What's up, guys! My name is Chris, this is actually my first post on this site, but I ran across this conversation and can't help but jump in...
I started reading this question
by devgcl regarding if EQ was necessary in a two-channel playback system...
I've been doing live sound professionally for a little while now, and recently decided that I needed to step up my home listening experience. The past year has seen the acquisition of a pair of B&W 685s and a Hafler TA1600. I run direct out of my PC (line level) into the amp, straight to the speakers. It's not the absolute cleanest hi-fi signal chain, but hey, I have student debt, car payments and beer to pay for and FLAC isn't that
So I'm curious to hear opinions on EQ applicaiton and how you guys go about it. I struck out this weekend to figure this out myself, borrowing some gear from work and twisting some knobs, here's what I did --
Using Rational Acoustics SMAART v7, a Sound Devices USBPre and Earthworks M23 measurement mic, I measured each of my 685s at 3'. I measured them at full-range, as well as woofer only and tweeter only to see what the passive crossover was up to... (I'm using last three digits of the S/N to keep track of which speaker is which...)
'Speaker 249' -
'Speaker 250' -
Basically, all of the data is the result of comparing 'Reference' input to the 'Test' mic input. SMAART is sending Pink Noise out of the USBPre, which is a two input/two output interface. One output is literally plugged straight into one of the inputs, this way you create a 'Reference' of what Pink Noise 'should sound like,' and then the second input is the M23 measurement mic.
The top graph shows impulse response. In large rooms you can see wall reflections and stuff here, for now we just want to be sure the mic and the reference loop are in time with each other.
The second graph is Phase Response, however it should be noted all traces have been 'aligned' for comparison to each other (normally Phase Response graphs look much different...).
The third graph shows Frequency Response and Coherence. Frequency Response is pretty obvious, it's the difference between the Reference Pink Noise and what the Test mic is hearing (amp+speakers+room acoustics). The line at the top is Coherence, which is basically data quality. I've set this so the lines disappear if the Coherence (read - Data Quality) is less than 60%.
You can see a color legend on the right to figure out which trace is which. Sorry for the funky naming, the computer I was working on wasn't allow me to type t's and a few other letters for some reason?
This is in my apartment, so speaker placement is less than ideal and there's a fair bit of backround noise (A/C, fridge, neighbor playing Call of Duty...). Nevertheless, we can see things aren't very flat. This is 1/24th octave smoothing, which is a much higher resolution than we're used to seeing (most spec sheets are 1/6 or 1/3 octave smoothing). In real life, things are never going to be a beautiful curving flat line, so we're looking for large frequency bands that seem out of place.
Both boxes are dealing with a crazy boost above 10 KHz, which makes sense, these do sound crispy sometimes. One more-so than the other, manufacturer tolerance? Acoustics? (Sorry, B&W if I'm I'm putting the magnifying glass on...). Things also seem week around 2.5 KHz, in the crossover region, only to hype back up 600 Hz. And there's this iinnssaannee rumble at 90 Hz or so -- really the box, not the fridge or something else in the room, that boost is coming from the system (most likely the speakers). It sounds like the Shuttle taking off if you isolate it...
So I used (on loan from work, so I gotta take it back...
) a Meyer CP-10 (http://www.meyersound.com/pdf/products/electronics/cp-10_ds.pdf
). This thing is sweet, watch the phase traces below and tell me if you see it making things all kinds of whack - it doesn't. A crappy EQ you can hear it working, it's that nasty hollow hiss that you hear (even when cutting), this thing, you can make it a super tight Q and cut 12dB and never hear anything except a certain frequency disappear - it's nuts.
I did the 'Live Sound Guy' thing and flattened the speakers out --
Speaker 249 -
Speaker 250 -
Some of the lines make it look like they are overall much less quieter than the others, this is because I wasn't correcting for EQ differences as much as I should have - things you learn after your done...
Ultimately, I high-passed them at around 120 Hz to get rid of that nasty low crap (you can't really expect these speakers to reproduce lower than they are meant to anyway...) I also shot and re-EQ'd based on the listening location, not a fixed 3' away. There's a lot more I need to work on to get this system 'optimized,' both on the electronic level and room treatment level, and more measuring and experimenting well be done - but coming from the Audiophile prospective, do you guys correct for room and speaker sound or stay true to the most direct, cleanest audio path possible???
My personal take is this - The 865s, like most speakers in the world, are 'hyped' the way their manufacturer likes. These guys defiantly have a low and high boost, and take out some mids (my ears tell me that more than the software). I'm not sure I like this 'better' or not, but only taking time to listen to a lot of music both ways will tell (currently Daft Punk's Random Access Memories, just got done with Pink Floyd's Momentary Lapse of Reason). It's certainly very different than it was before...
Now, I'm not about to put on a pair of headphones. I like the impact a pair of speakers in a room provides, and (for me) it puts me in the room with the music more than headphones. So to some extent, I'm probably going to like it when 100 Hz and below are up three or six dB (I *really* need to decide on a sub...) and when the highs cut a little more to stay on top. But if I'm trying to reproduce the music as the engineer wanted me to hear it, then the system has to be tuned for flat response, and this requires (in order of importance) proper speaker placement, room treatment and probably some (subtractive) equalization.
So what do you guys think? To EQ or not to EQ?