| Author | Message | ||
     ChrisC007 Rating: N/A |
Real time 2D->3D is a reality but it needs the most powerful computer yet known - the brain! Let me explain - several years ago there was a craze for "Magic Eye". Not everyone could see the 3D images, but those who could were impressed. I was pleased that I got the knack of viewing magic eye very quickly and I could pick up a picture and view it in 3D almost instantaneously. There are lots of clips and photos on the internet presented in side by side format that allow you to see in 3D using the same method, known as parallel viewing, making your eyes focus on a point beyond the screen to a point where the clips or photos overlap and become 3D. So what would happen if you used the same parallel viewing on a single 2D clip? Nothing? Well, believe it or not, no! The clip actually appears to be 3D. What happens is that when you focus beyond the screen, you get two out of focus images. The brain tries to reconstruct the image from the available information and in the process tries to reconstruct the 3D (as that is what the brain is used to seeing) and it does a pretty good job. I don't know if this will work for everyone, and dont expect to be able to watch a full movie like this but it is interesting to try out. Some scenes are much better than others as well, scenes with more information for the brain to process (eg. deep forest scenes in Predator and also some of the battle scenes in the Star Wars), quite bizarrely, seem to produce better 3D effect! If you couldn't get the hang of Magic Eye then I doubt that this will work for you. | ||
| Charles Rating: N/A |
ChrisC007, I think that the "3D effect" you see with your described technique is mostly due to a creative imagination. There's NO visual information in a single 2-D still image that the brain can convert into a stereoscopic image, regardless of how you look at it. The "Magic Eye" effect is totally different, because those images contain specific imbedded stereoscopic cues. If you are using this parallel viewing technique on a video clip during playback, you MAY be getting an occasional Pulfrich 3D effect as objects move horizontally across the screen (One eye sees the object slightly later than the other), but this couldn't produce a consistent, orderly stereoscopic image. | ||
| ChrisC007 Rating: N/A |
Charles, have you actually tried it? I appreciate it does sound a little far-fetched, but there is actually quite a lot of information in the 2D image/video - take this image for example: http://www2.ups.edu/campus/Cherry_Trees.jpg You can tell that the tree on the left is the closest even though it's a 2D flat image - you can tell that all the trees are closer than the path and that the path is nearer than the hedge - all this is information about the depth of the scene and the position of each element within that scene - the trees partially obscure the view of the path, therefore they are in front. Obviously it is much more than this, size, angle and relativety to other objects all combine to form quite accurate depth perception. For example, if you sit at your desk and close one eye, you are now only seeing in 2D. However, it is quite simple to reach out and pick something up off the desk, because your brain has enough information within the 2D view to be able to perceieve where the object is. It's not perfect, of course, as the brain can quite easily be fooled. Video does work better than still images, but this is nothing to do with Pulfrich effect. It is because motion/speed adds another clue as to the depth, this is not just horizontal movement either. I know that this is nothing to do with "Magic Eye". I only mentioned Magic Eye because this is how I discovered this strange phenomenon in the first place and also so that people who have mastered Magic Eye would know how to give it a try. | ||
| Charles Rating: N/A |
ChrisC007, Let me clarify my comments. I'm not saying that a 2-D image has no depth cues. As you have noted, there are many 2-D cues that the brain uses to determine relative depth (object size and position, overlapping and blocking, linear perspective, texture gradient, etc.). All of these cues are present in your "Cherry Tree" image. With video images, the depth cue motion parallax is also present, because near objects appear to move faster than more distant ones. However, all of these are 2-D depth cues which are in the image regardless of whether you are looking at it normally or with your parallel viewing method. Looking at the image with your eyes relaxed to produce two overlapping side-by-side images doesn't alter or enhance these cues. My original statement was that single 2-D images have no information that the brain can convert into STEREOSCOPIC images. Stereoscopic imagery (perceived actual spatial depth between objects) is the ONLY type of "true 3-D" that can mimic the brain's view of the real world. For this to be present, each eye must see a slightly different image with binocular horizontal parallax and accommodation differences between objects. Looking at a single image with relaxed eyes can't possibly produce this, unless that image has dual images hidden within it (the "Magic Eyes" method). Yes, I tried your method on several 2-D still images and videos and examined them for any increased depth effect, but couldn't see any. (I use the parallel viewing method frequently for scanning left-right image pairs, so I'm very familiar with how it works.) Since the resulting view is somewhat unnatural, the brain may be a little confused and come up with a "pseudo 3-D" impression, but it doesn't mimic real depth. | ||
| ChrisC007 Rating: N/A |
Charles, Clearly this method does not work for you. (I am not using identical images/videos in side by side format, merely a single image). I said it might not work for everyone. | ||
| Scott Warren Rating: N/A |
Chris007, What Charles is trying to say (and I agree with him) is that "Looking at a -standard- 2d picture via Parallel method" is no different than "Looking at a -standard- 2d picture NORMALLY", or even for that matter "Looking at a -standard- 2d picture with 1 eye". You don't get any NEW depth cues. You need a difference in the perspectives between each eye. You are also clearly using psuedo-scientific circular reasoning (read: BAD!) when talking about self-reported benefits. You want people to respect your discoveries, it had better be independently, objectively provable and repeatable. Otherwise, it's just more snake oil, and the 3D community doesn't need any more of that. Scott | ||
