Low bass in your space? Are you sure?

[sfogg]

"Now THAT'S what I said before,"

You also said if you looked at the low bass 'as described' by an instrument it wouldn't be the same frequency as the original.

That is 100% dead wrong.

" The problem is when the waveform reflects back onto its original path, in particular when the listening place is positionally less of the wavelength which is typically the case. Now you are listening to a compound waveform because of the overlying reflection. "

And the point you keep missing is this happens well above that point to.

When you hear a note playing during music do you honestly think it is one single cycle of that wave? A single cycle of a 1000hz wave is 0.001 seconds in duration.

Of course it isn't. It is a continuing wave that has many many cycles. Those reflect off the walls too (even more so then the bass) and interact with the later cycles of the same wave.

Not only that all the reflections are interacting will all the direct sound...... same or different frequencies... it doesn't matter it happens either way. The higher frequency reflections also have the problem of smearing imaging as they will alter the directionality of the apparent sound source. In the deep bass there isn't nearly as strong of a localization to the reflections.

You keep fixating on a single cycle which really doesn't mean much as you don't listen to music that is just a single cycle in length.

And when talking about bass when the wavelength gets long enough that reflection is actually a benefit.

Why?

Because the wave is so long the reflection is still in phase with the direct sound and the two 'couple' (like putting a speaker in a corner) to increase the amplitude of that wave. That is what room gain is and it can be used to extended the bass response of a well designed room/sub/woofer.

This is the same thing as what PWK wrote about in various DFH with regards to apparent/virtual bass sources on a corner loaded speaker. The wavelengths get so large all the boundries in the room can become virtual bass sources.

"You can't hear an unaltered uncolored (re: TRUE) low frequency waveform in a room shorter than the wavelength being produced or the modes take over and that is what you hear."

You do realize that room 'modes' are just changes in amplitude, not frequency....right?

When listening in an average room you are hearing reflections from pretty much the entire spectrum. So by the same logic you aren't hearing any of the music 'uncolored' if you listen in a room anyway.

Again, there is no argument that a room will alter the amplitude response of a speaker. What it won't do is to turn a 20hz wave into a 60hz wave.

Shawn

[D-MAN]

The unalterable fact is that in order to hear the true low frequency waveform without coloration it has to be in a space large enough to allow the full expression of the wavelength to occur without reflection.

That isn't to say that you would not hear any "bass". You just wouldn't hear the waveform as originally produced.

In other words, the sound heard would be COLORED by propagation effects that are not part of the original waveform.

DM

[sfogg]

"The unalterable fact is that in order to hear the true low frequency waveform without coloration it has to be in a space large enough to allow the full expression of the wavelength to occur without reflection."

No, the unalterable fact is if you want to hear it without room induced amplitude changes (what you call coloration) you have to listen in a space with no reflections at all. Quite likely you wouldn't like listening to music in stereo in that sort of setup.

Just because one cycle of a wave would fit in the space or not really makes no difference to what you are arguing for. The room may be large enough to fit one cycle of a 20hz wave but what about two cycles of that wave? You still get reflections... the first cycle of the wave will interact with the second cycle of the wave.

And again this is not limited to the bass range.

You will hear *exactly* the same amount of 'pure' (no room influence) sound from a 20hz tone as you would from a 20kHz tone before the reflections kick in.

Speed of sound doesn't change by frequency.... it takes just as much time for a 20hz tone to pass you, hit the back wall and reflect and travel back to you as it does a 20kHz tone to do the exact same thing.

Shawn

[dragonfyr]

Danger! Danger! Warning Will Robinson!

Watch out folks! If you are not extremely careful, you might discover what acoustical analysis and those TEF analyzers and time domain spectroscopy (TDS) and maximum length sequence analysis (MLS) tools are useful for! Not to mention modeling tools such as EASE/EARS, CATT-A, Ulysses and other...

