method to calibrate equalization?

[four4875]

Ive been thinking, and thats probably where i went wrong... but.. how do we know what sounds 'right' for equalization? is there a definitive method to properly adjust everything?

all i could really think of would be use a tone generator to go through the spectrum of audible frequencies and use a sound level meter and play the tone and adjust the equalization till it has a compleyely flat response all through. would that be effective? or would it cause like the mids to be too loud and such? i'd like to see how accurate reproduction of music compares to the average users settings, and how that all works.

so what do all of you SQ guys say? how should it be properly set? or is it just a personal taste kind of thing? thanks for the input.

 

[Prodigious]

Generally people set the equalizer based on their personal preferences and the type of music they listen to. That's... the whole damn point of a customizable equalizer.

"accurate reproduction" of equalizer settings for a song would only be valid for that song alone, or at most the album it belongs to. So unless you want an EQ preset for every song in your collection, find a happy medium for each basic genre of music you listen to.

 

[mortimer]

Use an equalizer with a pink noise generator and a microphone. Adjust to flat with the microphone next to where your head will usually be.

 

[Randyman...]

You can try and compensate for your Speakers' shortcomings with EQ, but the end result will not be the same as a truly "Flat" speaker in a good room. The above mentioned methood of Pink Noise and a Reference Mic will do just fine.

Flattening the Speaker's response is achievable, but EQ'ing the room is MUCH more involved. The room (any room) will have peaks and nulls that vary with where you are in the room (called "Modes"). You can EQ the room so it sounds OK in one spot, and then move a foot over, and your sound is all screwed up. Room modes can cause frequency response irregularities in the 20dB range (or higher) at extreme peak/null modes! This is easily observable by noticing how the low bass will be louder in some areas of the room, and quieter in others. EQ also adds phase shift (unless you use processor intensive "Linear Phase EQ's"), and this further distorts the waveform (you make the Frequency Response "better" at the expense of poorer phase response).

Early Reflections also have a HUGE impact on the overall sound of a system, but EQ can not do anyting to clean them up. Anytime you hear 2 sounds within ~20ms of each other, your brain decodes them as a single sound. You will hear the speakers' "Direct Sound" first, then you will also hear the speakers reflection off the walls and ceiling to your ear (which would take approx 1ms per foot), and your brain interprets these 2 sounds (actually a multitude of reflections) as a single sound, and the end result is an inaccurate reproduction of the source material. Even $50,000 speakers are affected by this phenomenon unless the room is properly treated and dimensionally appropriate.

The solution? Fantastic speakers, set up in a room that is dimensionally appropriate, and treated properly. No EQ will be wanted or needed in such a system

One thing I'd say, DON'T over-compensate! The consumer "Smile EQ" (boosted bass, boosted highs) is very common, and easily adaptable - but it is FAR from how things really sound. Once you get used to this "Smile EQ" sound, it is hard to listen to music in a "Natural" environment. Being an old Car Stereo "Bass-Head" from High-School, I was deeply into this type sound. Now, I'm an Audio Engineer in my spare time, so having a "True Reference Sound" is VERY VERY important to me. I had to re-learn how to listen to music on a "True" system - and I'm still shaking the old "Smile EQ" effects (I have been trying to shake it for over 5 years now). Most people mistake "Clear" for "Boosted Treble" - and that just ain't right . Boosted Treble is not "clear" - it is simply a distortion of the source material... "Clear" is more a function of good Tweeter design, fantastic DAC's, and lack of phase issues in the crossover region - NOTHING to do with EQ...

My Mackie HR-824's are flat, as are my Dynaudio BM-6p's. I LOVE the sound.

The're your ears - do what you want.

