Fairly off-subject but I wonder if anyone's heard of something like this being built?
In the DSP world we toss around concepts like convolution, room responses and the like, usually in the context of getting natural reverb but always note the input could be convoluted by any other signal. Taking it backwards: what if the other signal was a sample of someone saying, "hello?" And if this wasn't DSP, but rather a surface was prepared to give this as a natural echo in response to a clap or say a gunshot? Find a nearly straight cliff somewhere, or perhaps a smooth, flat, fossil lake bed, and machine it in a slight sawtooth pattern such that "vertical edges" of the sawtooth were in fact all facing the clapper perfectly. For 44.1kHz reproduction you'd need a vertical edge every 7mm or 1/4", though 1/10 that fidelity would still be phone-quality. The vertical extent of each of the sawtooth-edges would correspond to the strength of that "sample." The edges from near to far would appear from the clapper's position to be the same angular height and width, and the edges wouldn't be flat but in fact be subtly curved so the entire edge faced the clapper. This would mean that energy reflected for the first sample and last would be the same energy (assuming they were meant to be the same magnitude) and that energy would be focused at the clapper. Result: drive out to Hello Point, clap, and hear the recording. (A distant relative of this concept: in my youth there was some highway in Minnesota with grooves that caused your car to play a few bars of music as it drove by. It might have been the state song or something. That was simple monophonic square waves, and was more akin to a record paper, but is similar in that a surface was machined to reproduce sound under certain circumstances.) -- dupswapdrop -- the music-dsp mailing list and website: subscription info, FAQ, source code archive, list archive, book reviews, dsp links http://music.columbia.edu/cmc/music-dsp http://music.columbia.edu/mailman/listinfo/music-dsp
