[Emirikol]

No body knows about the existence of a multi focal Head-Mounted Display? Thanks

[mboeller]

you mean holographic HUD's like the one in use in fighters, some business jets and in some commercial jets?

[Emirikol]

Usually VR HUDs have a fixed focus,
I mean a HUD with multi-focal optics. The observer's eyes can focus near objects and far objects freely as in reality

[epicstruggle]

I did a paper for one of my CS classes about the need for a multi focal optics in a HMD unit. I described a system that would accomplish this. In my search for related works, I did not come up with anything like what i proposed. I figure in the last 6 months since i did this paper, nothing will have changed. The main obstacle to a MFO in HMD is the lack of good miniturized lcd displays. Sony which many thought would revolutionazy the industry, has dropped their plans to produce more of their HMDs.

Emirikol, if you do find anything, please PM me. I would be very interested in hearing what you find out.

[Emirikol]

I have found something about Super Multi View Images but it seems to me another holographic picture system, ... simply unconfortable

[MfA]

Originally Posted by Emirikol

Usually VR HUDs have a fixed focus,
I mean a HUD with multi-focal optics. The observer's eyes can focus near objects and far objects freely as in reality

As far focusing is concerned there might be one degree of far, but there is a continuous spectrum of near of course ;) Having a mix of multiple focus points might make some simulations a little more realistic, but isnt what we really need variable focus optics? (AFAICS with variable focus a single focal point would be sufficient together with a normal stereoscopic setup too, I doubt it makes a big difference wether something we are not looking at is truly out of focus or merely is made to seem so by the stereoscopic effect and some blurring.)

Marco

PS. I was assuming with multi focal you mean the system mixes images at different fixed focal points (presumably with a projection system in the spirit of "Pepper's Ghost").

[Emirikol]

Marco, the goal is to simulate reality the best we can.
Studies have dimostrated that the fixed focus is the main cause of head and eyes fatigue in wearing a VR o AR HeadMountedDisplay.

In real life eyes change focus tens times per minutes while in VR eyes are hold in almost the same position and at the same focus lenght.

The 3D effect in the human vision is dued to:

1. Stereoscopy
2. Perspective
3. Focusing

In HMD usually we see a stereoscopic perspective correct 3D image focused as if it was a 2-3m distant ahead wallpaper. This is not natural.

To produce an image with a progressive focus from infinite to 30cm is a primary effort of VR designers.

[mboeller]

Originally Posted by Emirikol

Marco, the goal is to simulate reality the best we can.
Studies have dimostrated that the fixed focus is the main cause of head and eyes fatigue in wearing a VR o AR HeadMountedDisplay.

In real life eyes change focus tens times per minutes while in VR eyes are hold in almost the same position and at the same focus lenght.

The 3D effect in the human vision is dued to:

1. Stereoscopy
2. Perspective
3. Focusing

In HMD usually we see a stereoscopic perspective correct 3D image focused as if it was a 2-3m distant ahead wallpaper. This is not natural.

To produce an image with a progressive focus from infinite to 30cm is a primary effort of VR designers.

IMHO multifocus HMD is not the answer.
IMO you need an eye-tracker so the system knows where the person looks (updated at >50Hz). This portion of the screen is updated and focused then so everything looks perfect.
The other parts of the screen should look slightly blurry at the same time to enhance the "feeling" to look.
It should be possible to use an low-energy laser to track the eye-movement (IMHO this has been done already, but I'm not 100% sure).

[Basic]

mboller:
That would make the out of focus blur look real, but it wouldn't remove the eye strain from conflicting focusing vs stereoscopy. Because you would still have an incorrect focus.

If the optics are made so the image is in perfect focus no matter how the eye is focused and combine that with what you said, then there's a chance that the eye strain is removed.
But it might not be so easy to construct such optics, even if it's at least theoretically possible. (By focusing all light through a point in the center of the pupil, making a "virtual pinhole camera". )

It might be easier to make optics that can change the focal plane physically with electrical motors, combine it with syntetic bluring for off-focus parts, and control both with an eye tracker.

