Where did I go wrong?

[jtheisen521]

Hi guys,
Wanted to share a recent experience and get some feedback. First a little about me, I've been lurking around here for a while and have done a couple of builds (Aviatrix MLTL's, OS) and have been wanting to try my hand at designing something from scratch. Santa brought me a UMM-6 mic this year and I started playing with it the other night. I have never really done any measurement before so there was a bit of a learning curve. I have a small office/hideout where I often like to listen, currently my Aviatrix are in there driven by a Denon AVR along with a small sub. After familiarizing myself with REW a bit I started measuring. First I measured the speakers individually, once I learned to use the gating and smoothing functions I was producing graphs pretty close to the ones in Curt's writeup for the speakers, ok, off to a good start.
Next I rigged the mic in my favorite chair right where my head would be (centered about 8 ft from each speaker with a slight toe-in), and started measuring. My first sweep with the current tuning (all done "by ear") produced a rather ugly looking graph, so I started tweaking. Getting a smooth transition from the sub turned out to be toughest. After over 2 hours of sweep, tweak, repeat I had managed to get a rather respectable pretty flat line from about 30 hz all the way up to 20,000.

Feeling pretty satisfied with myself I grabbed a strong adult beverage, dimmed the lights, kicked back, spun up The Dark Side of the Moon and waited to enter the audio nirvana that was sure to follow. However, after a few minutes of listening I came to the conclusion that it sounded...Boring, - clinical, lifeless, a lot like the overhead music at the grocery store. I enjoyed it much more before with all the ugly room reflections.

Are my ears just unsophisticated? will they become retrained if I keep it this way and give it some time. Or is that much sought after flat line highly overrated...you decide. Sorry for the long rambling post

[jtheisen521]

Just reworked it a bit, brought up the bottom end and gave it a slight downward slope. Liking it MUCH better now. Having it flat across the spectrum made sound a bit bright and fatiguing. I’ll try this for awhile. Sent from my iPhone using Tapatalk

[joshshetter]

There's a lot to EQ correction.

Are you employing a slope to the correction?

Are you correcting the entire spectrum or a limited bandwidth? Generally correction works best when limited to the lower frequencies, it depends on the room but correcting above 500-600hz can lead to a pretty sterile sound. Rooms are generally most detrimental from ~20-500hz while the narrowing of dispersion as you go up in frequency usually reduces the rooms interactions and you're mostly getting direct sound.

I usually end up with just 3-4 filters with my speakers, different speakers and different rooms even. That's generally really all you need, just a few here and there to tame the major room modes. But every space is different. I fix nulls with placement.

[Paul O]

Nobody really likes the sound of a perfectly flat speaker system.. whether they admit it or not Look at the fletcher munson curve and you will begin to understand why. I'm sure you have heard of the smiley face EQ, it' is a thing because it corrects for the natural curve of our hearing system.. plus a bit depending on your tastes. I prefer a shallow version with just a couple dB scooped out through the mids and a couple dB low freq boost, others go a llittle or a lot more radical. Depending on the SPL you like to lsten at you may find the highend a bit too bright with ruler flat response out to 20khz, so maybe try aiming for something like this.. at least to start with and tweak from there.

[LOUT]

It sounds like you were eventually EQing them to be flat from the listening position, but that's not (as far as I'm aware) usually recommended.
I believe you ideally aim for flat from the near-field on-axis...BUT this will typically give a natural downward slope (slightly elevated bass and lowered treble) at the listening position from room reflections and off-axis response, etc.

For deep base, I think it's more often recommended to use multiple subs in complimentary placement/positions to help smooth the bass response instead of trying to force out room modes for a single listening position.

[fpitas]

Flat at the listening position will be unbearably bright. I start with more or less flat in gated measurements, and adjust until I get the tonal balance I prefer. There's no one easy answer because it depends on the room, the speaker directivity, the listening level and your own preference. To make matters even worse, different recordings vary significantly in tonal balance. A lot of pop music was never intended for good speakers.

[jtheisen521]

Thanks for all the feedback, it was a learning experience for sure. I understand wanting the speaker to measure flat nearfield on axis, so I just thought that flat from the listening position was the next goal. Boy was I wrong! this really helped point out the importance of the room and gave me a lot to think about.

[fpitas]

The science can only get you so far, to maybe 90%. Past there the final voicing is kind of an art form.

[Steve Lee]

Don't think too hard though or you will chase your tail and get nowhere.

Listen and adjust til it sounds good.

Every song that you play will have been mixed differently and you will find yourself adjusting EQ, Q and everything for every song. (like I do).

Get something that is easy to listen to and forget the rest.

I find big woofers with low Q (tight), 6" mids out front and center (presence) and laid back tweeter are the most pleasing for listening.
IOW - boosted lows (BSC) wide flat midrange and downward-sloped highs.

Good luck.

[Chris Roemer]

I will add that it's next to impossible to listen/measure a pair of speakers; this is due to comb filtering. Only measure one at a time (a stereo pair is GREAT for listening, but it's the nuances in timing (distance) between the 2 channels that creates separation and imaging).

[Steve Lee]

This ^ is right on the money advice. Push those stereo speakers as far apart as it takes to hear depth in the music mix.

[Billet]

I take a different approach, I first tune my speakers to get the sound I like then I measure them to document the sound.

[LOUT]

Do you take nearfield, anechoic (or gated, or some other method) measurements...or do you usually measure full-room from listening position?
Have you noticed any interesting patterns in what you like and dislike?
Even if it happens to be personal preference, still sounds pretty interesting. I wouldn't trust myself to be consistent at all, lol.

[fpitas]

I measure to get close to what I know I typically like, then tweak from there. I use reasonably flat reference headphones (AKG K601) as a comparison to be consistent.

[Billet]

I measure the entire room from my listening position. I sometimes take a gated measurement with a short sweep also. I generally find the response measures relatively smooth in room after modeling, building, and tuning. There may be a bit of a rising response. I'm not sure the measurements mean much of anything really, there is always so much ambient noise.