[youngho]

I should note further that the example of the outdoor brick wall you provided in the alt.sci.physics.acoustics is an approximation of a standing wave because the direction of the reflection is opposite that of the original signal. Some, but not most, comb filter effects would do this as well. Again, this would depend on the definition of "standing wave" used. I believe that the reality is that most of what you refer to as non-modal "standing waves" in the bass region are so varied in frequency and level that they end up smoothing out so that they are not a concern in the bass region in typical room acoustics.

[Ethan Winer]

In the discussion you linked, you seem to be defining it as "ANY case where acoustic interference causes a deep null (we can define "deep" separately)."

Actually the waves stand still at peak locations, but I know what you mean.

A subwoofer with its phase set incorrectly creates standing waves with the main speakers.

Possibly.

Do these apply to comb filter effects?

Yes. Again, if "standing wave" means the wave is standing and not traveling, then the cause is irrelevant, no?

you're arguing that virtually all room acoustics are standing waves, with the possible exception of the direct signal from the loudspeakers, and if they're not exactly equidistant from the listener, then these are also standing waves, as well.

No, peaks and nulls caused only by different arrival times are not due to standing waves. In that case the waves from both directions are traveling.

--Ethan

[Ethan Winer]

most of what you refer to as non-modal "standing waves" in the bass region are so varied in frequency and level that they end up smoothing out so that they are not a concern in the bass region in typical room acoustics.

I don't know what you mean by "smoothing out" because SBIR peaks and nulls are as damaging as peaks and nulls caused by anything else. The only difference between modal and non-modal peaks and nulls is modal peaks are also resonant and have an extended decay time. And non-modal peaks and nulls can happen at any frequency. But the drop in volume on only some bass notes is just as audible and damaging whether the nulls are modal or not.

--Ethan

[bpape]

Agreed.  Non-modal frequency response abnormalities are very common and can be just as harmful.

Bryan

[youngho]

In the discussion you linked, you seem to be defining it as "ANY case where acoustic interference causes a deep null (we can define "deep" separately)."

Actually the waves stand still at peak locations, but I know what you mean.

A standing wave doesn't literally stand still when it comes to sound. The woofer still moves back and forth, causing periodic fluctuations in pressure. The fluctuations occur at a certain frequency, and these fluctuations are transmitted through air at a speed of approximately 1130 ft/sec. In standing waves, waves of the same frequency travelling in opposite directions can cause a characteristic pattern of constructive reinforcement and negative reinforcement characterized by periodic distribution of the maximal areas of each, which can be described mathematically. The standing wave is simply a pattern with certain characteristics, and the waves that create the standing wave pattern are travelling at the speed of sound. If a sound wave literally stands still, how can it be sound? How can it have a frequency and wavelength? How can it travel at the speed of sound?

Please provide your definition of standing wave when it comes to sound. Please. I copied and pasted your own words from the discussion you linked to. "By extension, it seems to me that ANY case where acoustic interference causes a deep null (we can define "deep" separately) has by definition also created a standing wave" were your own words. Can you comment on the bipolar radiation pattern? Do you consider this to be a standing wave?

Do these apply to comb filter effects?

Yes. Again, if "standing wave" means the wave is standing and not traveling, then the cause is irrelevant, no?

you're arguing that virtually all room acoustics are standing waves, with the possible exception of the direct signal from the loudspeakers, and if they're not exactly equidistant from the listener, then these are also standing waves, as well.

No, peaks and nulls caused only by different arrival times are not due to standing waves. In that case the waves from both directions are traveling.

Aren't the peaks and nulls of comb filter effects are caused by different arrival times? The reflection takes longer to travel to the ear because it has a long distance to travel. If certain frequencies arrive in phase with the original signal, it results in a peak; if other frequencies arrive out of phase with original signal, it results in a null; everything else that's in between is in between. Are you arguing that comb filter effects' peaks and nulls are "not due to standing waves" because the waves (of the original signal, as well as that of the reflection) are both travelling?

As for modal ringing, the extent to which this really occurs is controversial. Much of the long decay time is actually due to the high amplitude to start with, even if it's not truly a minimum phase phenomena. We've been through this before.