But of course its your choice! You can go back to first principles and invent them, or you can become familiar with what it is that they look at, how they look at it, and what useful information they provide, along with how that information is translated into a useful basis for affecting productive modifications.

Couldn't resist!! Have fun!

[D-MAN]

I haven't changed my response from last time, guys.

This is to back up what I said before, and I think it does.

The supposition I presented was that frequency changes with phase-related (reflected) overlayment results in modal effects which effects both amplitude and frequency when compared to the original.

Everyone knows that a given low frequency when mixed with another (even the same frequency but with phase differences due to propagation issues) will result in

1) new frequency modulations,

2) differing amplitude(s),

3) different over-and-under harmonics patterns.

All of these are frequency and amplitudinal changes.

AKA: comb-filtering effects. It's lumped into the term COLORATION, but it's all distortion of the original frequency.

DM

[dragonfyr]

----------------

On 6/6/2005 2:50:59 PM D-MAN wrote:

I haven't changed my response from last time, guys.

This is to back up what I said before, and I think it does.

The supposition I presented was that frequency changes with phase-related (reflected) overlayment results in modal effects which effects both amplitude and frequency when compared to the original.

Everyone knows that a given low frequency when mixed with another (even the same frequency but with phase differences due to propagation issues) will result in

1) new frequency modulations,

2) differing amplitude(s),

3) different over-and-under harmonics patterns.

All of these are frequency and amplitudinal changes.

AKA: comb-filtering effects. It's lumped into the term COLORATION, but it's all distortion of the original frequency.

DM----------------

Of course! The combination of the various wave forms with varying start/arrival times into a resultant wave form is called "superposition".

And what you observe are some of the resultant anomalies in the frequency domain!

Now you have decided that the problem exists in the frequency domain, primarily in the form of comb-filtering, room modes, and unintellibility.

Now, the $64 question is, how can SPL meters, RTAs and EQs break the resultant summed effects down into their constituant parts? And assuming they could, how would they 'tell' you the non-linear characteristics and relationships of those components, sufficiently enough that you could address these components and possibly correct for them?

The fundamental problem is...they cannot!

Hence your paradigm and frame of reference must change!

Welcome to the time domain! ...as eerie music complete with frequency and polar aberrations plays in the background...

[pauln]

I have to go with D-MAN on this one. Anyone who has heard the basses and cellos playing in an auditorium know that this sounds much "stronger" but not louder than the same stuff played at home in rooms. The "stronger" aspect is the long wavelengths given lots of room to propagate many wavelengths before reaching boundarys where they are reflected, distorted, and absorbed to various degrees.

Confining a long wave in a short room does cause anomolies, boominess and overemphsised bass. Go listen to an orchestra in a large space - the bass is low and strong - but not loud and boomy like in a residential room.

In the same way that you can't defeat the physics of the speaker in terms of bass response, you can't escape the physical limitations of the listening room either.

Paul

[D-MAN]

----------------

On 6/3/2005 8:15:53 PM sfogg wrote:

That in no way shape or form supports what you were saying in that it is impossible to reproduce a wave that has a wavelength that is longer then your rooms length.

Shawn

----------------

I don't believe that I ever said that. I said that it's impossible to experience a TRUE 40 Hz waveform in a room smaller than the wavelength in question in length.

If I did say that of which I am accused, I was wrong. The low frequency waveform will propagate into the space, but what happens to it is a matter of the properties of the enclosing space.

DM

[scriven]

----------------

On 6/6/2005 4:55:48 PM D-MAN wrote:

... I said that it's impossible to experience a TRUE 40 Hz waveform in a room smaller than the wavelength in question in length...

----------------

Why pick on 40 Hz? That is true of the full audio spectrum!

[sfogg]

"I don't believe that I ever said that."

You might want to re-read that thread. You claimed a room had a low frequency cut off below which reproduction wasn't possible.

http://forums.klipsch.com/idealbb/view.asp?mode=viewtopic&topicID=63516

"I said that it's impossible to experience a TRUE 40 Hz waveform in a room smaller than the wavelength in question in length."

