- Written by Ofer Laor
- Published on 19 July 2012
The Toshiba 55ZL2 3D without Glasses
I'll try and refrain from being overly technical. The idea behind Glasses-Free 3D is fairly simple and uses what we call the "lenticular" technique that you see in 3D postcards and bookmarks. A new layer is placed over the LCD panel called a "lenslet layer". This layer is made up of miniature lenses, which are placed very exactly so that a single lens is dedicated for each triplet of the 24,883,200 sub-pixels (which means that this layer contains approximately 8.3 million lenses). Each block of 3X3 lenses is identical and repeats over and over.
Each lens forms its own light column. That means that light from a specific pixel can be shown as long as you're in the space where it is active. If you move to the left or to the right of that viewing area, that pixel will appear to be turned off (in ideal conditions). Immediately after that pixel's view area, comes (with a small overlap) the pixel underneath it or next to it. This pattern is designed so that when one light field (or view) is active, its neighbors are not but when one ends, the next one immediately begins. This is repeated, until nine separate views are formed. Each view is approximately 6.3cm in width, and since eight usable views are made (the 9th one turns out to be unusable) you get 50.4cm for each viewing zone. The lenses are made so that the entire cycle is repeated - i.e., that very same pixel starts working again if you're exactly 50.4 cm away from the first point where you saw it.
Now, what is actually playing on each of these nine pixels? That's where the camera and face detection comes into the mix. The camera and appointed CEVO core determines which person's face is closest to the center of the screen. Since it has essentially two images to show that person, it will create them in such a way that the pixels pointing at them will show that picture. It will also try to optimize it so that adjoining pixels will also show that picture.
Once more people are detected; they'll have to move their heads to join in the fun. There are a few ways to adjust the tracking. The first tool is a picture-in-picture screen showing what the display is actually seeing from its point of view. There are a few green "slots" signifying the ideal positions for people to be seated. Faces in good spots will have dark blue rectangles in them (if you watch "Person of interest", this concept will be familiar to you.). People in areas that are "not good for 3D" are marked with incomplete light blue squares.
The next technique is location circles. When this feature is turned on, you can see three circles on the screen – on the left, center, and right top parts of the screen. These are very light and unobtrusive. I kept forgetting to turn them off during viewing and actually like when they are there. When you're seeing a good 3D image, you should only see 3 circles. However, if you're not in a good seating position, these will turn into arrows pointing in the direction where you should be located – forward/back (or angled), and left/right.
Pressing the tracking button (buried at the bottom of the remote) gets the system to rediscover where viewers are located. If people move, this is probably a good thing to try.
There is also a manual override option that lets you see a test pattern on screen and bypass the camera's detection. This is a cool feature, but probably less useable for normal people that move around.
Seating positions are critically important. Toshiba recommends seating 2.2m away from the TV and has very specific angular recommendations regarding vertical and horizontal angles where 3D works well. Go outside of these parameters and you will see blurry images rather than proper 3D.
Note that the 3D being produced is 720P per eye. I was actually expecting some dark areas (similar to what passive 3D displays produce – similar to CRT scan lines) but the image is complete – but the lower resolution and the equivalent of crosstalk are the main drawbacks here.
Crosstalk in active/passive 3D occurs when information from one eye appears inadvertently in the other. In this case, it occurs when one view "leaks" into another – causing your to see a fuzzy image or one comprised of several views.
3D Glasses-free is still far from perfect. The need to stay upright and steady really caught me off guard. I'm used to slouching on my couch and this is not the right TV to try and do that with. Some portions of the screen seemed too blurry to me and the 3D effect is much more subtle than I'm used to with active or passive 3D screens. Also, when using the 2D to 3D conversion, I felt that the display was way too subtle in the conversion – it took something quite dramatic (like an object against a far away background) to notice the 3D effect.
However, the effect is much more pronounced when viewing a Blu-ray than with Side-by-Side or converted content. Avatar 3D on Blu-ray, or Monsters vs. Aliens looked much better than converted content. The resolution loss was quite clear, though.
My feeling was that this technology is still rough around the edges and will require much more fine tuning before it can really become mainstream.