Valve/HTC Vive Discussion

[Chivas]

The HTC has suggested that the Vive was delayed because Valve made some sort of breakthrough design. Very interesting, hopefully its has something to do with improving the display resolution, or it could also be a very good tactic to make people second guess their decision to buy the Rift. OR more importantly slow down the hit their stocks are taking since the delay announcement.

Edited December 12, 2015 by Chivas

[Chivas]

HTC is suggesting that the consumer Vive changes will include better weight distribution. Currently the HMD weight is too far forward, so they are changing the strapping system etc for better balance. They are also changing the wands, for better balance. These changes can't be the technological breakthrough they've hinted at.

 

I'm assuming that Vive came to the realization that they couldn't compete with the Oculus price, and specs, so have delayed their headset to add whatever the tech breakthrough they've found to atleast compete, or dominate Oculus. depending on the quality of their tech breakthrough. Unless Oculus is really confident with their final specs, you can bet that Oculus is scrambling to find out what that tech breakthrough is.

 

Oculus and Vive have been experimenting with some sort HDR lighting display that will blow away any of the current HMD's. Valves techs say it gives an awesome experience before it melts down. Its a beach demo where the actually feel the heat of the sun, which could also be part of the problem. ;) So much so its very disappointing when they go back to the current headset tech. Interesting times.

[spacenavy90]

HTC is suggesting that the consumer Vive changes will include better weight distribution. Currently the HMD weight is too far forward, so they are changing the strapping system etc for better balance. They are also changing the wands, for better balance. These changes can't be the technological breakthrough they've hinted at.

 

I'm assuming that Vive came to the realization that they couldn't compete with the Oculus price, and specs, so have delayed their headset to add whatever the tech breakthrough they've found to atleast compete, or dominate Oculus. depending on the quality of their tech breakthrough. Unless Oculus is really confident with their final specs, you can bet that Oculus is scrambling to find out what that tech breakthrough is.

 

Oculus and Vive have been experimenting with some sort HDR lighting display that will blow away any of the current HMD's. Valves techs say it gives an awesome experience before it melts down. Its a beach demo where the actually feel the heat of the sun, which could also be part of the problem. ;) So much so its very disappointing when they go back to the current headset tech. Interesting times.

 

I would argue that the Vive is actually more advanced especially with its Lighthouse tracking and better wands. However, I have a feeling that the later release date, apparently lack of game support (from what I can see now at least) and the higher price tag will hurt them. I'll still pick one up though.

[Chivas]

I would argue that the Vive is actually more advanced especially with its Lighthouse tracking and better wands. However, I have a feeling that the later release date, apparently lack of game support (from what I can see now at least) and the higher price tag will hurt them. I'll still pick one up though.

 

 

Lighthouse may or may not be more advanced, and think there are pros and cons to each system. Both systems are still in flux so specs/features could easily change. Oculus current demos of the Touch inputs have put the two camera trackers more infront of the user, where it works better for the vast majority of current VR software content. There will be occlusion problems for any two component tracking system, unless Lighthouse can see thru the body, arms, hands, which it can't. Oculus looked all types of tracking systems, and I think they chose the camera system, because it made it easier ,and cheaper to track full body, and facial movement.

 

Personally I think the Oculus Touch inputs are better and more versatile than the Wands because of the added finger dexterity, and tracking feature. That said the Vive wands are going thru a major overhaul and could easily add a feature like that.

[spacenavy90]

Lighthouse may or may not be more advanced, and think there are pros and cons to each system. Both systems are still in flux so specs/features could easily change. Oculus current demos of the Touch inputs have put the two camera trackers more infront of the user, where it works better for the vast majority of current VR software content. There will be occlusion problems for any two component tracking system, unless Lighthouse can see thru the body, arms, hands, which it can't. Oculus looked all types of tracking, and I think they chose the camera system, because it made it easier ,and cheaper to track full body, and facial movement.

 

Except they ARE more advanced in terms of positioning, tracking and latency while also being significantly cheaper than the Oculus' cameras.

