frequency range for instrument vs speaker

I have spectral analysis software on my laptop (the OmniMic from Parts Express), and a surprising number of recording show sub-30Hz energy, especially Telarc CDs. If you're into contemporary jazz I think the 20-40Hz octave is critical for really hearing what's on the recording.

Over the past year I've become a huge fan of subwoofers, and not because I want to blow myself out of the room with bass, but because you can sometimes get much smoother bass by placing one or more subs in appropriate locations than you can with just a pair of stereo speakers. It is really the smoothness of the in-room response between 20Hz and 100Hz that determines bass quality.

If you get an OmniMic or a similar set-up and measure your in-room response you might be very surprised at the number peaks and dips you see, often exceeding +/- 10db. Every room has different modes, and moving your speakers even an inch one way or the other can often make a big difference.

I tend to think similar to Drew_eckhardt.

A speaker that is flat and can perform without breakup nor compression at real life volumes best will tend to extend similarly below the frequency range where the lowest and most common fundamentals occur even if not much really occurs down there. Not only does this best cover pretty muchanything we might hear that is there but it also is an insurance policy that the speaker has the bandwidth to deliver as needed where it is most likely to matter.

For lower volume listening, at levels safer to the human ear, this will all tend to matter much less or not at all.

FWIW, Squeezebox touch has a frequency spectrum display mode that can be switched to by touching the display area where album art is displayed by default I believe. Dunno how accurate, but it appears to be somewhat useful to get an idea of what the frequency content of what is playing is. A nice bonus feature! Also has a VU meter display for those that miss those on our modern gear. Very pretty!

Other points about low frequency extension are that

1) It's not a brick-wall. Extension to 50Hz generally means that output is at least 3dB down at that point with a noticeable 1dB of drop up much higher like at 80Hz.

2) The effects are worse than the numbers would suggest due to the equal loudness curve spacing at low frequencies. Once you get to 50Hz at moderate listening levels 3dB is like 6dB down at 1KHz which is 1/4 the acoustic power and quite significant (a full 10dB gets you a perceptual halving of volume).

3) The -3dB extension does not imply that the speaker is actually flat to that point. Many small speakers have a bump in frequency response (a few dB) which boosts the harmonics of low frequency instruments and can either give the impression the missing frequencies are still there or make the speaker sound boomy depending on what they're trying to reproduce. This design choice is more likely in speakers with less low frequency extension.

Drew: I understood that speaker frequency response is measured in an anechoic chamber so that consistent comparisons can be made with regard to speaker response performance. In the real world, the speaker sits on a hard or soft floor, close to or far from walls and ceilings. These variables, plus furniture, curtains and rugs create colorations in the music and then we must consider the type of amplifier and cables attached to the speakers and how they respond to the impedence characteristics. So speaker designers can create an instrument that looks perfect on paper when operated in an ideal environment, but how it behaves in our homes can be an entirely different matter, right? For example, just placing speakers near corners in a room can reinforce the base or adding curtains on the wall can dampen down the highs. I also learned in the past that too much dampening in the room kills the music. I have my speakers positioned a good distance from any wall. My speakers sit on spikes on a wood floor but I have a good sized rug placed directly in front of them. I haven't made a detail spectrum measurement but I believe I am close to a flat response curve as I could possibly get. Not saying that it is flat, just that I have acheived a satisfactory balance. I think about all of the parameters that speaker designers must consider both acoustically and electronically and then the nearly infinite variables of peoples' listening enviroment, associated cables and electronic gear and I can see why we have so many types and styles of speakers.