 

[four4875]

what i was thinking is that if my system were to accurately reproduce what is put into it it should sound as the studio recorded it. but then i wonder if studios do some of their own EQing to it to make up for common problems with sterios. and when i think of using an eq to flatte out a curve, i was thinking a few DBs at most, not trying to pull up the mids like 15 db or something.

im not too sure how the pink noise would work, and you said to use a mic... connected to what? basically anything with a level meter? and how do you isolate the individual bands?

and for the reflection kinda stuff, wouldnt anechoic wll coverings fix that? not like im going to do it or anything, up to a certain point i usually take SPL over SQ (basically till i can hear it distorting) so im not all that concerned, i was just trying to figure out how one would properly set it, unless they just go by ear and make it however they like it. im not sure if tis my mids or what, but for me from about 1 to 4 khz band seems insanely loud, like its going to rip my ears out, but nobody else notices it.

software EQ vs hardware ? better to impliment a free software one or a cheap analog one? although i will soon be using digital to a reciever so it will be kindof hard to keep the analog goin. what im tryin to ask is do the analog ones still have the phasing problems?

 

[Randyman...]

Studios usually mix to a reference system. The goal is to have a fantastic sounding production. They do compromise slightly for lesser systems, but making the audiophiles suffer so it sounds better on an iPod is not happening (at least not with EQ, but Compression is a WHOLE different ballgame ).

As far as the EQ and Pink Noise, there are 2 methods: Use an RTA (Real-Time Anylizer) with a mic input along with a reference mic (the mic that was calibrated to the RTA) and a Pink Noise generator - this allows a level meter for each band (usually 15 or 31 bands), OR you can use a "Sinewave Sweep Generator" that produces a variable tone, and a "Flat" omnidirectional mic, and a way to view the mic's amplitude - this way you sweep the Sine Wave across the spectrum, and watch the amplitude on the Mic's meter (more accurate). For the "Sweep" method, you can even use a Test CD with a sine-sweep with a known interval (EG: 0:00 is 20Hz, 0:10 is 100Hz, 0:30 is 500Hz, 0:45 is 1KHz, etc) in a pinch. A Sine Wave generator with a frequency display is usually easier to zero-in on the problem frequencies.

I would not recommend using any "Cheap EQ" - be it Analog or Digital. If you must keep the cost down, go with the EQ in your PC (whatever media player you have should have an EQ). A cheap analog EQ will have drastic phase issues, and a 10-band EQ is not fantastic for room shaping. For tuning a room, a parametric is probably a better idea (exact control of the frequency and bandwidth). You might need 5-10 or more bands of Parametric EQ.

If you do this, be prepared to see deviations on the order of 10dB-20dB or more (beyond an EQ's usual cut/boost range). The room has a DRASTIC effect on the sound, and room treatments are preferred over drastic EQ... Also, move a foot over, and the sound is no longer "Flat". Treating the room will lessen the dicrepancies as you move around the room.

Treating the room. A true "Anechoic Chamber" would yield zero reflections, but sounds VERY un-natural. If you have ever been inside one, you know it feels like your head is sucking into itself (an odd phenomenon ). You DO want a certain amount of RT-60 time (decay time) to keep the environment "Natural". The key is getting rid of ANY early refection that will reach your ears less than 20ms after the direct radiating sound, as well as making the room decay in a controlled manner (you want a even decay across the entire spectrum).

The first line of defense is placing absorption to hinder these early reflections. You can have a buddy walk along the perimeter of the room with a mirror flat against the wall. You sit at the "Listening Position". Anywhere you can see the speakers' reflection in the mirror is a place that needs absorption (including the ceiling!). You will want these panels to be fairly "Broadband" absorbers (should be effective to at least 200Hz or so all the way up). Owens-Corning "703" is a VERY common material for DIY'ing absorbers (it is 2' x4' rigid fiberglass; 2" thick - you can make a wooden frame, and cover with Burlap for asthetics). The material you cover them with should be "breathable" so the pressure can pass through to the fiberglass inside. A "tighter" fabric will still allow lows and mids to enter, but might reflect some highs. You can choose a covering that compliments your desired end result (you don't want too much high frequency damping). Spacing these panels 2" off the wall also helps increase their effectiveness at lower frequencies.

