Jump to content

Thoughts on "eq'ing" a room


Coytee

Recommended Posts

That might be a bad title so pardon me if I phrased it wrong.

Here's my thought/question...

Mike B was over once and used his Behringer to help find that I have a room node (mode?) issue at about 90hz. I brought out my dbx equalizer and it happened to have a slider control at 90hz so I knocked it down a bit there and the muddy sound cleaned up. (this was back when I had my Khorns)

Since I presume that to be a room issue, I presume it will also affect the jubilees, so I've kept the equalizer in the system.

As time has passed and I've become more familiar with my EV-Dx38, it dawned on me that I'm eq'ing in it for the speakers as well as then the dbx unit "for the room"... hmm...can I take my eq'ing OUT of the dbx and maybe put it into the DX and drop it down at 90 hz a bit since I have one or two unused slots?

So, that's what I've done and that is the thrust of my question/wondering...

I'm doing this right now, by ear. I've got a friend who's an engineer (EE) and said he'd bring his big honker room analyzer over for me to use while I play around with this.

That's the background and here's the basic question...

When I start playing around, is it usually better to start with a HIGH "Q" (which I think means its a tall, skinny effect) and then if need be, lower the Q, (which I think gives a wider curve) or is it better to start wide & go narrow?

in other words, from a practical point of view, what is the logical way to play with Q's and maybe how many DB's might I want to start with?

Though I don't remember the specific numbers, I think I've just slapped a -3db at 90hz but I've put the dial from both extremes from full 'q' to mimial 'q'

Just looking for some guidance on general ways to narrow my issue while it remains in this room (which I can NOT touch with other sound treatments)

Link to comment
Share on other sites

ONLY minimum phase regions can be equalized.

Otherwise anomalies must be addressed via signal alignment before you can effectively address equalization.

When will we move past the notion of equalizing room anomalies?

Heck, even for those who search the web for new sources of foam should have read that even Ethan has finally realized (in 2005) what we knew in 1987.[;)]

Room modes are resonances that are reinforced by room dimensions due to the large size of the wavelengths relative to the room size. A resonance persists in time. Think of a gong or bell for an image of resonance. These are optimally addressed with bass traps where you address only the room based problem while leaving the direct signal integrity intact. Thus you absorb the reinforced resultant resonances rather than screwing with the direct sound! EQing screws with the direct sound! So if you EQ, it assumes that your speakers are defective in their response instead of the room! And with EQing you are left with a goofy direct signal as well as the same room reinforced resonance!

AFTER you adress room modes with traps and signal align the sources in a minimum phase configuration, THEN you can EQ the direct signal. But this is generally only available to be done in regions as it is difficult to have the entire spectrum minimum phase.

May I suggest folks refer to the "What can we equalize" chapter in the 3rd Edition of Davis and Patronis' Sound System Engineering. And in the meantime in the next several weeks I will try to dig up a copy of the article addressing this topic.

Here is a brief description of the PEQ software module for the TEF:



Parametric
equalization made easy



There's a new equalization technique on the market, and it
offers sound engineers the very latest in sound analysis technology.
Introducing PrecisionEQ software, a program to use with the TEF analyzer and a
parametric equalizer. Sound Lab PEQ provides a simultaneous display of the
unequalized sound system frequency response, the inverse of the equalizer
electrical response and the equalized response in the room.



With PrecisionEQ, the sound engineer can equalize a sound system
to a high degree of accuracy in much less time than with traditional
equalization techniques. In addition, since only minimum phase regions can be
equalized, PrecisionEQ offers a mode allowing you to highlight those regions on
the equalized graph.









Incredible accuracy,
incredible ease



With PrecisionEQ software, a TEF analyzer, and a parametric
equalizer, you can quickly and accurately equalize a sound system,
accomplishing in a matter of minutes what may have taken hours with real time
analysis.



Using the precision equalizer, you match the response of the
inverse equalizer curve to the unequalized curve of the system. This results in
a much flatter equalized curve.









Precision and
flexibility



PrecisionEQ software provides you with greater accuracy than
traditional real time analysis. The software allows you to view the precision
adjustments capable with a parametric equalizer. The result is superior
performance of the sound system.



