2 other ways for making VR headset

This reminds me of castAR a little bit. Very cool and well done I like the presentation alot very professional looking and some neat ideas there too!

thank you baloniman :)

Thank you also to Oculus team to gave me the possibility to try VR at home. It is a old dream I have since I am kid :)
It is far from perfect but comparatively to DK1, it is playable.

There are so many show stoppers in this thing, I don't even know where to begin. Okay, let's start with the "high-speed mirrors". In the research I'm doing (ultra high speed OCT imaging of the eye's backside, through the pupil) we're using about the fastest galvano mirror scanners you can get; the limit is set by the mass of the scanner mirrors themself. And we can hardly get it beyound some 320 lines at 25 frames per second. And to get the scanners that fast you require really bulky amplifiers to drive the solenoids. Oh, and the things get quite warm, too.

And they are really, really uncomfortably loud. Ahhh, the noise... Did I mention the noise? Imagine having a high pitched tone emanating from a transducer at seriously high pressure levels? We're talking about 70dB(A) when being close to the scanners. When we operate out high speed scanner everybody wears hearing protection.

The next problem I see is your supposed mirror. Unless you're incorporating a really complex holographic optical system that diffracts the light appropriately, you'll have a hard time making that change in propagation direction. Unfortunately that works only for a specific combination of pupul and laser aperture position. Move the pupl and you have to move accordingly the emitter. For reflections angle of incidence == angle of exit.

"datenwolf" wrote:
There are so many show stoppers in this thing, I don't even know where to begin. Okay, let's start with the "high-speed mirrors". In the research I'm doing (ultra high speed OCT imaging of the eye's backside, through the pupil) we're using about the fastest galvano mirror scanners you can get; the limit is set by the mass of the scanner mirrors themself. And we can hardly get it beyound some 320 lines at 25 frames per second. And to get the scanners that fast you require really bulky amplifiers to drive the solenoids. Oh, and the things get quite warm, too.

I get the sense that he's talking about a mirror array, similar to something like a DLP chip. Each mirror is only responsible for a single pixel, so it can switch a much lower rate than a mirror that's responsible for the whole screen.

it would take a while for this to implement

this is after people know about VR

as of today VR is so complicated normal users would just stay away

maybe in 20 years would it be common to do that

Hopefully as we already at a feasible working point on retinal projection eg avant glyph. That clocking up to 2k resolution in each eye isn't far off along with increased FOV using a single screen will be a small stepping stone to great VR picture quality via retinal projection. We just have to be able to render this long term goal we all want for picture quality.

Still wonder on long term retina health using this method but they say it's ok.

** Note : I've yet to physical try out a glyph through numerous reading and reviews the tech sounds promising**

More I use the DK2 it's not a great comfortable experience as I'm always adjusting the headset to get comfortable.

"13thFloor" wrote:
Hopefully as we already at a feasible working point on retinal projection eg avant glyph.

Wow, virtual retina display… Sounds impressive. The problem is, it is not. In the end it boils down to imaging a small rectangular imager area to some rectangular area on the eye's retina. This essentially the very same as the Rift does.

But the Avant Glyph has the added challenge of what's called the entrance and exit aperture. You probably know the effect from looking through a telescope or a microscope: As long as your eye's pupil doesn't align up with a certain sweet spot in the optics your FOV is extremely limited. To overcome this the optics requires really, really large lenses, to increase the so called "Numerical Aperture" as far as possible. To put this into perspective: The eye interface lens in our OCT scanning system has an diameter of 3", and we have it easy, because we're using a telecentric optics. If we had to magnify the angular range of the scanners, we'd have to introduce an intermediary image lens of much larger diameter.

For the Avant Glyph and its requirement for compact optics this means that you either

- can not roll your eyes without loosing sight of the imager.

or

- that the interface lens must get very close to the eye (almost touch the cornea).

or

- largely reduce the FOV to trade this with exit aperture.

Either is not optimal for VR applications.

"13thFloor" wrote:

Still wonder on long term retina health using this method but they say it's ok.

It's okay, because effectively you're just looking at a (smallish) screen through magnifying glasses. As long as the screen is not illuminated with more than 1mW of power total this is perfectly safe. It would be a different kind of story, if that was a scanning spot system, like our OCT scanners.

Nice post datenwolf. I see you been reading up loads more into this.

Human nature is to always move eyes when looking around so that problem has to be solved I guess. As with Oculus approach and everyone debating screen res and GPU required retinal projectors are also in their infancy so great strides will be made in both.

"13thFloor" wrote:
Nice post datenwolf. I see you been reading up loads more into this.

Not so much reading but practical experience. A few years ago I actually tried implementing that VR glasses concept from the thread starter; albeit in very simple form. At some point I just abandoned it because of the optical problems: Even if you manage to make a well working holographic optical system, it only works at certain temperature or humidity, because as soon as that shifts, the holographic patterns skew and everything gets off. Frustrating. Oh, and it works well only monochromatically.

When it comes to eye interfacing optics. Well, that's part of my day job as a researcher. We're doing stuff nobody else has done before (ultra high speed 3D widefield OCT imaging of the human eye) and one of my PhD student colleagues spent the better part of the last year with designing a 120° FOV interface optics. Looks scary to use though.