For a TV to recover hue information from the I/Q phase, it must have a zero phase reference. It also needs a reference for amplitude to recover the saturation information. So, the NTSC signal includes a short sample of this reference signal, known as the color burst, located on the 'back porch' of each horizontal line (the time between the end of the horizontal synchronization pulse and the end of the blanking pulse.) The color burst consists of a minimum of eight cycles of the unmodulated (fixed phase and amplitude) color subcarrier. The TV receiver has a "local oscillator" which it synchronizes to the color bursts and then uses as a reference for decoding the chrominance. By comparing the reference signal derived from color burst to the chrominance signal's amplitude and phase at a particular point in the raster scan, the device determines what chrominance to display at that point. Combining that with the amplitude of the luminance signal, the receiver calculates exactly what color to make the point, i.e. the point at the instantaneous position of the continuously scanning beam.So not only would there be no coherent color signal, there would also be no 'reference' color burst at the beginning of the scanline. So the TV would have no idea what color to make the dot.
My guess with the lack of color snow is that there isn't any coherent 3mhz signal in there. Just a bunch of random noise, so it looks like a black and white signal to the TV.
Interestingly, most modern TVs just show a blue field when they don't detect a signal. Kind of ruins the beginning of Neuromancer: "The sky was the color of television, tuned to a dead channel". Makes it sound like a bright, sunny, cloudless day. Rather then an oppressive Grey.
posted by delmoi at 10:11 PM on April 8, 2009 [12 favorites]