ISee3D Completes CDR for Single Lens 3D Camera

[cybereality]

http://www.advancedimagingpro.com/web/o ... era/3$6084" onclick="window.open(this.href);return false;

Some company called ISee3D (pretty original, huh?) claims to have created a technology that can record a stereoscopic image from a single lens. I went to their website (http://www.isee3d.com/" onclick="window.open(this.href);return false;) and read the patent. It sounds interesting. Could be of use in the medical industry where they need very small cameras. It might also help me realize my dream of having a stereo3d spy camera.

[yuriythebest]

ISee3D - yeah "original"

it's cool cause apart from iz3d, samsung also has a similar sounding 3d technology I think it's "ic3d" or something- it's for autostereoscopic displays


**goes of to register a patent for "ayeSee3d" **

[mediavr]

This seem related to the V3 technology where an oscillating aperture in a cine lens provides subliminal amounts of parallax leading to enhanced depth clarity in videos. I know it sounds like smoke and mirrors but companies like Angenieux are supporting it.
http://www.inv3.com/technology.html" onclick="window.open(this.href);return false;

It is interesting I think that small amounts of parallax can create spatial enhancement in video material -- our brains must to coded to pick up like single pixel amounts of parallax -- or at least parallax oscillations.

cf. also light field photography which will be big in stereo in a couple of years I think
http://en.wikipedia.org/wiki/Light_field" onclick="window.open(this.href);return false;

Peter M

[cybereality]

This seem related to the V3 technology where an oscillating aperture in a cine lens provides subliminal amounts of parallax leading to enhanced depth clarity in videos. I know it sounds like smoke and mirrors but companies like Angenieux are supporting it.
http://www.inv3.com/technology.html" onclick="window.open(this.href);return false;

Actually, I think that *is* a bunch of smoke and mirrors. I looked at some of the samples, they look just like regular 2D videos to me. From what I gather they are supposedly getting some parallax information and using that to simulate depth. I guess this is similar to the "wriggle" stereo-photography but on a much smaller scale. Still, it reeks of B.S.

cf. also light field photography which will be big in stereo in a couple of years I think
http://en.wikipedia.org/wiki/Light_field" onclick="window.open(this.href);return false;

Interesting article. I am not exactly sure what it all means, but it sounds like it could be useful for making a holodeck.

[martinlandau]

MediaVR, interesting, I have some thoughts and would enjoy some feedback. First off relating to getting a stereo3d image from a single lens. I read a lot of charles darwin as a kid, http://en.wikipedia.org/wiki/Eye" onclick="window.open(this.href);return false; Reading this link explains why one eye and the evolution of the eye leads to advances that give us the eyes we have today, we were predators, but our eyes evolved wrong, octupus eyes don't have blind spot like our eyes do.

It seems to me nature is efficient and a creature would have evolved with one lens instead of 2 or more to get good depth measurements if possible, where are all the creatures with one lens? I did see this movie Krull once where they had cyclops people.

Secondly, http://www.mtbs3d.com/phpbb/viewtopic.p ... e&start=15" onclick="window.open(this.href);return false; Cyber bought some special pinhole glasses that DO something, and yuri elaborated on it from the link below - Cyber I would like for you to check out the link Yuri posted - those pinhole glasses have many more pinholes than the glasses you bought from ebay - do you know where I can get the many pinhole glasses? It seems for certain people, pinhole glasses can be very helpful.

http://en.wikipedia.org/wiki/Pinhole_glasses" onclick="window.open(this.href);return false;

Pinhole glasses, also known as stenopeic glasses, are eyeglasses with a series of pinhole-sized perforations filling an opaque sheet of plastic in place of each lens. Similar to the workings of a pinhole camera, each perforation allows only a very narrow beam of light to enter the eye which reduces the size of the circle of confusion on the retina and increases depth of field. In eyes with refractive error, the result is often a clearer image. Unlike conventional prescription glasses, pinhole glasses produce a clear image without the pincushion effect around the edges (which makes straight lines appear curved). While pinhole glasses are useful for people who are both near- and far-sighted, they are not recommended for people with over 6 diopters of myopia. It should also be noted that pinhole glasses reduce brightness and peripheral vision,[1][2] and thus should not be used for driving or when operating machinery.


