Join 3,432 readers in helping fund MetaFilter (Hide)


Equalizer Settings
August 11, 2005 4:24 PM   Subscribe

Equalizer settings on audio devices: essential or pointless?

Whenever I start playing around with the EQ settings on any of my media players, I'm invariably unhappy with my new settings -or- my original setting. Is there a general setting I can use that rises above the rest? Or, should I just leave it flat? Does it depend on the genre? Is it just personal preference?
posted by jgee to Technology (16 answers total) 1 user marked this as a favorite
 
well, the point of an equalizer is to give you a flat response across the frequency spectrum.

say you feed a single pure frequency to a particular speaker and then sweep the pitch from the lowest possible to the highest possible frequency. if you were to measure the loudness of the sound produced by the signal, and then make a graph of the loudness vs the frequency, you would obtain what's called a response curve. for a big speaker, your response curve would peak in the low frequencies and fall off towards zero in the high frequencies. vice versa for a small tweeter.

speakers aren't the only things that can be characterized by response curves either. amplifiers, microphones - pretty much everything that comes between the original sound source and your eardrum (which by the way, also has a response curve!) in general, the sound you hear is the sound produced, attenuated by the response curve of the first element, which is then attenuated by the second element, etc. the whole system from source to your ear can be characterized by a single response curve which is the product of all the individual response curves.

the idea of a graphic equalizer is that, if one knows the total attenuation of their system, they can compensate for this by boosting certain frequencies and attenuating others. if you take the response curve of your system, flip it vertically, and then make your graphic equalizer look like that, you should have a fairly accurate reproduction of the original sound.

that said, making a measurement of this kind of thing is only something a truly geeky audiophile would do. most of this is going to be dominated by your speakers -- even the crappy amp in your $20 discman has a pretty flat response compared to even some high-end headphones, for instance.

so probably the short answer for most people in most situations is: if you have small speakers, boost the bass range. if you have a nice stereo with lots of different sizes, leave it mostly flat. if all you have are big booming speakers, boost the treble end.

the even shorter answer is, yes, do whatever sounds good to you. (:

/nerd out
posted by sergeant sandwich at 4:47 PM on August 11, 2005


I think it's all about personal preference and the type of music you happen to be playing at the moment. For example, if I'm playing mostly talk radio that comes in on a very loud station, I turn down the midtones and treble, since, to me, the voices sound less tinny this way, and are more pleasant to listen to.

If I'm listening to the first track of Daft Punk's "Discovery," then I turn the bass all the way up, mid almost all the way up, and push the treble down, since I like the song better that way.
posted by odinsdream at 4:48 PM on August 11, 2005


pointless.
posted by Nelson at 4:56 PM on August 11, 2005


I find it essential on portable MP3 players, masks the lousy midrage pretty well.
posted by dong_resin at 5:11 PM on August 11, 2005


For exceptional circumstances (ie. bud earphones, strange sounding or bad quality source audio) it's useful.

For every day stereo use, it's not. To use it like it's supposed to be used, you need to blast some whitenoise into the room, use a spectrum analyzer to determine the frequency response, and adjust your EQ accordingly. Mind you, the 5-band EQ on most home stereos is pointless for this exercise anyway - it's not going to help you get rid of annoying resonances and standing waves.
posted by Jimbob at 5:22 PM on August 11, 2005


Not pointless.

Another thing that seems like it's most difficult for home audio or computer grade speakers to do is to appropriately output midtones, since they're usually just harsh piezo tweeters, a tiny pair of mids and a subwoofer.

Bass is easy, it's forgiving, it doesn't generally have a lot of detail - at least, not of the sort that most people would notice.

Treble is way too easy - at least to get loud, sufficient quantities of it. Getting excellent quality is a different story, but for most people they won't even notice bad treble unless it's too loud, or incredbily noisy, crackly, or "splashy"- where the treble is distorted to the point of noticing that it's even there in the first place. Treble needs to be "dry", like a good vodka martini. It should be able to handle lots of detail and complex sounds without balking, and it shouldn't be overbearing. In fact, you shouldn't even really "hear" it at all, it just needs to be there, unassuming, like the right amount of icing atop a cake.

But mids are where most of the actual listening tends to go on for most music and films.

Sergeant Sandwich's response is spot on and excellent.

But what to do without a spectrum analyzer?

Leaving it "flat" is an appropriate response. So is fine-tuning it to wherever sounds pleasant to you.

I like tuning down my bass a little, tuning down the treble a little, and bringing up my mids a little. This helps compensate some for crappy mp3s, and tiny midrange speakers in my 2.1 desktop speakers. My EQs generally look a little like the curve of the top of a loaf of bread. Just a subtle curve from the lowered bass up to the mids and back down to the highs. The mids are generally at "0" or just a notch above.