You added that in later but neglected (and still are) the fact that music is not a single cycle long.

Shawn

[WMcD]

My interpretation of the data is that room modes can result in boost. The bigger the room, the lower the modes possible. This is common knowledge.

But, any size room can contain transmit sound to your ear without modal effects. The sound curves shown probably just show normal roll off of the woofer being used to drive the system (room), not the system response below the peaks.

Were this not true, we would not hear any bass in automobiles. Obviously we do.

I've not ever seen any reports that room effects distort waveforms via clipping or anything else to create new frequencies of a single test tone.

What response there is to a single test tone in any room will be an addition of some reflections at different phases and magnitudes due to reflections and the attendant delay.

There may be an underlying issue which is questioned. It is legit to wonder.

Suppose you start with a pure sine wave. Then you add in the same frequency at different phase and levels (per reflections aka delays). The first thought is that there must be some waveform distortion from the original pure source due to the additions; and waveform distortion means different frequencies . . . or does it?

There is a lot of math behind it. However, the resulting sum is still a pure sine wave of the original frequency; it (the sum) may be altered in phase and magnitude, but that is all. It seems like voodoo book keeping; though true.

I've suggested this before. Put on a pure test tone and walk around the room. Try 200 Hz and below as a start. (I see the Audacity freeware generates tones.)

You will be able to detect, by ear, the anti-nodes where sound levels drop to zero as if someone turned off the amp . . . at that spot, and others. It is the start of realizing room effects first hand. It is the difference between reading about it and wondering, and knowing it first hand.

It is most easy to hear with bass notes.

Back to the original point: There is never any new treble frequency; or any lower frequency either.

Gil

[pauln]

"However, the resulting sum is still a pure sine wave of the original frequency; it (the sum) may be altered in phase and magnitude, but that is all."

Sounds like voodoo because it is untrue. Anyone can generate two sine wavesof the same frequency and shift the phase of one relative to the other and clearly hear the "eeaaoowwuueeaaoowwuu..." The shifting superposition is as if additional harmonic overtones are being injected and withdrawn from the signal. (A changing sound is not a pure signal).

Consider also that the "big wave in the small room" problem is much more complicated than drawing waves as if the world was flat and two dimemsional and amenable to analysis on graph paper. The 3D world (D as in D-MAN?) requires that variably phase shifted and variably level attenuated copies of the primary source sound impinge that primary sound in 3D space as said sound is propagating from virtually all 3D angles of attack so that the composite superposition of the wave is not a more wiggly line on a graph but a much more complicated pattern not unlike a three dimensional fingerprint in space constantly moving, changing, interacting with itself, and sounding different in different places at different times (long sentence - gasping for breath!)

I heard two large tankers fighting for right-of-way in the ship channel one afternoon while sailing in the bay. They were about 3 miles away and the sounds of their huge deep horns were amazing. Not that loud, but there is no way it could be reproduced to sound like it came across the open ocean the way it did in an enclosed room.

[WMcD]

Sure, you want to phase modulate a sine wave in time and mix it with other, fixed sine wave, you get oddball stuff.

Can you phase modulate a sine wave in time with a room having fixed walls? Not that I can see.

Gil

[D-MAN]

Phase changes with the angle of reflection and the corresponding comb-filtering that occurs when the reflected wave intersects the original. In the case of low frequencies in small rooms (say 15 ft to back wall), its still the same waveform; the reflected waveform will intersect the "still coming out of the speaker" waveform somewhere in the middle (i.e. 40Hz=32 ft long wavelength per cycle or 16ft positive, 16 ft negative phase).

Concerning reflection (quote):

As the angle of incidence approaches 90°, the angular wave impedance approaches infinity, so that B = -A, approximately. There is again total reflection, but in this case the phase of the reflected wave is reversed, as in a dense-to-rare reflection, and C = 0. (source Link)

Comb-filtering occurs when waveforms (even the same frequency) intersect at angles. The results are numerous apparently spurious with changing frequencies and all of them are detrimental (i.e., distortions of) to the waveform(s) being propagated.

