Bestand wählen

Holographic Video and 3–D Television

Zitierlink des Filmsegments
Embed Code

Automatisierte Medienanalyse

Erkannte Entitäten
right so I particularly want to thank
against small a PhD student in my group will be seeing him on screen for my life have been a couple of minutes when Smith which was a postdoctoral researcher Jim parables was also a PhD student and Stevie Benton and by a wonderful bit of serendipity I'm standing here and on my left I just beyond the lovely zebra hologram is an exhibit about documented in this work in the early 1960's and so are and how many of you know this Steve that as an undergrad worked in the docks lab at that time and Steve did what I guess you describe at MIT is a hat he made the sonar plotter print out Happy Birthday stock and doctor kept that taped to the wall of the lab uh for about the next 25 years for all I know it's still over there you know it's not center and if it's still floating around the museum you really should grab it because it is signed by the so my students and I collaborated with Steve his students on holographic television for about 12 years before Steve passing and I think the reason it works so well is that Steve was coming at all of the video as being the ultimate expression ultimate instantiation of holography and I was coming at it as the ultimate expression of ultimate instantiation television and so we found a very good meeting point in the middle and I'm going to use that theme as I talk about what the opportunities and 3 D television we've already heard alluded to earlier this afternoon I have to do with all video so it's probably worth looking back a little bit especially given Sean's talk to open the day will can we learned from
the 3 D cinema got nothing good or what have you and 1 thing you can learn is Howard Hughes through a whole lot of money and whether you have to be a certain age for Jr in 3 D to mean what it meant in the fifties that's Jane Russell for those who are of a certain age and this 1 does say Mickey Spillane is kind of furious savagery temptation and man woman violence in 3 D I need I say more but I can't let's acknowledge
that be perfectly honest upfront 3 these cinema and 3 D TV is characterized by high and tawdriness and it goes back a long time the 1st cinematic 3 D showing that I can track down of was from New York in 1915 the 1st color 3 D feature interestingly was autostereoscopic it was version of Robinson Crusoe that shown in the USSR in 1947 the most profitable 3 D film ever and again unless you were there at the time you get this wrong was the infamous thus stored as which was rated X and the fact that continues even now with you to in 3 D and everything else this still up out netted anything else that anybody has done if you converted to 2009 dollars that says something important so
what why when I went to the Consumer Electronics Show in Las Vegas in january which as you might have guessed was somewhat shrunken given the state of the economy was there an insane amount of 3 D television whether for entertainment gaming purposes all over the place well 1st of all because H these happen so the industry needs something new after the 2nd that most consumers don't remember Mickey spellings man woman violence and so there's another opportunity to try and get a better and we're all replacing our TVs and because we're all replacing TVs display technology is undergoing a really rapid technological change of a sort that hasn't happened since about the 19 thirties and so this presents a unique technological opportunities slip in all kinds of extra stuff there are a new delivery mechanisms we're not restricted to just broadcast television or cable or VHS tapes in particular we have little any which has a lot more bandwidth and media has a lot more flexibility in the kinds of content can carry and we have internet video which is completely unrestricted because the Kodak is in software receiving so you cannot only come up with a new representation you can send people the call that so they could decode assuming they have an appropriate use of receiving where the display and another point that a lot of people missed is that there is already a huge amount of 3 D friendly countries namely every video came out they can be re render for stereoscopic for multiview 3 D most of the CGI movies the databases are being archived for possible repurposing in 1 of the repurposing means is 3 so this is a huge opportunity and the other thing that happens is that you didn't notice it but when you bought your last computer you got a computer for free it's called the GPU energy use in the PCs and laptops and game consoles and now even phones have gotten so over developed that they cannot lead to high quality graphics in real time thinking too high quality 3 D graphics in real time because they have enough extra horsepower and we'll talk about that because it turns out that even more powerful than you think they're even more powerful than 80 in video thing which is illustrated by some work that when Smith engine Baracus just slipped back into the room I have been working on at the media lab and the rest kilometers those of
