about 20-30 years then
(i remember watching an oculus keynote or something that said this would take awhile)
 

Nzyme32

Member
I hope this has anything to do with the size of it.

I don't want monstrosities like this:


This is something I wold pay good money for:

They had pictures of the newer DK2 that "leaked" from their website the other day - although they could all be a part of a cunning ruse. The DK2 appears smaller than the original but it's no exactly easy to tell the scale of it.
 
You're correct to doubt.

Wireless really is out of the door for quite a while, though I can think of one possible candidate, it don't think it is foolproof. Take it from somebody that works with wireless devices all the time.

Optical maybe? Idk what kind of bandwidth you can get with that but it was something I was wondering.
I don't think it's possible to introduce any kind of wireless version at the last minute, you'd need to change too much. Expect it to be visor camera/sensors related.
 

tuxfool

Banned
Optical maybe? Idk what kind of bandwidth you can get with that but it was something I was wondering.
I don't think it's possible to introduce any kind of wireless version at the last minute, you'd need to change too much. Expect it to be visor camera/sensors related.

Optics are just EM, just like what one calls wireless. If you're going to go with a waveguide, there is fiber, which kind of defeats the purpose.
 

Seiru

Banned
Sorry, but it's really hard to take HTC seriously. All of their marketing up to and including the delay has been a disaster ("We're pleased to announce"....a delay!). And now, they come back and say, "Oh, it was because of a technological breakthrough!". Okay, HTC. Okay.
 

Speely

Banned
This is just speculation on my part. But I imagine 60Ghz UWB radios could potentially yield some results, maybe.

For a current headset, you would need a transfer rate of 2560 x 1440 x 90fps x 24bpp. This yields a data rate of approximately 8Tbps just for image data. In the next Wifi standard they're going to be integrating such radios but those only have a maximum throughput of 4.6Gbps.

Now we are also ignoring the fact that these tests are done in laboratory conditions, we are ignoring sound data (and positional data though that is extremely low bandwidth). Performing any compression or encapsulation of packet data will inevitably yield more latency and there can be no retransmissions or dropped frames.

So as I said, very doubtful, but this is the only possible candidate. It should be noted that my previous assessment was based on the fact that I was under the impression that current standards had a higher throughput. Calculating the numbers makes one considerably less optimistic. This isn't even considering that future headsets will increase in resolution. Who knows somebody may still think of something?

Thanks for the response. It's nice to know possibilities exist, even if they aren't foolproof. A wireless solution would really make the VR experience so much better.
 

hesido

Member
Eye tracking? Foveated rendering?

If they get foveated rendering to work, VR will have better graphics than "classic" renderings. Foveated rendering fits perfectly to a VR setup.

I'm assuming the breakthrough is wireless headset tech, though, maybe using Terrahertz waves? Expecting Vive to ship by the end of 2016 and brain tumors by the end of 2024.
 

viveks86

Member
Love the first page of guesses. Can't wait to see what the hell the breakthrough is. I'm guessing the new dev kits would incorporate it?

If it turns out to be nothing, I'm gonna be pissed
 

tuxfool

Banned
Thanks for the response. It's nice to know possibilities exist, even if they aren't foolproof. A wireless solution would really make the VR experience so much better.

Check my edit. My calculation has some faulty assumptions.

CW9qjoH.png


Still out of reach, but not nearly as far as based on my original calculation.
 

Nzyme32

Member
Sorry, but it's really hard to take HTC seriously. All of their marketing up to and including the delay has been a disaster ("We're pleased to announce"....a delay!). And now, they come back and say, "Oh, it was because of a technological breakthrough!". Okay, HTC. Okay.

Pretty much. If it came from Valve I'd at least be a bit more inclined to believe them, but HTC are much more dubious, so I expect overzealous marketing speak. Would be nice if it is genuine though
 

bee

Member
there was an image on reddit of a portal robot against a real life background, i'm betting some sort of AR with the camera
 

Danlord

Member
Well Sony has a 4K screen for their Xperia Z5 Premiun phone so I'm wondering if they've sourced similar screens or even that particular screen. Sony works with sourcing their camera sensors to Apple and Samsung so I don't see why they wouldn't do similar thing with their screen tech to HTC even though they're competitors.

I'm going to go with upgrade in screen tech. If they've not got 4K screen then I think they've increased the fps of their existing screens.
 

BigTnaples

Todd Howard's Secret GAF Account
SoftKinectic already doing this without a glove or controller. Sony aquired them in Oct.

https://www.youtube.com/watch?v=R53lEQq02wg

That is pretty awesome, although the limitation here is if your hand dropped out of view, you would lose whatever you were doing.


So for instance in the Vive Glove demo, he can hold something to the side, or say drop a gun to his side, and the game could still track that he has it in his hand. Where as here, in this case, it would "drop".
 

tuxfool

Banned
You just had it as Tbps instead of the actual Gbps, otherwise it checks out.

There are demos reaching almost 20Gbps bandwidth with wireless transmission, but of course that's a very controlled scenario and not even close to a realized product.

These are for longer distances, multiple users and are not in ISM frequencies. The advantages of 60Ghz is that it is extremely short distance, line of sight and ISM thus creating better wireless conditions for something like a consumer wireless VR headset. The cost of something like this must be staggering.
 

Eppy Thatcher

God's had his chance.
I had a lot of Wang jokes typed out and then deleted the whole thing out of shame. Still though.

Eye tracking would be cool... but is that really a Very big tech breakthrough?... it would be sweet if they had something on the outer case that could track your hands/arms so you didn't need any devices in your hands to do stuff in the VR...
 
Top Bottom