It is currently Mon Mar 25, 2019 4:37 pm



Reply to topic  [ 50 posts ]  Go to page 1, 2  Next
 what's DK2's "real" resolution (how much pixels discarded) 
Author Message
One Eyed Hopeful

Joined: Tue Jul 15, 2014 10:38 am
Posts: 6
Reply with quote
Hey guys,
I'm wondering what's the real resolution for DK2, I read an article that calculated for DK1, 73% of the actual screen space is used, and all other pixels are wasted.
http://www.previewlabs.com/would-a-4k-r ... -possible/

I'm not sure for DK2, what's the percentage of pixels discarded, would that number be lower because the design is different (I believe less light is blocked)?

The following post also raised the same question, but I found no definitive answer http://www.reddit.com/r/oculus/comments ... t_hint_it/

Thanks!


Tue Jul 15, 2014 10:52 am
Profile
Golden Eyed Wiseman! (or woman!)

Joined: Fri Jun 08, 2012 8:18 pm
Posts: 1329
Reply with quote
I don't think anyone received their preorder yet, so it's hard to say. However, the DK2 has a smaller screen than the DK1 AFAIK, so assuming screen size was the "issue" before, the DK2 should make better use of the available resolution.


Tue Jul 15, 2014 5:03 pm
Profile
One Eyed Hopeful
User avatar

Joined: Tue Jul 01, 2014 7:29 am
Posts: 34
Reply with quote
I ran my own numbers, and this is what I came up with... about 68% viewable, and 32% "discarded".

Image

Please correct me if I'm wrong.


You do not have the required permissions to view the files attached to this post.


Wed Jul 16, 2014 12:23 pm
Profile
Petrif-Eyed
User avatar

Joined: Sat Jan 09, 2010 2:06 pm
Posts: 2255
Location: Perpignan, France
Reply with quote
As I've posted on this blog and on the Oculus forums, the amount of viewable pixels depends on the IPD and the eye relief of the viewer. There are numerous examples of captures from Rift demos where there are no lost pixels at all or only a very small amount.

Examples

From VR Sickness, The Rift, and How Game Developers Can Help :

Image

From SDK v0.3.2 Preview 2 Discussion :

Image

Illustration of IPD influence over the viewable zone (with 2mm eye relief)

63.5mm IPD (equal to the separation between the lenses) :

Image

70mm IPD :

Image

75mm IPD :

Image

Illustration of eye relief influence over the viewable zone (for a 63.5mm IPD)

4mm eye relief :

Image

10mm eye relief :

Image


Wed Jul 16, 2014 12:45 pm
Profile WWW
One Eyed Hopeful
User avatar

Joined: Tue Jul 01, 2014 7:29 am
Posts: 34
Reply with quote
Hi Fred. I think we're talking about two different things. The question was how much of the screen was being used to display image data -- let's say it's about 70%. If I understand you correctly, what you're saying is, of that 70%, very little is being lost. I don't think anyone argues with that.


Wed Jul 16, 2014 1:30 pm
Profile
Petrif-Eyed
User avatar

Joined: Sat Jan 09, 2010 2:06 pm
Posts: 2255
Location: Perpignan, France
Reply with quote
No, what I mean is that 70% of viewable pixels is only a particular situation for a specific IPD and eye relief. What I mean is for horizontal pixels, not the total viewable area in horizontal and vertical dimensions because nothing can be done about this because of the way the optics works.

I think it's most probably the case for people having a high IPD (more than 65mm) and/or a quite prominent forehead. For the higher percentile of people with a smaller IPD (around 65mm) and a less prominent forehead there are no pixels lost in the horizontal dimension.


Wed Jul 16, 2014 2:03 pm
Profile WWW
One Eyed Hopeful
User avatar

Joined: Tue Jul 01, 2014 7:29 am
Posts: 34
Reply with quote
Hi Fred. I understand what you mean. I still think we're discussing apples and oranges, but that's cool. :D


Wed Jul 16, 2014 2:26 pm
Profile
Golden Eyed Wiseman! (or woman!)

Joined: Fri Jun 08, 2012 8:18 pm
Posts: 1329
Reply with quote
IanBruce,

Are you talking about the dark area along the edges, or the blackened areas in the corners? The corners can't be used without adversely affecting FoV.


Wed Jul 16, 2014 3:41 pm
Profile
One Eyed Hopeful
User avatar

Joined: Tue Jul 01, 2014 7:29 am
Posts: 34
Reply with quote
I'm talking about the areas on the actual, physical OLED display which aren't addressed because they're unnecessary.