you who work or vision people you know this stuff you know there are a lot of very important cues that make you and your brain and your eyes interpret the world 3 D there are also some funny secondary ones that kind of cute but we don't have time to talk about today and ideally if we're going to give you the ultimate perceptual experience were going to make you think you're looking at reality were going to give you all of those used in a way that is mutually consistent which is not used for the 3 D display still or video is true of a lot of well assuming we're just
trying to get to views to your eyes what are the mechanisms for doing what you can have the old red green anaglyph glasses you can have a time-multiplexed where you have shorter is this is what the so called 3D ready of plasma and LCD TVs have interestingly there are now LCD screens that refresh 120 hertz and even 240 hertz you can have full frame rate in both eyes and the price of the glasses has come down to below 40 dollars and you can do polarisation multiplexed that by the way is what stored as the and and and that was the 1st cinematic feature to use polarization multiplexed so it's not a particularly new idea that you can do angular multiplexing autostereo our last area of lens array alternately hollow of video now that requires a lot more pixels that requires a lot more computation and there's a lot more users you voted for the something really sophisticated because it gives you more degrees of freedom do represented as well you could just have
to TV channels you could have anamorphically squeezed the left and the right in 1 picture these of course are completely compatible with the mechanisms we already have distributing represent content but there some more opportunities because as you may
know video coders like and pick 2 1 and 4 represent video it's a predictive coder where since each frame of video was similar to the 1 before and 1 after it you can use 1 frame to predict the next 1 using vectors and to describe the differences and it turns out you can also if you have multiple views across multiple frames you can predict 1 view from the next in addition to 1 frame from the next frame and there is now a backward compatible extension to the H . 2 6 4 codec that supports multiple views and so this is a very efficient way of dealing with the but the bit rate still is proportional to the number of views and when you get to the point where you wanna give the problem using something like that so that there really is no transitions in your head around us that's not a particularly efficient thing to do and that is led people to look for more sophisticated things like sending a signal to the screen was a depth map and then using to sort of 2 and a half the rendering of the receiver to reconstruct the individual frames by changing the parallax of individual parts of the image based on the depth map of and that it turns out to be fit in and make to paid for by sticking the depth map and private data and ordinary to the TV the receiver will just be ignored ultimately and of course this is what
video games do you have a full 3 D representation that superficial not there isn't really a standard for transmission that way right now but if any of you plays a massive multiplayer online games knows that you managed to get all the polygons to your PC and R and B C manages to render them on on its display independent you can render 2 views you can render vertical parallax horizontal parallax whole video of various sorts hard to generate from real scenes of some people some of my students in the past past some over measures students are working on doing that for real scenes you need to come you need to decode it at the receiver with something like a GPU to render it like a video game the costly to render all the views in order to do that we
already heard mention of 3 D at home there's a whole bunch of other stuff 2008 was an incredibly active here this isn't even a complete list of everything that went on and standards by H . 2 6 4 now has multiple multiview video coding built into a 3 D at home all the usual suspects are involved in that the Blu-ray Disc association by is now in the process of finalizing a 3 D format for Blu-ray disks of other simply home-entertainment force on is making standards for 3 D content representations in the called a meeting last August and they had 80 company show so there's an enormous amount of money and technology going into this and there's some
stuff missing any of you here who has tried to do production of post production on even to view 3 D movie knows that the workflow in the schools are really even stereo friendly little multi-view from on and the other thing we forgot besides Mickey Spillane is we forgot what works and doesn't work in 3 D video and 3 D films are a lot of stuff that you would normally do in terms of shock composition in terms cuts in terms of transitions in a 2 D movie that look perfectly fine either destroy the 3 D effect or make the audience part or both what what many people going back to the 19 nineties technology discussion we had earlier this morning what people used to call the art sickness back in the nineties you know when when it's happening on something the size of an IMAX screens not really much fun so people have to rediscover the creative and technical aspects of how to maintain a reasonable 3 D illusion throughout the whole show we heard this