PrecisionEQ and a TEF system create a very powerful package
for sound analysis, making the equalization process fast, easy and accurate.



Link to comment
Share on other sites

ONLY minimum phase regions can be equalized.

Otherwise anomalies must be addressed via signal alignment before you can effectively address equalization.

When will we move past the notion of equalizing room anomalies?

But even it isn't 'proper' I still have to admit it SOUNDS better when I reduce the quantity at 90hz. So, even if it's not the BEST way (I can't do anything to this room) if I play with my PEQ's, what's the best way to go about it??

Link to comment
Share on other sites

I just put a couple of quarters, nickles and dimes on my speakers plus stick a few hockey pucks on the wall to EQ my listening room. Some chewing gum under the table helps....First you have to know where the room problem is before willy nilly slapping EQ both active or passive.....Ive walked into some rooms where it sounds like a anechoic chamber...others like the room is covered in glass.

Link to comment
Share on other sites

lol.. I've got someone who's willing to lend me his analylzer (is it called a room, sound, spectrum I don't know but it's for measuring sound in the room)

If it shows the same peak at 90 hz, and if I want to try to deal with that inside my Dx-38, then when I do, is it better to start with a wide or narrow "Q"?

My guess is narrow then widen as necessary??

Link to comment
Share on other sites

Richard, this is akin to saying, well I know the job requires a torque wrench, but I only have a hammer. How can I do the job with the hammer?

I am guessing that your friend has a real time analyzer - and RTA, which displays the SUM of all of the sound sources. It cannot resolve the individual components. And room modes are generated by the dimensionally reinforced resonance of the room. Your room is a tuned pipe just like in a pipe organ (although perhaps not as optimally configured).

I wish I had a good answer that circumvents the physics involved, so I guess my only suggestion would be to make sure that you have a really BIG hammer. It won't make the problem any better, but it might make you feel better.

[;)]

If you want to really solve this problem, install corner traps consisting of stacked triangular sliced Roxul Rockboard60 or OC705 wedges forming a 45 degree wedge in the corners of the room (full height in the rear and half height above the speakers in the front).

Link to comment
Share on other sites

Do room modes contribute linearly with rise in SPL or do they have less proportional presence at low SPL and increasingly more at higher SPL?

That is, depending on the room, is there an SPL level below which room modes would not be an issue? - An SPL below which the room would not "load" the modes? I think this must be true for the extreme case of very low level, but does the "no load" range extend high enough in SPL for practical listening?

Off hand I am thinking that below about 90dB there is not much room mode loading. The onset seems to come into play about mid 90's on up?

An implication (if true) would be that the extent of the room treatments or other actions might properly vary depending on the SPL required? Or do the treatments always work proportionaly with the SPL (linear)?

Link to comment
Share on other sites

When will we move past the notion of equalizing room anomalies?

Probably when no improvements are noticeable...

But you'd have to have your head in the sand to believe that no amount of EQ can compensate for room acoustics. Correct the problems? No, but every sound engineer in the world is going to run different settings on their FOH graphic EQ for every different venue. You can't just set the EQ for the system and expect it to work everywhere.

It's a realm where you move from the ideal world into the real world.

Link to comment
Share on other sites

lol.. I've got someone who's willing to lend me his analylzer (is it called a room, sound, spectrum I don't know but it's for measuring sound in the room)

If it shows the same peak at 90 hz, and if I want to try to deal with that inside my Dx-38, then when I do, is it better to start with a wide or narrow "Q"?

My guess is narrow then widen as necessary??

You want to match the Q of the peak. To do this, you will need a measurement tool that can detect the Q. A 31-band RTA is not going to be sufficient (if that is in fact what your friend is bringing). If you'd like, my measurement rig will automatically do all of this for you. [H]

When doing EQ, just keep in mind that it takes time for the resonance to build up all the way. If you notch out the resonance frequency so much that it measures flat, you're going to have a big hole at that frequency for the time before the sound builds up. So instead of trying to make it measure a flat frequency response, just notch it out until it's not annoying anymore. I'm trying to sit down and try to quantify what constitutes a "natural" reduction based on the measurements to see if it's possible to find some kind of psychoacoustic trend...I do know that making it measure flat never sounds good.
Link to comment
Share on other sites

Do room modes contribute linearly with rise in SPL

or do they have less proportional presence at low SPL and increasingly

more at higher SPL?