Proper EQ tuning for studio, reference, or live PA systems gets a hell of a lot more complicated then that and either requires an active reference tone and spectrum analyzer, or known "profiles" of the speakers and gear in the mix. Or just a good ear. You'll sometimes see some pretty wild EQ settings in these setups, with specific frequencies notched out of the mix or punched back into it.
posted by loquacious at 5:25 PM on August 11, 2005 [1 favorite]


There are certain audio sources with characteristic distortions that can be improved with an equalizer. For example, old time radio recordings usually have a lot of roll-off in the high frequencies, which can make speech harder to understand.
posted by Galvatron at 5:27 PM on August 11, 2005


First of all, this technical detail is really interesting!

My non-technical two cents: I basically don't touch the equalizer except on my travel MP3 player. I like to get a little more bass and "punch" when I'm walking to work and back on my crappy little headphones. So I put in a little "smile" curve that raises the bass and treble and leaves the ones in the middle flat. You audiophiles can now proceed to mock me. :) But that's the only setting I ever use.

It may not be a great idea if you have an Ipod, though, since the equalizer in those seems to make things sound poopy. (distortion?)
posted by selfnoise at 5:33 PM on August 11, 2005


Thanks for the answers, everybody.
loquacious: why leave the mids, which are at the top of your curve, at or just above zero? Why lower everything rather than raise it?
posted by jgee at 5:33 PM on August 11, 2005


Someone correct me if I'm wrong, but I believe the reason why it's better to lower everything rather than raise it is that lowering usually attenuates the particular frequency range under adjustment, while raising will amplify it. Since amplification is more likely to introduce such things as distortion than attenuation, attenuation should theoretically produce a cleaner sound. That's not always true, but it's probably true enough that it's better to only lower equalizer settings and then adjust the overall volume accordingly. The only reason to raise a setting is if you can't lower the other settings enough. E.g., the settings go from +10 to -10, but you need the midrange to be 12 higher than the bass. Even after lowering the bass to -10, you'd need to raise the midrange to +2 to get the delta you need.

I think that's accurate, but if I'm wrong, the other alternative is to center everything around 0. So, in my example above, you'd instead set the bass to -6 and the midrange to +6.
posted by EatenByAGrue at 6:05 PM on August 11, 2005


We know nothing about your listening environment, speakers (headphones?), musical tastes, etc., so an authoritative direct answer isn't really possible. EQ ain't pointless, and it doesn't need to be rocket science, but it isn't simple. There are several kinds of EQ, and they can be applied at different points in the signal chain, and they are useful for different purposes. EQ can be digitally simulated or analog. And EQ algorithms and circuits differ in quality, interface, bandwidth, envelope, and gain/attenuation power, among other things. So there's no simple answer to the best way to EQ other than "what sounds good to you." But you can learn a lot about how EQ works. Among other things, no one has directly mentioned Fletcher-Munson curves here -- though Sergeant Sandwich gave a heads-up about your ears' response curves. You hear the same frequencies differently at different amplitudes, which is what the good old "loudness" or "contour" button on stereos of yore corrected for, by boosting lows and highs progressively as the amp was turned down. Nowadays good receivers give you various finely tuned room EQs, but it's the same idea. Digitally, there are aural exciters aplenty providing real-time EQ effects, and simulated parametric eqs, noise-reduction plugins, and the good old 5-10 band graphic in all the media players. Knowing more about EQ makes it more fun to experiment with. The answer to your query, however, lies solely within you.
posted by realcountrymusic at 6:38 PM on August 11, 2005


Audio-purist philosophy might be adverse to fondling with the equalizer (the phase shift issue...). It's "always" better to remove frequencies than to add them (the distortion/phase issue). In the end it's the ears that should decide.
posted by noius at 8:15 PM on August 11, 2005


It sounds like part of your problem is not knowing what you like, and a great way to develop your own pair of "golden ears" is to listen to some really great systems. Aurant sells some very nice stuff, and is nearby. Go in, act like you're thinking about buying a system, and ask to listen to some different setups.

Then, remembering what you liked about those systems, you may have some more luck getting the most out of your system.

I don't think of myself as a purist, but I also generally avoid playing with EQ's. There are many things you can do with and EQ to make the sound worse, and few that will actually make it sound better.
posted by joshuaconner at 10:42 PM on August 11, 2005


Previous AskMe thread.
posted by fuzz at 4:13 AM on August 12, 2005


Fantastic stuff, sergeant sandwich and locquacious... fascinating!
posted by Chunder at 5:45 AM on August 12, 2005


I find that I really need different settings when listening to spoken voice and to music.

For music, I almost never change the standard settings other than occasionally fiddling a little with the bass.

But when listening to speech (Car Talk, This American Life, Garrison Keillor, etc), it's much more audible with lowest and highest frequencies turned down somewhat and usually middle to upper frequencies turned up a little.

If voice is lower (Garrison), need more mid to mid-low settings, if higher (Sarah Vowell, David Sedaris), more mid to mid-upper settings.

YMM definitely V, since I'm not an audiophile, just someone who listens to NPR a lot.
posted by marsha56 at 1:59 AM on August 13, 2005


« Older Have any of you gotten an ADD2...   |  Removing dog odor from carpet.... Newer »
This thread is closed to new comments.