Bryan, can you walk through SBIR effects in the bass frequencies for me? My limited understanding is that of all probable speaker locations, the number in which the bass driver is equidistant to multiple room boundaries is a small subset, plus because there are at least six room boundaries, typically, and because of the relatively omnidirectional nature of bass (meaning that the bass wavelength will meet different parts of the same boundary at different times), as well as the hemispherical nature of bass reflections, most of these would end up as negligible compared to the modal effects. What am I missing?

Thanks!

[*Scotty*]

Standing waves can seen in rivers and can be easily created in wave tanks or in your bath tub. The fact that the waves do not move is why they are called standing waves.
Here are some links to more information
http://en.wikipedia.org/wiki/Standing_wave
http://hyperphysics.phy-astr.gsu.edu/Hbase/waves/standw.html
http://id.mind.net/~zona/mstm/physics/waves/standingWaves/standingWaves1/StandingWaves1.html
http://id.mind.net/~zona/mstm/physics/waves/standingWaves/standingWaves.html
http://www.walter-fendt.de/ph14e/stwaverefl.htm
Scotty

[youngho]

Waves in water are transverse, as opposed to longitudinal. The water does indeed move, except it's up and down.

Sound is a transverse wave. The air particles are moving back and forth. At a node, they are moving with maximal velocity, which is why the pressure is minimal. At a antinode, they are moving with minimal velocity, which is why the pressure is maximal.

Look at your second link: "Under these conditions, the medium appears to vibrate in segments or regions and the fact that these vibrations are made up of traveling waves is not apparent - hence the term "standing wave"."

THE FACT THAT THESE VIBRATIONS ARE MADE UP OF TRAVELING WAVES IS NOT APPARENT

Look at your second link: "Certain sized waves can survive on the medium. These certain sized waves will not cancel each other out as they reflect back upon themselves. These certain sized waves are called the harmonics of the vibration. They are standing waves. That is, they produce patterns which do not move."

AS THEY REFLECT BACK UPON THEMSELVES. It is the pattern that does not move.

The third and fourth links are focussed on transverse waves.

The fifth link is subtitled "Explanation by Superposition with the Reflected Wave". When you watch the animation, do you see an incident wave moving forward and the reflection moving backwards? They are travelling. Together they produce the standing wave pattern.

Thanks for posting these links and reinforcing what I said. The fifth one does it ABSOLUTELY perfectly.

[bpape]

ALL cancellations are caused by waves arriving at your ear at the same time from different directions and/or out of phase.  Comb filtering is no difference.  SBIR is simply another form of comb filtering.  Yes - particles are moving back and forth and not truly travelling - however, the 'wave' does move.  One can have standing waves and travelling waves.  The travelling waves absolutely move in space.  This is easily seen in an untreated room by looking at waterfall plots with a very long time window.

Bryan

[youngho]

ALL cancellations are caused by waves arriving at your ear at the same time from different directions and/or out of phase.  Comb filtering is no difference.  SBIR is simply another form of comb filtering.  Yes - particles are moving back and forth and not truly travelling - however, the 'wave' does move.  One can have standing waves and travelling waves.  The travelling waves absolutely move in space.  This is easily seen in an untreated room by looking at waterfall plots with a very long time window.

If two waves of the same frequency arrive at your ear from different directions but in phase, they will not cancel out but rather add, right?

Comb filtering is fundamentally different from a standing wave pattern due to the pattern of the distribution of nodes and antinodes. "Comb filter" describes a frequency response at a particular location. "Standing wave pattern" refers to response at different locations for a particular frequency.

Again, in a standing wave pattern, the component waves at the corresponding frequency are traveling in opposite directions. Who ever argued that traveling waves don't move in space?

My limited understanding is that of all probable speaker locations, the number in which the bass driver is equidistant to multiple room boundaries is a small subset, plus because there are at least six room boundaries, typically, and because of the relatively omnidirectional nature of bass (meaning that the bass wavelength will meet different parts of the same boundary at different times), as well as the hemispherical nature of bass reflections, most of these would end up as negligible compared to the modal effects. What am I missing?

[bpape]

Correct on the phase.  However, it's still comb filtering - just additive.  SBIR is a combination of constructive and destructive interference of 2 waves colliding - yielding a sawtooth type response.