The room modes will occur in addition to the reflected wave-induced comb-filtering. The modes of course are frequency dependent.

DM

[dragonfyr]

----------------

On 6/6/2005 10:40:11 PM William F. Gil McDermott wrote:

Suppose you start with a pure sine wave. Then you add in the same frequency at different phase and levels (per reflections aka delays). The first thought is that there must be some waveform distortion from the original pure source due to the additions; and waveform distortion means different frequencies . . . or does it?

There is a lot of math behind it. However, the resulting sum is still a pure sine wave of the original frequency; it (the sum) may be altered in phase and magnitude, but that is all. It seems like voodoo book keeping; though true.

Back to the original point: There is never any new treble frequency; or any lower frequency either.

Gil

----------------

D-Man

"Comb-filtering occurs when waveforms (even the same frequency) intersect at angles."

Gil, superposition of waves does not always result in a sine wave, as has been pointed out.

But your example is, if it assumes the lack of reflected signals, the only case where the result of summing two sine waves is itself a sine wave exhibiting the same frequency and the same phase, but with greater gain. To illustrate the other extreme, 2 sine waves of the same frequency and 180degrees out of phase result in 'no' waveform.

But in the real world, this is compounded by the fact that the arrival times of the different wave forms, their phase, varies with respect to one another, resulting in complex waveforms exhibiting neither 'pure' sine wave nor consistent gain characteristics.

And then, two waves of different frequencies and phase variance as well as different gain structures present an infinite number of possible variables... At which point fourier analysis can be used to determine the fundamental component frequencies...

Oh, and D-Man, the critical relationship that results in the effects you describe are not the "angles", but rather the difference in time - the phase- of the wave forms. But this difference in phase is generally the result of the reflections of the signals that reflect (no pun intended!) a longer wave path and a correspondingly greater delay in time.

A pretty good illustration of this with visual examples is found at: http://www.kettering.edu/~drussell/Demos/superposition/superposition.html

[Erik Mandaville]

Dragonfyr:

"A pretty good illustration of this with visual examples is found at: http://www.kettering.edu/~drussell/Demos/superposition/superposition.html"

This was interesting! Shawn gave me another example using water as an analogy, and that helped make this clearer for me. I've been following along, and was curious how LF waveform movement(thus, the moving water analogy works well!)would appear visually. It's no wonder how truly critical a part of the equation the listening environment is!

Thanks,

Erik

[D-MAN]

----------------

On 6/7/2005 1:06:13 PM dragonfyr wrote:

Oh, and D-Man, the critical relationship that results in the effects you describe are not the "angles", but rather the difference in time - the phase- of the wave forms. But this difference in phase is generally the result of the reflections of the signals that reflect (no pun intended!) a longer wave path and a correspondingly greater delay in time.

----------------

Agreed. However, let us assume the wavefront is semi-spherical, and not planar. That would mean that the resulting interference pattern is "more" dependent on angle of intersection than a simplified planar intersection of two waveforms. The interference combing would occur differently dependent on the angle (path).

The analogy would be one of intersecting the leading surfaces of bubbles merging together at different angles?

Maybe not.

The point being is that the any angle other than the original soundwave path REPRESENTS or EMBODIES time and phase differences from the original soundwave. Even two speakers (mids/highs) with different pathways display time and phase differences between them although they may be physically close together and produce the same frequency.

DM

[dragonfyr]

----------------

On <?xml:namespace prefix = st1 ns = "urn:schemas-microsoft-com:office:smarttags" />6/7/2005 2:21:00 PM D-MAN wrote: <?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" />

----------------

On 6/7/2005 1:06:13 PM dragonfyr wrote:

Oh, and D-Man, the critical relationship that results in the effects you describe are not the "angles", but rather the difference in time - the phase- of the wave forms. But this difference in phase is generally the result of the reflections of the signals that reflect (no pun intended!) a longer wave path and a correspondingly greater delay in time.

