There was a seminar at lunchtime. The speaker talked about a new, portable, 3D imaging system they had made. People could draw up a design for a building, load into the computer, and the computer would project it onto the screens forming a cube. You could stand inside the cube and wander around the inside of the building using a hand-held gaming type console. You were given a special pair of 3D specs. After the seminar, several of us went for a tour around the real thing. I wondered from the presentation whether it was really 3D, i.e. stereoscopic vision, or whether it would be like mono-vision, basically like walking around a building with one eye shut. I didn't like to ask in case I looked stupid. Anyway, it turned out it was 3D. I wondered how they did this and thought they must have used polarised light so I asked the man. He said, what they did was to refresh the image about 120 times a second, alternating between the left eye vision and the right eye vision. The 3-D specs were synchronised with the screen refresh, so that right lens went opaque when the left eye image was projected, and vice versa. It strikes me that you could do 3-D television this way. Televisions have evolved with interlaced rasters. Twenty-five times a second the odd numbered lines are refreshed; twenty-five times a second the even numbered lines are refreshed. You could televise the left eye images on the odd lines and the right eye images on the right lines. It never occurred to me that you could do this. I remember old 3-D specs used to have red and green lenses, so I always thought in terms of wavelengths of light. Actually, that is another way of doing it. You could have the green pixels represent the right eye; red and blue pixels, the left eye. The raster system reminded me of my last job, which was developing code for television broadcast equipment. British and most European countries use a PAL system in which the screen in refreshed 50 times a second. The Americans used a different system, NTSC (iirc) which refreshed the screen approximately 29.997 times a seconds. It used to be 30 times a second, but they devised some strange way of including colour information in the additional split second. It used to make our job considerably more difficult. Computer and television technology was converging. I remember once Microsoft tried to persuade the broadcast industry that instead of using the raster system, everyone should switch over to progressive scanning (i.e. top left to bottom right), which was how computer screens are scanned. I thought that was rather cheeky of Microsoft, but actually, the image resulting from progressing scanning is crisper than interlaced rasters. Televisions grew up with rasters because the refresh rate was limited by the frequency of the A.C. supply, 50 Hz in Britain and Europe, 60 Hz in America. I think the reason they could not use progressive scanning from the start was that it took too long for each part of the screen to be refreshed, which would result in flicker. The tour around the studio (called the cave, or cube for the more portable system) reminded me of my first job after college. I worked as a graduate software engineer for a company that made aircraft simulators in the visuals department. Our visual systems used to render air ports. I flew around Hong Kong a few times. That was quite an expensive system, which I imagined at the time would become much cheaper as computer technology improved. There were two parts of the projection system from what I could remember. The main imagery was projected something resembling television screens, but runway lights were rendered as points of lights. This cube system was different in that it was stereoscopic (probably not that much of an advantage to be honest), and that CAD files could be loaded into the computer and rendered into a 3-D model that you could walk around. I had to say that the image quality was not brilliant. It was not exactly like walking around a real building, but that is largely due to the quality of the files. A more detailed, more lifelike data file would result in a more lifelike building to walk around. The speaker said that the system had already proved useful in spotting some 3-D clashes. In 2-D, some of pipework looked fine, but in 3-D it was found that several pipes occupied the same space. It also brought to attention that some important safety controls were not in reach.