 

You can't really say that Oculus "looked at all the types of tracking" either because what they accomplished with Lighthouse was thinking straight out of the box. I bet if Oculus knew about it before, they would've gone with it instead. But at this point they are dedicated.

 

This being said, Oculus still has the advantage in terms of mainstream awareness and game support. Though Valve is making good strides in reaching out to the community especially with OpenVR.

[Chivas]

Except they ARE more advanced in terms of positioning, tracking and latency while also being significantly cheaper than the Oculus' cameras.

 

You can't really say that Oculus "looked at all the types of tracking" either because what they accomplished with Lighthouse was thinking straight out of the box. I bet if Oculus knew about it before, they would've gone with it instead. But at this point they are dedicated.

 

This being said, Oculus still has the advantage in terms of mainstream awareness and game support. Though Valve is making good strides in reaching out to the community especially with OpenVR.

 

Interesting I read that the Lighthouse tracking was significantly more expensive than the Oculus cameras, which is one of the reasons that the Vive system will more expensive. I read this quite sometime ago, but will try to find a link.

 

Palmer did state they looked at the Lighthouse tracking, but wanted something that would more easily track full body movement, and facial expression. One reason shop simulator only shows the hands in their demos, is there is a problem in that only the hands are tracked, and any difference between how the body is displayed, and its actual movement causes major problems for the user. Interestingly they found when only the hands are displayed, users imaginations just filled in the missing parts.

Edited December 12, 2015 by Chivas

[Chivas]

I couldn't find the link, but if I remember correctly, part of the reason is the more expensive active sensors on the Lighthouse system, as opposed to the less expensive passive sensors on the Constellation system. BUT it could be a wash if the cameras are more expensive than the Laser boxes.

 

One system may have lower latency numbers than the other, but users have experienced no noticeable difference in latencies in either system, so if one is slightly faster, does it really matter.

 

It doesn't take much imagination to figure out why Palmer would prefer full body, and facial expression tracking, that's easier to implement with cameras.

Edited December 13, 2015 by Chivas

[vicx]

part of the reason is the more expensive active sensors on the Lighthouse system, as opposed to the less expensive passive sensors on the Constellation system

 

You have mixed up a few concepts in that statement. These videos should help you out.

 

How the lasers scan.

 

How the electronics work.

 

The sensors on the Lighthouse controllers and HMD are very simple components; no more complex or expensive than a basic photo-transistor. The sensor fusion on the controllers can be done by cheap micro-controllers. IMO the haptic part of the controllers is a way more complex than the positioning component and yet it too can be produced cheaply.

 

The Lighthouse emitters are not necessarily expensive to produce. The emitters I expect will have undergone serious industrial redesign and this is one area where I am very curious to see what the HTC engineers have come up with.

 

A Rift with one camera and xbone controller is going to be cheaper than a Vive with two lighthouse controllers and two lighthouse emitters. BUT a Rift with two cameras and two constellation touch controllers approaches parity with the Vive in terms of cost.

[Chivas]

You have mixed up a few concepts in that statement. These videos should help you out.

 

How the lasers scan.

 

How the electronics work.

 

The sensors on the Lighthouse controllers and HMD are very simple components; no more complex or expensive than a basic photo-transistor. The sensor fusion on the controllers can be done by cheap micro-controllers. IMO the haptic part of the controllers is a way more complex than the positioning component and yet it too can be produced cheaply.

 

The Lighthouse emitters are not necessarily expensive to produce. The emitters I expect will have undergone serious industrial redesign and this is one area where I am very curious to see what the HTC engineers have come up with.

 

A Rift with one camera and xbone controller is going to be cheaper than a Vive with two lighthouse controllers and two lighthouse emitters. BUT a Rift with two cameras and two constellation touch controllers approaches parity with the Vive in terms of cost.

 

My point still stands, the light points on the Rift headset/controllers don't pass data. The Lighthouse sensors do, and there are 19 sensors on each controller, and quite a few on the headset. The Lighthouse sensors could still be cheap, but not as cheap as the constellation passive diodes, or whatever they are on the Rift.

