It depends on what sounds you want to bring out hte most. As a general rule, cut rather than boost (amplification stages in EQ sections are not known for their quality) and try not to do weird, vast changes with the EQ (as it pushes each section out of phase with the others, causing the sound to become vaguely muddy).
As for which band sounds like what, I'll refer back to an issue of TapeOp from a few months ago:
30 Hz is barely audible, but with a decent sound system can add a lot of "oomph" to the sound.
80 Hz is what Graham Hick refers to in the article as 'rumble', and 100 Hz is just plain muddy.
200-700 Hz is the warm, very audible bass area.
1 -5 kHz is the area humans hear best, and where most of speech occurs. A slight boost at 1 kHz will make the entire signal seem a little louder, while 3 kHz will bring out speech and generally make words a little clearer.
5 kHz is where much of the "ess" sound occurs. Cut it to get rid of hiss, or boost it to add to some words - but be careful, boosting too much will sound just plain bad, no two ways about it.
8 kHz is a high sound that sounds high, probably not necessary on many male voices and can be cut somewhat for most voices period. Anything too far above that can probably be cut altogether, though it's worth doing so with a lowpass filter rather than a graphic EQ if possible because that'll screw less with the phase of the signal.
autophage, that is a very good consise way of explaining it, equalizing is one of the things i find hardest to teach people because it relys more on your ear than on any technical rules, but i like your way of explaining it in a physical sense, i hope you dont mind if i print that and give it to new techies to help them understand what you do when you equalize.
I don't equalize for a specific type of program so much as for the room. Most of my work is punk rock concerts, but the same idea works for any live sound gig.
The first time a band has to sound-check with me, they think I'm out of my mind. I don't wait for feedback to happen and try to fix it. Instead, I intentionally cause feedback... during the sound check. After the show, they usually thank me - they could actually hear what they wanted from the monitors.
What I'm doing is finding the frequencies that the room itself reinforces, then notching those frequencies out with the graphic EQ.
I start with the EQ set flat, both for mains and for monitors, and set a good but fairly quiet main mix... channel faders set to 0dB, submasters set to 0dB, main faders set to about -15 dB, monitor master knobs (my board has uncalibrated rotary knobs for monitors) just above 50%. I use the input gain knobs to set an initial level for each channel (faders are for tweaking relative levels to sweeten the basic mix as necessary), then use the channel-strip monitor sends (and requests from the musicians), to build good, but somewhat quiet, monitor mixes.
Then I'll start increasing the overall level on one of the monitor master knobs until it just barely starts to feed back. I'll make a best guess as to the frequency of the feedback and pull down the fader for that frequency on the graphic for that monitor mix. If that kills the feedback, I guessed right. If not, I put the fader back up and try the ones on each side of it, one at a time. If necessary, I keep spreading out from my original guess until I hit the one that stops the feedback.
Then I go back to the gain and push it up a little farther, until it starts feeding back at another frequency. Again I'll make a best guess, with the advantage that I know whether it's higher or lower than the one I just got done with. I repeat the procedure until I've got four or five frequencies notched out.
Then I go back to the first frequency and push it up, slowly, until it just barely starts to feed back, then pull it back down the tiniest fraction, until the feedback stops. Repeat for each frequency I notched out originally, then back the monitor master down by about 10%.
Then repeat the whole procedure for the next monitor mix (I can run up to four, but usually only do two... and that's often better than most of the bands I work with get from anybody else).
Finally I do the same with the main mix. By causing feedback during the sound check, I've found the frequencies where feedback is likely to occur and cut them down in the mix, both for the mains and the monitors. The result is that I can push the system, especially the monitors, a lot harder without causing feedback during the show. It works well enough I've actually had the occasional musician ask me to turn the monitors down - something you almost never hear a musician say.
While it may be time consuming at first, with practice you get to the point where your first guess at the feedback frequency is usually right and it goes pretty quickly. And not only is it louder, it sounds better, because you've compensated for the room accoustics.
thats an interesting process, unfortunately if i do a concert i have at least 8 bands and not enough time to do that even quickly. anyway even if i did they would look at me like i was crazy and probably walk of stage. anyway ill try that next time i have a band thats willing to wait knowing that they will benefit from it (probably my brother's)
DMXtools, I like your method but it only corrects for feedback rather than what the mix sounds like - although both are of equal concern.
Another piece of advice is to play with lots of EQ units other than the one you usually use - you'll get to know quickly which bands feel good to you personally, which bands your board's onboard equalizer doesn't isolate well, that sort of thing.
