EQ approaches??? [Archive] - Audio & Video Forums

PDA

View Full Version : EQ approaches???



Feanor
04-11-2006, 04:17 PM
I currently own an equalizer, a Behringer T1951 and am considering upgrading to a Behringer DEQ2496. My questions is what is the best way to measure the un-EQ'd speaker response.

For one thing, I suppose it is best to measure and EQ each speaker separately, (see Mike Anderson's thread on the DEQ2496). Agreed?

The other question is trickier. At what distance should one measure the response of each speaker (and why)? Mike A. measured his speakers from his listening position. On the face of it, this is very reasonable as the EQ will level not only the inherent speaker respnse, but also the room response at the location where it's most important.

But I've read a couple of things that make me wonder whether listening position is the whole answer. First, I've read some papers from Harmon International, (Dr. Floyd Toole, et al.), wherein it was determined employing blind, listener panel testing, that the direct sound from the speaker is more important to the perceived "good" sound of a speaker than the reflected sound.

Second, I read DEQX's white paper on EQ, and phase and time correction that their high-end stuff does. DEQX is very clear that one must measure the speakers near-field, or "quasi-anechoically" to properly establish EQ as well as the other aspects, for the speaker or each driver comprising the speaker of DEQX's crossovers are used. After that is done, further correct for the room correction may be done, however DEQX also most treats this as an after-though!

When we measure from the listening position with ordinary equipment such as the Behringer, we are obviously basing our results on the combined direct and relected sound. But is this the way to go? Or should we base our EQ primarily on near-field measurements, perhaps correcting for gross room effects after that?

Resident Loser
04-12-2006, 07:19 AM
I currently own an equalizer, a Behringer T1951 and am considering upgrading to a Behringer DEQ2496. My questions is what is the best way to measure the un-EQ'd speaker response.

For one thing, I suppose it is best to measure and EQ each speaker separately, (see Mike Anderson's thread on the DEQ2496). Agreed?

The other question is trickier. At what distance should one measure the response of each speaker (and why)? Mike A. measured his speakers from his listening position. On the face of it, this is very reasonable as the EQ will level not only the inherent speaker respnse, but also the room response at the location where it's most important.

But I've read a couple of things that make me wonder whether listening position is the whole answer. First, I've read some papers from Harmon International, (Dr. Floyd Toole, et al.), wherein it was determined employing blind, listener panel testing, that the direct sound from the speaker is more important to the perceived "good" sound of a speaker than the reflected sound.

Second, I read DEQX's white paper on EQ, and phase and time correction that their high-end stuff does. DEQX is very clear that one must measure the speakers near-field, or "quasi-anechoically" to properly establish EQ as well as the other aspects, for the speaker or each driver comprising the speaker of DEQX's crossovers are used. After that is done, further correct for the room correction may be done, however DEQX also most treats this as an after-though!

When we measure from the listening position with ordinary equipment such as the Behringer, we are obviously basing our results on the combined direct and relected sound. But is this the way to go? Or should we base our EQ primarily on near-field measurements, perhaps correcting for gross room effects after that?

...a calibrated pink-noise source (in my case a Crown test disk) and an analog RS SPL meter...connecting the meter on a camera tripod (to minimize any reflections/interference) observing the instructions on how the meter should be faced, located at my preferred listening position, one speaker at a time...Using 1kHz as the reference to set the overall level, l'd just measure at each frequency point and plot it on suitable graph paper and when completed, applied the inverse to the equalizer...I did it a number of times to fine tune it and still wound up gently rolling off everything above 10k....sounded smoother and more natural...of course I was using an analog, graphic EQ, but the fundamental things apply...

Each speaker will probably measure a bit differently due to each one's unique room position, so individual measurement is best IMO...while it's probably not impossible to do so, trying to EQ every point in a given room will result in no point being spot on and I prefer that spot to be where I sit. It's sorta' like a Swiss Army knife...sure it does many things, but none of them particularly well.

When you say near-field do you mean like within a meter? How far away is your normal position, two or so meters away? That's still pretty near I think...since we don't live in a quasi-anechoic environment, what's the sense of using unreal measurement points? That would be good if you are trying to establish dispersion patterns of individual drivers or as a system in toto, but as we all know even minimal changes in either the speaker's or the listeners position can cause dramatic perception changes...as does volume...as does source material...

jimHJJ(...little sense in trying to hit a moving target...)