But my favourite idea is to have optics for the "virtual pinhole camera", use multiple light sources to move the pinhole position over the pupil, and find a LCD display with a few hundred Hz refresh rate (ouch ). With such a system you could render one frame from each position of the pinhole, and send it to the eye to the right pinhole position.
The result is kind of like T-buffer (or any other acc-buffer) depth of field, but the mixing will be done in the eye according to how you're actually focusing. Pretty much the real deal.


But to answer your question Emirikol:
No I don't know of any such HMD.

[Dave B(TotalVR)]

Ive been harking on (though not on forums) about using LCD displays to produce a holographic diffraction grating for ages. Then all you need to do is shine a laser through the back of the LCD and you have a full 3D image that you can focus on the front and back of perfectly. Not only would this absolutely rule in terms of game immersion it would also not really be much more expensive - if any more expensive at all - than current flat panel displays. All you need is an LCD screen capable of refreshing at 120hz for 60hz effective refresh, or 225hz for 75hz effective as you need to do each colour channel separately. Switching the laser at that speed is no problem.

Imagine the pink demons in Doom 3 actually runnig right up to you face before you knock them a couple of metres back with your dual barrels of justice...

[Basic]

That would need an *ahem* rather high res LCD.

[MfA]

Emirikol : ok, so I was assuming wrongly (apparently, you say little explicitly, since you now say progressive focus I assume you meant that all along). Multi suggests to me multiple but discrete, so I thought you meant something else ...

Marco

PS. <A HREF=http://www.google.it/search?hl=it&ie=UTF-8&oe=UTF-8&q=%22Device+and+Method+for+Viewing+Three-dimensional+Computer+Graphic+Displays%22&lr=>a little googeling</A> turned up they did a research project on a MEMS based deformible lens for a variable focus HMD (a system with motors and normal lenses is probably impractical, at least that assumption gave me the query which found the link :) at Berkeley, but I cant find any related publications ... so it might have been found to be utterly infeasible, at least at the time. Or maybe the publications are still pending review, if you are really interested you could always mail mr. Banks.

PPS. couldn't virtual retinal displays "easily" emulate variable focus BTW?

[SA]

Variably focused HMD displays is still an unsolved problem. As was mentioned, it is the leading cause of eye fatigue, headaches, etc. when using HMDs, and probably a major reason they have not become more popular even for professional uses.

Direct retinal scanned HMDs (where a multicolor laser is scanned directly into the eye, see http://www.mvis.com) and eye tracking with conventional HMDs offer the most promising avenues to solve this problem.

[Emirikol]

I have studied the laser scanner display but the only way to obtain variable focus is to change the focus by modulate optics for EVERY pixel during the scanning process....But This meccanism should operate at about 1024x768x30=23MHz!!!, We must exclude sliding lens...MEMS system that deformate a lens may, perhaps, be a possibile solution, I dont know.
Another possibility is to perturbate the angle of the scanned laser beam to simulate, on a pixel basis, the optics' convergence or divergence, but the needed accuracy is very high and I doubt a today MEMS chip can be able to do that.

[MfA]

Wouldnt it be easier to make the laser project the right colour for the spot on the retina the scanner is aiming at at a given time, instead of trying to aim the original pixels at the right spots?

[Nagorak]

Ok if eye fatigue is such a big deal, can someone please explain why there is no such fatigue when staring at a computer monitor (with the exception of one that's running at crap refresh rates)? I mean I know supposedly there is fatigue from looking at a computer monitor, but I've never experienced it and I spend hours in front of mine, much of which is spent reading and writing text, etc.

So that being the case couldn't your eyes just "get used" to VR displays, and if so isn't all this nonsense about making it more realistic on the eyes, just that: nonsense?

[Captain Chickenpants]

Partially because your screen is further away than VR headset would be, and your eye has to work harder to focus on things close up.

Partially because even without thinking about it you will be looking around
at other things.

It is actually recommended to take a short break every so often (about 50 mins I beleive), and look at things at a different distance so as to relax the eyes. (At least thats my excuse and I am sticking to it).

[SA]

There are two primary problems with HMDs.