Image


Wed Jul 16, 2014 4:59 pm
Profile
Golden Eyed Wiseman! (or woman!)

Joined: Fri Jun 08, 2012 8:18 pm
Posts: 1329
Reply with quote
FredZ's posts address the area within the red boxes, at least the left and right edges. The top and bottom area I believe are due to the optics, their distance from the user's eyes, and the screen size. If the screen size is smaller in the DK2 and all other things being equal to the DK1, that should reduce the unused area along the edges. It's the corner areas just outside where the image is warped that inherently cannot be used without reducing the FoV (or perhaps displaying a weird border).


You do not have the required permissions to view the files attached to this post.


Wed Jul 16, 2014 8:36 pm
Profile
One Eyed Hopeful
User avatar

Joined: Tue Jul 01, 2014 7:29 am
Posts: 34
Reply with quote
Thanks MSat. I agree... Oculus is striking a balance between eye relief, focal length, IPD, available screen sizes, and the decision to use a single display for both eyes.


Wed Jul 16, 2014 10:37 pm
Profile
Petrif-Eyed
User avatar

Joined: Sat Sep 17, 2011 9:23 pm
Posts: 2218
Location: Menlo Park, CA
Reply with quote
The DK2 rendering looks quite a bit different than the DK1 because the lens center lies close to the 1/4 and 3/4 points of the screen. So the render circles are near-symmetrical and don't overlap as much in the center. The outside edge is not wasted the same way that it is in DK1 and the rendering "fills" the screen to the edges better. I'm sure there are pictures from E3 around that convey what I am describing.


Wed Jul 16, 2014 11:17 pm
Profile
One Eyed Hopeful
User avatar

Joined: Tue Jul 01, 2014 7:29 am
Posts: 34
Reply with quote
Thanks Brentlew. I guess I'll have to wait to see what you mean. Anyone who's worked with these systems knows that the chief determining factors for display rendering are imposed by the limits of spherical and chromatic aberration -- particularly in single-lens systems like the Rift. Oculus has some of the best people working on this, but there's only so much finagling you can do before you run up against physical laws that have been with us since the Big Bang. ;)


Thu Jul 17, 2014 6:24 am
Profile
Petrif-Eyed
User avatar

Joined: Sat Sep 17, 2011 9:23 pm
Posts: 2218
Location: Menlo Park, CA
Reply with quote
Yes, chromatic aberration is stronger in the DK2 lens and requires more aggressive correction.


Thu Jul 17, 2014 10:04 am
Profile
One Eyed Hopeful
User avatar

Joined: Tue Jul 01, 2014 7:29 am
Posts: 34
Reply with quote
Hi Brantlew. I have to assume they're using an achromat doublet to fix (or at least reduce) the abberation. It also occurs to me that there may be a solution in the rendering... manipulating the chromasity at the outer limits of the displayed image to compensate for this effect. I haven't seen the the DK2 myself, so I really have no idea what they're doing. They're pushing the limits at 90-100° FOV.


Thu Jul 17, 2014 11:56 am
Profile
Petrif-Eyed
User avatar

Joined: Sat Jan 09, 2010 2:06 pm
Posts: 2255
Location: Perpignan, France
Reply with quote
IanBruce wrote:
Anyone who's worked with these systems knows that the chief determining factors for display rendering are imposed by the limits of spherical and chromatic aberration -- particularly in single-lens systems like the Rift.
There shouldn't be any spherical aberration with the Rift since they're using aspheric lenses.

IanBruce wrote:
I have to assume they're using an achromat doublet to fix (or at least reduce) the abberation.

Chromatic aberration correction is done in software.

Btw brantlew is part of the Oculus VR team if you didn't know.

IanBruce wrote:
I haven't seen the the DK2 myself, so I really have no idea what they're doing. They're pushing the limits at 90-100° FOV.

The DK1 already has more FOV than that for some people. The author of the Oculus SDK has a horizontal FOV of 109° for his left eye. If his right eye is similar that would give him a 117.4° horizontal binocular FOV.


Thu Jul 17, 2014 2:24 pm
Profile WWW
One Eyed Hopeful
User avatar

Joined: Tue Jul 01, 2014 7:29 am
Posts: 34
Reply with quote
Fredz wrote:
There shouldn't be any spherical aberration with the Rift since they're using aspheric lenses.

Well, yes... One corrects for spherical aberration by using aspheric lenses.

