The world is full of simulators. That’s why simulation is definitely in my top 3 of most interesting things in Engineering.
An important insight which I had a few years back is that a lot of our world is based on simulations. Take a cell-phone, for example. The job of a cell-phone (in essence) is to simulate the complex, organic machinery of a human voice apparatus (which includes everything from the lung to the tongue) using nothing more than a vibrating piece of metal, which is its speaker.
Or take a photograph, which must simulate the incredibly complicated interaction of light and matter using nothing more than a few inks and paper.
These simulations are of such high quality that we do not stop, even, to think about them; for many people today, the distinction between live music and recorded music almost doesn’t exist, and blindfold, perhaps, many would not even be able to distinguish whether the source of music is a band in front of them or a set of good speakers.
But at the core of it, making simulation work is a task of engineering genius.
Take color, for example. We know the physics basis of color: that it is equivalent, somehow, to the wavelength of light. So the color ‘red’ could be nominally defined as light with a wavelength of 700 nanometers.
But if you see a bright red car, for example, the effect is very different from that of a monochromatic ray of red light. The car is a complex shape with many different paints and pigments on it. At every point in the car, a combination of the pigments used, and the surface on which the pigments are used, define a way by which that point on the car interacts with light that falls on it – absorbing, reflecting different wavelengths to different extents and in different directions. Each point, then, is an infinite tangle of parameters that govern how it interacts with light.
And the light that falls on it is another glorious tangle of parameters. It has a spectrum – defined by a nearly infinite number of wavelengths, each having a certain intensity. And this light may not even be uniform; for example, it may be a spotlight, that is bright in the center and dim on the sides, or sunlight, which is (for most practical purposes) uniform white.
This light falls on every point in the car and produces an ‘effect’. The result of this effect is that a now-modified light falls in our eyes, focused by our eye-lenses onto our retina, which ‘perceives’ it.
And in that perception lies the secret key that allows us to simulate the effect of all this complex intermingling of many materials, light and matter, with nothing more than a few inks on a photograph.
Consider this stunning photo, from a BBC article on Animal colors through animal eyes:
The left side is how we perceive it; the right side is how a bird’s view of it is enhanced. This is because our eyes have only 4 parameters for ‘perceiving’ light: the rods, that sense the brightness of the light, and three types of cones, that each sense the level of a different color.
And – this is the interesting part – a bird’s eyes have four types of cones instead of three: they can not only view what we can, but also light in the ultra-violet spectrum.
The fact that we have three types of cones ensures that all light we see – any spectrum, i.e. any mixture of wavelengths – eventually is sensed by us as merely a set of 3 color-intensities.
And that gives rise to that beautiful idea: that there may be two different spectra of light that stimulate the cones in the same way – and therefore look identical to us. Thus is born the whole field of ‘reproducing color’ – photography, painting, printing, displays, monitors, projectors.
And we are able to simulate the colors of the entire universe, really, with nothing more than 3 paints, or inks, or filters.
More tomorrow (this is the first of a 3[?]-part series of ruminations on simulation).