Feanor
04-12-2006, 08:37 AM
...
When you say near-field do you mean like within a meter? How far away is your normal position, two or so meters away? That's still pretty near I think...since we don't live in a quasi-anechoic environment, what's the sense of using unreal measurement points? That would be good if you are trying to establish dispersion patterns of individual drivers or as a system in toto, but as we all know even minimal changes in either the speaker's or the listeners position can cause dramatic perception changes...as does volume...as does source material...

jimHJJ(...little sense in trying to hit a moving target...)

RL, the point of near-field measurements, (1 meter or a little more depending on setup), is that perceived sound quality, (per Toole, et al.), is based much more direct-from-the-speaker sound than it is on total sound which includes reflected. (Note that this is one of Toole's principal criticisms of Consumer Report's testing method.)

This is the basis for my question. Of course, total room response at your listening position will be different from near field, but perhaps total room reponse less important than direct response even at the listening position. :confused5:

Your ear subconciously distiguishes direct from reflected sound, but a standard SPL meter doesn't. Hence to register mainly the direct sound, you need to move your meter in closer to the speaker than your listening position.

Resident Loser
04-12-2006, 09:43 AM
RL, the point of near-field measurements, (1 meter or a little more depending on setup), is that perceived sound quality, (per Toole, et al.), is based much more direct-from-the-speaker sound than it is on total sound which includes reflected. (Note that this is one of Toole's principal criticisms of Consumer Report's testing method.)

This is the basis for my question. Of course, total room response at your listening position will be different from near field, but perhaps total room reponse less important than direct response even at the listening position. :confused5:

Your ear subconciously distiguishes direct from reflected sound, but a standard SPL meter doesn't. Hence to register mainly the direct sound, you need to move your meter in closer to the speaker than your listening position.

...what you ultimately hear is a combo of direct and reflected sound isn't that what should be measured?

If you are trying to plot the EQ of a number of different loudspeakers for some sort of comparison to each other, near-field response does remove room effects, but that's why anechoic response has been the preferred, if a bit non-real-world, method used.

Proximity affects many things, frequency included. It's been a while but as I recall, mags like High Fidelity and Stereo Review would use spliced FR graphs...the lower freqs were miked at one meter to remove boundary effects and added to the upper freq plots which were, I believe, measured anechoically...may be good for initial design baselines or going speako a' speako but certainly not ideal in the average listening room.

Plus I think the time factor between the earliest and latest sounds the ear hears may be less discretely apparent or at least less determinable (although measureable I'm sure) as there is some augmentation and rarefication involved in the process...a bit more of smeared sound as a result of direct, early and later reflections. Wall treatments anyone? LEDE rooms?

Even if you eliminated reflections with room treatments would the SPLs produced @ 1m be proportional and balanced at a greater distance...now I think we are moving from FR per se to dispersion patterns and such...which goes into imaging and soundstage and since little that's recorded has naturally occurring points-of-reference upon which to base things, it really goes off into uncharted and highly subjective territory...that's one reason my EQing has timbral balance as it's goal, anything else that might happen is whipped cream and a cherry on top.

jimHJJ(...where is jneutron when we need him?...)

Feanor
04-12-2006, 12:19 PM
...what you ultimately hear is a combo of direct and reflected sound isn't that what should be measured?
...
jimHJJ(...where is jneutron when we need him?...)

To some degree you are disagreeing with Floyd Toole, (which is fine :cornut: ).

Not to put words in his mouth, but I think Toole was implying that room effects, in general, are less important than generally supposed, (e.g. Consumer Report), think.

I dare say he'd conceed that the sooner a reflection arrives at the ear and the louder it is, the more signficant it is. So severe 1st relections are a problem that should best be dealt with using room treatments, etc.. (Toole has written about room treatments too, and is an advocate of them.) But given moderate or controlled reflections, direct sound is more important: viz. total sound, including early and late reflects does not correlate as well as direct sound with what his listen panelists considered to be best sound.

In my own case, since (a) early relections aren't severe in my listening room, and (2) my Radio Shack SPL meter is semi-omnidirection and can't distinguish direct sound from reflected sound, I think I ought to consider near-field measurements as a basis for EQ. Your situation might be different.