Bryan

[youngho]

Did I ever say that SBIR is not a comb filter effect? I simply said it's not a standing wave effect, as I understand a standing wave to be defined.

Can you comment on the audibility of comb filter effects after reading Floyd Toole's book?

To repeat myself, "I believe that the reality is that most of what you refer to as non-modal "standing waves" in the bass region are so varied in frequency and level that they end up smoothing out so that they are not a concern in the bass region in typical room acoustics" and "My limited understanding is that of all probable speaker locations, the number in which the bass driver is equidistant to multiple room boundaries is a small subset, plus because there are at least six room boundaries, typically, and because of the relatively omnidirectional nature of bass (meaning that the bass wavelength will meet different parts of the same boundary at different times), as well as the hemispherical nature of bass reflections, most of these would end up as negligible compared to the modal effects. What am I missing?"

[bpape]

It can still introduce the same effects as a standing wave.  A severe (sometimes) peak and/or null at a set of frequencies.  SBIR is a set of standing waves - both constructive and destructive.

You do not need to have a driver equidistant from multiple surfaces to experience these issues.  In fact, you can have 2 different sets of SBIR effects where some of them interact and cancel each other through the same comb filtering effects - while other portions of one or both remain.

I do not believe that most bass anomolies just blend into oblivion and are not heard.  Some may, some do not.  IMO, it has to do largely with the width of the null in relation to the spacing of frequencies at the octave in question.  If you have a 5hz wide null at an octave between 20 and 40Hz, you'll hear it.  If you have a 5Hz wide null at an octave between 1khz and 2kHz, you won't.

Bryan

[youngho]

A standing wave has two effects: constructive interference and destructive interference. I've said this all along.

Comb filtering has two effects: constructive interference and destructive interference. No argument there.

Just because two things are composed of the same components do not mean that they are the same thing. White light and a refraction spectrum/rainbow are not the same thing, even though they might contain all the same wavelengths of light. Just because two things can produce the same effect does not mean that they are the same thing. A pistol and a revolver are not the same thing, even though they both produce a bullet travelling at high velocity.

Adding a second speaker results in constructive interference. If it's not exactly equidistant, it will result in a pattern of constructive and destructive interference. This frequency response pattern is a comb filter. It is not a standing wave. How is a single loudspeaker reflection modelled in psychoacoustic experiments? By a second loudspeaker. Specular ray-tracing models are used to show where the virtual speaker would be from a single reflection, and then if the wall is eliminated, an actual second speaker can be placed in the location of the virtual speaker. By controlling the loudness of the second speaker, you can examine things like the audibility of single reflection at various levels. Ethan drew a distinction between a reflection versus a second loudspeaker interacting with the original source. In fact, they are essentially the same phenomena. The thing is, in most typical listening rooms, there is more than a single reflection because there is more than one wall or boundary. Most rooms have six or more. Plus windows, doorways, and HVAC complicate things in the bass region. Most people listen to two or more loudspeakers. If the room and loudspeaker setup are not EXACTLY symmetric, then you'll have another set of reflections and interference patterns to hear.

A 5 Hz wide null in the low bass region has a VERY different Q than a a 5 Hz in the midrange, so you're comparing nulls of different width, but you're right, narrow interference dips should be less noticeable than wide interference dips of a similar amplitude.

[*Scotty*]

youngho,do a simple experiment in a wave tank with two sources and you will always see an interference pattern regardless of spacing or symmetry of the sources.
Your comments about loudspeaker setup symmetry are valid but not because doing so avoids another set interference patterns.
As far comb filter effects and standing waves are concerned it doesn't really matter if they are not exactly the same when the solution is the same,room treatments.
Scotty

[youngho]

Scotty, I absolutely agree that multiple sources produce interference patterns. I failed to indicate that I was talking about the signals reaching the ear when I was discussing the second loudspeaker. Sorry about that. In any case, these interference patterns are not necessarily standing wave patterns. Just about all room acoustics involve constructive and destructive interference from reflections (the exception being anechoic chambers, of course), but these are not all standing waves. You linked to this website earlier. Why don't they call this pattern (http://id.mind.net/~zona/mstm/physics/waves/interference/twoSource/TwoSourceInterference1.html) a standing wave pattern? I believe it's because it does not fit the mathematical description of a standing wave, with the periodic and linear, not radial, spacing of nodes and antinodes.

