In this section, we are sharing instructions to building more sophisticated glasses-free 3D displays, including those using multiple stacked liquid crystal displays (LCDs) or combinations of LCDs, lenses, diffusers, and polarizers. You should have a good understanding of the fundamental building blocks before you get started on this site. Below, we show a few of our prototypes – once you have your building blocks in place, you can build any of them as well. Please use the menu to browse through different projects; at the moment, we provide instructions to building static Layered 3D displays using inexpensive printed transparencies, dynamic LCD-based Polarization Fields displays, and also dynamic Tensor Displays. We have a few new projects for super-resolution display, displays supporting accommodation, and high dynamic range (HDR) displays in the pipeline and will update the blog as soon as those are published.
Using the instructions on fundamental display building blocks, you should be able to turn your computer monitor into the base-display required for all other compressive displays. The base display is basically just a monitor with a rigid frame attached that holds the rails on which you can mount all other components.
With the base display in place and an additional LCD (as described here), the easiest thing to build is a dual-layer Franken Display. Depending on the software you use to drive it, you can run this as a simple parallax barrier display or a Tensor Display. We have source code and software tools available that you can use to calibrate and drive it as a parallax barrier display. That is basically what the Nintendo 3DS is doing (congratulations, you built yourself a big 3DS already). Unfortunately, parallax barriers are very dim and also low resolution. The loss in resolution will be severe if you work with multi-view displays (as opposed to the 3DS, which is just stereo). We also provide source code for calibrating and running this piece of hardware as a Tensor Display, increasing resolution, brightness, dynamic range, and depth of field.
This is getting more interesting: a three-layer Franken Display. In glasses-free 3D mode, there is no conceptual way of using parallax barriers to drive this. Two possible solutions for a software driver are computed tomographic light field synthesis (as used in Layered 3D) or nonnegative light field tensor factorizations (as used in Tensor Displays). Source code is available for both. The depth of field (“3D pop-out” effect) and the field of view within which an observer gets a high-quality glasses-free 3D experience of the three-layer Franken Display are much higher than for the two-layer version.
We have also experimented with a combination of LCDs and lenslet arrays. This device can be used as a lenslet-based 3D display (source code for calibrating and running it is available) or as a light field probe for computational illumination (I used it for experiments in Schlieren photography and refraction object reconstruction).
This sandwich of two LCDs and a lenslet array is a little more exotic. I’ll let you know when I found a good application for it.