[ekovalsky]

Just thought of another interesting thing about my room/set up.  The 10" woofs of the RM30s fire toward each other, which run full range, for whatever that is worth.

That should enhance the bass in the center of the room.  The primary waves will collide and enforce each other, at least if they have the same signal coming from them.

[John Casler]

In any event, where I was originally going with my front back "push pull" was ultimatly seeing if any DSP system could take readings from the listening position and then adjust the fronts (alone) to function maximally, then integrate the rears with whatever phasing/delay/advance/amplitude/FR to achive the ultimate bass.

I would think that all the technology and measurment sophistication is already available to create such a system "within" a DEQX or TACT system, or even a Behringer.

It would be a matter of creating the needed software and measurment means.

Because in the end, phase is just a timing/action measure and this will vary depending on the relationships and distances, just like the phase relationships between the subs and the mains.

Maybe, you should not put a back wall on your new room?  Then you wouldn't have to worry about sound reflecting off it.  What a blessing that would be.

[youngho]

Youngho,

It's not the listener to sub distance that is important with the out of phase, dipole setup.  It is the DIFFERENCE in distance from the listener to the in phase and out of phase subs that is key, however, that is complicated further once you get off axis because there are 4 subs involved.  If his subs were all equally distant from the wall, then a central listening position would have almost zero SPL from the subs with them wired out of phase.  Dipole takes the irregularities in room shape out of the equation to a great extent.

John, I don't think that John Casler's system would act as a dipole(s), given that the listener is closer to each than the poles are to each other and that the listener occupies more than a point in space. As I understand it, the directivity pattern of a dipole results in cancellation when the listener is further away from the dipole than the poles are to each other, but the radiation patterns of each driver overlap in an infinitesimally thin line when one gets so close to the dipole itself.

[youngho]

Just for clarity's sake, I understand the Harman solution to represent two subwoofers in a rectangular room, driven in phase, located at the midpoints of two opposing walls, with the listener located near the middle of the room, as suggested in this white paper: http://www.harman.com/wp/index.jsp?articleId=1003.0

Thinking through the mathematics of the DBA on my drive to work this morning, I think I understand the Harman solution better, as well as its limitations. Although the Harman solution is relatively simple, it only eliminates first-order modes. By locating the subwoofers in the node for odd-order width-related modes, it does not energize these modes. By using mode cancellation by having a driver on each side of the length-related mode, it cancels the odd-order length-related modes. However, it does not seem to address even-order width- and length-related modes.

The lowest axial mode freqency f might be calculated by the equation 1130/2x=f, where x represents the axis is question. Thus for a 28 ft long room, this resonant frequency would be 20 (and multiples thereof). The duration of one cycle would 1/f.

The length of time t that it takes a sound wave to traverse that dimension would be x/1130. Substituting for x, we see that 2t=1/f. In the case of a single subwoofer, this would mean that the first wave front reflects off the rear subwoofer, reverses phase, reflects off the front wall, reverses phase again just as the subwoofer is entering into its second cycle. t should also represent the delay that one would need to set for the DBA.

Thus, in the Harman solution, when the wave front of the lowest axial mode frequency from the front subwoofer is reflecting, the rear subwoofer is exactly 1/2 way through its cycle, meaning that it cancels out this odd-order mode and its multiples. When the wave front of an even-order axial mode frequency is reflecting, the subwoofer is beginning a new cycle, meaning that it does not cancel this mode.

The primary advantage of the DBA would be that it cancels both odd- and even-order modes. If done with only a pair of subwoofers, the listener would still have an issue with width-related even-order axial modes, but done with two pairs of subwoofers, with the subwoofers located at the nodes of even-order modes and symmetrically flanking the central node of odd-order modes, this should address all width- and length-related axial modes, leaving only the height-related modes as potential problems.

The primary disadvantage of the DBA, then, would be the relative sophistication of the method itself, at least for now, as most listeners do not use electronics capable of setting delay so precisely. Also, it seem as only neighbors on the other side of the rear wall would benefit in terms of transmitted noise, no?

[ekovalsky]

youngyo,

You seem to have a pretty good grasp at it.  Reduction of SPL outside the listening room will mainly be at the rear wall as you say, but since reflections (and modes with SPL peaks) are reduced there should be some improvement along all walls.

In my case, the only connection between the dedicated audio room and the rest of the house is the rear wall -- the other three walls are all exterior.  So reducing transmitted SPL through that rear wall allow me to do my serious listening, which is mostly at night after the family is asleep, at higher levels.  Because of the size of our house and the doubled rear wall and insulated sliding doors I can listen and very satisfying levels currently, but sometimes I have wished I could crank it up a bit higher!