In any case, you're right, it's a semantic argument.

[youngho]

However, this discussion does raise some interesting questions for me. In a "man cave," when thinking about absorption cost-effectiveness, bass absorption typically is positioned to primarily address modal concerns. We know that placing broadband absorption on the front wall and rear wall is beneficial. Wouldn't it make sense, if aesthetic concerns be damned, to start recommending two absorption panels, attaching them to each other with a hinge, and mounting them at, say, 90 degrees to each other and 45 degrees to the wall? Obviously the angles could be different. I have something like the RPG Variscreen or the RealTraps MiniGobo in mind. If you have panels or superchunk traps straddling the corners, anyway, it just continues that theme along the front and rear wall if you have a series of them. Sure, it wouldn't look great, but you could hide it behind a shoji screen, or even better, modify easily to look more like a shoji screen by using light-colored fabric covering, allowing the wooden edges to be visible, maybe adding decorative/stablization bars, and radiusing the sides before attaching the hinges or else cutting one edge corner at at angle so that you can just glue/attach two panels and make them look like they share an edge. It seems like this would better address bass absorption because of the variable air space. Any thoughts?

[Ethan Winer]

If I addressed everything raised in this thread since yesterday it would be all I do today. So I'll just hit the high points from youngho:

In the discussion you linked, you seem to be defining it as "ANY case where acoustic interference causes a deep null

I probably did say null in that discussion but I was wrong. Eric Desart explained (later) that I had the polarity reversed in my thinking. The waves collide at a peak location, defined as a point where the pressure is maximum. So two positive-going waves arriving from opposite directions doubles the pressure (6 dB peak), and that's the point where the waves stand still.

A standing wave doesn't literally stand still when it comes to sound. The woofer still moves back and forth

Sure, the speaker is moving, but that's not the location in the room where the waves are standing still. Even in the case of a resonant standing wave (which you recognize as a standing wave) the speaker is still moving air!

If a sound wave literally stands still, how can it be sound? How can it have a frequency and wavelength? How can it travel at the speed of sound?

There is sound because the pressure rises and falls. But there's no wave movement at the point. Or something like that. I admit I'm not a physicist.

Aren't the peaks and nulls of comb filter effects are caused by different arrival times?

Yes, but it depends on whether the waves arrive from opposite or similar directions. When you sit two feet in front of a wall, with the speakers in front of you, reflections off the rear wall behind you come at your ears from a direction opposite that of the waves from the speaker. But when you hear mono music from two stereo speakers, and you're pretty far back from both speakers, the waves can arrive at different times but not be traveling in opposite directions. Do you see the distinction? In both cases you get peaks and nulls because there are two arrival times, but the waves stand still only in the first example of the rear wall. This aspect too was discussed in depth a few years ago, though I forget where. It might have been in the Sound On Sound magazine forum.

As for modal ringing, the extent to which this really occurs is controversial. Much of the long decay time is actually due to the high amplitude to start with, even if it's not truly a minimum phase phenomena. We've been through this before.

I'm not aware it's controversial! The extended decay is due to conservation of energy, no? If the walls had no losses the sound would continue forever.

More tomorrow if needed.

--Ethan

[bpape]

A standing wave has two effects: constructive interference and destructive interference. I've said this all along.

Comb filtering has two effects: constructive interference and destructive interference. No argument there.

Just because two things are composed of the same components do not mean that they are the same thing. White light and a refraction spectrum/rainbow are not the same thing, even though they might contain all the same wavelengths of light. Just because two things can produce the same effect does not mean that they are the same thing. A pistol and a revolver are not the same thing, even though they both produce a bullet travelling at high velocity.