And remember that it varies based on what room you're in, not just for feedback but for everything. I have to add a LOT of 2 kHz to the stereo in my room, for instance, for it to approach sounding good - I'd suppose that's the phase cancellation between left and right - but that issue moves down to about 900 Hz in my living room (which is, sure enough, about twice the size of my room).
autophage... you have an eq on the stereo in your room! thats sooo cool... i just use the eq controls in my media player software, but i usually dont even do that... dont really need it for pre-mixed songs
For EQ'ing a room, a good way to go about it is to take a cd that you've listened to ALOT. You should know the sound of the disc inside out. Put that in and stand in the middle of your room, with the volume at about concert level (maybe quieter to save your ears). Play around with the frequencies till it sounds right to you.
But only cut the frequencies you don't want, don't boost.
For theatre gigs try and find a cd that is very vocal oriented (I'll even reccomend some rap for this). For other types of gigs (rock, punk, symphony etc..) pick a cd wtih the corresponding genre.
You need to
1) Adjust for the speakers and room. If you hear any peaks in frequencies, take them out so they match the rest, thereby removing any frequency's that drown out others.
2) EQ to the mics. This usually involves feedback removal, the rest can usually be done from the desk. Bring up the level on the mic you want to eq untill you hear feedback. Take the level back down so that it's close to getting feedback but not quite. Bring up bands on the EQ untill you get one that feeds back without applying much gain, and take it out about 4-6dB.
In other words, if you haven't gotten this from the earlier posts,
THERE IS NO "RULE OF THUMB"
I tour into a different venue every week, and sometimes two in one week. Each venue has a very different EQ curve to get the show sounding the same (or as close to the same as possible given the limitations of each venue) from day to day and room to room.
What is the same, however, is the general curve that the system puts out, as measured with somethign like Smaart (or, arguably, an RTA, but these aren't nearly as good since they don't take time into account).
But, to get that same response curve from room to room requires a drastically different EQ setting each time, depending on what frequencies are lacking in the room and what are overly resonant.
TassieBogan pretty much hit the nail on the head, though. It's not easy, and takes lots of practice to develop your ears to do it really well. Heck, I'm still developing my ears, and I've been doing this professionally part time for the last 6 years and full time for the last two! And I will be for decades to come )
As you get into doing it professionally, there are tools like Smaart (or SpectraFoo if you're a Mac guy) that can help make it quicker and more accurate by giving you a graphical readout of what's happening, but even then, your ears are the final arbiter. If the computer shows you something that should sound good, but you think it sounds bad, trust your ears, not the computer.
In my current situation, I actually have it really nice in this aspect. I am able to use parametric EQ in the console to tune my system to flat with the help of Smaart, and then a graphic EQ to tweak from flat to the actual tone curve I want (a parametric would work great for that, too, perhaps even better, but I use what I've got to work with).
Now, you could of course do it all in the one EQ (and there's an argument to be made that the less processing on your signal, the better), but in a situation where I need a consistent curve in many different rooms and have limited time in which to do it, I've found it wonderful to be able to set my tone EQ exactly how I want it to be, since I don't want the system to be quite flat, and then use the computer to help me quickly flatten it out. Of course though, it still comes down to listening to my sound check tracks and seeing how it really sounds with music playing through it that determines whether it's good or not.
I'm getting sidetracked though, and probably way beyond your needs and means, so I'll go back to saying that TB pretty much hit it right on the head )
Andy is right... there is no rule of thumb. Equalization has long been an argument amoung sound engineers since EQ's were developed. What is the proper method? In reality, it depends on you and your ears.
For example, you stated that you don't use the EQ settings in Media Player because you don't need it on premixed tracks. Well, lets say I take that premixed track and try to play it on a super cheap laptop with tiny speakers. In that case, you might tweak the EQ to bring out parts of the music that you couldn't hear before.
Or, let's say you take a mix tape (or cd) into your friends car, and the bass is so loud you can't hear the rest of the music. You turn the bass down, right? Well, you've adjusted the EQ.
Setting the EQ for theatre is similair - you want to adjust the sound so that it sounds as natural as possible while trying to cut as much feedback from open mics as possible. It's an art that takes a long time to perfect - and no matter what, someone will always tell you they weren't happy with the sound. Why? Because you can't please everyone. You just have to try to please as many people as possible, including yourself!
All very good posts and I am just going to offer you this suggestion.
If you have the time, set the EQ flat (in the middle, no cut, no boost) and then in turn cut and boost each frequency 1 by 1. This will allow you to hear what each frequency contributes to the mix. I found this helpful when first starting out.
Now this may sound silly but also close your eyes as well. Our eyes can often confuse what we are hearing and shutting out this stimulus can help you to hear.
600Hz and 1.2kHz are common frequencies for feedback so they might be the two to reach for in a panic to try to pull out of the mix.
So, as TassieBogan and Andy_Leviss have already pointed out, there is no "one setting suits most applications" setting that you can apply to an EQ and that your ears should be the guide.
My advice is to take all of the techniques discussed here and take the time to experiment. This can be a time consuming process and one that you should give adequate time to.