Furthermore Cyber from your previous thread:

I imagine this would require the correct pinhole setup at the correct interocular width with the correct sized screen at the correct viewing distance.

and

It enhances atmospheric perspective (which is one of the main ways we determine distance, and relative distances of objects) It enhances color seperation [sic], so that in certain scenes containing the right combinations of red and blue, or orange and green
it creates a deeper effect, by making one color seem farther back, similar to the way Chromadepth looks.

So cyber, tying all that together, I am betting if we get the right sized screen, with the right kind of pinhole glasses, at the right viewing distance, and with source material that is the right combinations of red and blue or orange and green, we are going to have something SWEET! I notice from your review some scenes looked more psuedo3d than others, perhaps those scenes had the better combinations of red and blue or orange and green? I am sure the science of this can be determined to maximize this effect for the general population.

If we can use this "parallax" tech but refine it with pinhole glasses tech to make it more focused and hit each cell in our eyeball properly instead of just scattering all over, should help no?

http://www.mtbs3d.com/phpbb/viewtopic.p ... ure#p29163" onclick="window.open(this.href);return false;

I met with the guys at the post above at siggraph, this is kinda related. Adaptive coded aperture projection, they GREATLY cleared up images projected onto curved or dome screens with this technology. Cyber I don't know about the tech you posted, but I saw the presentation at siggraph by these guys, and thier tech made a huge difference to the clarity and color.

s http://www.uni-weimar.de/medien/ar/research.php" onclick="window.open(this.href);return false; were showing off:

Adaptive Coded Aperture Projection

Displaying dynamic aperture patterns on a liquid-crystal array at a projector's aperture plane together with inverse filtering supports projector defocus compensation, high quality projector de-pixelation, and increasing temporal contrast of projected video sequences. Such adaptive coded apertures are a step toward next-generation auto-iris projector lenses.

[mediavr]

Somewhat related to these unconventional methods of producing improved depth clarity is monocular stereopsis -- a long studied phenomenon in visual psychology. The idea is that sometimes, in the right conditions, a 2D photo or painting can look really 3d -- indistinguishable in fact from the solid depth sensation we usually associate solely with parallax based stereo viewing. And it is not just a mental sensation, it actually is conducive to physically measurable increased accuracy in photo interpretation (for camouflaged scenes etc). V3 mention this photointerpretation application of their tech too and support it with a psychology research paper.
Monocular stereopsis doesnt necessarily mean viewing with one eye, it includes devices, like the synopter (a Zeiss version of the concept) where two eyes look at a scene through a single objective (like a twin eyepieces for a microscope except for viewing normal scenes). The idea was that people would buy this to look at paintings in galleries and see them as 3d. I write about it a little in this blog post
http://www.mediavr.com/3d/?p=40" onclick="window.open(this.href);return false;
How I think of it is this: we look at a painting on a wall, it seems flat, we can see it is flat, stereo information tells us it is flat. If we can remove the stereo information of flatness our natural tendency to see the world (and images) as a volume, not a cyclorama surrounding a bunch of 3d objects in the foregound can emerge. There is work being done by psychologists interested in virtual reality today on improving the synopter device and they report significant progress.
http://www.liv.ac.uk/VP/rob.html" onclick="window.open(this.href);return false;

Peter M

[martinlandau]

Ok I read more on isee3d patent. The picture showing the wood grain in parallel versus toe in speaks more to me than 1000 words. Cyber or anyone, the 2 virtual cameras in the nvidia driver or iz3d driver - are they doing the right toe in as you focus on far away versus close object? Or are they simply parallel too? Is this what autofocus feature help with in iz3d driver? This is something that needs to be addressed by neil maybe with new s3dga standards?