Mike Anderson
04-13-2006, 07:39 AM
A couple points:

1) I should point out that my speaker configurations is quite assymetrical with respect to the room. It was crucial to EQ each speaker independently, because the automatic room adjustment would give me crazy results otherwise. Independently, one speaker will have a peak in frequency where the other speaker has a dip. There's no way to resolve this without EQing indpendently; otherwise, you're trying to force the EQ to do something impossible. In my case, the results weren't pleasing!

2) In my mind, the most important fix is flattening the bass resonances. There, the room is really everything, so I think it'd be a big mistake not to put your mic a far enough distance from the speakers to deal with it.

Resident Loser
04-13-2006, 07:48 AM
...are only the two of us playing here...

Floyd the Toole man eh? Familiar with the name and some of his stuff, but notta whole lot...JBL white papers? anywhich-way...

This is the way my mind works...if you are using a measurement system that would base it's findings on a single impulse, RTAs and Fast-Fourier Transforms perhaps, I would guess the timeframe would be short enough to negate some room effects...at least I think I recall that being the case...

If however you are in the stone-age and using the old pink-noise source and an SPL meter there is a substantial amount of time involved...the slice of noise at each measured frequency point lasts 10-20 seconds maybe longer (it's been a while since I did it)...and the needle jumps around a bit and you have to sorta' average out the needle sweep and fluctuations visually and plot it on the old graph paper...I would hazard a guess that would be of a sufficient length of time to excite whatever reflections are going to be agitated enough to be of some significance and would therefore be required to be included in the overall measurement.

As I said earlier, perhaps when trying to compare one speaker spec-wise to another one might be a more valid approach, but for room equalization there may be other things to consider

Music isn't all pizzicato; there are legatos and basso continuo involved...

jimHJJ(...and that doesn't just apply to classical...)

P.S. Sorry MA didn't see you come in...

Resident Loser
04-13-2006, 08:02 AM
A couple points:

1) I should point out that my speaker configurations is quite assymetrical with respect to the room. It was crucial to EQ each speaker independently, because the automatic room adjustment would give me crazy results otherwise. Independently, one speaker will have a peak in frequency where the other speaker has a dip. There's no way to resolve this without EQing indpendently; otherwise, you're trying to force the EQ to do something impossible. In my case, the results weren't pleasing!

2) In my mind, the most important fix is flattening the bass resonances. There, the room is really everything, so I think it'd be a big mistake not to put your mic a far enough distance from the speakers to deal with it.

...have to be a particularly radical asymmetry...I went through great pains to give my speakers similar (if mirror-imaged) environments...no matter how carefully you may try to balance things out, there are other factors e.g. furnishings, doorways, windows...that can cause measureable differences even though they may seem to be of little consequence in the grand scheme of things.

jimHJJ(...and then of course we have the dog and where he choses to park himself and his bed at any given moment...)

Feanor
04-13-2006, 08:44 AM
A couple points:

1) I should point out that my speaker configurations is quite assymetrical with respect to the room. It was crucial to EQ each speaker independently, because the automatic room adjustment would give me crazy results otherwise. Independently, one speaker will have a peak in frequency where the other speaker has a dip. There's no way to resolve this without EQing indpendently; otherwise, you're trying to force the EQ to do something impossible. In my case, the results weren't pleasing!

2) In my mind, the most important fix is flattening the bass resonances. There, the room is really everything, so I think it'd be a big mistake not to put your mic a far enough distance from the speakers to deal with it.

About measuring the bass. If I do go ahead and try the close-mic for the speakers, I won't use that as the basis for bass EQ, but rather measure the bass from my listening position.

teledynepost
04-13-2006, 09:29 PM
Are EQ's commonly made with mic in, for an auto EQ type thing? Someone I know has something like this, I can't remember the brand now it was a while ago. Is this like receivers with 'auto room correction'? I don't see how this could be any good if it can't measure at the listening position.

Feanor
04-14-2006, 05:22 AM
Are EQ's commonly made with mic in, for an auto EQ type thing? Someone I know has something like this, I can't remember the brand now it was a while ago. Is this like receivers with 'auto room correction'? I don't see how this could be any good if it can't measure at the listening position.