The first has to do with the problem mentioned, lack of focal variation. The brain uses depth cues (primarily from the stereo disparity) to decide how far away each image point is. It then uses this distance information to automatically focus the eye. As the eye focuses correctly, it correlates the resultant newly focused image as a feedback mechanism with the depth input data to know when the eye is properly in focus. Automatic focus cameras do something similar. The problem with HMDs is that the focused image never matches what the brain expects from the stereo disparity thereby creating the headaches, etc. Stereo glasses used with a monitor cause the same type of problem.

The second problem is head tracking latency and accuracy. This usually causes motion sickness. The brain is very sensitive to slight differences between the motion the inner ear senses and the motion the eyes see. Head tracking latency and accuracy is much easier to solve, and many high end professional HMDs have solved this problem.

[SA]

The simplest solution is to use a combination of eye tracking and pupil size tracking from sensors as input and process the 3d image using the proper depth of field calculations for focusing at the given 3d image point with the given pupil size. To further reduce the disparity between what the brain expects and what it sees, including proper bright light processing such as glare, etc. would also be of some benefit.

[Dio]

I'm one of the ones who has a trouble with conventional HMD's - I tend to get headaches after 5-10 minutes. I find it difficult to cope properly with the different 'stereoscopic focus' vs. 'image focus' - my eyes have a tendency to look in the wrong places.

I found this while experimenting with a HMD about ten years ago now. I suspect I would see an improvement if I got used to it - since I used to have similar problems for the first half-hour or so after I changed from my glasses to my contact lenses, but nowadays there's no switch time or headaches at all.

[Dio]

A second note: I find stereoscopic glasses (w/projector or monitor) approaches don't create the same problem for me, probably because of the aforementioned thing that 'far is far, but there is a lot of near'.

[Emirikol]

Originally Posted by SA

There are two primary problems with HMDs.

The first has to do with the problem mentioned, lack of focal variation. The brain uses depth cues (primarily from the stereo disparity) to decide how far away each image point is. It then uses this distance information to automatically focus the eye. As the eye focuses correctly, it correlates the resultant newly focused image as a feedback mechanism with the depth input data to know when the eye is properly in focus. Automatic focus cameras do something similar. The problem with HMDs is that the focused image never matches what the brain expects from the stereo disparity thereby creating the headaches, etc. Stereo glasses used with a monitor cause the same type of problem.

You are right SA.

But I must disagree with your second post.
The observer pupil size is not correlated to focus depth.
To change the deep of field of the image doesn't improve the vision quality. Because the image is always focused as if it was a screen XX cm away from the observer eyes. When the eye see an object apparently near (sensation given by the stereoscopic effect) the brain instantly tries a "near" focus. When the eye see an apparently far object, the brain tries to focus the eye to infinity....but in an usual HMD the image is always at the same focus, so changing the observed object forces the brain to try a wrong focus with conseguent fatigue

[mboeller]

Originally Posted by Emirikol

But I must disagree with your second post.
The observer pupil size is not correlated to focus depth.
To change the deep of field of the image doesn't improve the vision quality. Because the image is always focused as if it was a screen XX cm away from the observer eyes. When the eye see an object apparently near (sensation given by the stereoscopic effect) the brain instantly tries a "near" focus. When the eye see an apparently far object, the brain tries to focus the eye to infinity....but in an usual HMD the image is always at the same focus, so changing the observed object forces the brain to try a wrong focus with conseguent fatigue

IMHO it should be possible to track the thickness of the pupil with an laser and use this info to focus the picture in the right distance using an holographic display (or retina display).

[MfA]

I think by the time we have a holographic display we will also have the computational power to use it for full 3D ... just using it to put a 2D image with artificial DOF effects at the correct focus depth seems a bit of a waste :)

[mboeller]

Originally Posted by MfA

I think by the time we have a holographic display we will also have the computational power to use it for full 3D ... just using it to put a 2D image with artificial DOF effects at the correct focus depth seems a bit of a waste

Sorry;

I ment holographic HUD's like the one in use in fighters like the new F16 models or more modern fighters. I didn't ment real holographic displays.
Holographic HUD's project the image so that it appears to be in infinite distance ( or a few meters away, depending on the requirement ).