As for electronics a Behringer Ultracurve will do the trick, it is less than $200 on Ebay.  It also has extensive parametric EQ capabilities -- its primary function.  Of course four subs are also needed, that is where most of the cost will be.

[JohninCR]

Youngho,

I agree with you that calling it a dipole when you are listening between the poles probably isn't an accurate term, however, I do believe many of the same principles apply especially with regard to room interaction.

The bottom line is that using an equal number of in phase and out of phase subs puts zero net acoustic acoustic energy into the room.  This means less reflected energy and less transmission outside of the room, which in turn means that more of what you hear is directly radiated sound.  The trick is placement that results in a satisfactory response.  You may also have to deal with a sloped response, but a 1st order filter takes care of that simply and effectively.

The alternative is to run everthing in phase and try to correct everything through processing.  In it's typical form this processing only flattens the peaks and valleys in response.  As you explained, in more advanced forms you can even cancel modes on one axis.  IMHO, this is all just a lot of tail chasing because you still end up with a highly energized room with all those reflections and their destructive influence.

[John Casler]

I'm starting to get to the point where I think the best listening room might be an outdoor "tent" with acoustically treated lining.

The walls would be too thin to create bass problems and Mid and HF reflection could be easily treated out.

Maybe after I win the Lottery, and pick up the "Malibu Compound", I'll build the "Listening Tent" and see if it flys.

Or maybe a Band or Concert Shell (Minature of the Hollywood Bowl) with "NO" walls in the listening area.

Hmmmmm.

Can you imagine, a whole new Audiophile Movement of "Outdoor Audiophiles"    

Oh and to be sure, I never called "MY" Push/Pull set up a dipole.

[youngho]

Oh and to be sure, I never called "MY" Push/Pull set up a dipole.

John, thanks for the clarification. This is actually why I referred to you in the third person when discussing your setup with JohninCR, but I did not make it clear.

However, I don't know that I would refer to your setup as a push/pull setup, either, as I believe that this traditionally refers to an arrangement where the front of one driver and the back of the other driver are facing the listener. In addition, my understanding was that the drivers need to be acoustically coupled, which implies close proximity. Perhaps I am mistaken.

I agree with you that calling it a dipole when you are listening between the poles probably isn't an accurate term, however, I do believe many of the same principles apply especially with regard to room interaction...The bottom line is that using an equal number of in phase and out of phase subs puts zero net acoustic acoustic energy into the room.

I'm sorry, but I believe that this is not accurate. Room interactions will be affected to different extents, depending on the frequencies, the distances between the subwoofers, and the distances between the subwoofers and the listener.

Perhaps I am mistaken, but take my previous example of two subwoofers set up according to the Harman method but driven out of phase--one subwoofer in phase located at the midpoint of the front wall, and one subwoofer out of phase located at the midpoint of the rear wall. What actually happens? All odd-order axial modes should be amplified due to constructive interference. This would hardly represent zero net acoustic energy into the room for these frequencies, right?

[JohninCR]

Youngho,

In your example, 2 subs each in the center of the opposing wall firing out of phase and at the same time would cancel out all of the modes on that axis, wouldn't it?  In fact in the center of the room, you would have zero amplitude regardless of frequency and where all of the modes would occur would be nulls instead.

Yes, zero net acoustical energy is input into the room.  If you need proof, just put 2 subs facing each other fairly close together and they won't put out any sound no matter how loud you play them.  Just because you move them around so you can hear sound doesn't change the fact that it's a net zero input.  That's not to say you can't stimulate modes depending upon your placement, but the net input will still be zero, which logically has to stimulate room modes overall to a lesser extent.

[youngho]

Yes, you're right. I've even forgotten what point I was trying to make.

[klh]

That was funny

[John Casler]

Yes, you're right. I've even forgotten what point I was trying to make.

That seems quite easy when we are trying to "wrap our heads around" just what is going on.

Here are some questions:

1) If you have a single subwoofer playing a sustained test tone, and the initial wave has reflected off the rear wall, returned, and now has reflected off the front wall, what does it do to the direct wave as it "sums/syncs" with it as it travels with the direct wave to the listener again?  Will it amplify? or nullify?

2) We have two subs on the front wall staggered so that one is behind the other.  They both play the same tone.  The front sub is "delayed" to account for the distance of stagger, so that they are both "in phase" with each other.  Will their "summed" outputs amplify?  or Nullify?