I think that depends on how one defines linear. I think a really good

analogy to use is that of striking a bell. The harder you strike it,

the longer it rings. However, I think it's pretty safe to argue that

the rate of decay for the ringing is always the same, regardless of

SPL. It's just that when you strike it harder, that it takes more time

for the ringing to drop below the noise floor of the environment. In

light of how we percieve sound, I would call it "linear with SPL", but

that's only because the terms are vague.

Off hand I am thinking that below about 90dB there is not much

room mode loading. The onset seems to come into play about mid 90's on

up?

Absolutely not true. IF there were a threshold, it would be much much

lower in volume. Heck, I measure my room at 75dB and the room modes are

insanely strong.

If you'd like, it would be trivial to provide some measurements showing

the trends. In fact, that might actually be educational...If I don't do

it tomorrow, I'll do it 2 weeks from tomorrow. I can measure all the

way between the noise floor and the max output of my system (or more

likely the maximum input of my measurement rig). It'll be interesting

to note how the waterfalls change.

Link to comment
Share on other sites

You want to match the Q of the peak. To do this, you will need a measurement tool that can detect the Q. A 31-band RTA is not going to be sufficient (if that is in fact what your friend is bringing). If you'd like, my measurement rig will automatically do all of this for you.

I must admit you've got me intruged... I'd bet lunch that Mike would love to see it in action too (probably take someone like him to decipher to me what its telling me)

I DID just speak to the wife tonight about adding some bass traps downstairs (I'm not EVEN going to try upstairs). If you saw the look on her eyes about putting them downstairs, you'd understand why I'm not even going to bother with the current room.

I think it's going to be a herculean feat for me to even get sound treatments downstairs as she thinks it sounds absolutely absurd that anyone would "treat a room for sound" and build "bass traps"

[*-)]

Link to comment
Share on other sites

lol.. I've got someone who's willing to lend me his analylzer (is it called a room, sound, spectrum I don't know but it's for measuring sound in the room)

If it shows the same peak at 90 hz, and if I want to try to deal with that inside my Dx-38, then when I do, is it better to start with a wide or narrow "Q"?

My guess is narrow then widen as necessary??

You want to match the Q of the peak. To do this, you will need a measurement tool that can detect the Q. A 31-band RTA is not going to be sufficient (if that is in fact what your friend is bringing). If you'd like, my measurement rig will automatically do all of this for you. [H]

When doing EQ, just keep in mind that it takes time for the resonance to build up all the way. If you notch out the resonance frequency so much that it measures flat, you're going to have a big hole at that frequency for the time before the sound builds up. So instead of trying to make it measure a flat frequency response, just notch it out until it's not annoying anymore. I'm trying to sit down and try to quantify what constitutes a "natural" reduction based on the measurements to see if it's possible to find some kind of psychoacoustic trend...I do know that making it measure flat never sounds good.

Based on what you wrote a few months ago, Mike, that's what I did and it's working well. I treated the room with "bass traps" made of Cor-Plas in two corners and beside the bookcase in the third corner. There's no fourth corner, since the room opens to a corridor on that side. Then I connected my old EQ to the subwoofer, not the main speakers, since since I figured the La Scalas weren't having much input into the annoying bass resonance at 30-35 Hz.

As you suggested, I just dialed down the response at 30 Hz by 4 dB, although the peak at 35 Hz was 11 dB above reference level (the EQ has sliders for 30, 60, 120, 240...Hz). After trying a few settings, that was the one I settled on, because it gave the most natural-sounding bass. There's still a narrow peak in that area, but there are dips above and below it, so pulling it down further would just make the dips worse.

Perfectly flat response was not possible, so I settled for the best-sounding compromise and now the room only shakes if the recording engineer wanted it to. It's no longer annoying and I'm happy with that.

