h a l f b a k e r y
I CAN HAZ CROISSANTZ?
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
or get an account
The dust holographic display (henceforth DHD) consists of a
chamber enclosed by glass, and containing a few million
particles of dust. The dust particles are sufficiently
in shape (and partially transparent) that light hitting them is
scattered at random. A gentle fan keeps the
moving, though not very fast. [EDIT - on reflection,
moderately fast random motion would be better.]
In the base of the chamber is at least one laser unit. The
consists of one low-power infra-red laser and one higher
visible laser, with the beams collinear with one another (this
can be done). The combined beam is directed by, for
instance, a MEMS mirror so that it scans the entire volume of
Most of the time, only the low-power IR laser is active,
out a modulated signal. If it strikes a dust particle, the
scattered light is sensed and used to calculate the distance
from the laser to the dust particle. Since the angle of the
beam is known, this means that the exact position of the
intercepting dust particle is also known.
At that instant, the visible laser is turned on for a few
microseconds. The intensity of the visible laser is adjusted,
illuminate the dust particle with the appropriate brightness
(or, if the particle is in a "dark" part of the 3D image, the
visible laser is not fired at all).
With the right density of dust particles and a fast enough
it should therefore be possible to build a complete 3D image
by illuminating the appropriate dust particles at any instant.
Obviously, three lasers (R, G, B) would allow colour
and it might be desirable or necessary to have multiple
to achieve higher scan speeds.
The density of dust particles also needs to be right. If they
too sparse, the image will be made up of too few points of
light, and will be dim and speckly. If they're too dense, the
lasers will seldom be able to hit dust particles on the far
of the box, since nearer ones will intervene.
[2 fries shy of a happy meal, Sep 22 2019]
||Nice, Even just the dust, case and pump might make a beautiful art piece with the right mineral specs.
||Cool. One of my earlier ideas might help to give the laser light consistent dust particles to reflect from. [link]
||//throw a large number of face-tracked parallax enabled video feeds //
[bigsleep] That's some processing right there or would there be a set of pre-rendered angled views which can just be displayed when needed? So not a true different view for each person.
||I think you'll find that this won't work very well, though I'd be happy to be proven wrong.
||With dense enough dust to get a bright image, I'm pretty sure that your laser beams will strike multiple
particles at any given time, each particle being at a different distance from the projector, due to the laser
beams having nonzero width. This will cause radial blurring centered on the projector. Maybe, if your
rangefinding uses something like a TSADC instead of a TDC, you can detect when multiple motes are in the
beam and avoid illuminating them in that case. Either way, though, to get a radial resolution of 0.3 mm
(comparable to 2D displays), you need a ToF resolution of 1 ps.
||Initially, I was expecting this idea to be based on time of flight or two-beam intersection somehow.
Actually, you might be able to do a hybrid of those two techniques. Being a biologist, you might be
familiar with fluorescence lifetime imaging microscopy, but that's not really all that relevant to my
suggestion, just something it reminded me of. Anyway, could you make fluorescent dust that is charged by
one flash of light, and then has a very narrow time window in which it will emit light only if triggered by
another flash of another wavelength? Then you could charge the whole volume of dust with a flash from a
lamp of some sort (which might as well be a laser or LED as well, for a narrow enough pulse width),
followed by triggering selected regions of the dust with a laser projector, with range selectivity achieved
by the relative timing between the charge flash and the trigger flash. Therefore, the lamp and the
projector would have to be mounted at different locations, so that their pulses don't just follow each
other with constant time difference across the width of the volume, which would defeat the selectivity.
||//your laser beams will strike multiple particles at any
given time, each particle being at a different distance from
the projector, due to the laser beams having nonzero
||Not so. Obviously, it will be desirable to have as thin a
beam as possible. However, if the system detects multiple
reflections of the IR "probe" beam (ie, if two or more
particles are both in the beam at the same time, and the
nearer particle does not completely occlude the further
one), the visible laser will not fire. If it detects a single
reflection (ie, there is only one particle in the beam, OR the
nearest particle completely occludes the further particle),
the visible laser will provide illumination appropriate to the
position of that single (or nearest) particle.
||In reality, given a finite beam width, there will have to be a
compromise in the density of the dust: too little, and the
image will be sparse because only the dust particles act as
points of illumination to the viewer; too much, and the
system will seldom have a clear shot at only one particle,
and hence it won't illuminate.
||What you'd probably want is a low density of particles,
moving randomly and moderately fast, and a visible laser
that can deliver a very powerful, very short pulse.
||I said most of that. And the last sentence suggests that the
particles might need to be replaced on a continuous basis.