On 23/12/2013 09:23 AM, scott wrote:

> If it does take off then there are significant opportunities for further
> development, thinner, lighter headsets, higher resolution/framerate,
> wider FOV, wireless connection etc.

I spent all of last night reading about holographic displays.

[No, I do not mean volumetric displays. I mean a display where you 
compute the interference patterns of an optical wavefront, then use some 
kind of display device to produce that pattern.]

The big plus with this is that objects at different distances have 
different focal depths - exactly like they do in the real world. The 
machine basically generates the same light rays that the real object 
would, so your eyes can respond to it the way they normally do already. 
No weird glasses required. Multiple viewers possible. Etc.

The big downside is... you need an *insane* pixel pitch. And then you 
need a small supercomputer if you want to draw this stuff in realtime.

I did see a demo of a commercial offering which is supposed to be coming 
to market soon - but it uses eye-tracking to slash the amount of 
processing power required. (I.e., it won't actually work in the real 
world and so will never be commercially viable.)

Post a reply to this message

> [No, I do not mean volumetric displays. I mean a display where you
> compute the interference patterns of an optical wavefront, then use some
> kind of display device to produce that pattern.]

A friend from university did this:

http://lightblueoptics.com/videos/holographic-laser-projection-technology/

Judging by the (lack of) recent updates it didn't really take off.

FWIW before this, his research project at university was to modify a CD 
burner to burn holograms onto a CD. I guess the "pixel pitch" on a CD is 
enough for it.

> I did see a demo of a commercial offering which is supposed to be coming
> to market soon - but it uses eye-tracking to slash the amount of
> processing power required. (I.e., it won't actually work in the real
> world and so will never be commercially viable.)

Yes, eye tracking always introduces further issues. One demo I saw, 
which I'm surprised hasn't taken off, is to use the video feed from a 
camera as an environment map texture. This demo was running on a laptop 
using the inbuilt webcam and the way shiny surfaces reflected your face 
and what was behind you in the room was very believable.

Post a reply to this message

On 23/12/2013 01:58 PM, scott wrote:
>> [No, I do not mean volumetric displays. I mean a display where you
>> compute the interference patterns of an optical wavefront, then use some
>> kind of display device to produce that pattern.]
>
> A friend from university did this:
>
> http://lightblueoptics.com/videos/holographic-laser-projection-technology/
>
> Judging by the (lack of) recent updates it didn't really take off.

Indeed.

I also found a website with software for computing interference 
patterns. They claim that if you take something like a 600 DPI laser 
printer and print the pattern onto a transparency, you can shine a laser 
pointer through it and get a very fuzzy, very grainy hologram. (Assuming 
your printer doesn't try to interpolate or otherwise alter the precisely 
computed pixel patterns!)

> FWIW before this, his research project at university was to modify a CD
> burner to burn holograms onto a CD. I guess the "pixel pitch" on a CD is
> enough for it.

CD? Perhaps not; CDs don't use visible light, they use infra-red, so the 
dot pitch might not be small enough. A DVD, on the other hand, uses a 
red laser, so it certainly ought to be able to do a red-light hologram 
without difficulty. (Provided you can convince the drive to put the dots 
where you want them!)

>> I did see a demo of a commercial offering which is supposed to be coming
>> to market soon - but it uses eye-tracking to slash the amount of
>> processing power required. (I.e., it won't actually work in the real
>> world and so will never be commercially viable.)
>
> Yes, eye tracking always introduces further issues. One demo I saw,
> which I'm surprised hasn't taken off, is to use the video feed from a
> camera as an environment map texture. This demo was running on a laptop
> using the inbuilt webcam and the way shiny surfaces reflected your face
> and what was behind you in the room was very believable.

When I was at university, we were told that several companies had 3D TV 
technology that was "nearly ready for market". That was ten years ago. 
(I suppose they were probably talking about the laughably primitive 
lenticular lens technology that you occasionally see in shops and stuff.)

Post a reply to this message

> I also found a website with software for computing interference
> patterns. They claim that if you take something like a 600 DPI laser
> printer and print the pattern onto a transparency, you can shine a laser
> pointer through it and get a very fuzzy, very grainy hologram.

Oooh that sounds interesting, might have to give that a go. I got one of 
those cheap $20 green laser pointers, if you focus it it's enough to 
light a match or melt your name into anything plastic :-)

> CD? Perhaps not; CDs don't use visible light, they use infra-red, so the
> dot pitch might not be small enough.

Well if a 600dpi print just about works, then by my rough estimations a 
CD should be about 20000 dpi. Maybe not perfect, but it works.

> A DVD, on the other hand, uses a
> red laser, so it certainly ought to be able to do a red-light hologram
> without difficulty. (Provided you can convince the drive to put the dots
> where you want them!)

You can't without modifying the drive. IIRC they had to bypass pretty 
much the whole signal path and just drive the laser directly (well not 
quite, but you get the idea).

> When I was at university, we were told that several companies had 3D TV
> technology that was "nearly ready for market". That was ten years ago.
> (I suppose they were probably talking about the laughably primitive
> lenticular lens technology that you occasionally see in shops and stuff.)

10 years ago the company I used to work for was already selling a laptop 
and mobile phone with a parallax barrier 3D screen (that could be 
switched between 2D and 3D mode). That technology has developed into the 
dispay in the Nintendo 3DS and LG Optimus 3D today. That type of 3D is 
fine for single-user applications (because no glasses are needed), but 
for multi-viewer it obviously doesn't work.

