Do resonators trap harmonics ?

[PaulP]

If a resonator trap is tuned to a particular frequency will it also trap harmonics
of that frequency ?

Put another way : modes start at a base frequency and then repeat at
multiples of that frequency. If a resonator is used to target the base
frequency will it also act on the multiples or would separate resonators be
required for each of them (which would be impractical).

Paul P

[SAC]

Yes, If the fundamental driver is damped, assuming no additional forces are active, the harmonics are effectively damped as well.

[Ethan Winer]

Originally Posted by PaulP

If a resonator trap is tuned to a particular frequency will it also trap harmonics of that frequency?

No. A wood panel bass trap is basically a two-pole filter. So it has a center frequency where it absorbs the most, with the absorption falling off at higher and lower frequencies.

--Ethan

[SAC]

?

Whether the trap (any trap) itself is tuned to the fundamental and all of the harmonic intervals is a moot point. (...but you are correct in the sense that a resonant trap is normally only tuned to the fundamental.)

But that being said, the practical affects of damping the fundamental resonance is that, (unless there is another from another source or a variety of special cases not applicable here), the harmonics of the fundamental resonance are damped as well.

[PaulP]

Originally Posted by Ethan Winer

No. A wood panel bass trap is basically a two-pole filter. So it has a center frequency where it absorbs the most, with the absorption falling off at higher and lower frequencies.

Strike one...

What about for helmholtz resonators ?

Paul P

[SAC]

For any absorber or trap or whatever - if the fundamental driving function is damped, the harmonics resulting from that function are effectively damped.

[SaSi_SiDi]

I'd say the correct statement would be that a tuned absorber only traps the frequency it's tuned at.

It's an obvious deduction that harmonics will most likely be attenuated as well but that should not be considered axiomatic.

Simplest opposite scenario is a room with dimensions A x 2A x C where one dimension is twice the other. Killing the 2A mode will not do anything for the A mode as there is a separate "source".

[Ethan Winer]

Originally Posted by PaulP

What about for helmholtz resonators ?

I have no experience with those, but I can't see why a Helmholtz trap would absorb octaves away from where it's tuned either.

--Ethan

[PaulP]

Originally Posted by Ethan Winer

I have no experience with those, but I can't see why a Helmholtz trap would absorb octaves away from where it's tuned either.

Strike two...

Drat. I had this idea that if I killed a mode with a resonator trap it would
take care of the higher order modes/harmonics as well. Hit the three main
modes and modal problems would be solved without losing an enormous
amount of useful energy to absorption.

Targetting several modes for each room dimension would require quite an
assortment of resonators and if several are required at each frequency to
achieve enough effect then there won't be much space left in the room.

Paul P

[Ethan Winer]

Originally Posted by PaulP

Targetting several modes for each room dimension would require quite an assortment of resonators and if several are required at each frequency to achieve enough effect then there won't be much space left in the room.

Exactly, which is why broadband absorption is the best solution in most rooms. Further, peaks and nulls don't occur only at modal frequencies. Comb filtering happens at any frequency, and as soon as you move your head or a microphone the frequencies change. The only time I'd consider a tuned trap is to target a single large peak at a very low frequency. And I'd build it in such a way that a broadband absorber could be placed in front to handle all the higher bass frequencies.

--Ethan

[SaSi_SiDi]

My room has a nasty mode at 29,5Hz. I did some tests using pure sine waves and at that frequency both the air and the couch I was sitting gave me a nice full body massage.

I have also measured a rise around 60Hz, but nowhere near the effect apparent at 29,5Hz. At 29,5 the peak measures at about 25dB above the nominal frequency response of the speakers, compared through the whole octave.

I am not sure if the broadband treatment will have any serious effect down at 30Hz but that's something to measure while mounting traps to take care of the higher frequencies, flutter echo and decay time.

I think that once these issues are addressed, then it's time to consider tuned absorbers at any remaining offending frequencies. And down at 30Hz, in a small-medium room, diffusion is not practical both in terms of diffuser size and distance from the listening position.

[tINY]

For those frequencies, multiple subs are called for. The only way you are going to deal effectively with that mode is to not excite it.

Bag End has a product that may be fo some use for this.