 

Both systems can have five parts, a headset, two controllers, and two trackers. All else being equal if one company sells their hardware near cost, as Oculus has suggested they will, then its quite likely the Vive system could be at least a hundred dollars more expensive. Not to mention Oculus partnership with Samsung could avail them custom displays at a cheaper price point than Vive. HTC does not manufacture their own displays, and would have to contract them out. No word yet, on Vives display manufacturer, or what type of contract/partnership they have with them.

 

Vive could obvious also sell their hardware near cost, but they haven't suggested they will. Considering HTC current problems, and the double digit loss in share price since the delay announcement, its unlikely they are in the mood to sell their hardware near cost.

 

If I were a betting man, I'd say Oculus had a considerable advantage. That said, Vive's delay, and the suggested reason for the delay could easily change the betting line. I'm not so sure, but hoping that the first consumer versions of VR headsets are good enough for flight sims. Hopefully whatever breakthrough Vive/Valve have made will make decent VR for flight sims more likely in the first consumer versions.

[vicx]

My point still stands, the light points on the Rift headset/controllers don't pass data. The Lighthouse sensors do, and there are 19 sensors on each controller, and quite a few on the headset. The Lighthouse sensors could still be cheap, but not as cheap as the constellation passive diodes, or whatever they are on the Rift.

 

You aren't making ANY valid points about the controllers at all. Both the Vive and Oculus touch controllers require microcontrollers, as well as the standard 'instantaneous' motion sensors, batteries, battery charging circuits, wireless transceivers and haptics drivers. The cost of a dozen or more photo-diodes is a matter of a few cents.

Edited December 13, 2015 by vicx

[Chivas]

You aren't making ANY valid points about the controllers at all. Both the Vive and Oculus touch controllers require microcontrollers, as well as the standard 'instantaneous' motion sensors, batteries, battery charging circuits, wireless transceivers and haptics drivers. The cost of a dozen or more photo-diodes is a matter of a few cents.

 

 

http://metaversing.com/2015/03/24/examining-the-valvehtc-vive-ecosystem-basic-sensors-and-processing/

Not so sure why its so hard to understand. The Oculus Rift is just using dumb markers on the headset and inputs. They aren't connected to anything and don't have to transmit data. The Vive markers {Sensors}, must see, and send the angle, and timing of the laser pulses from the trackers. What would be more expensive, sixty sensors, or sixty dumb markers.

[Chivas]

https://youtu.be/WJcjWCNkuCU

 

Interesting Video, if your a Rift fanboy, you might not want to watch, but in fairness he appears to have tried the CV1 without inputs. VR gets the biggest reaction when our hands and movement are in the game. I understand that movement is a presences multiplier in VR, but as primarily a flight simmer I don't entirely buy the concept that you have to move around to have a great VR experience.

 

Again I don't think he was using a CV1, but a prototype that the CV1 will be based on, but not necessarily the same specs/features.

[hannibal]

i saw this video. not a fair comparison, it was bash of oculus product just because the rift did not have room scale implementation.

[Chivas]

Sixty sensors at a few cents each. Will make no difference money-wise, but a big difference in quality of experience.

 

I doubt your right on either count.

 

Smart sensors that can detect, send the exact angle, and timing, are likely to be more than a few cents.

 

The quality of the experience has yet to be determined. People are experiencing problems with both systems, as expected with new tech.

[S3NTRY11]

Travis Deyle has a PhD in Electrical and Computer Engineering

 

Great stuff, thanks for sharing!

 

Someone just posted a great collection of Alan Yates' quotes. Take his analysis with a grain of salt, but the quotes are legit.

[vicx]

Two or more base station visibility lets you track a single sensor, which may be on a non-rigid object. (ref)

 

Aka a fingertip :)

[spacenavy90]

Aka a fingertip :)

 

I remember in an interview, they were saying finger tracking would be difficult but hopefully they can make it work!

[Chivas]

No big surprise your opinions disagree with the experts, but sixty sensors at a few cents each will make no difference to cost, but a big advantage for Vive.

 

This solution is elegant for a few reasons:

• 
  
• The computation overhead is minimal, especially compared to image processing.
  