Post a reply to this message

Am 23.12.2013 15:14, schrieb Orchid Win7 v1:

> I also found a website with software for computing interference
> patterns. They claim that if you take something like a 600 DPI laser
> printer and print the pattern onto a transparency, you can shine a laser
> pointer through it and get a very fuzzy, very grainy hologram. (Assuming
> your printer doesn't try to interpolate or otherwise alter the precisely
> computed pixel patterns!)

There are even people out there who, using nothing but a compass with a 
sharp tip, cut patterns into acrylic glass to produce simple 
stereographic images (point clouds actually).

And yes, it appears to actually work.

Post a reply to this message

> There are even people out there who, using nothing but a compass with a
> sharp tip, cut patterns into acrylic glass to produce simple
> stereographic images (point clouds actually).
>
> And yes, it appears to actually work.

You can often see holograms in the paintwork of cars on a sunny day, it 
looks like an odd lens flare effect that is hovering some distance above 
the paintwork. It is caused by poor washing technique which drags grit 
across the paint creating tiny (but nothing like the wavelength of light 
tiny) grooves in the paint.

Post a reply to this message

On 23/12/2013 02:37 PM, scott wrote:
>> I also found a website with software for computing interference
>> patterns. They claim that if you take something like a 600 DPI laser
>> printer and print the pattern onto a transparency, you can shine a laser
>> pointer through it and get a very fuzzy, very grainy hologram.
>
> Oooh that sounds interesting, might have to give that a go. I got one of
> those cheap $20 green laser pointers, if you focus it it's enough to
> light a match or melt your name into anything plastic :-)

Knock yourself out:

http://corticalcafe.com/software_onlineCGHinstructions.htm

They demo it with a red laser; I don't know if that's crucial. (The 
laser wavelength is probably adjustable in the program...)

>> CD? Perhaps not; CDs don't use visible light, they use infra-red, so the
>> dot pitch might not be small enough.
>
> Well if a 600dpi print just about works, then by my rough estimations a
> CD should be about 20000 dpi. Maybe not perfect, but it works.

Yeah, I've been sat here trying to figure out what the actual DPI of one 
of these disks is. I guess it requires knowing how many inches per meter 
there are. :-P

>> A DVD, on the other hand, uses a
>> red laser, so it certainly ought to be able to do a red-light hologram
>> without difficulty. (Provided you can convince the drive to put the dots
>> where you want them!)
>
> You can't without modifying the drive. IIRC they had to bypass pretty
> much the whole signal path and just drive the laser directly (well not
> quite, but you get the idea).

Doesn't LightScribe help here? Or does that only allow you to specify a 
grey level, which the device itself then dithers?

Post a reply to this message

Am 23.12.2013 16:18, schrieb scott:
>> There are even people out there who, using nothing but a compass with a
>> sharp tip, cut patterns into acrylic glass to produce simple
>> stereographic images (point clouds actually).
>>
>> And yes, it appears to actually work.
>
> You can often see holograms in the paintwork of cars on a sunny day, it
> looks like an odd lens flare effect that is hovering some distance above
> the paintwork. It is caused by poor washing technique which drags grit
> across the paint creating tiny (but nothing like the wavelength of light
> tiny) grooves in the paint.

Yup, that's essentially the method they use for it - except that they do 
it on purpose :-P

Post a reply to this message

> Knock yourself out:
>
> http://corticalcafe.com/software_onlineCGHinstructions.htm
>
> They demo it with a red laser; I don't know if that's crucial. (The
> laser wavelength is probably adjustable in the program...)

Bookmarked - will definitely try that. I wonder if an ink-jet will work 
just as well...

> Yeah, I've been sat here trying to figure out what the actual DPI of one
> of these disks is.

I estimated 7e9 bits of data (probably an underestimate due to error 
correction bits etc) and an area of 17 square inches. sqrt(7e9/17) is 
about 20000 dpi.

> Doesn't LightScribe help here? Or does that only allow you to specify a
> grey level, which the device itself then dithers?

Maybe, but you'd probably still have to hack the hardware significantly 
to allow the lightscribe functionality to work on the data side of the 
disc at a much higher resolution.

Post a reply to this message

>> Knock yourself out:
>>
>> http://corticalcafe.com/software_onlineCGHinstructions.htm
>>
>> They demo it with a red laser; I don't know if that's crucial. (The
>> laser wavelength is probably adjustable in the program...)
>
> Bookmarked - will definitely try that. I wonder if an ink-jet will work
> just as well...

I suspect you'll find it nearly impossible to convince an inkjet to put 
the dots exactly where *you* asked for them.

>> Yeah, I've been sat here trying to figure out what the actual DPI of one
>> of these disks is.
>
> I estimated 7e9 bits of data (probably an underestimate due to error
> correction bits etc) and an area of 17 square inches. sqrt(7e9/17) is
> about 20000 dpi.

...or you could just look up the pit width figure from the spec sheet, 
and work out how many times that fits into an inch. ;-)

>> Doesn't LightScribe help here? Or does that only allow you to specify a
>> grey level, which the device itself then dithers?
>
> Maybe, but you'd probably still have to hack the hardware significantly
> to allow the lightscribe functionality to work on the data side of the
> disc at a much higher resolution.

That sounds like just a firmware nudge to me...

Post a reply to this message