Custom displays are a lot of fun to look at, but this one is something we’d expect to see at a trade show and not on someone’s kitchen table. [Taha Bintahir] built a 3D volumetric display and is showing it off in the image above using a 3DS file of the Superman logo exported from Autodesk. In the video after the break you can see that the display is a transparent pyramid which allows a viewer to see the 3D object inside from any viewpoint around the display. Since first posting about it he has also added a Kinect to the mix, allowing a user to control the 3D object with body movements.
There’s basically no information about the display hardware on [Taha’s] post so we asked him about it. It works by first taking a 3D model and rendering it from four different camera angles. He’s using a custom designed prism for he display and the initial renderings are distorted to match that prism’s dimension. Those renderings are projected on the prism to give the illusion of a 3D object floating at its center.
We’re hoping to hear more details about how this was designed and what hardware is being used. We’ll post a follow-up if [Taha] shares more information.
[youtube=http://www.youtube.com/watch?v=LuCbx5h1ajU&w=470]
That’s awesome, hope a tutorial or something gets put out
Our future televisions will be pyramids!
You can see the ‘secret’ of the technique for a few frames in the 0:27 to 0:28 range. The top of the box is a computer monitor, so you are seeing a reflection of the S, one for each side of the pyramid, to give you the illusion that the S is suspended inside the volume.
However, while the S will apear inside, all parts of it will just be at the same depth. And it isn’t really 3D either; there’s no stereoscopic effect generated by this system. What appears 3D in the video is an artifact of the model being rendered as spinning. The illusion would break down quickly if it was static and you walked around. You would only see four views, one for each face of the pyramid.
A very neat demo and an interesting way to display something a depth ‘inside’ an object, but not nearly what the title claims. Great work though!
This is badass.
Excellent keep it up Taha! Good to see NUI groupers coming up with more homebrew tech. Would be great to scale up a bit.
Awesome. Presumably the clever bit is actually in the software skewing the projected images so that they reflect to look like they’re in the middle of the prism.
Time will tell, hopefully they’ll release a tutorial!
That’s interesting, but as far as I know, that’s no 3D volumetric display. That’s just the reflections of flat images. It does give the illusion that the object is floating, but there’s no depth. Yes, you can see it from 4 different angles, but that’s it.
It’s a cool thing, but don’t call it 3D volumetric as it is not (or I’m missing something…)
This is fantastic. As you said we could really do with a guide on how to make this. I also wonder how big the 3D object in the prism can be made he doesn’t seem to use a lot of it’s volume.
Looks like there is a company making product display cases like this since 2008.
http://www.engadget.com/2008/10/30/realfictions-dreamoc-3d-holographic-display-demoed-on-video/
http://www.realfiction.com/en/
I’m guessing that there are 4 projectors in the top projecting the 4 movies down onto the 4 faces of the pyramid.
@Truth,
A single flat panel mounted above with 4 images on it.
Sorry but no videos of a STATIC 3d object and then moving around it.
It’s not a 3d display , it’s a 3d illusion. the rotating objects mask that it’s not a real 3d display.
I could be wrong, but until he posts video of a STATIC object, and walks around it videoing the object from all sides, I’m calling it not real.
I saw something similar at CES 2011 and thought it was pretty funky.
http://www.youtube.com/watch?v=eUffesb_nhg
You could use a few tiny screens to display the static object from a different angle for each panel, or one large screen that renders different angles to different quadrants I suppose.
Sure, it’s not truly 3D, just 4 different angles, but it’s still cool nonetheless.
Yes pretty cool. You can do the same using a cylinder, just do the image transformations accordingly.
Cylindrical Holography
It’s been quite some time since this thread was started, but all the naysayers here are wrong my friend. It’s called Cylindrical Holography, and it is gaining ground and already becoming cheaper.
http://www.rayhologramart.com/2011_2we.html
http://www.rayhologramart.com/2012_2w.html
http://design-milk.com/hanging-around-at-the-philadelphia-museum-of-art/
http://www.holonorth.com/new.htm
http://www.globalspec.com/reference/34375/160210/chapter-4-4-2-scanner-devices-and-techniques-implementation-of-holographic-scanners
“We investigate the computer-generated cylindrical rainbow hologram. Since the general flat format hologram has a limited viewable area, we usually cannot see the other side of the reconstructed object. There are some holograms to solve this problem. A cylindrical-type hologram is well known as the 360-deg viewable hologram. There are two kinds of cylindrical holograms, a multiplex hologram and a laser reconstruction 360-deg hologram. Since the multiplex hologram consists of many 2-D pictures, the reconstructed image is not truly 3-D. In contrast, a laser reconstruction 360-deg hologram has a true 3-D effect. However, since the spatial resolution of the output device is not enough and the calculation amount is huge, there are few reports on computer-generated cylindrical holograms. In our previous study, the computer-generated cylindrical hologram was realized as a Fresnel hologram. The calculation amount was too huge and took about 44h in total calculation time, though we had used several PCs. We now propose the rainbow-type computer-generated cylindrical hologram. To decrease the calculation amount, the rainbow hologram sacrifices the vertical parallax. Also, this hologram can reconstruct an image with white light. Compared with the previous study of the Fresnel type, the calculation speed becomes 165 times faster. After calculation, we print this hologram with a fringe printer, and evaluate reconstructed images.”
