New Rift prototype looks sweet!

Post Reply
MSat
Golden Eyed Wiseman! (or woman!)
Posts: 1329
Joined: Fri Jun 08, 2012 8:18 pm

New Rift prototype looks sweet!

Post by MSat »

And here I was thinking Oculus would just be sitting around since the release of DK2.

I'm psyched about the extended motion tracking and integrated audio. Also mentioned was the higher resolution, which if I was a betting man I'd say is courtesy of the Note4 panel.
User avatar
cybereality
3D Angel Eyes (Moderator)
Posts: 11406
Joined: Sat Apr 12, 2008 8:18 pm

Re: New Rift prototype looks sweet!

Post by cybereality »

Yep, it's pretty sweet.
budda
Sharp Eyed Eagle!
Posts: 388
Joined: Thu May 22, 2008 9:13 pm

Re: New Rift prototype looks sweet!

Post by budda »

Hi,

As the screen refresh rate of the display keeps going up, it only seems a matter of time before shutterglasses will be incorporated into the Oculus Rift.

Some Samsung TV's have 240 Hz displays working very well with shutterglasses.

Oculus uses Samsung displays and screen technology.

So it seems logical this is the direction they should go in.

As long as the shuttering does not detract from the latency and motion artifacts aspects, then its an easy way to double horizontal screen resolution and greatly increase field of view without the divider blocking the inside view.

This has all been said before. It will happen.

Thanks.
Endothermic
Binocular Vision CONFIRMED!
Posts: 284
Joined: Tue Jun 26, 2012 2:50 am

Re: New Rift prototype looks sweet!

Post by Endothermic »

I don't think it seems logical, you have to consider the design of the RIFT.

With shutter glasses on TVs you don't have optics having to focus the TV onto your eye and things like the CastAR using shutter glasses have a separate display for each eye making for simpler optics.

Unless they change the whole design of the RIFT from using a single large display to using two micro displays I don't see it working.

The optics involved with focusing the whole panel that size to an eye that close which is now significantly off center to the display wouldn't be simple or cheap (if it's even doable) which is what the design of the RIFT is.
User avatar
Dilip
Certif-Eyed!
Posts: 692
Joined: Sat Dec 13, 2008 9:23 am
Location: Ahmedabad//INDIA

Re: New Rift prototype looks sweet!

Post by Dilip »

Shuttering would be last thing i would ever want on any HMD. For god's sack, its horrible thing.....

1) Cuts brightness to Half.
2) Messes with display vibrancy.
3) Causes eye strain.

Even Sony is phasing out shutter glasses.Their Best 4K TVs are passive only...

http://www.sony.co.in/product/kd-84x9000
http://www.sony.co.in/product/kd-65x9004a

even though that means half of resolution due to inherent limitation of passive while done in interleaved mode. You could get soft interleaved but still there are lines. Only single company walked
on path of sandwich panel to offer full HD passive 3d that was IZ3D which is now no more.

One complaint from CRESCENT BAY, why headphones are left with such supper fragile covering and by look they resemble those disposables offered in most airlines inflight. They may sound good but why look/aesthetics totally ignored.
Endothermic wrote:Unless they change the whole design of the RIFT from using a single large display to using two micro displays I don't see it working.
Rift evolved on single panel concept as that's magical ingredient,which makes rift an affordable affair. This was made very clear from beginning of project by Palmer they have no intentions going dual screen due to sync as well as pricing limitation. So i don't foresee such thing in near future.
User avatar
PatimPatam
Binocular Vision CONFIRMED!
Posts: 214
Joined: Thu Jun 28, 2012 1:31 pm
Location: Barcelona

Re: New Rift prototype looks sweet!

Post by PatimPatam »

Yep I can confirm that it is pretty sweet indeed :-) I'll be more than happy if the CV1 HMD is as good as this!


I did try the Crescent Bay at Oculus Connect and was very impressed, especially with the tracking, very accurate and stable, and with greatly improved capture space. Also i was given the demo by Dov Katz himself (Computer Vision Engineer at Oculus), that was very cool, really nice guy. I would have liked to be able to speak with him for a bit longer but i didn't want to bog down the line of people waiting.

Anyway very happy to see that Oculus have finally come around and added some LEDs at the back! I bet it wasn't easy to get it working without having a rigid HMD.. One important implication related to this that i haven't seen mentioned so far is that now the setup is going to become much simpler in comparison to the DK2, since we will no longer need to place the camera at the front; if you have a small room for instance you could simply place it at the side or at the back.


The only slight complain i have about the demo is that the distortion didn't seem quite right on some scenes, you could clearly notice wobbly vertical lines here and there. Also i think my technical mind prevented me of just letting go and completely enjoying the experience, i was constantly testing the tracking, checking the pixel structure, aliasing, smearing, etc


On a side note, it was great to finally meet Cyberality (although very briefly, i was stressed out trying to find a Wifi to reserve my demo spot), and WiredEarp, who is a really knowledgeable but down-to-earth chap.

And of course it was totally awesome being able to chat with Carmack for a few minutes (we talked about the problems of bringing positional tracking to the Gear VR) and also meeting Palmer, although unfortunately i picked a really bad time; just a few moments after we started talking his birthday "show" began, with everyone singing happy birthday, and then he went nuts with the Oculus cake and jumped to the swimming pool :-D Hahah, maybe another time :-P
User avatar
brantlew
Petrif-Eyed
Posts: 2221
Joined: Sat Sep 17, 2011 9:23 pm
Location: Menlo Park, CA

Re: New Rift prototype looks sweet!