Fredz wrote:
Chromatic aberration correction is done in software.

Thank you. I thought that might be the case, but it wasn't evident in the raw output I've seen.

Fredz wrote:
brantlew is part of the Oculus VR team if you didn't know.

I did not. Thanks for the tip.

Fredz wrote:
The author of the Oculus SDK has a horizontal FOV of 109° for his left eye. If his right eye is similar that would give him a 117.4° horizontal binocular FOV.

I was quoting Oculus's own published numbers. Have these changed?


Thu Jul 17, 2014 11:06 pm
Profile
Petrif-Eyed
User avatar

Joined: Sat Jan 09, 2010 2:06 pm
Posts: 2255
Location: Perpignan, France
Reply with quote
I think Oculus has been quite conservative on the FOV values on their product page to not oversell their HMD. Not anybody can experience such a wide FOV with the Rift though, if you wear glasses, have a high IPD or a big forehead the FOV should be less. And the FOV in the consumer version may be quite different, so it makes sense to stay a bit vague about this before it's launched.

But still, the values I gave are in the official SDK, so not really hidden. At least not to developers.


Fri Jul 18, 2014 11:47 am
Profile WWW
One Eyed Hopeful
User avatar

Joined: Tue Jul 01, 2014 7:29 am
Posts: 34
Reply with quote
I think we can all agree that anything beyond 90° FOV is enough to get you past the window or "looking glass" effect -- inside looking out, vs. outside looking is. Delivering that for everyone was -- I would assume -- their primary goal. If so, mission accomplished. A few degrees here or there shouldn't diminish the immersive experience -- one that is far and away superior to anything else in the market.


Fri Jul 18, 2014 12:43 pm
Profile
Petrif-Eyed
User avatar

Joined: Sat Jan 09, 2010 2:06 pm
Posts: 2255
Location: Perpignan, France
Reply with quote
Yes, according to Michael Abrash in a recent talk, presence even starts to work somewhere around 80° and improves significantly at least out to 110 degrees (the widest he has tested).

Also, Eric Howlett who created the LEEP lenses for the first commercial HMDs said that presence could work with a 70° FOV but that the wider the field of view, the greater the illusion of presence was, up to the 270° mark. He also said that to provide an immersive VR experience, HMDs should show as much of the 270° field of view as possible.

So we are already there as you said, but there is still room for improvement.


Fri Jul 18, 2014 7:16 pm
Profile WWW
One Eyed Hopeful
User avatar

Joined: Tue Jul 01, 2014 7:29 am
Posts: 34
Reply with quote
I actually knew Eric and was at his lab in Massachusetts several times. I still have a couple of his demo LEEP optical sets in storage. My own optical system was created to improve upon what he started.


Fri Jul 18, 2014 8:54 pm
Profile
Petrif-Eyed
User avatar

Joined: Sat Jan 09, 2010 2:06 pm
Posts: 2255
Location: Perpignan, France
Reply with quote
Ah nice, can you tell us a bit more about your optical system ?


Fri Jul 18, 2014 11:18 pm
Profile WWW
One Eyed Hopeful
User avatar

Joined: Tue Jul 01, 2014 7:29 am
Posts: 34
Reply with quote
Sure. You can Google "Ian Bruce Lockheed patent" (no quotes) and you'll find what I'm referring to. Lockheed Martin just had a bunch of their claims on their Gen 2 and 3 helmet mounted display system for the F-35 denied because of my earlier patent -- hence the Lockheed reference.

In practice, a personal 160°h x 115°v FOV system would look something like this:

Image


Fri Jul 18, 2014 11:54 pm
Profile
Cross Eyed!

Joined: Sun Jun 02, 2013 4:21 pm
Posts: 175
Reply with quote
Are there 1.8" OLEDs with that higher a resolution to warrant the level of detail needed ?


Sat Jul 19, 2014 3:07 am
Profile
Petrif-Eyed
User avatar

Joined: Sat Jan 09, 2010 2:06 pm
Posts: 2255
Location: Perpignan, France
Reply with quote
Oho, very nice ! I need to read that patent.

Do you think a wide FOV could also be possible with the same optical system but with microdisplays (0.7-0.9" diagonal) ?.

Can you share a bit more about your background also if it's possible ?


Sat Jul 19, 2014 8:10 am
Profile WWW
One Eyed Hopeful
User avatar

Joined: Tue Jul 01, 2014 7:29 am
Posts: 34
Reply with quote
Attreyu: Yes, certainly. OLED displays even smaller than 1.8" are being produced in quantity at WUXGA resolutions (1920 x 1200 per eye). Even higher resolutions are available now, or on the immediate horizon. By comparison, WUXGA would be just over 3 times the resolution of the DK2.