-tINY

[PaulP]

Originally Posted by tINY

For those frequencies, multiple subs are called for. The only way you are going to deal effectively with that mode is to not excite it.

I think you may have covered this in another thread recently but could you
give a short description of how this works ? It seems to me that modes are
a function of the room and not of the number of speakers and their placement.
Won't more subs just exicite the room more ?

Paul P

[tINY]

If you place a speaker in the pressure anti-node (half-way between the two walls creating it), the mode is never excited. Think of it as the wave bouncing off of the walls and coming back to cancel itself out.

Turns out that you can do the same thing using two speakers placed 1/4 of the way between on either side as well. As a benefit with more subs, you also get rid of the comb-filtering-like boundary effects because there is more than one transmitter of the sound. You have to run them in mono, which usually isn't a problem below 120Hz or so.

The "subwoofers" article here is good:
Harman International - White Papers




-tINY

[PaulP]

Thanks tINY for the link to the subwoofers acticle. I just finished reading
it and found it very interesting as it appears to describe an effective way
of dealing with modes. And all harmonics as well given enough woofers.
Looks like the Holy Grail.

I can't quite grasp how it works though, there is very little information in
the document as to how, other than "Loudspeakers located at a null point
of a given mode will not couple to that mode, resulting in no excitation of
that mode."

You say something similar :

Originally Posted by tINY

If you place a speaker in the pressure anti-node (half-way between the two walls
creating it), the mode is never excited. Think of it as the wave bouncing off of
the walls and coming back to cancel itself out.

I understand standing waves and how the reflection cancels things out at
a null (and causes double levels in between). I don't at all see how a
woofer positioned in a null and generating the appropriate frequency would
not cause the mode to exist. Once the sound leaves the woofer won't it
cross the room, hit the wall and bounce back on itself like it always does ?
What is different when the woofer is in the null ?

Turns out that you can do the same thing using two speakers placed 1/4 of the way between on either side as well. As a benefit with more subs, you also get rid of the comb-filtering-like boundary effects because there is more than one transmitter of the sound.

For this to work properly, using subs, do you have to high pass the regular
speakers so they don't produce the modal frequency ?

You have to run them in mono, which usually isn't a problem below 120Hz or so.

What would be the usual way of producing mono bass from two stereo
channels ?

Paul P

[tINY]

It's an impedance and phase issue. If you have a speaker 1/4 wavelength from a wall, then drive that frequency, the reflection is 1/2 wavelength delayed so the speaker cancels out the wave. Think of it as a positive pressure returning just as the speaker produces the negative pressure part of the wave.

In terms of impedance, the driver is fighting itself, so it sees an infinite impedance.

As far as a mono subwoofer, many systems are available to do this. You sum Left and Right (or Left-center-right and the rear channels) and run it through a low-pass filter. On each of the other channels, a complementary high-pass filter is used.



-tINY

[Victor_Stoian]

Originally Posted by tINY

If you have a speaker 1/4 wavelength from a wall, then drive that frequency, the reflection is 1/2 wavelength delayed so the speaker cancels out the wave.

so instead of a peak we'll get a deep dip at "that frequency"?

[PaulP]

Originally Posted by tINY

Think of it as a positive pressure returning just as the speaker produces the negative pressure part of the wave.

Just did a few sketches and finally clicked on this. This is fantastic, the mode
is notched right out of existence. Well, in 2D anyway. Things probably aren't
so perfect in the 3D world.

Paul P

[SAC]

Easily stated...a bit more difficult to do.

1/4 wavelength - from which surface? And at which frequency?

While wavelengths less than 1/4 wavelength will couple, you still have the additional reflective surfaces! In fact, you have at least 5 of them - and you must deal with those reflections as well! - despite the fact that, in addition (for starters), the study conveniently chooses to ignore the oblique and tangential modes for simplicity, as if they would never sum to cause a problem...And after all, your head never changes vertical elevation - well except when you stand or sit - and modes don't apparently don't operate vertically as well. Wait, you mean they do?

And the commonly drawn conclusion that 4 subs are 'best' is questionable even by their own mention compared to two subs (none of which can be localized...or so they claim!!!! another assumption). Two subs provide substantially less loss of low frequency response than four at 1/2 the cost, with only a slightly greater increase in (only axial) modal activity - while ignoring the frequency response between nodes(!) and much more as the reader is expected to ignore much more in terms of assumed conditions.

