Dual vs single sub

mijostyn wrote: "I think Noble is on the right path here..."

And I think you’re on the right path too. Your idea for spacing four subwoofers along the front wall will greatly improve their interaction with that end of the room.

But here’s the thing: The rest of the room matters just as much. Here’s why:

Unless you have a very large room, by the time you even BEGIN to hear the bass energy, reflections off the rest of the room’s boundaries are fully in play. So Noble’s approach of modal smoothing throughout the room is imo better aligned with known psychoacoustic principles.

That being said, once you have four subs, there are a LOT of different things you can try.  See the writings of Earl Geddes and Todd Welti. 

Duke

"One question concerning 4 subs - How are they connected to the amp?"

The amp I supply with the Swarm is the Dayton Audio SA-1000, part number 300-811 at Parts Express.

It has a single channel of amplification but two sets of output binding posts wired in parallel.

The subs are connected in series-parallel, such that the four 4-ohm subs present the amp with a 4 ohm load.

Two subs are connected in series (via binding posts on the bottom), forming a two-sub "series string". Then the other two are also connectged in series, forming a second two-sub "series string". Then the two "series strings" are each connected to one of the sets of binding posts on the back of the amp, which puts them in parallel with one another.

Going into a bit more detail, the first sub in each "series string" has two sets of binding posts. One set is connected to the amp, and the other set is connected to the second sub in the "string", which only has one set of binding posts.

So you’d need four sets of speaker wire: Two sets go between the amp and each "first" sub of a "series string", and the other two sets of speaker wire go between the "first" and "second" subs of each "series string".

Some people use two amplifiers, typically one for each series-wired string, for a bit more flexibility.

"Stacked spades/bananas?  Seems like a lot of hardware to me."

No stacked spades or stacked bananas.

I guess it’s a lot of speaker wire. Unless the runs are really long, sixteen gauge wire works just fine. You end up with the equivalent of thirteen gauge wire because of the series-parallel topology.

Duke

Editing timed out before I could fix one sentence; here’s how it should read:

"The array you described only tries to provide significant smoothing in one dimension, and it primarily does so at one end of the room (your four subs only approximate a line source; each sub still behaves as an omnidirectional point source).

Duke

"If you create a linear array very little energy disperses off the top and bottom or in this case to the side walls minimizing that interaction."

The array you described only tries to provide significant smoothing in one dimension, and it primarily does so at one end of the room (your four subs only approximate a line . The same number of subs arrayed differently can result in significant smoothing in two or even three dimensions (if you elevate one of the subs so that it’s closer to the ceiling than to the floor), and this smoothing will extend throughout the room.

"The only thing that really works here is just get rid of the back wall. I am fortunate in that I designed my media room without a back wall."

Relatively few people have that option. I have a hard enough time selling four subs without also requiring customers to remove their back wall! In other words, regardless of how well what you are doing works for you, it is not a generally-applicable solution.

"I am not sure if spreading other sub woofers around the room would minimize this or just create more complex patterns."

More complex patterns of modal behavior is how modal smoothing naturally occurs in a large room. The reason a large room has smoother (faster, better, more natural-sounding) bass than a small room is that the larger dimensions result in greater de-correlation of the in-room bass energy. Greater decorrelation = smaller and more numerous peaks and dips. A distributed multisub system mimics large-room modal behavior in a small room.

Duke

"The best you can hope to do in any enclosed space approximating a livable room is to control primary reflections and since they have the most energy that represents 90% of the battle."

I disagree with this statement as far as the bass region goes.

From a perceptual standpoint, speakers + room = a minimum phase system at low frequencies (according to Toole and Geddes and Welti, among others), which means that the ungated in-room frequency response is what matters. Focusing on the first reflections is at least paying attention to room interaction, but imo it’s not focusing on solving the problem that matters the most to the ears in the bass region.

"Putting subs all over the room simply creates more primary reflections."

Primary reflections in the bass region are not the primary problem that needs solving. But having many primary reflections contributes to solving the primary problem.

"... room control [EQ?] frequently has to correct troughs of up to 10 db and more..."

With a decent distributed bass array, you will not begin to have troughs of 10 dB or more. If your technique results in troughs of 10 dB or more then it is not solving the problems that matter. I don’t claim my customers will necessarily get plus or minus 3 dB in-room in the bass region, but many of them have reported that.

"Oh, by the way Duke, I am not trying to sell anything."