[martinlandau]

"ow I think of it is this: we look at a painting on a wall, it seems flat, we can see it is flat, stereo information tells us it is flat. If we can remove the stereo information of flatness our natural tendency to see the world (and images) as a volume, not a cyclorama surrounding a bunch of 3d objects in the foregound can emerge."

I work with disabled war veterans sometimes, and the brain is an amazing thing. People that have had thier limbs blown off, sometimes the limb still feel like it is there, and even itches or burns or gets cold. I agree, you remove certain things, the brain will compensate and make information that is not there originally. I recently saw a show on the science channel with that Katu guy, also had ray kurzweil in it, it was about superhumans and amazing human powers. One guy run a marathon in freezing temperatures and was A OK, most humans would die from exposure to cold like that. They had another guy from germany that could do all kinds of advanced algorithms in his brain, compute really large numbers - but he was special because most savants only can do 1 algorithm, but he could rewire his algorithm to do any kind of computing. Anyways they had this one guy who had been born without eyes. Now back in psychology glass my professor asked me once to describe a red apple to a blind man who had never "seen" anything. I could not figure out how to describe the color red to a blind person who had never experienced sight - can you do it mediavr? Well this blew my mind, this blind guy who had never had eyeballs, he was from italy I think, he was painting pictures, with accurate representations of the physical objects - like a ship on the water - and with proper colors! I still don't understand how he did it - it was fascinating. I can only speculate that genetically something has been passed to him that gave him the neurons necessary for sight, even though he never had eyeballs and couldn't develop sensory experience that way, so the sensory stuff must have been passed from his parents somehow. Anyways the scientists wanted to see if he would accurately paint this 6 sided building and do the convergance right on the 3d aspect of it. He DID! It blew their minds too, they don't understand how with no eyeballs he could properly understand sterescopic stuff like that. Anyways they put his brain through an MRI and saw some parts of his brain were working like a guy who had been born with eyeballs, but they couldn't explain it, because he never had eyeballs that stuff never should have worked in his brain, he was "seeing" without ever having his own eyes, perhaps cellular memory from his ancestors possibly? Anyways he said he didn't like how the MRI made his brain feel and would never do it again.

edit: here is the show

http://www.discoveryhd.ca/shows/castdet ... 0&sid=4608" onclick="window.open(this.href);return false;

http://www.discoveryhd.ca/shows/castdet ... 5&sid=4608" onclick="window.open(this.href);return false;

Esref Armagan

The Blind Painter"
Esref Armagan - 'The Blind Painter'
The Superhuman and Quest: Esref Armagan, of Ankara, Turkey, is a 53-year-old blind painter. Blind since birth, Armagan is a gifted visual artist who can draw and paint in three dimensions; drawing comparisons to Renaissance architect Filippo Brunelleschi, the first artist to master three point perspective. Armagan paints houses, boats, birds and butterflies, even though he's never actually seen any of these things. He paints with lively colours and has even learned to draw in perspective, yet his brain has never detected hue, light or shadow. Over the years, Armagan developed his own methods for creating his artwork and no one has taught him or described what techniques to use. He started with pencil and paper, and by 18 he was painting with his fingers, first on paper, then on canvas with oils. Nowadays, he works primarily with fast-drying acrylics. After displaying his work at more than 20 exhibitions in Turkey, Holland, the Czech Republic and China, Armagan's disarmingly realistic work and his abilities have revolutionized our knowledge of how much congenitally blind people can understand about the layout of space. Dr. John Kennedy, a psychologist and Director of Life Sciences at the University
of Toronto, researches the psychology of perception and cognition in both sighted and blind people. He put Armagan through a battery of tests in which he successfully drew a series of solid objects, including a cube, in three-point perspective. Further tests, at Harvard University's Neuroscience laboratory, tested Armagan while drawing and revealed that as he drew his visual cortex wasn't lying dormant - it had been recruited by his other senses and lit up as though he was seeing. For the ultimate challenge, Dr. Kennedy takes Armagan to Italy to recreate Brunelleschi's perspective masterpiece - the Baptistery in Florence.