Rather than a broad general study, this is actually a very narrow and constrained study.

Folks need to review the Toole sub study more closely. It provides a very general guideline within a set of very constrained conditions. It does not provide as much information as some think nor the broad conclusions that some draw! - in fact it provides much less when you realize the conditions assumed and what they simply ignore! And they choose to ignore (and assume!) ALLOT.

Oh, and ANY attempt to apply this study to the vertical use of subs violates the very conditions assumed by the study as well!

[legezer]

In a small room you can sort out those sub frequencies with two or three helmholtz boxes . I did it in my last room. I half damped them as my construction was not perfectly accurate. one was 60x80x20 cm on the wall, and 2 floor standing larger ones by 2 other walls . took it from impossible to lovely sounding. Surely makes more sense to sort the room out than add more subs.

now i'm in a larger room its going to take a lot more than 3 boxes!
just taken delivery of a whole load of RS45

[PaulP]

Originally Posted by legezer

In a small room you can sort out those sub frequencies with two or three helmholtz boxes . I did it in my last room. I half damped them as my construction was not perfectly accurate. one was 60x80x20 cm on the wall, and 2 floor standing larger ones by 2 other walls . took it from impossible to lovely sounding. Surely makes more sense to sort the room out than add more subs.

I'd have thought so too but I've found very little in the way of practical
information regarding building them. And have wondered at their efficiency.
What source did you base your design on ? Have you done any measurements
to know how much of an effect they had ?

Paul P

[PaulP]

Originally Posted by SAC

1/4 wavelength - from which surface? And at which frequency?

My understanding is that the 1/4 wavelength only applies to the first order
axial mode that has a null in the middle of the room. This null is a plane
across the center of the room. To attack it properly would probably
require a bipolar planar speaker the size of the room. Given this is somewhat
impractical the sub on either side of the plane is a compromise. Apparently
still effective enough to have a beneficial effect.

While wavelengths less than 1/4 wavelength will couple, you still have the additional reflective surfaces! In fact, you have at least 5 of them - and you must deal with those reflections as well!

I don't see what reflections at other frequencies has to do with anything.
The only ones we're interested in are the ones related to the modes.

- despite the fact that, in addition (for starters), the study conveniently chooses to ignore the oblique and tangential modes for simplicity, as if they would never sum to cause a problem...

You must admit that they will be lower in level compared to the axial
modes. Perhaps they aren't as bothersome.

And after all, your head never changes vertical elevation - well except when you stand or sit - and modes don't apparently don't operate vertically as well. Wait, you mean they do?

I noticed this lack as well. It seems to me that it would be better if the
four subs were mounted halfway up the wall. Perhaps they found it impractical
(or figured an end-user would find it impractical) to mount a sub in the
middle of a wall, not to mention in the center of the ceiling and in the
middle of the floor. They wouldn't be the first to ignore the vertical.

And the commonly drawn conclusion that 4 subs are 'best' is questionable

Actually they said that 5000 subs is best.

even by their own mention compared to two subs (none of which can be localized...or so they claim!!!! another assumption). Two subs provide substantially less loss of low frequency response than four at 1/2 the cost, with only a slightly greater increase in (only axial) modal activity

They did say as much.

Folks need to review the Toole sub study more closely. It provides a very general guideline within a set of very constrained conditions. It does not provide as much information as some think nor the broad conclusions that some draw! - in fact it provides much less when you realize the conditions assumed and what they simply ignore! And they choose to ignore (and assume!) ALLOT.

My understanding of the study is that they were looking for a practical
method that caused a noticeable improvement in the modal reponse of a
room. I don't believe they say anywhere that they solved all the problems.

To me it looks to have more potential than absorption for the treatment of
modes. How much absorption would you need to tackle a 25 hz mode ?

Oh, and ANY attempt to apply this study to the vertical use of subs violates the very conditions assumed by the study as well!

How is that ? You don't think Toole thought of this ? Seriously ?

