| Author | Message | ||
     Abraham Shpitz (Abraham) New member Username: Abraham Post Number: 1 Registered: 8-2008 Rating: N/A |
Hi ,I am looking for an explanation to the reason that the same 3d image (with negative parallax),has a different depth perception when viewed as a side by side parallel or in the page flipped mode; in the page flipped mode I obtain detail coming out in front of the monitor's screen, while in the parallel mode the same detail stays on the CRT's surface. For those interested I can send sample images. Please send E-mail to ashpitz@yahoo.com | ||
| Charles Arrants (Charles) Intermediate Member Username: Charles Post Number: 62 Registered: 5-2006 Rating: N/A |
Abraham: I've also noticed this phenomenon with 3-D parallel images viewed with the relaxed-eye method versus with shutter glasses. Apparently when the brain processes the stereoscopic imagery in relaxed-eye viewing, it sets foreground objects at zero parallax compared to the surrounding screen, because it interprets this as a more "natural" view. (Wth page-flipped images, the horizontal parallax is fixed externally, and the brain can't overide it.) This is an educated guess on my part. If you or anyone else can find a more researched explanation, please post it. If my reasoning is correct, foreground objects should also appear at zero parallax when right-eye-first stereoscopic paired images are viewed with the cross-eyed viewing method. (I've never been able to use that method successfully, so other readers of this forum would have to comment on that.) | ||
| Andrea (Heandrea) New member Username: Heandrea Post Number: 8 Registered: 5-2006 Rating: N/A |
Viewing a parallel 3d image with relaxed eye you cannot perceive any object popping out of the screen. Infact the two images are competely separated, so to align any object present in both images, your left eye must point to it on the left image and the right eye will point to it on the right image. So your eyes must converge less than they would do looking at the border of the screen and you perceive the object behind it. Instead, when you use anagliph or shutter glasses (or some HMD), the two images overlap and the objects with zero parallax make your eyes converge with the same angle as the screen; varying the parallax you can move the objects behind or in front of the screen. If you are interested, I post a message in this thread about a similar topic: http://www.3dphoto.net/forum/index.php?topic=1157.0 | ||
| Charles Arrants (Charles) Intermediate Member Username: Charles Post Number: 63 Registered: 5-2006 Rating: N/A |
Andrea: I just viewed several of my parallel-format images, and you're correct: The closest foreground objects appear to be just behind the screen, not quite at zero parallax. Your explanation that this is a consequence of the lack of ocular convergece with parallel viewing makes sense. But then with the cross-eyed viewing method, the eyes converge strongly, with the point of convergence well in front of the screen. So can you tell me if the field of depth in a cross-eyed image appears to be slightly in front of the screen as a result? | ||
| Charles Arrants (Charles) Intermediate Member Username: Charles Post Number: 64 Registered: 5-2006 Rating: N/A |
Andrea: Nevermind -- I finally managed to force my eyes to fuse cross-eyed images correctly, and discovered the answer for myself: The hyper-convergence places the entire image depth of field forward of the screen. | ||
| Abraham Shpitz (Abraham) New member Username: Abraham Post Number: 2 Registered: 8-2008 Rating: N/A |
Hi Charles and Heandrea My tentative (hopefully educated) explanation: In the page flipped mode ,both the left and right lmages are dragged by the software to the center of the screen until the perfectly overlap. This forces the right eye to the left and the left eye to the right until they coverge at the center of the screen. this is an equivalent of a virtual negative parallax that brings forward the whole stereo image ,so that the stereo window and the zero parallax detail of the image stay on the surface of the CRT (actually on the phosphore where the image is created) and negative parallax detail come out in front of the screen. In the parallel viewing mode the convergence of the eyes due to centralization of the image does not occur the only convergence of the eyes is due to detil with negative parallax,therefore the stereo window stays within the monitor and the negative parallax detail stays on the surface of the screen. I believe that the page flipped mode of viewing gives a better approximation to what happens in real life viewing; when we look forward at an object , or a scene both eyes converge towards the image and focus on it. This is what happnas in the page flipped mode on the center of the screen (with the stereo window and zero parallax detail). When we view 3D images that have no negative parallax, both modes of viewing give the same result with the stereo window at the surface of the screen and the rest of the image behind it. | ||
| Abraham Shpitz (Abraham) New member Username: Abraham Post Number: 3 Registered: 8-2008 Rating: N/A |
Hi Charles and Heandrea My tentative (hopefully educated) explanation: In the page flipped mode ,both the left and right lmages are dragged by the software to the center of the screen until they perfectly overlap. This forces the right eye to the left and the left eye to the right until they coverge at the center of the screen. this is an equivalent of a virtual negative parallax that brings forward the whole stereo image ,so that the stereo window and the zero parallax detail of the image stay on the surface of the CRT (actually on the phosphore where the image is created) and negative parallax detail come out in front of the screen. In the parallel viewing mode the convergence of the eyes due to centralization of the image does not occur the only convergence of the eyes is due to detil with negative parallax,therefore the stereo window stays within the monitor and the negative parallax detail stays on the surface of the screen. I believe that the page flipped mode of viewing gives a better approximation to what happens in real life viewing; when we look forward at an object , or a scene both eyes converge towards the image and focus on it. This is what happnas in the page flipped mode on the center of the screen (with the stereo window and zero parallax detail). When we view 3D images that have no negative parallax, both modes of viewing give the same result with the stereo window at the surface of the screen and the rest of the image behind it. |