talked about here it is a physicist or not you know photons don't just go halfway across the room and make a left turn but back or Pepper's ghost is what most people think pyelography is that a lot of things we need to fix that what it really looks like something
more like this it looks like a hologram except that it's all it's a video screen and that's not really good photo we can construct the object behind the display screen straddling the screen around in front of the screen
and what's special about it is that if you do it right it makes all the depth cues in a consistent manner with reality of because ultimately the goal of good holography fall apart from being the visual indication medium is to reconstruct the light wave fronts that were coming from the object not just a of neurons it's actually trying to send you the physically right stuff and so you're I brain to respond correctly and the other thing since as we said it can reconstruct the object in front of the display device which means in the viewer's space as when Kluzniak explored for a PhD thesis introduced knocked out requirement for which he was same way and sure she explored the idea that you can have a haptic interface to the object is the object hanging out there front of the screen you can hold it with a phantom and you can have a data glove and with and not a whole lot of 3 D display technologies support work that kind of interaction but also not clear how many things it's good for but at least it gives us the opportunity to try these new ways of interacting unitary so what we need for whole a
video we need funds and we've talked about that a little bit we need a distribution mechanism we talked about that a little bit we need the computation for the display all address that moment and then we need the hard thing the electrooptical display we need the whole the there are no used to
joke but you could have a meeting of everybody in the world that ever built the whole of the you displayed in my office and still have room for the MIT football team I and have a very big office but in the past couple of years a number of additional people have emerged who were working on all of your actively in is a much longer list of people who have done it at 1 point or another before the funding round and in particular there is a commercial venture called serial it's another commercial venture that I won't name because they are actually doing all of the just marketing with the use of Yale University of Arizona including 1 of our alumni University Kilgo neon University Tokyo University of Agriculture and Technology we think the unique standing behind what we're doing at the media lab right now is that we have an emphasis on making this a consumer technologies to be cheap it's gotta be fast it's going to work with the infrastructure of the content we have now which is not a problem so that the other people present just so hard to do it at all the same and it's gonna cost 200 box and it's just plug into your PC without changing your PC is an extremely aggressive thing from you and of course using
diffraction to reconstruct the light wave fronts this is drawing actually by Betsy based on some of the lectures slides
In order to do that you need your displayed have pixels of the size of a wavelength of light so that means if you're screen is a meter wide you need 2 million pixels per scan that's a lot of pixels and and you also need a million Scanlon that makes it really hard for so you can reduce the problem of that by doing horizontal parallax only in part of it about that before that's what we're spending most of the time trying to do right now although ultimately would like to do for their and you can transmit the diffraction 1 of the things that people got really long the when they started thinking about holographic television even conceptually back in the sixties and seventies they would look at how much data was in a diffraction pattern and this was just too hard to transmit so all of that he was announced on and our interest has been right some other representation and compute the diffraction pattern at the receiving end and we're now doing that with standard you use
and so what I'm going to do is I'm going to take you with them smaller than a little tour of an antique more than 14 years old at this point and I know 7 John lusting after this thing they want to hear although it's they actually had a keep it alive why they might think differently about we're going to go over to my lab and talk about the system which was built by Steve bed and then by his particularly his grad student here say somewhere in the early and mid 19 nineties this was useful work to so can we come over to the 2nd unit the all right so we get to go they're Minami
tha so it's going alright so that's stand while using making as a military we can indeed Don can you hear me fantastic so now this is the marked you holographic video display is that both said this is a
an NCH already from the mid 19 nineties many continues to be important because it in conspires isn't farms our current work and holographic video and the things that I want to get from this tour is that it works number 1 but that's it's big and expensive and loud as you can probably tell so let's go ahead and start here with the computer the PC that this we have originally driven by a super computer