Link to comment
Share on other sites

Pauln,

It is easy to demonstrate that lower levels also contribute to the resonance. Go into a larger public restroom, building stair well, and hum. Change the pitch, and you will start hitting frequencies that resonate. You aren't anywhere near 90db. I know a vocalist who would do a sweep with his voice while doing a sound check, and he knew close enough to be able to tell the sound tech what to change on the eq to get rid of the larger peaks, reducing the chances of feedback. It worked, too. The also reinforces what Mike said about sound engineers setting eq for each venue. It may not be the best idea, but it still helps.

Bruce

Link to comment
Share on other sites

When will we move past the notion of equalizing room anomalies?

Probably when no improvements are noticeable...

But you'd have to have your head in the sand to believe that no amount of EQ can compensate for room acoustics. Correct the problems? No, but every sound engineer in the world is going to run different settings on their FOH graphic EQ for every different venue. You can't just set the EQ for the system and expect it to work everywhere.

It's a realm where you move from the ideal world into the real world.

GOOD sound engineers aren't trying to use EQ for correcting room anomalies! They use it for what it does well and that is to adjust the direct sound. But GOOD engineers know this.

Link to comment
Share on other sites

When will we move past the notion of equalizing room anomalies?




It's a realm where you move from the ideal world into the real world.

If you put the system and room in the minimum phase condition with signal alignment, then a parametric EQ clears everything, but any divergence causes problems.

I'll have to remember that physics evidently lives in the realm of the ideal and audiophiles live in the world of the real.

The irony is that this is a tired 20 year old debate that has effectively been laid to rest 15 years ago...well, except in the realm of audiophiles who find more value in multi-kilobuck CD players and passive crossovers lacking signal alignment than signal alignment and room treatment. So, you can get mad at me for serving as messenger and making a correct statement valid for both the ideal and real realms, or you can become familiar with that to which I refer and effectively address the situation.

Heck, even Ethan has finally discovered that EQ doesn't work to correct room problems. [:D]

Let me jump back for a second and provide a bit of context. Most are
familiar with electrical impedance. A system where the source and load
impedance are perfectly matched (or 'terminated'} results in a system
where all of the source energy is absorbed and no energy is reflected.
In acoustics, we also deal with impedances. All surfaces have an
acoustic impedance. An anechoic chamber is a room where the wall
surfaces match the acoustic energy source impedance and all of the
acoustic energy is absorbed and none is reflected. And depending upon
the surface, the acoustic impedance varies, resulting in various
frequencies being absorbed more or less than others. Thus with each
reflection, the character of the sound is modified relative to the
source signal. Likewise, they are modified in their travel time or
phase from the direct signal simply because their travel distance is
greater.

Step back and remember what a standing wave/room mode is! It is a resonance that is predominate at a particular frequency(ies) determined by the dimensions of the room where the wavelengths are large relative to the room. While we talk of a peak at a particular frequency, the peak which is predominately created by the summing of multiple resonances, the real problem is the persistence of the signal in time.

Think of a gong or a bell. Rather than a distinct thunk, we experience an extended ringing which is the resonance of the signal that persists in time.

Now, the problems with room modes is not so much the gain peak, it is the muddiness, the boominess, created by the persistence in time. So while we may refer to the mode at a particular frequency, the REAL problem is the resonance's persistence in the time domain. And as this resonance is created by the dimensions of the room reinforcing the frequency, it is in effect a tuned pipe, much like an organ pipe.

So it may be a bit confusing to realize that while we refer to the problem of a room mode being a particular frequency, it is not so much the gain of the frequency, it is the persistence in time at a particular frequency that is the problem.

Equalization ONLY effects the gain of the initial/direct signal, it does NOT effect the rate at which the signal decays!

So, if an EQ does not fix this persistence in time referred to as resonance, what does? Good question!

Let me relate this in a similar but slightly different point of view that some may relate to more easily. Some are familiar with the use of a port to tune a speaker cabinet. Here you are effectively creating a Helmholtz resonator where the dimensions of the box and the size of the opening create a resonance. This tuned resonance creates a peak that is used to reinforce a frequency.