Portent: For centuries, it was held that the brain was a fixed entity and hard-wired for each independent function, incapable of adapting itself after injury. Armagan's story has revealed that the brain has the potential to adapt and rewire itself according to individual needs. The brain's ability to reorganize its functions based on new information and experiences is defined as neural plasticity. Dozens of medical therapies have been developed as a result of breakthroughs in thinking about plasticity - specifically strokes, autism, schizophrenia, spinal cord injuries, epilepsy, chronic pain and many other previously "untreatable" conditions. The next steps lie in learning enough about plasticity to harness it for individual needs. But in the next stage of our evolution, when we have the technology to program human life and fix any identifiable defects, would people like Armagan be eliminated before they are even born? While most people acknowledge the differences between coercive and elective forms of eugenics, will there be room in this future world for those of us who are considered "genetically unfit?" In this vision of the future, we must look not only at what we will gain, but also at what might be lost.

[cybereality]

Alright, this is all interesting stuff but at the end of the day you are not getting a stereo image without two video streams. Stereo, by its definition, requires two views. Anything less is psuedo-scopic and basically junk-science. The synopter doesn't really make sense and any effect it might have is likely to be marginal at best. Those 3D TV-Eyes glasses were straight-up bunk. I said they had *some* effect, but it wasn't 3D. My guess is that the way the holes where lined up caused a horizontal parallax shift, such that fast horizontal motion will appear more 3D similar to the Pulfrich effect. But really, you might as well just get a pair of X-Ray Specs!

Ok I read more on isee3d patent. The picture showing the wood grain in parallel versus toe in speaks more to me than 1000 words. Cyber or anyone, the 2 virtual cameras in the nvidia driver or iz3d driver - are they doing the right toe in as you focus on far away versus close object? Or are they simply parallel too? Is this what autofocus feature help with in iz3d driver? This is something that needs to be addressed by neil maybe with new s3dga standards?

Yes, that image was a very nice illustration. From my understanding the stereo drivers give you the option to control this (which would be what the convergence setting is doing). So to achieve a similar result you would need to set separation extremely low and adjust the convergence to match the object of focus (which is how I always configure my settings). I am still confused how this can be achieved with a dual-camera rig but it seems simple enough to implement with virtual 3d cameras.

[martinlandau]

Alright, this is all interesting stuff but at the end of the day you are not getting a stereo image without two video streams.

Did you try wearing those pinhole glasses with your polarized glasses on the zalman? Maybe using both makes things better.


" Stereo, by its definition, requires two views. Anything less is psuedo-scopic and basically junk-science."

There you go again Cyber with that negativity, people used to tell nikolai tesla to stop with the silly AC stuff, it would never take off and was junk next to DC transmission.


" The synopter doesn't really make sense and any effect it might have it likely to be marginal at best."

A couple marginal increases here, and a couple more marginal increases there, and pretty soon your are talking about something really sweet!


"But really, you might as well just get a pair of X-Ray Specs!"

Really, if god meant for people to fly, they would evolve wings Mr. Wright!

"Yes, that image was a very nice illustration. From my understanding the stereo drivers give you the option to control this"

So as I focus on far away mountain, and then on sword in my hand, the virtual cameras in the game adjust toe in? I thought the convergance setting on nvidia was set - parallel - as the pdf at the link you posted says is bad and makes even marines get nauseated and sick.

"separation extremely low and adjust the convergence to match the object of focus (which is how I always configure my settings)."

I can't imagine you sitting playing oblvion, and look way out to the mountains in the distance, and adjust convergance to get right toe in, then you look down at the sword at your feet and adjust convergance again to adjust toe in for close in focus, need something to do that automatically.