and we replaced it with a PC a few years
ago right so a is now running after this
GPU this a commodity well I am graphics card that's where the yeah holographic friend is a computed in changed to analog signals the signals a sense into this Saracco of electronic where that signaling is up-converted
amplified as travels then Sudia optics Stable
1st see the acousto-optic
modulator the where that
RF signal in terms of the z piezo transducers that actually pounds acoustic waves across a
stock that treats Siena dynamic holographic pattern and the like traveling through that pattern action diffracts into the holographic image requests the pattern is
moving at the speed of sound in the crystal so we need this addition Electronics Inc. is Bank of fast scanning mirrors to make the
output appear stationary nap actually comes a great great here and force they were not able to see it with this camera and but
were you know st agree with me that that is
there so all right so I wanna point
out we show this still from a personal point out 2 things but most of what's on that table is being replaced by a chip
that then is making we're going to go over
to another lab and take a look at that and that shit cost about 10 dollars
and you see that the racket like products over on the left that's being replaced by 4 little printed circuit boards each about
half the size of playing for us so that
that rack is about 6 feet high 19 inches
wide and about 2 feet and that's been replaced
by about 10 dollars worth of electronics that sits
between the VGA connector DVI
connector on your PC and the input to the new world you display of OK so we're going to see how
fast you guys can run over to the other
room and I'm going to talk through a couple of
slides while you're doing that so will be back to in a minute then that so can we
cut back to the black top ch
so I want Quinn and Jane have demonstrated that it's possible to take a standard and NVIDIA GPU of the kind that's in your PC if you bought your PC last year and you can make a standard definition TV horizontal parallax only holograms at about 15 frames per 2nd with a single GPU for 3 D scenes of about the complexity of a PlayStation 2 video-game that presented at the SPIE conference in January and this picture at the bottom is not all that illustrative list showing the effective range of views of an object and then the hologram that's actually generated from them and that's what gets sent to the light modulator so
you need as Dan eluded a device called a spatial light modulator into which you can not only put the whole around but you can rewrite the hologram at 15 or 30 or 60 frames per 2nd so you get smooth motion there a lot of things you could do for that mean you could take an LCD display the problem is an LCD display the pixels are so big that the best angle you can get is very small so about all you can make is all on-axis holograms which are a whole lot of fun to watch a TV but you could likewise use digital micromirror device is at university Arizona is using rewritable photopolymer material which is effectively imagine a photopolymer hold race electronically almost instantaneously and of course you still need a way of writing it again I to some degree they push the complexity just a little far back farther back in the system rather completely solved the problem by the photopolymer material can be quite large which is very attractive but what we used in the old overdue displays a mark 1 more true that were developed and Stevens group with some collaboration from my students were acousto-optic modulators the crystals of tellurium dioxide we're now using a much higher bandwidth and cheaper technology called a guided wave devices don't really have time to talk about that in detail but for any of you who knows how a sort devices surface acoustic wave filter works so this is 1 of those except that it's made out of a transparent material so the light travels through the material the acoustic waves travel across the surface and the fractal light in the wave guidance materials these are very cheap and they're easy enough to make that we can make them here on MIT campus very quickly and so
this is a photograph of 1 day and uh made about a year and a half ago and I'm going to trust that at this point they've managed to run down the hall and the set up in the other room where Don will just show you quickly this set up to where we test and calibrate that even at what can we use switch to the
chair Numidia hang on worst all switching the and yeah certainly there before could hear the rules on the but the thing the and
it that we are we stopped and then wrote in years we can hear you we just can't see you
know doing it can might come the overhead kept at a chair
and I let me
get here now but
switch laptop and try switching back
the so this device but was about the size of your
thumbnail very small OK then you're
on you can pay me we can see you
alright and there are what you the pretend
and the the unique in
Kilkenny you can see here device we can
see that I people about right so it could be a
guided wave acousto-optic modulator and
to think about it on the other room but will go through a little quickly wouldn't within a you this way