Okay. Maybe I’m missing something. How is that relevant? Please clarify because I don’t want to jump to conclusions.

Thanks.

Duke

Several posts above I wrote: "your four subs only approximate a line source; each sub still behaves as an omnidirectional point source."

Presumably in response, mijostyn wrote: "4 subs at 4 foot intervals with the outer two in corners does not approximate a linear array. It is most definitely a linear array up to 125 Hz or so."

Okay, but a line array is not a line source. A line array *approximates* a line source, which is what I said.

"If you review the acoustical properties of linear arrays you will note that they throw very little acoustic energy off their ends almost totally killing any refection off the side walls of the room

Imo their behavior at their ends isn't a potential issue; their behavior out in front of their ends (along the sidewalls, in this case) is a potential issue.

I know how to model line arrays. The spacing you used is not optimal. I can explain why if you are interested.  Because its propagation is not perfectly perpendicular to the line, your array will have increasing sidewall interaction with distance. I’m not saying that’s a fatal flaw, only that it happens.

"No other array does this in any dimension."

A planar array does this in two dimensions. In fact, I think a planar array would make more sense than a line array for your situation, given that your room doesn’t have a back wall (which imo offers you a unique opportunity). Again, I can explain if you are interested.

Duke

"Duke had told him OB/Dipole subs would be a better choice for use with his loudspeakers."

"when I told Duke what drivers I was going to use he said mine will kill his."

Okay that’s it. I’m firing my marketing department!!

Seriously thank you both, and thank you too Tim. I try to promote the distributed multisub concept when I post here rather than my specific product, but that distinction probably doesn’t always come across. My product is really just one possibly way of doing it. I would rather spread a good idea (which Earl Geddes developed and freely shared with me) than sell a lot of subs. (This isn’t as altruistic as it might sound - I have enough orders on the prosound side of my business to keep me busy full time.)

Nor is a distributed multisub system necessarily the most sensible approach for every situation.

If mijostyn takes my posts with a grain of salt, that’s fine. He has every right to be skeptical; I would be if our positions were reversed.

Duke

"I don’t believe that virtually all CDs have bass frequencies summed."

That’s okay. I didn’t either and argued with Earl about it. I forget the details of what he said but he had been sufficiently thorough in his investigation to change my mind.

It’s easy enough to do stereo bass anyway if you want to.

Duke

"The Ford Motor Company?"

Ford hired Dr. Earl Geddes as a consultant.  One of the questions he wanted to know the answer to was how common is it for CD's to have stereo bass content.  He asked everyone in the department to bring in their CD's.  After testing about a hundred CD's and not finding a single one with stereo content in the bass region, he stopped testing.  So no Ford did not test virtually all recordings, but the sample size was large enough for Earl to reliably conclude that stereo bass on CD's is quite rare. 

Duke

"I don’t believe it. No offense to you personally."

Could you clarify?  What exactly do you not believe? And, no offense to you personally either.

"I’ll take the stereo bass. It sounds better."

If you would be willing to list some recordings that do have true stereo bass, I would appreciate it very much.  I would really like to try some known-stereo-bass recordings on my system and see how they sound. 

What if you could get a realistic sense of immersion/envelopment from recordings that do not have true stereo bass, yet could easily adjust to take full advantage of true stereo bass, in about ten seconds?  Might that have some utility?

Duke

 "If we agree with Duke's statement that 2 subs in a room provides twice the bass smoothness of 1 sub and 4 subs in a room provides twice the bass smoothness of 2 subs..."

I was paraphrasing my mentor, Earl Geddes. 

Not only does the frequency response get smoother at a given location, but the frequency response variations from one location to another within the room also decrease (which follows from the first statement, but it may not be obvious). 

Duke

@phusis wrote, about the distributed multi-sub approach: "... not drawing on the potential benefit of high-passing the mains (from ~80Hz or so on up to make the most of it."

The distributed multi-sub approach has nothing to do with whether or not the mains are highpassed. Those are two separate decisions. I’m probably the one who caused the confusion because I’m a distributed multi-sub advocate but not an advocate of automatically high-passing the mains in every situation. I think it depends on the specifics.

"...it’s not my intention to bash the multiple sub approach, which I know can sound great, but for some to ease up a bit on this being the only proper way to implement subs in your home setup."

The distributed multi-sub approach apparently works well in many situations, but is not the only approach that works well, and in some situations it would not be the approach I'd choose.  It depends on the specific situation. 

Duke