Sat Jul 19, 2014 8:28 am
Profile
Cross Eyed!

Joined: Sun Jun 02, 2013 4:21 pm
Posts: 175
Reply with quote
IanBruce wrote:
Attreyu: Yes, certainly. OLED displays even smaller than 1.8" are being produced in quantity at WUXGA resolutions (1920 x 1200 per eye). Even higher resolutions are available now, or on the immediate horizon. By comparison, WUXGA would be just over 3 times the resolution of the DK2.


Care to share a link providing a model/type/producer ?

Because on www.panelook.com there are none.


Sat Jul 19, 2014 9:17 am
Profile
One Eyed Hopeful
User avatar

Joined: Tue Jul 01, 2014 7:29 am
Posts: 34
Reply with quote
Hi Fredz:

Yes, but display size has a direct impact on FOV. A 0.9" display would only yield around 80 to 90° h. The form factor would be about the same, so there's no real advantage to using smaller displays.

Image

I'd previously worked with Apple's Advanced Technology Group as part of the Worldboard team. Worldboard was a specification for geo-coding the internet for use with with Augmented Reality devices on a global scale. My responsibilities were system user interface, application concept development, and the AR hardware device itself. This was in 1999, and needless to say, the program was way too far ahead of its time. http://service-science.info/archives/2060

The only really tangible outcome of this effort was my patent for near-eye, wide-field stereoscopic displays. The chief advantage of this system is its very small size. BTW: the proof-of-concept models were built for me by United Technologies Adaptive Optics division in Cambridge -- the same guys that built the corrective optics fix for the Hubble Space Telescope.


Sat Jul 19, 2014 9:19 am
Profile
One Eyed Hopeful
User avatar

Joined: Tue Jul 01, 2014 7:29 am
Posts: 34
Reply with quote
Hi Attreyu: Try this: http://www.emagin.com/oled-microdisplays/


Sat Jul 19, 2014 9:27 am
Profile
Petrif-Eyed
User avatar

Joined: Sat Jan 09, 2010 2:06 pm
Posts: 2255
Location: Perpignan, France
Reply with quote
Thanks for the information. I read the patent and it seems it expired in 2005, is that true ?


Sat Jul 19, 2014 4:11 pm
Profile WWW
One Eyed Hopeful
User avatar

Joined: Tue Jul 01, 2014 7:29 am
Posts: 34
Reply with quote
Correct. It wasn't necessary. Considerable improvements have been made to the design since, warranting a new set of more comprehensive applications. These improvements include disclosure of a see-thru-the-lens (VR/AR) system leveraging some of the same general principals, but relying on enabling technologies that either weren't available, or practical until quite recently. Temple University in Philadelphia is providing me some technical assistance with this effort.


Sat Jul 19, 2014 5:06 pm
Profile
Golden Eyed Wiseman! (or woman!)

Joined: Fri Jun 08, 2012 8:18 pm
Posts: 1329
Reply with quote
IanBruce, this is great stuff!

I'm intrigued by the fiber optic collimator. Seems like it would be a tricky part to manufacture. I wonder if it could be substituted with something like a fresnel lens comprised of elements the size of each pixel, possibly made via photoetching/photolithography.


Sat Jul 19, 2014 9:09 pm
Profile
One Eyed Hopeful
User avatar

Joined: Sat Nov 03, 2012 11:41 pm
Posts: 41
Reply with quote
I was curious, so I took the tuscany DK2 screenshot and whited out the rendered part.
Image
I wrote a program with opencv to count the number of black pixels in an image and print the info out. https://gist.github.com/xef6/402118a27ace8247478f
The result for the whited out tuscany image was:
Code:
658628 black pixels out of 2073600 total (31.763%)


So it looks like the black border occupies about ~32% of the screen space.


Sun Jul 20, 2014 2:52 am
Profile
One Eyed Hopeful
User avatar

Joined: Tue Jul 01, 2014 7:29 am
Posts: 34
Reply with quote
MSat wrote:
IanBruce, this is great stuff! I'm intrigued by the fiber optic collimator. Seems like it would be a tricky part to manufacture. I wonder if it could be substituted with something like a fresnel lens comprised of elements the size of each pixel, possibly made via photoetching/photolithography.