Paul P

[Pelonis]

Hey Guys, Lots of good advice going back and forth here. I think what you all are talking about is room resonance and the affect it has on the amplitude response of the room. If untreated, the response will vary depending on the position of the speaker/s and listener. Since room resonances are complex and comprised of a variety of conditions(height, width, depth, etc.),most rooms will have problems at a variety of frequencies depending on how evenly or unevenly the modal distribution is.. Things will change as more stuff( furniture, etc.) and people enter the room as the volume of the enclosure is being adjusted. First thing is to locate the pressure zones where the resonance is occurring. These areas are contributing as sort of secondary sources. Usually you will find that acute angles(corners) are a primary problem. Broadband absorption in the LF is really the best approach since it will be affective as the volume or conditions change and there can be more than one occurrence at any given location. It is true that if you take out a fundamental, there will be no harmonic since the harmonic is the result of the fundamental. If you were to completely absorb the fundamental of one dimension, say the length, you would eliminate that element from the modal response thus having less complications with modal distribution. less harmonic activity will result in less potential for peaks or nulls.

In summary, Room ratios play an important roll as well as how the LF is treated. Ultimately, the goal is to balance this all to where the response is even in as large an area as possible. This way your not moving thru nodes as you slide around in your chair. The balance is a bit delicate since if you absorb in the wrong area you will deplete the efficiency of the room and possibly not help the problem/s. The LF is the most difficult area to treat but once you get it, the rest of the range is a cakewalk. All LF occurrences can be dealt with by using well designed broadband absorbers in the correct position.

I hope this helps.

[Bubbagump]

29.5 Hz is pretty flipping low. Why not notch it out and move on? That situation seems like a good candidate for EQ as you have isolated the one nasty low end frequency and the room is otherwise well treated. Though, I don't know how you are using the room, but for general mixing purposes (heck, and even HT) I would not feel bad personally making 29.5hz go away.

EDIT: NM, getting the OP confused with a reply. Paul doesn't state what problem frequency he is chasing.

[tINY]

Originally Posted by Victor_Stoian

so instead of a peak we'll get a deep dip at "that frequency"?

No, you get the original signal and no ringing or reverberation at that frequency.

This effect is not perfect. But it does work well.



-tINY

[tINY]

Originally Posted by SAC

Folks need to review the Toole sub study more closely. It provides a very general guideline within a set of very constrained conditions. It does not provide as much information as some think nor the broad conclusions that some draw! - in fact it provides much less when you realize the conditions assumed and what they simply ignore! And they choose to ignore (and assume!) ALLOT.

Oh, and ANY attempt to apply this study to the vertical use of subs violates the very conditions assumed by the study as well!

The only way to study a phenomena is to constain the conditions - while some people can talk endlessly about other contribuitng factors and how the assumptions ignore this or that. That is the point of a study.

That all the sub are mounted on the floor in this study is one of the simplifying assumptions. Most people understand this and can still learn something about more complex systems from simple ones.

Move the acoustic center of the subs to the midpoint between the ceiling and floor and you will cancel out the odd-order modes of the floor-ceiling boundaries as well.

Smart people with real problems will notice that you can take 2 or 4 subs and move them around to deal with the worst of the peaks and nulls. Experimentation is usually required....




-tINY

[Pelonis]

It is absolutely true that moving multiple subs around is a valid technique for working out some LF kinks. There are ways to calc where the subs go to attempt to fix room anomalies but experimentation is likely going to come into play. Same goes for locating a single sub. It can be calculated but doing it in real time is easy enough.

EQ was mentioned and there again is another very valid approach to fixing things up. Sometimes it's the best solution especially if the problem is not to severe. A little convincing of an area or two to behave with a good clean EQ is not a bad choice

By far my preference is to get the room under control rather than try to fix a bad room with sub juggling or equalizing. When I get a minute I'll dig up some graphs of a room I designed for Sony with a single 18" sub working down to around 10Hz. I took measurements with the mic in the cone and then at the listening position. It's flat in both locations. In fact there is little if any change pretty much anywhere in the usable are of the room. This provide a most pure experience of the LF in my humble opinion. It obviously takes a lot more thought and understanding but if you can handle it, it's a joyful outcome.

I really love to see you guys having these discussions about LF. My passion for acoustics started there about 30 years ago. All of my designs start with considering this first. Every single one. Hundreds of them.

Cheers

.