Sony VR Headset/Project Morpheus/PlayStation VR
If it's 15th, then it's probably on 16th on my country. Blargh. More waiting.
D
Deleted member 11852
Guest
I hope they stream it in 3D. Exclusively in 3D. Screw all the peasants with 2D equipment, they're not worthy!
Pfffft... You can't force me to watch it in 3D, all I have to do is close one eye!I hope they stream it in 3D. Exclusively in 3D. Screw all the peasants with 2D equipment, they're not worthy!
Last edited:
Good point, I don't think the OR setup is designed for room scale, I thought at the time that PL was just saying that to make it appear CV1 would not have any major weaknesses compared to Vive. Until OR can demonstrate a CV1 setup that works at room scale and looks realistic then I think we should be skeptical. The technology just doesn't appear reasonable for it, while the Vive does.
it should work fine with 3 cameras in a V formation. But you'd have to run a 3rd cable all the way to the back of your room
You would only need two cameras, on opposite corners, just like vive. The occlusion pattern would be the same.
The only difference is the processing required (vive is almost free), and the imaging sensors being discreet pixels, limiting range (doubling the usable range would require 4 times the resolution, quickly hitting the limits of usb3). Vive resolution is time-based instead of spatially based, it's limited only by electrical noise, sensor rise time, and sensitivity. It should be easily extensible with upgrades up to warehouse scales. Which, admitedly, is fucking awesome.
Trying to get 50+ feet of usb3 with repeaters could be problematic, but possible. Anyway, room scale is nowhere near a mass market proposition, I don't think it will be for quite a while because it needs a dedicated room.
The only difference is the processing required (vive is almost free), and the imaging sensors being discreet pixels, limiting range (doubling the usable range would require 4 times the resolution, quickly hitting the limits of usb3). Vive resolution is time-based instead of spatially based, it's limited only by electrical noise, sensor rise time, and sensitivity. It should be easily extensible with upgrades up to warehouse scales. Which, admitedly, is fucking awesome.
Trying to get 50+ feet of usb3 with repeaters could be problematic, but possible. Anyway, room scale is nowhere near a mass market proposition, I don't think it will be for quite a while because it needs a dedicated room.
The Lighthouse base stations have a wide enough FOV on both axes that they can get full room coverage from both positions, the Rift cameras I believe are something like 70deg vertical, so you're not going to get the full corner to corner, floor to ceiling coverage no matter how you place them. Also I don't think we can reliably say that both systems are equal to the same degree when handling occlusion or how quickly they reacquire position after losing tracking. Conceivably the lighthouse system requires minimal visible tracking points over a single cycle to establish a position, while Oculus requires picking up an encoded strobe ID from the individual leds to know which are which and *then* do the image processing for position. I think overall it's just that Valve's tracking system had a very different design goal from the get-go, while Oculus has been expanding a system that was originally conceived to handle a single HMD in narrow space.
edit:
In that sense it's interesting to compare the differing needs of tracking an HMD and hands. With an HMD you have a device that almost never loses tracking because of the large amount of tracking points distributed over a large area (difficult to occlude it entirely), and limited amount of movement (it's less common for your head to leave the camera frustum than it would your hands.) With hands though you have the opposite - a small device with few tracking points increasing instances of occlusion, and your wingspan causing a lot of potential for reaching outside the frustum. Plus I can't imagine it being any easier to identify strobing leds over a series of frames when they're attached to a fast moving hand.
edit:
In that sense it's interesting to compare the differing needs of tracking an HMD and hands. With an HMD you have a device that almost never loses tracking because of the large amount of tracking points distributed over a large area (difficult to occlude it entirely), and limited amount of movement (it's less common for your head to leave the camera frustum than it would your hands.) With hands though you have the opposite - a small device with few tracking points increasing instances of occlusion, and your wingspan causing a lot of potential for reaching outside the frustum. Plus I can't imagine it being any easier to identify strobing leds over a series of frames when they're attached to a fast moving hand.
Last edited:
Ah yeah, I forgot about the added latency from strobing id...
I don't know what they did, but it doesn't look like Sony is strobing since the id is color+shape (from their patent filings, and richard marks research on ps2 eye toy). Wouldn't the ids be already closely predicted by the previous frames plus sensor fusion, and even the averaged gravity vector, which leaves a rather simple algorithm to trangulate against two cameras? It seems hard to actually lose the IDs to the point it needs to strobe them. It should only strobe as a desperate measure after some major occlusion event. Not just losing a point that reappeared where it should.
I don't know what they did, but it doesn't look like Sony is strobing since the id is color+shape (from their patent filings, and richard marks research on ps2 eye toy). Wouldn't the ids be already closely predicted by the previous frames plus sensor fusion, and even the averaged gravity vector, which leaves a rather simple algorithm to trangulate against two cameras? It seems hard to actually lose the IDs to the point it needs to strobe them. It should only strobe as a desperate measure after some major occlusion event. Not just losing a point that reappeared where it should.
I think this is true and gets to the nub of the issue.
The current lighthouse is rather 'jittery' at "warehouse" (20*20m) scale; you can see it demonstrated in the videos from the youtube channel "node guys". In any case, lighthouse is the superior room scale tracking option, in everything; cpu requirements, usb (cable, ports, processing) requirements, set up, tracking 'resolution', everything. There is not a single thing in which the Oculus tech excels, not a single thing. This is the very reason for the delayed motion control launch; they hope enough people, like eastman, are sold on the lie that oculus will have comparable tech, but that will never happen. Then "pride of ownership" takes over, and eastman and the likes will convince themselves "Room-scale is a gimmick, seated VR is where it's at! Who needs precise room-scale tracking? We still have cables anyway lol, I don't want to trip over! haha ".