" I am still confused how this can be achieved with a dual-camera rig but it seems simple enough to implement with virtual 3d cameras"

The article you linked, the PDF file there says IMAX uses special motors on thier cameras to adjust toe in, converging the cameras dynamically as the focus goes from a few centimeters out to about 10 meters or so, then after that everything is at infinity anyways. This is not practical for a surgeon wanting 3d with dynamic adjustment of toe in.

[martinlandau]

Holy Smokes, ok I went to spencer gifts today at my local mall, that shop with all the crazy goth and gag stuff. THey had some pinhole glasses, colored white. 12 bucks, went home and with my laptop lcd, 2 feet back and this video http://www.youtube.com/watch?v=DGfM052xJms" onclick="window.open(this.href);return false; Something sweet is happening. Take the glasses off, flat, put the glasses on, psuedo3d that is much better than I thought it would be, holy smokes. My brain is being tricked. I see what you mentioned Cyber. Lots of motion in the scene makes the effect work better like in that roller coaster sim. Furthermore, something is happening with my eyeballs and maybe someone can help me understand. The first time ever I wore eyescream stereo3d glasses, I started getting headache and eyestrain very quickly playing tomb raider. This does not happen for me with the iz3d polarizd solution or the reald glasses at the theater. Well I used these pinhole glasses on some static images, nothing happened and all was A Ok. But then I fired up that roller coaster simulator and wore the pinhole glasses and within about 10-20 seconds that irriation/stretching/burning strain that happened the first time I wore eyescream is happening now, it is making my eyes do something new and not normal. The pain is on both upper sides of the frontal part of my head. I have had to take off the pinhole glasses now because bad headache is starting, but I am excited, very good psuedo s3d effect for the 20 seconds I could wear it. I am gonna take some aspirin and try some other videos and see what happens, maybe some people have better effect with this than others.

[cybereality]

Woah! Did you time-travel into the past to get those glasses? The time-stamp on that picture says 2006!!!!

In regard to the other comments: I am not saying that any of the things discussed would have no effect. Only that the effect would not truly be stereoscopic and thus not really 3D. I mean, if I told you I had a pair of glasses that would make a black & white TV look color would you take me seriously?

And just FYI: I did try using the pin-hole glasses with my VR920 HMD, it didn't work.

[martinlandau]

Woah! Did you time-travel into the past to get those glasses? The time-stamp on that picture says 2006!!!!

Darn, you are onto me, now everyone will find out there are time travellers amongst us! Who in a million years would ever believe kris krisstofferson and cheryl ladd of charlies angels would make a time travel movie!?

In regard to the other comments: I am not saying that any of the things discussed would have no effect. Only that the effect would not truly be stereoscopic and thus not really 3D.

The poopooing you did of the technology I feel was not fair, it does help, much better than I thought, with further refinements, it might help a lot more. Whatever is causing the 3d effect, needs to be studied and refined to make it work better. You are making vast assumptions Cyber, the top neuroscientists in the world do not fully understand the brain or the visual system, as the discovery channel show I linked to above proves, the top researchers are still being astonished by what they find. You do not know how all the neural receptors and molecules and atoms and such work together in our quantum minds. Your whole mindset really confuses me, you have the inspiration to buy these glasses and test yourself, but then kill whatever enthusiasm you may generate in others with cruel ridicule. There are a lot of good ideas out there needing to be discovered, and anything that delays that hurts us all.


" I mean, if I told you I had a pair of glasses that would make a black & white TV look color would you take me seriously? "

Well see, it's different for me, I am a time traveller from the future, and I know everything to know about the human brain and visual cortex and how physics and dark matter and string theory and quantum fluctuations all unify, so I would have to take you seriously on those glasses.

And just FYI: I did try using the pin-hole glasses with my VR920 HMD, it didn't wor

But your pinhole glasses and brain are not the same as mine, IDIC - what does it mean?