their produce the being the travel to the
prism and couple then to my device at the direction the thought of it a of fiber-optic cable kind of in diffused onto the surface so the I have a thought of right that the RF comes into the cable and interact with transducer fears that kind of paint that substrate and create little ripple the travel across the end of the chapter beam of light and those Republican fractal-like incre the holographic image that they exit that state is secondary interaction agree rotate the polarization so added catalyst is polarized the Olivier book the nite of the women it so that that hope you hope we can take in the dot yes and is the apathetic the vertical everything back and forth and to try that it is that the nificant thing and then and then that crime back across and that the fundamental element of that fraction creating so a good thing to note here we think that going from the market here to the marque reconfiguration including the divide it by going from mechanical to electrical television where that might have been in the cabinet what we're looking at here would be the key 2 and the knowledge base and powerful in addition although when is the near the and so on a table in the other room have analog integrated optical and that the integrated optic divide we envisioned taking more and more about the stuff on the table and putting it on to the the right toward the goal of this whole video on a chip and that's really yeah the goal here and and what we even look at the thanks very much that you might wanna keep listening for the applause during and a couple of minutes and have had had thank you so no matter what should build it's never quite as good as you'd like and so therefore I if you want to make the image Figure you can't hire you can use multiple of these devices you can scan of the device is really fast you can imagine more than 1 place of You can steer this is actually would see real does you can track viewers eyes and you can only by other reconstructing the hologram but that will go through the viewers pupils realize of course that adds a lot of complexity it doesn't scale for lots of yours but each of these
ads cost and complexity everybody doing whole video is doing 1 or more of these and this is the way that these displays will become bigger so I'm going to close but in
tribute to Steve with an apology unapologetic of for holographic imaging for those of you who haven't already seen this of this final this was the book that Steve was working 1 when he passed away and Gini talked me into continuing to work on with a whole lot of help from a lot of Steve's alumni and it finally came out last April at the book launch party here at the museum thanks guys for letting us do that we had up in the gallery and I think actually the launch of this book was 1 of the rallying points to caused people to get a bit more serious about what was going to happen to the library collection here and so not only is it a really wonderful capture of speeds voice and the undergraduate class that Steve fault it's also look forward into areas like holographic printing holographic video and all of the different ways of doing something pyelography and it was a good excuse for people to get together here at the museum and say where do with all these holograms any we're going to do going forward so I'll stop there thank you for holding out to the end of the day it's been a very long day if fj bouquet Michael thank you very much now mn few minutes for questions and the good the the direction in the middle 1 kiss now what kind time-scaling we see this becoming that have really commercial product well if you'd asked me a couple of years ago I would have said at least 5 to 10 years but given the huge excitement about distributing 3 D content even as 3 D models I would say that at this point is really just at any given our recent successes in using off-the-shelf GPU so it's it's really just a matter of making the monitor and everything else exists and so I would say that this is potentially a 5 years last the tha and time sec a Chicony tell a bit more about the nature of the song how it works in assist possibility for scaling it up with the limitations of that well that's really dans area of expertise it's what is working on for his doctorate but effectively you need a material that has a couple of characteristics it needs to be of piezoelectric and it needs to be transparent and it needs to be such that you can create a wave guide inside and and lithium iodate turns out to work pretty well for that what we're looking at other kinds of spatial light modulators as well and this 1 there's a limit to how big you could make that device so you probably would have to do something like make a bigger piece of lithium iodate input multiple devices on it rather than making a single 1 and then there were some multiplexing and wiring issues that need to be worked out we have a lengthy discussions about these and I'm not prepared to say much more about what our plans are in that area except to say that we are looking to do 1 all of those just a book but we're question can is a monochromatic is the image monochromatic the image right now is monochromatic any of view it turns out that unlike regular television color is trivial and big is heart of which is to say