You could produce that deep bass if you knew exactly where the listeners are. Throwing some computer vision/ML in the mix could solve the problem.
PS: It can't be patented now, cause I mentioned it on HN first :)
PS: It can't be patented now, cause I mentioned it on HN first :)
I'm unsure what you are trying to say. The problem with deep bass, as far as I know, is not that it only works in certain places; it is that it can't do it at all. See the frequency response chart in the page I linked, which I suspect if anything is probably overselling its bass capabilities. As I said, I don't know if it's merely that nobody has cracked the problem, if it is somehow fundamentally impossible with this technique, or I'll add a third one here, if it's possible but impractical (i.e., "requires a 100x larger piezoelectric array" or something).
If you've really never heard one, they have a very characteristic tinny sound. Usually when you have a conventional tinny speaker, the speaker has a relatively narrow range and is missing both the low and the high ranges. With the ultrasonic speakers, the bass is entirely gone, but the response up through the higher frequencies is pristine pretty much all the way up the human range. So you get voices with no bass at all, but that are extremely, extremely clear and easy to understand because all the higher harmonics are intact, if not outright emphasized. Very distinctive sound, even if you don't notice the highly directional and sharp sound immediately the acoustic signature is unmistakable.
First of all, to produce a sound that is identically loud to the human ear, the sound wave pressure amplitude has to increase tremendously as you go down in frequency. At 40 Hz, which is not very deep bass, you're talking a 40 dB SPL increase necessary over 1000 Hz.
Furthermore, as the frequency drops, the sound waves become longer, meaning that with constructive interference, the volume of space in which you have to produce significant sound pressure increases with the inverse of frequency.
Taken together, this means that at 40 Hz, this type of speaker must output 16*25 = 400 times as much sound energy as at 1000 Hz. To reach deep bass territory of 20 Hz, you're looking at a 1600 times increase in the energy required.
That would be an interesting forum for geeks / problem solvers AND would kill stupid/obvious patents as a side effect, and act as a repository of prior art!
Since such discussions might articulate at least the broad ideas (even if they don’t get to a specific implementation) this might discourage at least overly broad patents.
PS: Did I just make this idea un-patentable by anyone else? :D