wjm (04/06/83)
I've been reading the comments about equalizers and equalization on the net
for the past week and will add the following general comments. An equalizer
is a very useful tool for improving the response of one's stereo system.
It is especially useful for "tuning" speakers to room response idiosyncracies.
Several caveats: 1. you need some sort of calibrated noise source, either a
test record (the Telarc Omnidisk has excellent 1/3
octave pink noise bands for this purpose) or a calibrated
noise generator - built in to some of the better ($500+)
equalizers
2. In order to avoid problems with standing waves, you
should use pink noise rather than pure tones.
3. You should locate the mike of your analyzer or your
ears in the location where you plan to do most of your
serious listening - since EQ settings will vary from
point to point in the room.
The best device to use for setting an EQ is the real-time
analyzer (although it can be done by ear, I strongly
recommend that it not be done that way). Again, several
of the better models have it built in.
As for specific EQ models, I'm quite happy with my Audio Control C-101
It is an octave band EQ, and real-time analyzer with a LED readout. It also
has a built-in pink-noise generator and comes with a calibrated mike. What
I especially like is that the L and R sliders for a particular band are next
to each other, rather than on opposite sides of the chassis. This makes EQ
in the generally un-directional bass region much easier.
I've noticed the effect of EQ sound sounding "brighter", and I suspect this
is due to the loss of highs in most recorded material - and the FCC edict
limiting FM stations to a maximum broadcast frequency of 15KHz. I'm hoping
that CD's will change this.
With my present speakers, I've found that I have to put a substantial cut
at 8KHz and a very large boost at 15KHz, which I suspect is not uncommon.
Of course, one should be careful that one's power amp and speakers can handle
the effects of such a boost. I'll repeat my recommendation to use speaker
fuses (in series with the hot terminal of the speaker) unless the speaker
comes with its own overload protection (as do some KEF models).
Bill Mitchell
Bell Labs - Whippany, NJ
(whuxk!wjm)jj (04/06/83)
I guess I have to put in my two cents (or scents, perhaps) about equalizers, speakers, and room accoustics. I've used equalizers for sound reinforcement systems often, mostly to correct for a bad (but forced) speaker location, or for shortcomings in the hall, but I've come to the conclusion that equalizers aren't that much use in the home, partly because they are tricky to set (especially by ear, there are some remarkable psychological processes going on), and they usually don't match the kind of frequency response alteration that is needed in most homes. Another method of changing the room's effect on the sound is to move the speakers, how and where depending on what you need (or want) to do. In most cases, a good set of speakers can be placed somewhere in the average room (although I will certainly agree that some shapes and entrance configurations defy this) that will have the effect that is needed. Equalizers, especially third octave eq's, have bands that are (while convenient to build) not centered where you want them, and usually not broad enough to match the aberation being removed. If the room has (reflection) peaks that are much sharper(at higher than midbass frequencies), and very distinct, I suspect the presence of too many hard surfaces. Octave eq's of course, have even more problems with center frequencies, unless you are very lucky. Parametric eq's, of course, will answer both of these problems, if you have the patience, time, and equipment available to set them properly. <There's nothing quite like a parametric eq that's set up completely wrong!> I do realize that some people, with very dead rooms, or with speakers that NEED equilization, will benefit from an equalizer, however, in general, the degradation from wildly changing phase <which maps to group delay, of course> response, added (and very strange forms of) IM distortion, especially on percussion and other impulsive signals, and potential misuse, seem to outweigh the benefits (which may be great, but are usually marginal). If you ARE using an equalizer, I can only echo the comments that have already appeared concerning bandlimited pink noise. (Or, just pink noise, if you have a real time analysis capability.) Any sufficiently correlated signal (and 1/3 octave pink noise is starting to qualify as correlated, at least in the sense that matters here.) will excite room modes, reflections, and so forth, and will likely distort the analysis to the point of near uselessness. Good luck. Comments and flames to /dev/null on your own machine, please. Good wishes and the like may be forwarded to: rabbit!jj