EQs don't necessarily come with a microphone. The one I currently have doesn't and I have measure response with a separate SPL meter. But I guess quite a few, such as the Behringer DEQ2496, have a microphone input and real-time analysers (RTA)

Yes, the idea is quite a lot like the 'auto room correction' found a many AV receivers. But EQs like the Behringer offer much more precise equalization than the recievers by virtue of having 1/3 octive and/or many parametric points.

Did you read my posts in this thread? If did, you'd understand why I've been talking about microphoning the speakers from close-up rather than from listening position. But if not, don't feel bad: Resident Loser, for one, completely missed my point. Granted, as Mike Anderson discussed, bass correction should be done from listening position.

Woochifer
04-14-2006, 04:01 PM
Keep in mind that Dr. Toole's findings about direct and reflected sound have to do with correlating the measured characteristics with how people form preferences. I don't think he was negating the importance of reflected sound, but rather emphasizing the preeminence of the direct sound in determining the speakers that people prefer. These double-blind listenings that he conducts typically take place in acoustically treated rooms, and the speakers are identically positioned for these listenings. This means that the reflectivity properties that affect each speaker set don't change, so the primary variation between listenings will be the direct sound properties and the off-axis response. Also, remember that another one of Dr. Toole's conclusions has been that listeners generally prefer speakers that have less variation in their off-axis vs. on-axis response.

With regard to how you would want to measure for EQ'ing purposes. I would aim more towards measuring at the listening position. For one thing, that's the only way that you can accurately correct for the more extreme room-induced effects that you get in the lower frequencies. With a close-miked measurement, the EQ would do nothing for the wave interactions that create cancellations and peaks throughout the bass range -- effects that will differ throughout the room.

Behringer is not really a home audio company, since their products are generally directed to the pro audio markets. So, I'm not so sure how relevant their white papers would be home use, particularly if you note that the room correction function was treated as an afterthought. If these white papers discuss quasi-anecholic and close range measurements, that would make sense if you're talking about near-field monitoring in a studio setup, since studio monitors generally have a narrow dispersion pattern so that they can minimize the room interactions. I use the Behringer Feedback Destroyer DSP1124 for subwoofer equalization, and the manual for that device (which is primarily about feedback suppresion with a live concert audio rig, with the parametric EQ function very much treated as an afterthought) was also irrelevant for my intended use.

If you really want to do a quasi-anecholic measurement, then you could always measure outdoors, so long as there are no hard boundaries nearby.

superpanavision70mm
04-14-2006, 08:22 PM
Lot's of people have technical know-how and also tweak things to the mathematical exacts, but for some of us...we like to use an old fashion method called THE EAR. You see...the ear works quite well...you simply listen to various sources that you know well. I typically go with something that is minimal in nature (classical music works well in some cases, or something like a solo guitar). I like to use my own ears for EQ levels. I guess it has something to do with my preference in wanting to hear things the way that I like to hear them. I also have a good understanding of what it 'should' sound like and try to reach that without colorizing the material. Everyone is going to hear things different anyway, so my suggestion is to find a level that you like and work around that.

Mike Anderson
04-14-2006, 09:24 PM
^^^ If your sole goal is to make something sound the way you want it to sound at a particular point in time, I suppose that's fine.

But most people here are generally trying to get something close to a flat frequency response, or like me, a baseline flat response with a bit more bass added in.

I seriously doubt you could get a flat response simply by using your ears, especially using music. It would be hard enough to do with pink noise, but music? Forget it.

Before I had an RTA unit, I used to twiddle with EQ all the time, trying to get something that sounded good to my ears. The problem is that it would constantly change from song to song. And room-driven peaks in the bass range were constantly a problem.

Once you get a flat response that takes care of your room problems, you can set it and forget it. The vast majority of the music I listen to sounds fantastic that way.

And if you get a piece that sounds poorly EQ'd, fine, fool with it, but at least you know then that the problem is in the mastering, not in your system.

Feanor
04-15-2006, 02:14 PM
Keep in mind that Dr. Toole's findings about direct and reflected sound have to do with correlating the measured characteristics with how people form preferences. I don't think he was negating the importance of reflected sound, but rather emphasizing the preeminence of the direct sound in determining the speakers that people prefer. These double-blind listenings that he conducts typically take place in acoustically treated rooms, and the speakers are identically positioned for these listenings. This means that the reflectivity properties that affect each speaker set don't change, so the primary variation between listenings will be the direct sound properties and the off-axis response. Also, remember that another one of Dr. Toole's conclusions has been that listeners generally prefer speakers that have less variation in their off-axis vs. on-axis response.
...
With a close-miked measurement, the EQ would do nothing for the wave interactions that create cancellations and peaks throughout the bass range -- effects that will differ throughout the room.