[martinlandau]

Cyber, at IBM where I used to work, there was a researcher who was told by every scientist on the planet that if the silicon chip got wet, it wouldn't do what it did when he got it wet. We all live a better life today with a higher quality and standard of living because of his "dumb luck" discovery that everyone else had stopped researching because it couldn't work that way - everyone knew that was stupidity - do you know who it was?

[mediavr]

cyberreality says:
"The synopter doesn't really make sense and any effect it might have is likely to be marginal at best. "

Here is a popular article on recent research work with the synopter and how it significantly increases the accuracy with which people can judge depth in pictures (2D pictures).

http://www.tudelft.nl/live/ServeBinary? ... Koende.pdf" onclick="window.open(this.href);return false;
If people can see depth more clearly with the synopter maybe they are actually seeing solid depth -- the sensation we normally associate just with parallax stereopsis.

Think of an old fashioned illuminated slide viewer, how bright colorful slides looked kind of 3d in it, and sometimes the reds came forward. If you ever worked with large format enlargers with big condensor sets and used the condensors as a giant magnifer on a light box with a bit transparency -- that looks even more 3d. Monocular stereopsis is a real phenomenon and optical arrangements can make it compelling. The relevance for regular 3d is that the same factors that lead to monocular stereopsis (removal of stereo information of flatness of the screen), ie. removing the frame somehow, eliminating convergence cues to flatness and closeness, removing constancy of parallax (with microwiggles perhaps like 3V) -- can be used sometimes to enhance regular 3d display.

Peter M

[mediavr]

cyberreality says:
"The synopter doesn't really make sense and any effect it might have is likely to be marginal at best. "

Here is a popular article on recent research work with the synopter and how it significantly increases the accuracy with which people can judge depth in pictures (2D pictures).

http://www.tudelft.nl/live/ServeBinary? ... Koende.pdf" onclick="window.open(this.href);return false;
If people can see depth more clearly with the synopter maybe they are actually seeing solid depth -- the sensation we normally associate just with parallax stereopsis.

Think of an old fashioned illuminated slide viewer, how bright colorful slides looked kind of 3d in it, and sometimes the reds came forward. If you ever worked with large format enlargers with big condensor sets and used the condensors as a giant magnifer on a light box with a bit transparency -- that looks even more 3d. Monocular stereopsis is a real phenomenon and optical arrangements can make it compelling. The relevance for regular 3d is that the same factors that lead to monocular stereopsis (removal of stereo information of flatness of the screen), ie. removing the frame somehow, eliminating convergence cues to flatness and closeness, removing constancy of parallax (with microwiggles perhaps like 3V) -- can be used sometimes to enhance regular 3d display.

Peter M

or if you prefer here is an audiovisual account of the synopter -- section 16
http://archive.nmc.org/content/visual_l ... ze/bendis/" onclick="window.open(this.href);return false;

[cybereality]

Guys, if you really love this synopter so much, here is a solution: Buy one of the many HMDs which no longer have any support from Nvidia. I'm sure you can find one cheap. Then load up your favorite game, and marvel at how amazing it looks in 2D. You would be getting identical images in both eyes, so it would be a modern electronic synopter. Let me know how that works out.

[mediavr]

Guys, if you really love this synopter so much, here is a solution: Buy one of the many HMDs which no longer have any support from Nvidia. I'm sure you can find one cheap. Then load up your favorite game, and marvel at how amazing it looks in 2D. You would be getting identical images in both eyes, so it would be a modern electronic synopter. Let me know how that works out.

Or you could buy one of the old fashioned sort from Phantascope
http://www.phantascope.co.uk/pages/appl ... opter.html" onclick="window.open(this.href);return false;
... they call it a cyclopter -- just to be different I guess --
but it a bit expensive -- like 500 pounds

Peter M

[mediavr]

re: the synopter -- which was mentioned also in recent news here --
the article in the New Scientist discussing the new version by English psychologist Rob Black
which prompted the news reports is now online
http://www.newscientist.com/article/dn1 ... aches.html" onclick="window.open(this.href);return false;

PeterM