• It's super-low latency. Unlike image tracking techniques, this system doesn't need to wait while data-intensive images are transmitted, processed, and analyzed. The microcontroller count can be quickly and accurately mapped to angles in (basically) one instruction cycle. This is super-critical when it comes to VR and AR, as angular errors or latency delays can create problematic artifacts that make them unusable.
  
• It relies on the high(ish) time resolution at the receiver to determine angles. This has major benefits over systems (like Raskar's, which I'll discuss later) that use a similar technique, but have limitations due to spatial resolution rather than time resolution.
  
• The receiver hardware is stupidly cheap: e.g., a $0.01 photodiode. It's also very small and light weight, so it could be included in almost any object.
  

http://www.hizook.com/blog/2015/05/17/valves-lighthouse-tracking-system-may-be-big-news-robotics

 

Travis Deyle has a PhD in Electrical and Computer Engineering

 

That's interesting, and you may have finally been right about something, even if its just the cost of the particular photodiode . Congratulations.

[Chivas]

I'm not so sure that placing sensors on all body parts is the best solution, in the long run. These tracking solution could end up being a combination of lasers, and cameras, or some other tech altogether.

[NineLine]

Keep it civil or I will have to shut it down.

[Chivas]

That would truly be extreme and in some cases inappropriate. A glove with a few cheap sensors would be great though.

 

Exactly why I thought a VR tracking system that would combine a Lighthouse type laser system with an Oculus type camera system. At some point VR camera tracking might not require markers at all, much like facial recognition cameras might work. Personally I'm not a big fan of gloves, and would rather just program my Hotas, but the VR space is moving quickly. We could see more than a few new tech ideas on that front next year.

[spacenavy90]

Exactly why I thought a VR tracking system that would combine a Lighthouse type laser system with an Oculus type camera system. At some point VR camera tracking might not require markers at all, much like facial recognition cameras might work. Personally I'm not a big fan of gloves, and would rather just program my Hotas, but the VR space is moving quickly. We could see more than a few new tech ideas on that front next year.

 

How will facial recognition work if half your face is covered by VR goggles. :P

[Chivas]

How will facial recognition work if half your face is covered by VR goggles. :P

 

I know your joking, but our bodies won't be covered by VR goggles. For facial recognition/expression problems, Oculus is working on solutions that sense facial muscle movement under the mask, etc etc.

 

One problem with current laser systems is that it would have to transmit positions wirelessly or wired from the different body parts. The problem is that cameras and laser can map body parts, but moving body parts are much more difficult.

[S3NTRY11]

I know your joking, but our bodies won't be covered by VR goggles. For facial recognition/expression problems, Oculus is working on solutions that sense facial muscle movement under the mask, etc etc.

 

One problem with current laser systems is that it would have to transmit positions wirelessly or wired from the different body parts. The problem is that cameras and laser can map body parts, but moving body parts are much more difficult.

IR tracking of discrete body parts is way off.

Fingers in Touch may work, but I don't think that can easily be extrapolated to the entire body, and there's no way that will be gen1 VR, in any case.

 

The bandwidth required for transmission of location data from Lighthouse controllers would be laughably small.

 

Alan Yates has stated that non-rigid-body tracking can be achieved by sensors with line of sight to more than one base station, which is promising.

 

I imagine gloves will be the only other way (aside from wands or equivalent) for the foreseeable future when it comes to cockpit control in VR, if you don't want to use your HOTAS for everything.

Edited December 17, 2015 by S3NTRY11
Typo

[Chivas]

IR tracking of discrete body parts is way off.

Fingers in Touch may work, but I don't think that can easily be extrapolated to the entire body, and there's no way that will be gen1 VR, in any case.

 

The bandwidth required for transmission of location data from Lighthouse controllers would be laughably small.

 

Alan Yates has stated that non-rigid-body tracking can be achieved by sensors with line of sight to more than one base station, which is promising.

 

I imagine gloves will be the only other way (aside from wands or equivalent) for the foreseeable future when it comes to cockpit control in VR, if you don't want to use your HOTAS for everything.

 

It should be obvious to anyone that this wouldn't possible in the first generation of VR, or in subsequent generations for more than a few years.