Behringer is not really a home audio company, since their products are generally directed to the pro audio markets. So, I'm not so sure how relevant their white papers would be home use, particularly if you note that the room correction function was treated as an afterthought.
....


As ever, thanks for your comments, Wooch.

Yes, no doubt Dr. Tool considered reflected sound significant ( --as do I for that matter); I have read some of his publish comments where he strongly recommends room treatments to overcome the worst effects. From what I've heard, room treatments, not equalization, is the way to overcome room effects.

I agree you and Mike that bass equalization is only useful and the listening postion. My question/contention only applies to the mids and up.

BTW, to be clear, the white paper I referred to was not by Behringer, but by DEQX that is very much a home audio company. There recommendation very clearly is to correct the speaker using near-field measurements. Granted, they are talking about time delay as well as frequency response. They definitely recommend room treatments too. And they do allow for listening EQ after near-field equalization and room treatment. See here ...
http://www.deqx.com/downloads/What-is-DEQX-Technology-Whitepaper.pdf

Woochifer
04-16-2006, 01:20 AM
Lot's of people have technical know-how and also tweak things to the mathematical exacts, but for some of us...we like to use an old fashion method called THE EAR. You see...the ear works quite well...you simply listen to various sources that you know well. I typically go with something that is minimal in nature (classical music works well in some cases, or something like a solo guitar). I like to use my own ears for EQ levels. I guess it has something to do with my preference in wanting to hear things the way that I like to hear them. I also have a good understanding of what it 'should' sound like and try to reach that without colorizing the material. Everyone is going to hear things different anyway, so my suggestion is to find a level that you like and work around that.

As Mike already pointed out, the purpose of measuring is to establish a consistent reference point. Where you depart from the reference point is where your personal preferences come into play.

With equalization, I suppose that it's fine to use your ears if we're talking about the old school graphic EQs. But, if you're trying to tune the low frequencies with a parametric equalizer to correct for room-induced problems, forget about doing it by ear. For one thing, the parametric equalizer varies not only by the center frequency but the bandwidth as well. With my subwoofer, the room boundaries created an 8 db peak between 36 Hz and 32 Hz. Without a SPL meter or test tones, I would not be able to identify the exact frequency where this problem occurs. The only way you can precisely and accurately set the center frequency and the bandwidth so that they correct for a specific room-induced problem is to measure and calibrate.

The other justification for measuring is that it's simply faster and the results can be replicated consistently. With a two-channel system, you can get away with level matching by ear. But, in a 5.1 setup, the ear is simply not precise enough to simultaneously match multiple speakers. Using test tones and a SPL meter, you set it, tweak it, and forget it.

superpanavision70mm
04-17-2006, 11:10 PM
so if there ear cannot match levels and various frequencies because it can't handle or identify the incredible sensitive levels...then how will you notice it anyway? In other words...if you can't use your ear to calibrate than how can your ear hear the difference when it is calibrated or not? Just food for thought.

I have heard systems that have been EQ'd via computer and they didn't sound all that great. Some people like more surrounds, more fronts, more sub, more low end, more high end, more mid range, etc etc etc. Preference matters too.

superpanavision70mm
04-17-2006, 11:17 PM
http:/gallery.audioreview.comshowphoto.phpphoto=1635&password=&sort=1&cat=500&page=1

Here is the link to my DVD collection. 1100+ including 70 laserdiscs.

superpanavision70mm
04-17-2006, 11:18 PM
http://gallery.audioreview.com/showphoto.php?photo=1635&password=&sort=1&cat=500&page=1


here is the link to my DVD collection

Feanor
04-18-2006, 05:22 AM
.... But, if you're trying to tune the low frequencies with a parametric equalizer to correct for room-induced problems, forget about doing it by ear. For one thing, the parametric equalizer varies not only by the center frequency but the bandwidth as well. With my subwoofer, the room boundaries created an 8 db peak between 36 Hz and 32 Hz. ....