----------------

Agreed. However, let us assume the wavefront is semi-spherical, and not planar. That would mean that the resulting interference pattern is "more" dependent on angle of intersection than a simplified planar intersection of two waveforms. The interference combing would occur differently dependent on the angle (path).

The analogy would be one of intersecting the leading surfaces of bubbles merging together at different angles?

Maybe not.

DM<?xml:namespace prefix = v ns = "urn:schemas-microsoft-com:vml" />

You would mention this! (Youll pardon me as I have a big grin, as this is exactly the type of issue that enters into the real world and is often overlooked by simpler modeling methods!!!)

Yes, Wave fronts are classically modeled as spherical, and easily viewed as such in 2space wave tanks as a few of you have already discovered! (Wonderfully illustrative techniques arent they!), with the velocity of the wave propagation relative to a point being uniform. But this model becomes a bit more complex in reality.

Suffice it to say, that we more generally deal with the superposition of waveforms at a point in space, and not distributed on a plane in space. This alleviates the complex modeling which will render such intersections to be very complex non-linear differential waveforms, as the variations for a particular waveform are not evenly distributed throughout the frequency spectrum due to various reflective and absorptive non-linearities. Does this make any sense!? If not, let me know and I will try to be a little clearer! But this area very quickly becomes very complicated and will quickly necessitate that the conversation be conducted in the necessary evil world of complex non-linear differential calculus! Unless, that is, you have access and can properly use some pretty neat analytical tools! And let me mention that they do not render the knowledge of the mathematical foundation unnecessary, but they do solve the issue of having to do many of the calculations! And that sounds pretty good to me!

(And if that isn't enough, as we begin to deal with larger spaces there are additional effects imposed by such additional variables as temperature gradients - acoustic shadows, etc. - although in small rooms acoustical temperature anomalies are not usually a problem!)

Sorry, but the tangents do occur! So you can see why it is nice to be able to limit our mention of these realities, or at least to address the resulting characteristics graphically!

Ok, so those are a few of the complexities that we have to deal with! So how do we make this complex system a little more manageable?

In practice the propagation of waves in 3space is generally denoted by the polar response. This is in effect the linear measurement of the summed point measurements displayed within the horizontal plane, or vertical plane, or in 3space by volumetric 'balloon' image projections. Likewise there are a variety of methods for visualizing such sound fields using modern techniques. Attached is a file that shows several of the more common ways of modeling and representing this. And while you will see that pictures are easily worth a thousand (or more!) words, it is difficult indeed to express this complex data in words - hence our more common 'flat' representation.

The more complex acoustic modeling tools utilize advanced graphical modeling capabilities to display this complex model graphically via 3space volumetric balloon projections and with intensity projection plots that appear similar to a topographical map.

And many of these tools allow us to take a resultant 3space sound field, and using identical methods to those used in analyzing an MRI scan, we can slice the volumetric image in to planes corresponding to our focus of concern.

And this becomes important as we are not only dealing with the quality of sound in one point in a room, but with the room response at many points. And as we increase the number of points we begin to deal with larger fields. And this is exactly what is dealt with as the room or space being analyzed increases in size and listening positions and our model becomes more complex.

I will also post a couple of other papers that will address various related aspects of representing sound fields and how signals interact, but this discussion is at the point in where everything explodes into the various tools and frames of reference from which the various phenomena can be measured and observed, and thus it is not easy to reduce the subjects scope! In fact, this is where it really begins to become interesting and the complexity and elegance of acoustics makes itself known!

PolarDataAltViews.pdf

[dragonfyr]

Here is another useful tool and modeling method for displaying varying wavefronts .

FarFieldCriteriaLdspkrBalloonData.pdf

[dragonfyr]

sorry, I guess I got so excited that I posted the same attachment twice

FarFieldCriteriaLdspkrBalloonData1.pdf