~http://opticalengineering.spiedigitallibrary.org/article.aspx?articleid=1089238
Abstract
A fast calculation method for computer generation of cylindrical holograms is proposed. The calculation method is based on wave propagation in spectral domain and in cylindrical co-ordinates, which is otherwise similar to the angular spectrum of plane waves in cartesian co-ordinates. The calculation requires only two FFT operations and hence is much faster. The theoretical background of the calculation method, sampling conditions and simulation results are presented. The generated cylindrical hologram has been tested for reconstruction in different view angles and also in plane surfaces.
© 2010 Optical Society of America
~https://www.opticsinfobase.org/oe/abstract.cfm?uri=oe-18-25-25546
Telepods
~http://www.thestar.com/business/2012/05/04/star_trek_comes_to_life_lifesize_3d_hologram_created_by_queens_university_researcher.html
~http://www.kurzweilai.net/life-size-3d-hologram-like-telepods-may-revolutionize-videoconferencing
~https://www.youtube.com/watch?v=wa5I-D5_sM0
I think this would be appropriate:
WANT!
@ MrX
Agreed, but a cone, not a cylinder.
Think of it the same way you learn to approximate integration. As you increase the number of faces you decrease the error in approximation of a 3d object. In this way, a cone would be of the correct shape with infinate faces and would show a 3d object correctly.
@Nate and MrX,
What you are seeing is a reflection from 4 images above the “panes”. This requires that you use a faceted design, with each angle being rendered and “proejcted” from above. A smooth cylinder would break the illusion, a faceted cylinder (or cone) would greatly reduce your display area, while increasing the angles to be rendered.
@Caleb
So your saying I’m right?
@Nate,
loosely. setting the infinite faces aside, lets just say that you had a cone with LOTS of faces. this would mean that each face would be tiny (unless you built the thing the size of a stadium). This means your projected object would also be tiny.
So I guess, yeah, you’re right but it wouldn’t be practically sized.
@ nate
as you approach infinite surfaces and infinite viewing angles you need infinite images projected without visual overlap. Difficult.
ah, tis true
As Phroon points out, this is not a volumetric display. @Truth mentions the company marketing a product that is similarly designed and described as a “holographic” display (which it is arguably not). This type of system is much closer to smoke & mirrors type parlor trick than holographic or volumetric display. Unfortunately the average consumer has no clue what a holographic display is, and if the object is rotated, as done here, it is harder to spot the illusion.
@Nate, gcat122 and Caleb
I was considering something like this:
http://www.youtube.com/watch?v=qQ1AWAzJEH8
You can make a 360º panoramic view of an object but that’s all – still no depth:
http://www.youtube.com/watch?v=GkWBQsMOa6k&feature=related
And yeah, the effect is not as good as what I was hoping for..
@MrX,
That doesn’t account for perspective changes in the object. You could put something together to detect where the person’s face is, and change the image to be the appropriate angle, in which case yours would work quite well. otherwise, you simply can’t change the perspective (no facets to hide the image change)
The 4 sided pyramid does this mechanically (without tracking you), by supplying 4 perspectives statically.
THis isn’t 3d, and isn’t volumetric.
@Caleb Kraft
You are right about the perspective changes. But yeah, with an active setup (tracking sensor) it is even possible to emulate the horizontal parallax.
Since we are on it, here it is another cool effect:
http://www.youtube.com/watch?v=lFqlQiTTHRs
This one is truly holographic by nature. However we still need to do phase-modulation of light in the place of the source object (they use the real object in this video) if we want to produce computer generated images which is not very convenient.
Thanks for topic!
The only really decent “real” 3d display I’ve seen is this one http://gl.ict.usc.edu/Research/3DDisplay/ It actually presents full 3d images, the concept seems scalable, to large display volumes, and the image is mainly only constrained by the capabilities of the device projecting the slice onto the spinning platter (which admittedly, you’d probably need a fast and precise projector, especially to achieve full color.
I saw this in “Forbidden Planet”, what is this guy, a Krell?
@Luke
I remember that project! Too bad I do not see it as scalable. Imagine a 50″ spinning mirror in cali when theirs an earthquake… Talk about severing family ties!
@uMinded
I don’t see a problem there as mirrors don’t have to be as fragile as used in everyday life. With modern material science I see it as scalable, but one question comes to mind… Where would a display as such be practical?
awesome! with the object wasn’t always spinning though. can’t really tell if you’re changing perspective from moving the camera, or if it’s just the rotation.
Combine this with johnny lee’s head tracking + a 3D monitor and you would have a display that felt like a true 3D hologram.
Anyone have a spare prism, 3D monitor, and a couple Wii remotes?
Isn’t this a variant of Pepper’s Ghost (like that used in the ballroom of Disney’s haunted house)? A pane of glass at 45 degrees reflects an image illuminated (or in this case, projected) from above/below. This technique dates back to the mid-1800s.
I see this type of 3d display in Moscow in new year, in which was playing animated CocaCola advertisment.
A train goes around coca cola bottle.
I’m no expert but it seems to be a clear example of an old stage illusion called ‘Peppers Ghost’. Google it. The trick is so old that many people have forgotten about it. The 45° glass panels are a giveaway.
NEEED TUTORIAL
@nate,
could u suggest me the way how i can do the same….