I measured my response across the whole spectrum and was surprised both by the magnitude of irregularities and their location. For example, at my listening position, I discovered a substantial peak a 1kHz with troughs on either side. I was able to level these quite well with my Behringer T1951 parametric -- the sound was a very noticable improvement in the naturalness of accoustic instruments and voices too with most recordings.

To have achieved this would by ear would have taken a long time at best, since I would be guessing at too many variables; most likely I would have achieved only a "preference" rather than accuracy. My own philosophy is that accuracy, not euphony, is the proper goal of reproduction.

Mike Anderson
04-18-2006, 07:39 AM
so if there ear cannot match levels and various frequencies because it can't handle or identify the incredible sensitive levels...then how will you notice it anyway? In other words...if you can't use your ear to calibrate than how can your ear hear the difference when it is calibrated or not? Just food for thought.

The problem with using your ears isn't that you can't hear the difference, it's that you have no reference point.

For example, whatever piece of music you listen to is going to have its own EQ built-in, because of the way it's produced. How do you know what the original recording sounded like (assuming it was recorded with microphones and not synthetically generated)? You can't. When it was recorded, produced and mastered, all kinds of adjustments were made to the EQ.

I can't speak to other peoples' systems/rooms, but I absolutely guarantee you that I can hear a HUGE difference between the un-EQd sound of my system, and the EQ set to give a flat response. It's a slam-dunk, and a vast improvement.

As Feanor says, acoustic instruments sound much more realistic and live (I noticed the same thing when I first did it). Second, the room resonances in the bass range are gone. That means a walking bass line that used to sound like:

"Bm BM Bm Bm BOOM Bm BM Bm BOOM Bm BM Bm BOOOOOM"

Now sounds like:

"Bm Bm Bm Bm Bm Bm Bm Bm Bm Bm Bm Bm Bm Bm"

Now that part you *might* be able to fix with your ears fairly well, assuming you had a good piece of reference music to start with, and provided you had a very good feel for where on the frequency spectrum a give note resides.

But the mids and high frequencies? Forget it. There's no way you can listen to a piece of music and say, "Ah there's a peak at 12kHz and a trough at 7kHz" or somesuch.

Resident Loser
04-18-2006, 08:01 AM
so if there ear cannot match levels and various frequencies because it can't handle or identify the incredible sensitive levels...then how will you notice it anyway? In other words...if you can't use your ear to calibrate than how can your ear hear the difference when it is calibrated or not? Just food for thought.

I have heard systems that have been EQ'd via computer and they didn't sound all that great. Some people like more surrounds, more fronts, more sub, more low end, more high end, more mid range, etc etc etc. Preference matters too.

...IMO only carefully produced, minimally processed, two-channel stereo material has even the remotest chance of presenting natural sounding audio...this excludes most modern effluvia and just about ALL HT...particularly 5.1, 6.1, 7.1 ad infinitum...

Most is a product of the lab...the lab, in this case, being the recording studio or workstation...contrived and measured for maximum "WOW!" factor...

To a certain point the ear, using a test disc, can be useful in the EQing, but it is a tedious process...constantly trying to compare each band of noise to the reference 1kHz signal @0dB (your arbitrary pre-set that is)...at some point Fletcher-Munson-itis takes over at the spectrum extremes and it becomes a fool's errand...

Using music is of a like task...there is no touch-stone, no baseline...

'Way back, in the stone-age, I purchased a half-octave graphic...despite all that I had read regarding the folly of trying to do any meaningful system EQing, with anything over a five-band, ganged unit, by ear, my hubris got the better of me (after all I was a musician and an audiophile wasn't I?) I decided to give it a go...

I was able, for example, to bring a snare drum into what sounded like sonic perfection...unfortunately it was at the detriment of nearly everything else in the particular track...Moving to the next track with a snare, the drum sounded horrible...I tried repeatedly to balance everything using various disks, but it was a definite no-go...With variances in miking technique, processing and without any true spatial information, I realized my error...I special ordered a Crown test disc from a pro audio consulting firm in NYC and tried to use my ears in lieu of an SPL meter...Can you say dead-end?

Suffice it to say, armed with a borrowed SPL meter, within an afternoon or two of taking readings and plotting, I arrived at the flatest response for my favorite listening position...Subsequent subjective evaluation using various source material over the period of a few weeks revealed to me that "flat" response was a tad too bright for my taste and the net result was the gentle roll-off of freqs @10k and above...Yes, my ears did the relatively minor fine tuning, but the foundation was provided by strict adherence to unambiguous, objective measurements.