| Joe Dunfee Rating: N/A |
Let me share an experience that may relate to the discussion. I've been doing 3D computer design and drafting since 1989. Years ago, when I heard that the first Immersive video game was coming to south Florida, I immediately went to see it to try it out. I was somewhat disapointed. While there was certainly a benefit to the immersive stereoscopic helmet, I think the benefit was greatly diminished for me because of my past CAD experience. I was already used to looking at a 2D display and forming a 3D map of it in my brain, so the stereo display wasn't as much of a benefit for me. I think this is the same "3D image from 2D picture" that you are talking about. It may also be the same effect that occurs a trained architect looks at blueprints for a house, and knows what it will be like to stand in the entryway. But, the non-professional just sees a flat drawing, and cannot get a good feel for what it will be in real life. In addition to the above, there are certainly some photos that seem to "feel" very 3d to everyone. Very old b/w photo portrates, done with very short focal lengths seem to look more 3d because of the way some items are in focus or out of focus. Another 3d-from-2d-photo effect seems to occur when taking pictures of nature in fog. I've seen several forrest scenes like that. I think the fog gives an additional set of 3d clues that just make the image feel like it has more depth than a similar photo without the fog. Joe Dunfee Joe Dunfee | ||
| blueskull Rating: N/A |
ChrisC007 your method is nothing more than just messing up your vision by focusing beyond a 2D image and your brain taking the two UNFOCUSED images from your eyes and "TRYING" to make a 3D image(our brain naturally does this when trying to focus on anything in the "real" world) and by doing this...you MAY or MAY NOT get a psuedo 3D effect based on parts of the 2D image that are similar to eachother....very hard to explain in text but the sample picture of the trees you gave, you can see that the background has an pretty straight landscape, a FLAT SURFACE (the landscape on the left of the trees is similar to the right), which when focused beyond... your left eye may match up the left part of the landscape and the right part of the landscape which your right eye is viewing(all from the same 2d image)and because the left section of the landscape "LOOKS LIKE"(psuedo) the right section of the landscape, your brain will take the two psuedo landscape images and it will actually look pretty 3D(PSUEDO) but it is only an "effect" by your brain(drugs can do pretty gnarly things with your brain too). ROFL | ||
| Hornet Rating: N/A |
This efect had name PULFRICH PHENOMENA 3D stereoscopic are the moving parts of picture, or complete picture (when the camera is moving), the effect is no so bad, have tested with stereoscopic player ,shutterglasses and projector, and some movies are better than 2D. | ||
| bob Rating: N/A |
hehe, just read this.. and it's true! Of course you see a weird, doubled image, but scenes really seem to appear in 3d, even though the effect is lost when trying to focus on a object. I don't think this is anyway like the pulfrich effect, its just as ChrisC007 said: "The brain in reconstructing a feel of depth." Unfortunately there is no application for that kind of fake depth. | ||
| Charles Rating: N/A |
Any "depth" that appears with this viewing method is based 100% on the viewer's imagination, unless the brain incorrectly takes two different objects in the image (such as two different trees) and treats them as single left / right views of the same object. In that specific situation, you're getting a false stereoscopic effect. To repeat what has been stated in earlier postings above: There are absolutely NO added depth cues with the above method. | ||
| Milan Pollé (Lamerdeluxe) New member Username: Lamerdeluxe Post Number: 1 Registered: 5-2006 Rating: N/A |
I agree with ChrisC007's findings, I discovered a similar effect myself. One evening I was lying on the couch watching TV at a very strange angle. Only one of my eyes could properly see the TV, while the other's view was blocked by my nose. In effect my brain didn't have any proper depth information and therefore started to make up depth information based on experience and logic. Suddenly the images on the TV got depth similar to stereo 3D. The more depth information my brain could gather based on camera motion or perspective, the more convincing the depth effect was. Recently it occured to me that you shouldn't have to use a strange way of viewing the TV, you just need to convince your brain to create depth information even though it's obvious the TV's display is completely flat. You do this by simply concentrating on trying to see depth in the TV's image. At first this will be pretty hard and will probably give you a minor headache (which is usual when learning something completely new). But now I noticed I often automatically see depth in TV images, as well as computergames. The effect is quite cool and convincing, though obviously slightly less so than when using something like shutter glasses. Your brain is quite used to making up depth information when actual information is too sparse. People with a lazy eye still see depth and when you shut one eye, everything doesn't look completely flat all of a sudden. One funny thing I noticed was when watching the news, there was a news-related rectangular image at the wall behind the newsreader. Strangely it appeared much further away than the wall, then I realised my brain interpreted it as a window view  Also, once during a particularly boring concert I noticed I could focus on moving parts in my view only, non-moving things were blurred. Quite a strange sight, a sea of moving arms of violin players. That must be how a cat always sees things. It seems your brain can do different kinds of 'image processing', most people never use it though. |