3) We have two subs. One on the front wall, the other directly behind the listener facing the rear wall, with the rear sub "in phase" with the front sub. (this means both woofers move in the same direction simultaneously) Will their output amplify, or nullify?

4) We have two subs.  One on the front wall, the other on the rear wall facing the listener, but "out of phase" (again this means both woofers move in the same direction simultaneously)

5) We have two subs. One on the front wall, the other directly behind the listener facing the rear wall, with the rear sub "out of phase" with the front sub. (this means both woofers move in different directions simultaneously) Will their output amplify, or nullify?

6) We have two subs.  One on the front wall, the other on the rear wall facing the listener, but "in phase" (again this means both woofers move in the opposite direction simultaneously or launching toward each other)

Hope that is easy enough to follow

[csero]

John,

Subs are omnidirectional speakers and  box speakers are generating pressure waves. It does not matter in which direction (relative to the listener) the cone moves, only matters if it creates a compression or a refraction. Subs are "in phase" when they both create compression at the same time.
Arrival time differences (of the direct vave and the multiple reflected waves) at the listener position create interferences depending the total path length and attenuation of the arriving waves.
Speculating without the knowledge of exact dimensions and absorption/reflection properties are pretty much useless. If you know these, the problem could be handled much easier if you treat every reflection as an omni virtual sound source with the proper distance and amplitude in freefield.

[JohninCR]

1.  Only with regard to those rear and front wall reflections, it depends upon the distance between walls.  Where the tone wavelength = 2 times the distance between the walls (distance = 1/2 wavelength) , those reflections will be exactly in phase and reinforce each other repeatedly.  That is your primary mode.  Multiples of that frequency will do the same thing along that reflecting axis.  Placement will be the key to what happens because the peaks and troughs are both reflecting in phase.  If you cut a hole in the wall and mounted the driver at edge of the wall you would create a standing wave at the mode frequency .  Even though everything is reinforcing the wave the wave peaks and troughs would be at the same point and you'd have no sound.  Move the sub away from the wall and the peaks and troughs wouldn't line up and you'd have big peaks and nulls at different points along that axis at different modal frequencies.

2.  The modes still come into play.  Most of the 2 waves is in phase and travelling the same direction and they sum, so you are going to create a standing wave at the room mode frequency for a big chunk of the output.  

The thing to remember is sound from a sub radiates in a sphere.  When you put one against the wall, all of the energy radiates into half a sphere (ignoring the floor).  The one out from the wall will radiate in a sphere, so some of the delayed output travels with the wall sub's output, in phase, and some goes and reflects off of the wall before travelling in the same direction delayed by double the distance from the wall.  It wouldn't be the same as 2 subs against the same wall in phase.  Similar, but different.

3.  Difficult to say.  Whether it's facing front or back doesn't really matter except facing back makes it the cab depth farther from the listener.  The sub right behind you will be louder because it is closer.  The sub against the wall will be louder than if it was into the room because it receives the boundary reinforcement discussed above.

4.  I think where you are getting confused on this is the moving the same direction simultaneously.  It doesn't matter what direction a sub faces.  One could face left and there would still be the same result.  The key is one is compressing the air, creating a wavefront, and in unison the other rarifies the air and vice versa (there is no net change in the air pressure of the room at any point).  2 out of phase wavefronts are created and would net to zero at some point even if the walls were perfect reflectors and air did not dissipate sound at all.  With 2 subs in phase they would keep building.

The bottom line is that 2 subs against opposite walls, out of phase, would have zero output at the exact center of the room.  As you move away from the center, sound would increase.  What you hear would be very frequency and position depedent.

5.  Same as number 4. except only one sub has boundary reinforcement.  With one so much closer, it will probably dominate what you hear.  Now if instead of behind you, you put it in front of you.  Then you have a dipole.   The SPL would be easier to predict and would depend upon frequency and the distance between the subs.

6.  This is just 2 number ones just more complex because you have the same thing happening but maybe in a different location.

It's all really too complex to try to explain simply and I have greatly simplified some things and ignored others, so I should probably just delete everything above.

Multiple subs, firing in phase increases the output into the room.  They will stimulate modes in all 3 directions.  Because they are originating from different directions this will have the tendency to smooth out the modes.  The peaks and nulls are still there, but they're happening in a bunch of different locations and overlapping.

Multiple sub with half firing out of phase, can still stimulate modes along the axis of the sub placement, but the side wall and floor/ceiling modes would be stimulated much less because the 2 out of phase outputs would net to much less at those perimeters of the room.  You would also get no room gain because there is no change in pressure in the room.

I hope I haven't stated anything erroneous and I hope at least some of it makes sense.