A radio signal has no picture
A study for Distributed Spectrum on how to draw a radio frequency. Not a chart of one, a symbol for one: something that reads as signal at the size of an app icon and still holds up as a full-screen field. Twelve motifs, each built from the actual physics rather than from what a wave is supposed to look like.
Everyone draws the same wave, and it means nothing
Ask for a graphic that says "radio" and you get a sine wave, or the three arcs off a wifi icon. Both are fine as pictograms and neither carries any information. They are drawings of the idea of a signal, and once you have seen one you have seen all of them.
The brief was a company whose whole business is detecting and locating radio transmissions. A generic wave would have been a poor fit: the interesting thing about their work is not that signals exist, it is what a signal looks like once you measure it. So the study started from the instruments rather than from the icon.
The set
Each one is a real display, not a texture
None of the twelve is decorative. The radar sweep is a plan position indicator, the format every rotating aperture radar has used since the 1940s. The waterfall is the standard time-frequency display, time scrolling down while frequency runs across. The phasor is an IQ vector, the in-phase and quadrature decomposition that every software radio works in. The standing wave shows what happens when a line is not impedance matched and part of the power comes back at you.
Building them from the equations rather than from reference images had a practical payoff. Anything driven by real physics animates correctly for free. The standing wave oscillates in place instead of travelling because that is what the maths does. The phasor trail decays at the rate the integration period implies. Nothing had to be art-directed into looking plausible, which is usually where this kind of graphic falls apart.
A mark and a field are different problems
The set deliberately spans two ends. At one end are motifs that survive being shrunk: the radial pulse, the diamonds, the phasor. Concentric or radial, one clear centre, legible at sixteen pixels. At the other end are the ones that only work large, where the whole point is density: the ripple warp, the arc stripes, the waterfall. Those are fields, not marks. They fail as an icon and they are the ones you would want across a wall.
Keeping both in one study was the useful part. A single motif that tries to do both jobs ends up mediocre at each, and the decision about which end a given application needs is much easier to make when you can see the two extremes side by side.
What I take from it
The study was not adopted, which does not change what it was for. The method holds: when you need a symbol for something invisible, go to how the thing is actually measured. The instruments are already a design language, refined over decades by people who needed to read them quickly, and they carry meaning that an invented shape never will.