Post by brantlew »

I'm sorry I missed you guys at the conference. I talked to Neil but didn't know that some of the other MTBS alumni were around.
User avatar
brantlew
Petrif-Eyed
Posts: 2221
Joined: Sat Sep 17, 2011 9:23 pm
Location: Menlo Park, CA

Re: New Rift prototype looks sweet!

Post by brantlew »

PatimPatam wrote:The only slight complain i have about the demo is that the distortion didn't seem quite right on some scenes, you could clearly notice wobbly vertical lines here and there. Also i think my technical mind prevented me of just letting go and completely enjoying the experience, i was constantly testing the tracking, checking the pixel structure, aliasing, smearing, etc
Ironically the distortion calibration is probably better on that device than any other headset we release. DK1 distortion correction (especially pre 1.3) was absolute crap but that low-res smear-fest basically just masked it. We have continued to improve the distortion process with each new hardware iteration but it simply becomes increasingly easier to see distortion with each new device. Every improvement in hardware (resolution and latency in particular) amplifies distortion inaccuracies and the limitations of the optics design. Also headset alignment becomes increasingly important - even 1 mm pupil vertical offset from lens center causes a swaying effect. And that's just the Y offset. X offset and Z distance also influence distortion. Its a hard hard problem and as was noted in the talks - optics and calibration continues to be one of our most difficult problem to solve.
User avatar
cybereality
3D Angel Eyes (Moderator)
Posts: 11406
Joined: Sat Apr 12, 2008 8:18 pm

Re: New Rift prototype looks sweet!

Post by cybereality »

Yeah, it was awesome meeting some forum veterans in person. Good times.
User avatar
PatimPatam
Binocular Vision CONFIRMED!
Posts: 214
Joined: Thu Jun 28, 2012 1:31 pm
Location: Barcelona

Re: New Rift prototype looks sweet!

Post by PatimPatam »

brantlew wrote:We have continued to improve the distortion process with each new hardware iteration but it simply becomes increasingly easier to see distortion with each new device. Every improvement in hardware (resolution and latency in particular) amplifies distortion inaccuracies and the limitations of the optics design. Also headset alignment becomes increasingly important - even 1 mm pupil vertical offset from lens center causes a swaying effect. And that's just the Y offset. X offset and Z distance also influence distortion. Its a hard hard problem and as was noted in the talks - optics and calibration continues to be one of our most difficult problem to solve.
Hmm interesting. The first thing that comes to mind is having an internal camera to track the exact pupils locations in relation to the headset (which could also be used for automatic IPD calibration and maybe even gaze tracking), but i understand that is easier said than done!
User avatar
brantlew
Petrif-Eyed
Posts: 2221
Joined: Sat Sep 17, 2011 9:23 pm
Location: Menlo Park, CA

Re: New Rift prototype looks sweet!

Post by brantlew »

PatimPatam wrote:
brantlew wrote:We have continued to improve the distortion process with each new hardware iteration but it simply becomes increasingly easier to see distortion with each new device. Every improvement in hardware (resolution and latency in particular) amplifies distortion inaccuracies and the limitations of the optics design. Also headset alignment becomes increasingly important - even 1 mm pupil vertical offset from lens center causes a swaying effect. And that's just the Y offset. X offset and Z distance also influence distortion. Its a hard hard problem and as was noted in the talks - optics and calibration continues to be one of our most difficult problem to solve.
Hmm interesting. The first thing that comes to mind is having an internal camera to track the exact pupils locations in relation to the headset (which could also be used for automatic IPD calibration and maybe even gaze tracking), but i understand that is easier said than done!
There is a misconception that lens distortion correction is a fixed function of the lens. Calculate the inverse and you're done. But distortion is also a function of where the pupil is located in 3D space behind the lens. It's a dynamic system that is different for each person's facial geometry and also importantly - where they are currently looking. Using collimated light helps a lot because it reduces this variance but there is still a great deal of it. Distortion correction is an interplay and compromise between optics and eye positions.

Good eye tracking and foveated rendering would be a revolutionary feature for a number of reasons - most often cited for the ability to generate radially decreasing render precision and lower bandwidth. But another big win would be dynamic foveated distortion - altering the distortion correction based on the pupil position. That would lead to a much more stable and "grounded" virtual world with the simple optics of the Rift.
User avatar
PatimPatam
Binocular Vision CONFIRMED!
Posts: 214
Joined: Thu Jun 28, 2012 1:31 pm
Location: Barcelona

Re: New Rift prototype looks sweet!

Post by PatimPatam »

brantlew wrote:But another big win would be dynamic foveated distortion - altering the distortion correction based on the pupil position. That would lead to a much more stable and "grounded" virtual world with the simple optics of the Rift.
Yep i didn't explain it as well as you did, but that's exactly what i was thinking about.

This wouldn't require accurately tracking saccades would it? I mean you could probably get by with lower camera refresh rates / higher latency tracking than what is needed for foveated rendering? Anyway it would be amazing if you guys can pull it off!
User avatar
brantlew
Petrif-Eyed
Posts: 2221
Joined: Sat Sep 17, 2011 9:23 pm
Location: Menlo Park, CA

Re: New Rift prototype looks sweet!

Post by brantlew »

Well saccades are interesting because you don't really fully process and "see" during the saccade. But head/eye counter-rotation is tough because the eye moves really freakin' fast while maintaining near perfect focus and conscious sight on a single point.
Post Reply

Return to “Oculus VR”