In a room, we have resonances created by the room dimensions. The way we treat the resonances is to use devices that have an acoustical impedance that is equal to the acoustical source impedance at a specific frequency (or band). Thus we use bass traps and /or Helmholtz resonators tuned to frequencies and placed at the location of the resonant standing wave peaks.

By matching the acoustical source which corresponds to the room reinforced resonance with a device which matches the acoustical impedance of the resonance, we are able to effectively and surgically terminate and absorb all of the acoustical energy of the resonance. Thus the persistence of the acoustic energy in time is absorbed, and by addressing this, the summation of the various modes is reduced and he gain issue is subsequently addressed as well.

In a closed room, all of the various mode peaks coincide in the corners, thus the reason for the effectiveness of a high density 'thick' corner bass trap. For mode peaks that may occur in other parts of the room (typically at the 1/2 and 1/3 points along each wall), straqtegically placed Helmholtz resonators act as terminators.

Note, in an open room such as a great room or a room that opens into another, the peaks do not occur at the open corners! In an open pipe, nulls occur at the open end of the the pipe. But peaks still occur in the originating end, and thus bass traps placed above corner mounted speakers are always advantageous, be it in an open or closed room.

OK, this is the 25 cent tour, and there is much more to resonances and treating them if one is interested. But before you go running about trying to solve a problem, it is always advantageous to first identify and quantify the nature of the problem. Once the precise nature of the problem is identified, it is possible to identify the proper tool appropriate to address the nature of the problem as well as to design an appropriate response and solution.

And with resonance, we address the persistence of a signal in time to resolve it. Simply trying to address gain of the direct signal, independently of the modified reflected signals an d the resulting subsequent superposition that further changes the nature of the acoustical signal(!!!) fails to do this. Thus an equalizer fails to address the basic problem.

Link to comment
Share on other sites

Pauln,

It is easy to demonstrate that lower levels also contribute to the resonance. Go into a larger public restroom, building stair well, and hum. Change the pitch, and you will start hitting frequencies that resonate. You aren't anywhere near 90db. I know a vocalist who would do a sweep with his voice while doing a sound check, and he knew close enough to be able to tell the sound tech what to change on the eq to get rid of the larger peaks, reducing the chances of feedback. It worked, too. The also reinforces what Mike said about sound engineers setting eq for each venue. It may not be the best idea, but it still helps.

Bruce

Adjusting gain before feedback levels and the modifying the direct sound from a monitor is not the same thing as addressing room resonances! Two COMPLETELY different animals!

And from the FOH position all you are doing is using the LC component of the LRC filters to cause phase changes that result in the polar anomalies being shifted slightly, dumping the peak or null on someone else! And it causes all of the rest of the spectrum to shift as well, so other frequencies are effected too!

In current state of the art SR systems, the once common banks of 20+ channels of EQ have been replaced with several for vocals and 1 or 2 for the house. And really sharp systems use SMAART to identify the minimum phase realms of the response and parametrically adjust those areas only and thats it for EQ.

Link to comment
Share on other sites

When will we move past the notion of equalizing room anomalies?

Probably when no improvements are noticeable...

But you'd have to have your head in the sand to believe that no amount of EQ can compensate for room acoustics. Correct the problems? No, but every sound engineer in the world is going to run different settings on their FOH graphic EQ for every different venue. You can't just set the EQ for the system and expect it to work everywhere.

It's a realm where you move from the ideal world into the real world.

GOOD sound engineers aren't trying to use EQ for correcting room anomalies! They use it for what it does well and that is to adjust the direct sound. But GOOD engineers know this.

Good sound engineers usually don't get to change the venue they're in and must work within the means of the equipment to arrive at as good as sound as possible. But just to stress the point - I purposely never said anything about "correcting" for room anomalies. I specically said "compensate" because you're working within a huge system of compromises. The only reason I made the analogy is to point that there can be subjective improvements. And just for the sake of argument, I'll remind ourselves that at home we're typically only aiming for good sound in a single listening position anyway.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...