There is a vast difference between the ears' ability to hear music's (or a particular instrument's) subtleties and it's ability to provide any real measure of accuracy when asked be a substitute for proper test equipment...

jimHJJ(...FWIW...)

Woochifer
04-18-2006, 10:57 AM
so if there ear cannot match levels and various frequencies because it can't handle or identify the incredible sensitive levels...then how will you notice it anyway? In other words...if you can't use your ear to calibrate than how can your ear hear the difference when it is calibrated or not? Just food for thought.

I have heard systems that have been EQ'd via computer and they didn't sound all that great. Some people like more surrounds, more fronts, more sub, more low end, more high end, more mid range, etc etc etc. Preference matters too.


As Mike said, it has nothing to do with preference or whether or not your ear can detect differences. The whole point of measuring is to precisely identify where to make the EQ adjustments and how much you should adjust. Equalizers have discrete frequency points and use decibel values to indicate the level changes, so why shouldn't that be reflected in how the equalization gets set?

And keep in mind that the equalizers that we're talking about in this thread are parametric equalizers, and something tells me that you've never used one before. Like I mentioned earlier, parametric equalizers are different from the old 20+ band graphic equalizers because they use a variable center point and a variable bandwidth. Listening alone will not allow you to quickly and consistently make any needed adjustments. Using measurements does not preclude you from making adjustments by ear once you've established a reference point. Almost all of us who use measurements to set our system settings will tweak with the settings afterwards.

Doing everything by ear is the equivalent of trying to cook without measuring any of the ingredients or even knowing what the contents are. You can try an extra pinch of this or season "to taste" but good luck trying to replicate it consistently.

Sir Terrence the Terrible
04-18-2006, 11:10 AM
Feanor,
How you measure is largely depend on what you are trying to accomplish. Are you trying to measure the speakers performance, or correct its room response. Each has a different way of accomplishing that goal.

If you want to measure the speakers response, the only way to do so is to gate (remove) the rooms influence. This require a quick up close "snapshot" of the speakers response so as not to allow any room resonances to influence the measurement.

If you are attempting to counter the influence of room resonances on the overall output of the system, then you should measure each speaker from the listening position.

Floyd Toole also recommends NOT using eq on frequencies above 300hz. That is because at those frequencies room treatments are much more effective at controlling slap echo, and strong early high frequency reflections. When speaking of upper frequencies which are in are most sentsitive area of hearing, it is always better to alter the room (treatments) than to alter the speakers frequency response via EQ. All EQ's create some level of noise. In the lower frequencies this noise is mostly masked by the signals, and our relative insensitivity to the frequencies. As you go up in frequency, especially above 500hz or so, the noise begins to become more noticeable with the effect of grain, grit or glare imparted on the signal itself.

Feanor
04-18-2006, 11:57 AM
Feanor,
How you measure is largely depend on what you are trying to accomplish. Are you trying to measure the speakers performance, or correct its room response. Each has a different way of accomplishing that goal. [ Both maybe, but my main issue here is speaker performance. ]

If you want to measure the speakers response, the only way to do so is to gate (remove) the rooms influence. This require a quick up close "snapshot" of the speakers response so as not to allow any room resonances to influence the measurement.

If you are attempting to counter the influence of room resonances on the overall output of the system, then you should measure each speaker from the listening position.

Floyd Toole also recommends NOT using eq on frequencies above 300hz. That is because at those frequencies room treatments are much more effective at controlling slap echo, and strong early high frequency reflections. When speaking of upper frequencies which are in are most sentsitive area of hearing, it is always better to alter the room (treatments) than to alter the speakers frequency response via EQ. [ ... for the purpose of room correction: Noted!! ] All EQ's create some level of noise. In the lower frequencies this noise is mostly masked by the signals, and our relative insensitivity to the frequencies. As you go up in frequency, especially above 500hz or so, the noise begins to become more noticeable with the effect of grain, grit or glare imparted on the signal itself.

I would construe that Dr. Toole's recommenation against using EQ for room correction (above 300Hz) is quite consistent with his emphasis on direct sound, versus total room response, as the major determinent of subjective speaker quality that I have been alluding to throughout this thread. Likewise, it is consistent with DEQX's approach wherein they recommend correction of the speaker's inherent resonse using "quasi-anechoic" measurements; they recomment speaker placement and room treatments in preference to using EQ for room correction.