Adding a second speaker results in constructive interference. If it's not exactly equidistant, it will result in a pattern of constructive and destructive interference. This frequency response pattern is a comb filter. It is not a standing wave. How is a single loudspeaker reflection modelled in psychoacoustic experiments? By a second loudspeaker. Specular ray-tracing models are used to show where the virtual speaker would be from a single reflection, and then if the wall is eliminated, an actual second speaker can be placed in the location of the virtual speaker. By controlling the loudness of the second speaker, you can examine things like the audibility of single reflection at various levels. Ethan drew a distinction between a reflection versus a second loudspeaker interacting with the original source. In fact, they are essentially the same phenomena. The thing is, in most typical listening rooms, there is more than a single reflection because there is more than one wall or boundary. Most rooms have six or more. Plus windows, doorways, and HVAC complicate things in the bass region. Most people listen to two or more loudspeakers. If the room and loudspeaker setup are not EXACTLY symmetric, then you'll have another set of reflections and interference patterns to hear.

A 5 Hz wide null in the low bass region has a VERY different Q than a a 5 Hz in the midrange, so you're comparing nulls of different width, but you're right, narrow interference dips should be less noticeable than wide interference dips of a similar amplitude.

Adding a second speaker can not only introduce constructive interference, but also destructive - no matter the symmetry.  You're correct that it isn't technically a standing wave.  It reacts the same and can be controlled in the same way with the same mechanisms though. 

Bryan

[youngho]

I probably did say null in that discussion but I was wrong. Eric Desart explained (later) that I had the polarity reversed in my thinking. The waves collide at a peak location, defined as a point where the pressure is maximum. So two positive-going waves arriving from opposite directions doubles the pressure (6 dB peak), and that's the point where the waves stand still.

Sure, the speaker is moving, but that's not the location in the room where the waves are standing still. Even in the case of a resonant standing wave (which you recognize as a standing wave) the speaker is still moving air!

There is sound because the pressure rises and falls. But there's no wave movement at the point. Or something like that. I admit I'm not a physicist.

It's because the sound waves don't stand still. They constructively and destructively interfere. It's the standing wave PATTERN that stands still. http://www.walter-fendt.de/ph14e/stwaverefl.htm illustrates this beautifully.

Yes, but it depends on whether the waves arrive from opposite or similar directions. When you sit two feet in front of a wall, with the speakers in front of you, reflections off the rear wall behind you come at your ears from a direction opposite that of the waves from the speaker. But when you hear mono music from two stereo speakers, and you're pretty far back from both speakers, the waves can arrive at different times but not be traveling in opposite directions. Do you see the distinction? In both cases you get peaks and nulls because there are two arrival times, but the waves stand still only in the first example of the rear wall. This aspect too was discussed in depth a few years ago, though I forget where. It might have been in the Sound On Sound magazine forum.

I assume we're still talking about bass. If bass comes from two opposite directions and arrive at the listening position in phase, they will constructively interfere. The incident and reflected waves do not stand still.

I don't see the distinction you're trying to make, but this is a semantic argument at this point. For you, most interference patterns seem to represent standing waves, although I can't tell because you've never provided your definition. Let's agree to disagree on this issue.

As for modal ringing, the extent to which this really occurs is controversial. Much of the long decay time is actually due to the high amplitude to start with, even if it's not truly a minimum phase phenomena. We've been through this before.

I'm not aware it's controversial! The extended decay is due to conservation of energy, no? If the walls had no losses the sound would continue forever.

In his book and white papers, Toole shows examples where reducing the amplitude of a bass mode through equalization reduces the decay time of the mode to a level similar to non-modal bass frequencies. There was a long discussion on DIYaudio about how this is not truly a minimum phase phenomena, but rather it approximates one. If I'm not mistaken, you had, in the past, argued very strongly that equalization might reduce the amplitude peak for a single listening position but doesn't affect modal ringing at all. The data seems to demonstrate that the longer decay time is due primarily to the higher amplitude compared to non-modal frequencies but that there isn't a significant prolongation of decay at modal frequencies compared to nonmodal frequencies. In other words, bass ringing is bass ringing, it's not particularly modal, it can occur at modal or nonmodal frequencies (hence, it's not modal ringing but rather bass ringing), etc. I'm not looking to reopen this discussion, so I'll leave it at that. No more from me on the subject.

[Rob Babcock]

Bass traps can help a lot, but moving the speakers and/or your listening position can also make a huge difference.  Modes are a function of your room and are based on math- you can't really change them.  But moving the transducer or the listener can change the way the room is engaged, if that makes sense.