If I was ever going to build a PC again (which is such a joke these days); then I would certainly go with Valve/ HTC. Better tech, games, everything. They don't have a lying CEO running the place for one thing. And that is something which is rare today trust.
On the topic of trust, I trust PSVR will deliver tonight. For me it's like the old iPhone VS the rest; nowadays iPhone CPU's always completely destroy the competition, but in the time of iPhone 4 VS Galaxy S1; a lot of people thought the Galaxy S1 would be faster, and better, and more future proof based on the specs. It would have better games because of a supposed faster processor and GPU, which on paper, was true.
The reality is that iPhone received the best, and the most games. It also had the fastest browser, calendar, OS and everything else. Also looking at updates and support it outlasted the Galaxy S1, S2 and even the S3. I think PSVR can be like that underpowered iPhone. Mocked by fans of the competition, but it will outlast them all.*
Lasting support and talented developers for VR are way more important than plus-baseline tech anyway
*Disclaimer: I have not bought an iPhone since 2010
If I was ever going to build a PC again (which is such a joke these days); then I would certainly go with Valve/ HTC. Better tech, games, everything. They don't have a lying CEO running the place for one thing. And that is something which is rare today trust.
On the topic of trust, I trust PSVR will deliver tonight. For me it's like the old iPhone VS the rest; nowadays iPhone CPU's always completely destroy the competition, but in the time of iPhone 4 VS Galaxy S1; a lot of people thought the Galaxy S1 would be faster, and better, and more future proof based on the specs. It would have better games because of a supposed faster processor and GPU, which on paper, was true.
The reality is that iPhone received the best, and the most games. It also had the fastest browser, calendar, OS and everything else. Also looking at updates and support it outlasted the Galaxy S1, S2 and even the S3. I think PSVR can be like that underpowered iPhone. Mocked by fans of the competition, but it will outlast them all.*
Lasting support and talented developers for VR are way more important than plus-baseline tech anyway
*Disclaimer: I have not bought an iPhone since 2010
The issue is when there are no 'previous frames', which is what we're talking about when you lose external tracking from occlusion or moving outside the frustum bounds. The other unknown here is how well an Oculus camera is able to handle acquiring those IDs when the LEDs are smeared from very fast motion. That's not to say that Oculus's tracking tech is inherently bad or will even necessarily be bested in practice by the Vive - the old adage of 'the proof of the VR is in the using' still holds here, so I'm willing to reserve judgement on Touch until it's released. ...I'm just not willing to wait until Touch's "H2" launch to be excited about working with VR again, so I'll have a Vive at least until then. That and I've got a nice 13'x16' empty space staring at me making my gears turn with ideas.
You make it difficult to talk frankly about technical shortcomings if you're going to use it to steer the conversation like that.
I concur. Lots of us are bitter and cynical and angry, but we aim to leave it at the Firewall when posting on B3D...
I'm sorry for wording it like that, but that's how it is; they simply delayed their motion controller solution do avoid direct comparison, and to keep Oculus fans / potential Oculus buyers in check. I really want to believe them, but if they even had a working room scale solution, the SDK would reflect that. Instead, 'fans' and tech enthusiasts like myself are left in the dark. Palmer tweeted a photo of 2 cameras in 1 room and that's it. He is also extremely vague in interviews about it. Implying that their tech has the same capabilities..
Remember when people asked 1 week before the 599 reveal: "You promised a 350 dollar VR solution, are you still on track?" ?
He simply responded: "yes, we are extremely happy at where we are right now" or something to that extend.
Take into account he is even more vague about 2 camera room scale vr tracking capabilities, and I am pretty sure that it will never come to fruition.
Sony as well as Valve have been really open in comparison; giving presentations on their tech, how they solve issues, answering questions in a honest way. There is nothing dishonest about their technology; it is proven, and demonstrated in real life, real use scenarios, by press end end users as well.
Remember when people asked 1 week before the 599 reveal: "You promised a 350 dollar VR solution, are you still on track?" ?
He simply responded: "yes, we are extremely happy at where we are right now" or something to that extend.
Take into account he is even more vague about 2 camera room scale vr tracking capabilities, and I am pretty sure that it will never come to fruition.
Sony as well as Valve have been really open in comparison; giving presentations on their tech, how they solve issues, answering questions in a honest way. There is nothing dishonest about their technology; it is proven, and demonstrated in real life, real use scenarios, by press end end users as well.
The problem with the Rift and its tracking is simple , you need more cameras. They can correct this by simply adding a second camera mounted in a V postion instead of the I postion the single camera is in now.
Tommorow the embargo lifts so we should see more info on the headset and touch since there were a bunch of games using it.
For problems with needing more cameras you can simply run a pass through usb port at the base of each one. I doubt one of the cameras needs the full bandwidth of usb 3
Tommorow the embargo lifts so we should see more info on the headset and touch since there were a bunch of games using it.
For problems with needing more cameras you can simply run a pass through usb port at the base of each one. I doubt one of the cameras needs the full bandwidth of usb 3
What time is is now in the PSVR gdc region? Any countdown site?
I feel that room scale tracking is a cool feature to have now but its going to be a gimmick for most users until the headsets can go wireless.
By the time that happens the screens will be so much better that these first headsets will be long forgotten.
In my opinion the htc headset is going to struggle, surely occulus with all that Facebook money is going to have a much bigger marketing run than anything htc will come up with.
By the time that happens the screens will be so much better that these first headsets will be long forgotten.
In my opinion the htc headset is going to struggle, surely occulus with all that Facebook money is going to have a much bigger marketing run than anything htc will come up with.
I don't believe its being streamed anywhere. So we will have to wait for news to get out and who knows if there is an embargo or not
Similar threads