Woochifer
04-18-2006, 01:47 PM
...IMO only carefully produced, minimally processed, two-channel stereo material has even the remotest chance of presenting natural sounding audio...this excludes most modern effluvia and just about ALL HT...particularly 5.1, 6.1, 7.1 ad infinitum...

Most is a product of the lab...the lab, in this case, being the recording studio or workstation...contrived and measured for maximum "WOW!" factor...

Of course that would depend on whether "natural" sound is a system goal to begin with! Most instruments nowadays are amplified, which makes the reference points elusive at best. Not every recording is done for "wow" factor (most are pretty standard issue, and certainly tamer than in the early days of multitracked stereo recordings with all the extreme channel segregation of instruments and side-to-side panning effects), but most are certainly not trying to replicate an acoustic perspective either.

But, with regard to what actually sounds natural, IMO the best multichannel music recordings I've heard produce a more open and natural sound than any two-channel tracks I've heard. Ironically, the multichannel recording that probably gives me the best "audience" perspective for a large hall performance is probably my SACD of the St. Louis Symphony performing Gershwin's orchestral/piano works, which was originally recorded for quad LP release using a minimal miking technique. The two-channel version has long been one of my best evaluation discs, but the multichannel version gave me a sense of space that the two-channel mix never projected. Now I'm curious about the acoustics of Powell Hall, because the multichannel playback certainly gave me a feel for how the orchestra would potentially sound at that venue.

The San Francisco Symphony's Mahler cycle is another great set of multichannel recordings that convey the sound off the stage very well, but they aim more for a "podium" perspective that seems to convey the mic position very well. Davies Hall has a hard sound coming off the stage and does not have very reverberant acoustics, and the multichannel track does a much better job at conveying this than the two-channel track does.

Another multichannel recording that will take the audience perspective better than anything I've heard from a two-channel playback is Bucky Pizzarelli's Swing Live DVD-A which was recorded in a jazz club. There too, I've heard both the two-channel mix, which is already astonishing in how well it takes the listener into the club, and here too, the multichannel playback simply takes things to a different level because the spatial relationship between the stage and the audience is more clearly defined by the playback. Plus, you hear the audience members all around you, just like you would in a crowded jazz club setting. And if you're really obsessive, that DVD-A has an additional six-channel track available that redirects the subwoofer and center channels into two high mounted speakers to further convey a sense of height and ambiance

The caveat of course is that proper multichannel playback is trickier to setup, with more pieces that have to fall into place. The alignment has to be done right, and you have to use the right settings. But, once you have the details nailed down, the results can be startling. And the thing to keep in mind is that most demo rooms are not properly setup for multichannel music -- from the speaker alignment to the level matching to the delay timing to the system settings. So, unless you've spent a lot of time around multichannel systems and taken the time to set them up properly, you likely have not heard multichannel music as it should be done.

Resident Loser
04-19-2006, 05:53 AM
...the quote you cited was intended to be the foundation for what followed...i.e. the folly of relying on most recorded music as any sort of basis (in lieu of calibrated signal sources and measurement devices)for system equalization...much like the wisful thinking of soundstaging in what for the most part is a contrived and manipulated sonic landscape, totally devoid of the required cues and artifacts that occur in nature.

The result may in fact be the aural equivalent of trompe l'oeil and pleasant to listen to, but given it's overall "iffyness" hardly a reliable representation upon which to base what should be an objective process.

jimHJJ(...at least at the outset...)

Feanor
04-19-2006, 06:04 AM
....

The result may in fact be the aural equivalent of trompe l'oeil and pleasant to listen to, but given it's overall "iffyness" hardly a reliable representation upon which to base what should be an objective process.

jimHJJ(...at least at the outset...)

RL, I like the quotes in you signature.:)

Deception and self-deception are a big problem for audiophiles.

superpanavision70mm
04-22-2006, 12:46 AM
People act like multichannel sound is something new. Meanwhile quadphonic recordings go back to the 70's. Sound is a 360 degree experience and while I am not a big fan of certain playfulness that goes into some 5.1 remixes, there are some prime examples (mostly in the SACD format) where the original masters sound more like they should. Plus with SACD you can choose 2.0 or 5.1, which is always a plus!