Another HUGE issue is bass trapping. Since the bass wavelengths are so long, they cause the most issues as far as constructive/destructive interference. Bass raps are usually large, and expensive. In a recording studio, 25% or more of the room's surface area may be dedicated to Bass trapping alone! The odd thing is, you will loose some SPL in a "Flat" room. Tis is due to reducing the amount of constructive (as well as destructive) interference. You sacrifice some irregular SPL dB's for the sake of a "True" low-end.

Unless you want to spend cash - just use your ears - and be conservative IMO. Both of my "Reference Systems" are relatively flat (no extra EQ's are added). Good speakers should do OK in a normal room w/o EQ, but a good room will really take it to the next level. Crappy speakers? EQ to your heart's content

 

[four4875]

right now the speakers arent all that great, but i have plans for putting together both a really loud SPL targeted PA system, and a home mid range decent sounding system. but basically i want to know what sounds 'right' as compared to the smiley sound that im used to haha.

 

[Randyman...]

That is the hard part - "Re-Learing" how to listen. I'm still amazed at what some people call "Good Sound" (EG: Bose ). In your case, it will be hard to know "What sounds good" with crappy speakers. I re-trained myself by investing in good speakers (Mackie HR-824's, and Dynaudio BM6P's), and leaving the EQ flat. And believe me, I LOVE bass (had eight 18" subs in my Mazda Pickup in High School, and I still have four 10"s in my Tacoma). The trick is tight, flat & balanced bass. I still have the urge to kick in the bass boost and shake my guts out, but only for a few minutes at a time...

I'll make a strong suggestion: Concentrate on ONE system, and make it the best you possibly can. You'd be suprised how loud something like a Mackie HR-824 can get and still remain "Flat" (8.75" woofer 150Watts RMS, 1" Tweeter 100Watts RMS, 42Hz-20KHz +/- 1.5dB). These are self powered, and you can hook them up directly to your PC! I use an RME soundcard to drive mine at +4dBu and over 107dB S/N. Not "Cheap", but NOWHERE near the "Top of the Line"! (~$1400/pr).

The speakers are the weakest component in ANY system. Invest every penny you can into the speakers, and you will be heading in the right direction.

I have a fairly large PA we use for Band Practice and for Outdoor Concerts we do on our 30-Acre plot. I try and EQ it fairly flat, but I still have slight low-end enhancement going on (Two JBL 18" subs in custom 32Hz tuned enclosures w/ 600W RMS each). The cool thing about a PA in a "Band Setting" is you have seperate control over each track at the board. I can keep the Bass Guitar "Meaty" w/o being "Muddy", and still get the lovely ~45Hz fundamental out of the kick drum by EQ-ing each channel seperately. EQ in this scenario is your friend. You have to carve room for other instruments, and eliminate offending frequencies that build up on a per-channel basis. Graphic EQ's are also used to kill feedback in the stage monitors (speakers that close to an open mic will feedback if not addressed/EQ'ed appropriately).

Ahh - I love audio engineering and production

Good luck. The sooner you start kicking the "Smile EQ" habits, the sooner you will REALLY be able to enjoy good Audio Productions. It is like listening to the music for the first time all over again. I still get goose-bumps when I can hear all of the "Ear Candy" going on in a mix that was previously buried by over-bearing Bass and harsh "tsssss-tssss-tsssss" treble.

 

[four4875]

my dad just spent $400 on a cheap PA, 200 watts output. i plugged it into the lappy to have a little play on it, and it seemes to jump off a cliff after about 60 hz, i was disappointed in it. i was thinking like some crown amps to a couple dayton 18s tuned really low, then some cabs with a 12 and a couple horns in each. not sure on mixer tho, and source would likely be the laptop. but that will be after i have a decent setup in here.