Hacker News Reader

Comment by cushychicken | Hacker News Reader

Could likely compensate that with digital EQ, though.

Bass frequencies are inherently impossible with that kind of physical design. To move enough air at low frequencies, you need more physical depth (higher amplitude).

So what? The peaking is at 10kHz. Way outside of bass range.

Wouldn't the accumulated effect be viable though? I mean, air molecules are tiny, but there are lots of them vibrating in a coordinated way to make bass..

Just compare the surface area and displacement of the actual tiny domes with the surface area and displacement of a bass-capable loudspeaker membrane. If you do the math, I don’t think there’s any chance.

Could you get a balanced sound if you cap the volume? Like maybe the domes can't do 60dB of bass, but what about 20?

Maybe, but then you lose the power efficiency benefit of this particular technology.

True, but if the frequency response already drops off that heavily at 1kHz you're going to need some very serious amplification to use it for anything other than speech, even after applying DSP - which I'm guessing defeats the whole purpose of the speaker in the first place.

20db is a lot of compensate (most consumer grade EQ are limited to 12db in either direction, and human ears are more sensitive at higher frequencies). To work with most existing amplifiers, there would likely have to be an in-line circuit to attenuate the voltage at the higher frequencies.

This comment is almost totally incorrect.

12dB in either direction was standard issue in, like, the 70s. Plenty of modern eqs can do almost 40dB of damping.

Human ears are, in fact, way less sensitive at higher frequencies. Review the Fletcher Munson curve. Takes an order of magnitude more power to generate same perceived loudness at 10kHz as 1kHz.

This is a piezo speaker, which typically have really different driving requirements than a typical transducer, so it's probably not suited to work with most existing amplifiers anyway. Getting the EQ right is table stakes for getting this production ready.

Source: me, I have about a decade of experience in consumer audio.

If, by "an order of magnitude", you mean "10db", which works out to be a bit more than 3x the sound pressure required. For comparison's sake, I consider an order of magnitude to be around 10x or more.

Exaggerations aside, I could very well be wrong about that part - I was basing my observation off the response curve of reference headphones, which all fall off drastically starting a bit before 10kHz. But since they sound the same throughout the frequencies, I inferred that hearing is more sensitive slightly before 10kHz, since even age-degraded hearing can go above the 14kHz range.

> 12dB in either direction was standard issue in, like, the 70s.

Look at the EQ provided with your phone's audio application. 12 dB plus and minus. Look at the EQ on your computer. Unless you're using custom software, it's more-than-likely going to be 12db (Spotify, iTunes, WMP, etc.).

40dB of dampening is custom hardware your average sound bar won't have. That your average consumer-grade home theater system won't have. That even a quality headphone amp/dac won't offer.

> Getting the EQ right is table stakes for getting this production ready.

Here we agree, as I pointed out in the post you're responding to.

Sound : https://en.wikipedia.org/wiki/Sound

The Decibel : https://en.wikipedia.org/wiki/Decibel

The Decibel Watt aka dBW : https://en.wikipedia.org/wiki/Decibel_watt

Sound Pressure : https://en.wikipedia.org/wiki/Sound_pressure

Sound Level Meter aka Sound Pressure Level Meter (SPL) : https://en.wikipedia.org/wiki/Sound_level_meter

>the Fletcher Munson curve< : https://news.ycombinator.com/item?id=31215503

Equal-Loudness Contour : https://en.wikipedia.org/wiki/Equal-loudness_contour

Dear Mods,

If you are going to break the Context tree, please include the replied reference post;

ie. >the Fletcher Munson curve< : > https://news.ycombinator.com/item?id=31215503 < Added reference. Thanks

Reply on news.ycombinator.com