Thanks MSat. It's actually not that difficult with the right equipment, and we've already identified several companies capable of producing this in volume to the new spec in several different optical materials. While validation of the collimator design can be done by 3D prototyping the transfer element in VeroClear (using the default support material as cladding), the results are far too coarse for a high-resolution display. The light channels needed are on the order of a few microns, and that's only possible using established means of production.

I can't judge the merits of your own suggestion -- only to say that you're probably better off avoiding any scheme requiring micron-scale positional and rotational alignment of several optical matrix elements. Ours does not. In fact, the final focus elements are user applied. This is important as we'll be supplying 3 sets of those elements for normal, near, and far focus. Fortunately, they are quite inexpensive to make -- remarkably so.


Sun Jul 20, 2014 4:38 am
Profile
One Eyed Hopeful
User avatar

Joined: Tue Jul 01, 2014 7:29 am
Posts: 34
Reply with quote
xef6 wrote:
I was curious, so I took the tuscany DK2 screenshot and whited out the rendered part... The result for the whited out tuscany image was 658628 black pixels out of 2073600 total (31.763%). So it looks like the black border occupies about ~32% of the screen space.


Yes -- This is nearly identical to my own analysis. That puts put the total addressed area at roughly 840 x 840 per eye, with the highest image concentration in the centre.


Sun Jul 20, 2014 4:49 am
Profile
Petrif-Eyed
User avatar

Joined: Sat Jan 09, 2010 2:06 pm
Posts: 2255
Location: Perpignan, France
Reply with quote
I'm curious, what are your IPD and eye relief xef6 and IanBruce ?


Sun Jul 20, 2014 6:55 am
Profile WWW
One Eyed Hopeful
User avatar

Joined: Tue Jul 01, 2014 7:29 am
Posts: 34
Reply with quote
Fredz wrote:
I'm curious, what are your IPD and eye relief xef6 and IanBruce ?


Eye relief should be set to about 1cm (no glasses). The image plane's foci is the eye's rotational axis, so your mileage may vary. IPD adjustability will conform to a 95th percentile standard, male/female: ages 12 to adult.


Sun Jul 20, 2014 8:22 am
Profile
Golden Eyed Wiseman! (or woman!)

Joined: Fri Jun 08, 2012 8:18 pm
Posts: 1329
Reply with quote
IanBruce wrote:
It's actually not that difficult with the right equipment, and we've already identified several companies capable of producing this in volume to the new spec in several different optical materials.
Very interesting. Does this end up being a relatively cheap part?

IanBruce wrote:
..you're probably better off avoiding any scheme requiring micron-scale positional and rotational alignment of several optical matrix elements. Ours does not.

Good point, and interesting that yours doesn't require it. I'm guessing each fiber optic element is much smaller than a pixel then..

I'm really curious to know more about your product, as it's looking very slick. Mind mentioning the company (perhaps via PM)?


Sun Jul 20, 2014 9:59 am
Profile
One Eyed Hopeful
User avatar

Joined: Tue Jul 01, 2014 7:29 am
Posts: 34
Reply with quote
MSat wrote:
Very interesting. Does this end up being a relatively cheap part?


Yes. Part of the expanded IP is weight and cost reduction in the production process. Eliminating field distortion is also important, but much less of a problem since the throw distances are so short ( ~1 to 8mm ).

MSat wrote:
I'm guessing each fiber optic element is much smaller than a pixel then.. I'm really curious to know more about your product, as it's looking very slick. Mind mentioning the company (perhaps via PM)?


You're correct... the transfer arrays are resolving the image at a sub-pixel level. Individual pixels are being resolved as averages of transfer cell groups -- even at resolutions greater than 1920 x 1200. This effectively eliminates stage 1 artifacts completely without compromising the image data.

We've been focused on the technology and IP, and less about the best way to bring this to market. We have attracted the attention of several companies -- both here and abroad -- but have made no commitments in that regard. I really can't tell you anymore at this time. Sorry.


Sun Jul 20, 2014 11:21 am
Profile
Golden Eyed Wiseman! (or woman!)

Joined: Fri Jun 08, 2012 8:18 pm
Posts: 1329
Reply with quote
I hope you have success in bringing this tech to market some way at a consumer-reasonable price. That form factor is fantastic!


Sun Jul 20, 2014 1:48 pm
Profile
Display posts from previous:  Sort by  
Reply to topic   [ 50 posts ]  Go to page 1, 2  Next

Who is online

Users browsing this forum: No registered users and 4 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Group
Designed by STSoftware.