> but the article does not justify that in any way. In a way, they did. They said that the 1kHz tone was "high quality", which most likely means it reproduced

> but the article does not justify that in any way. In a way, they did. They said that the 1kHz tone was "high quality", which most likely means it reproduced the waveform they sent fairly accurately. Of course, it's not a complete answer, and I too would appreciate the actual paper. "When 25 volts of electricity were passed through the device at 1 kilohertz (a rate of 1,000 cycles per second), the speaker produced high-quality sound at conversational levels of 66 decibels." Not sure if it matters, but the fact that it's producing ~10x (2^3.333) the sound pressure (which is around 4x louder by human perception) at 10kHz vs 1kHz is vaguely concerning. It would require a fair bit of additional resistance to try and get a "flat" loudness across the spectrum, especially since human hearing is more sensitive at higher frequencies. Absolutely doable, but it means the speaker film can't just be used out of the box.