3 ms·
> A tiny box with tiny drivers and not a lot of power simply can't reproduce lower frequencies well. I'd like to expand on this. The tradeoff has three variab
by dietrichepp 13y ago
> A tiny box with tiny drivers and not a lot of power simply can't reproduce lower frequencies well.
I'd like to expand on this. The tradeoff has three variables: enclosure size, amplifier power output, and low frequency response. The reason is because as the audio frequency goes below the resonant frequency, the speaker efficiency drops (and eventually you run into Xmax limitations too). Resonant frequency scales with enclosure size.
Low frequency noise sounds good so a lot of speakers have deep resonances that exaggerate a specific low frequency at the expense of accuracy. Cheap, and impressive in the showroom, as long as you don't listen too carefully.
One speaker designer took the tradeoff in a completely different direction: mount a 10" speaker in a small cube, and use a very high-power amplifier to drive it below its resonant frequency. The speaker had to be specially designed with extremely high excursion (Xmax) and a voice coil that won't melt. The amplifer had to be specially designed to absorb a powerful back EMF as the moving speaker sent current back to the amplifier.
Apparently, the result was a subwoofer that satisfied not only the designer but his wife, who didn't want more "furniture".
- raverbashing 13y agoInteresting On the premise of Bose equipment being "good enough" it should be measurable (in real units, not "audiophile unicorn units" Makes me wonder if it would be possible to have a better bass response exploiting non-linear effects and doing some "magic" (that is, some heavy signal processing and feedback loop including a mic)
- beloch 13y agoWith the right room you could probably take advantage of modes to boost the bass response of smaller speakers at specific frequencies but, ultimately, to achieve a reasonably flat frequency response down to 20-30 Hz you're simply going to need the ability to move some air. There's no getting around that. What you're suggesting is a bit like trying to turn a 4-banger into a V10 by tweaking the spark-plug and fuel injection parameters.
- dietrichepp 13y ago> Makes me wonder if it would be possible to have a better bass response exploiting non-linear effects and doing some "magic" (that is, some heavy signal processing and feedback loop including a mic) Non-linear effects of what? Speakers? Air? What would the microphone do? (Keep in mind that sound has to travel through air, which takes time, before it reaches the mic.) You can already get better bass response by equalizing the signal fed to the amplifier, but the trade-off here is that it requires a bigger amplifier (and a speaker with higher Xmax).
- raverbashing 13y ago"Non-linear effects of what? Speakers? Air?" Air. Like this http://en.wikipedia.org/wiki/Sound_from_ultrasound http://en.wikipedia.org/wiki/Sound_from_ultrasound (yes, the article is not very good)
- dietrichepp 13y agoThat requires a large amount of power, so you'd still have a big amplifier.
- raverbashing 13y agoWell, not really. See how big automotive amplifiers are for a given power. What makes home appliances big is usually a huge transformer inside (which for audio quality is good, even though switched power sources are getting better) AND the feeling that light/small is not good A modern consumer audio system is around 80% empty space inside.
- stcredzero 13y agoHow about using the little microphones on a Google Glass like device to provide information for negative feedback distortion cancellation? That would work better for lower frequencies. For higher frequencies, have a program adjust the audio signal in response to the user's position in the room.