that you can run these devices fast enough that you can do colors sequential 1 of aligned basis so you don't see any rainbows I you can stack 3 of the devices and run on 1 1 laser through each 1 and now there are relatively inexpensive green and blue lasers it's another thing I didn't talk about there's been an explosion of interest in using lasers for consumer video projectors and so you can buy quite inexpensively of powerfully enough to be dangerous red-green-blue semiconductor laser tryouts about which is suitable for this application so we're hoping to take advantage of that indeed the original market 1 0 a video display ran as RGB for about the last year has existence so there is some expertise around the media lab in doing this in color Michael I. could you say something about the advancement and content that you might imagine I mean this is hi significant with regard to the the world of television that we've known and where we might go with regard to 3 D in a certain and or maybe had and certainly that perhaps is an opportunity here for that the content to change as well we you have anything said what what I'd like to do is I'd like to build this many of these as possible and get them into the hands of as many creative people as possible and make this be like the early days of the Internet in the early days of holography where people did all kinds of damn fool things and some of them turned out to inspect us so we're not proposing any individual trajectory for content rather we're proposing to make something that's accessible inexpensive and relatively easy to use and then teach people how to use it and see what they come up with the just only on that and going 1 step further there may be some lessons to learn uh in that area from the experience of IMAX 3 D uh I used to be responsible mouthings things running a 3 D and IMAX theater and it is interesting that when the 3 D uh came in at IMAX uh moralist blew away to the films and I found in the theater we're running so we had a great deal of difficulty selling to the film when audience because they were so keen on 3 D but I max from the beginning has faced up content problem it's turned out to be and the effects both 2 D and 3 D of of large-screen format and see not seems to been taken a long time even further and experienced filmmakers to learn how to exploit the technology to some great advantage in the commonest criticism of IMAX including 3 D is great technology uh rubbish fills and and I know that may or may not be a problem for 3 D video or 3 D TV I don't know what we do have some recent examples of 3 D West somehow seemed hard for people to exploit the technology to great artistic or cinematic effect do you see that as a potential problem for this technology well I'm not going to go too far into my opinions of IMAX but all know that the way that I max was marketed with all due respect I as an experience rather than as a storytelling medium for the most part and the fact that it's not a particularly accessible medium to typical filmmakers already put a constraint on and advance and that's the sort of thing we like to avoid we don't wanna make this be a specialized medium for which nobody has access to the tools are so I hope would avoid some of the troubled by will be good for Michael thank you very much and fj
Vervollständigung <Mathematik>
Kollaboration <Informatik>
Topologischer Vektorraum
Ordnung <Mathematik>
Charakteristisches Polynom
Tabelle <Informatik>
Demoszene <Programmierung>
Weg <Topologie>
Endogene Variable
Streaming <Kommunikationstechnik>
Prozess <Physik>
Natürliche Zahl
Computerunterstütztes Verfahren
Element <Mathematik>
Komplex <Algebra>
Regulärer Graph
Schnitt <Graphentheorie>
Maschinelles Sehen
Speicher <Informatik>
Arithmetisches Mittel
Inverser Limes
Bildgebendes Verfahren
Leistung <Physik>
Physikalisches System
Objekt <Kategorie>
System F
Statistische Hypothese
Elektronischer Programmführer
Metropolitan area network
Güte der Anpassung
Motion Capturing
Rechter Winkel
Klasse <Mathematik>
Automatische Handlungsplanung
Content <Internet>
Spannweite <Stochastik>
Produkt <Mathematik>
Inhalt <Mathematik>
Stochastische Abhängigkeit
Materialisation <Physik>
LASER <Mikrocomputer>
Demoszene <Programmierung>
NP-hartes Problem
Kartesische Koordinaten
Arithmetischer Ausdruck
Einheit <Mathematik>
Freier Ladungsträger
Figurierte Zahl
Funktion <Mathematik>
Zentrische Streckung
Prozess <Informatik>
Technische Optik
Overhead <Kommunikationstechnik>
Nichtlineares Zuordnungsproblem
Interaktives Fernsehen
Digitale Photographie
Diskretes System
Automatische Differentiation
Mapping <Computergraphik>


Formale Metadaten

Titel Holographic Video and 3–D Television
Serientitel Photons, Neurons and Bits: Holography for the 21st Century (MIT Museum 2009)
Teil 11
Anzahl der Teile 13
Autor Bove, Michael
Lizenz CC-Namensnennung 3.0 Unported:
Sie dürfen das Werk bzw. den Inhalt zu jedem legalen Zweck nutzen, verändern und in unveränderter oder veränderter Form vervielfältigen, verbreiten und öffentlich zugänglich machen, sofern Sie den Namen des Autors/Rechteinhabers in der von ihm festgelegten Weise nennen.
DOI 10.5446/30389
Herausgeber River Valley TV
Erscheinungsjahr 2012
Sprache Englisch
Produktionsort Cambridge

Inhaltliche Metadaten

Fachgebiet Informatik

Ähnliche Filme