Multiple Small Subs - Geddes Approach

[soongsc]

Originally posted by markus76


Where's the problem? Use a RTA and you know what the noise floor is and what impact it has on your waterfall diagram. In REW it's just one click ?it doesn't get easier than this.

Heh. I'm afraid I don't want to go OT to discuss that in detail. I Don't believe the manual explains anything to that much detail to ensure it will provide data that I had explained. Of course I can't expect people to understand this (as usual).

[soongsc]

Originally posted by gedlee


...
Simulations can only tell you so much about real rooms, because rooms can easily violate the assumptions. Rooms like mine will clearly violate the models and its exactly those violations that I would contend offer the greatest gain. Its those things that make a room NOT act like a simple room - a modelable room - that improve it the most. You don't learn this from a model.

If the real situation is better than the simulation, and simulation will show good results already, then it seems to be what we expect from simulations. This is what we look for in any engineering situation. If there is something measured that contradicts this, then we just see how simulation can be realistically improved. Really no big deal.

[gedlee]

Originally posted by soongsc
I think you are just afraid that the analysis reflects the reality of you recommended setup, and thus providing more credibility to john k's analysis.

Soongsc - I try and be patient with you but if you are going to level insulting accusations at me then I'm going to say to learn what you are talking about first and then you can insult me. You usually get things wrong, which you did here too.

[gedlee]

Originally posted by cap'n todd


ok. I thought we agreed that the "direct" term was a fudge, but acceptably small error. That is, assuming a (relatively low) order modal calc is made, then the direct is not included, therefor a simple additional direct term is ok.

Todd My point has always been that there is a right way and a wrong way to add in the near field "fudge". You and Walker do it wrong. If its a small effect then why include it at all, and if its worth including then its worth doing right.

[cap'n todd]

Originally posted by gedlee


Todd My point has always been that there is a right way and a wrong way to add in the near field "fudge". You and Walker do it wrong. If its a small effect then why include it at all, and if its worth including then its worth doing right.

But Earl, how do I include it without using a very high order, which takes too long? I tried up to 35 and didn't see much evidence of direct sound appearing (at a distance of 4" fromt the source). I did read the other day that using an area source rather than a point makes the summation converge faster. I'm not sure how to do that, but...

Also, I'm interested in having a seperate direct term, so thus the fudge. In any case don't forget that this is all a fudge if/since the absorption is non 0! THe model a fudge anyway, and probably in other ways too if you go through the derivations with a fine tooth comb.

[soongsc]

Originally posted by gedlee



Soongsc - I try and be patient with you but if you are going to level insulting accusations at me then I'm going to say to learn what you are talking about first and then you can insult me. You usually get things wrong, which you did here too.

Well, if you do not care to address the Green's function vs. FEA in SoundEasy part, I fully understand your position, which is that others whom don't agree with you are wrong regardless.

It's a simple fact that all the best analysis in the world cannot generate result that match real world situation. But we still use it to try to get as close as possible. I simply can't see why when john k is putting himself on the line to do the simulation/analysis but you are not willing to provide your reference setup basic information and compare simulation versus measured data. This just draws some speculation doesn't it.

[gedlee]

Originally posted by cap'n todd



But Earl, how do I include it without using a very high order, which takes too long? I tried up to 35 and didn't see much evidence of direct sound appearing (at a distance of 4" fromt the source). I did read the other day that using an area source rather than a point makes the summation converge faster. I'm not sure how to do that, but...

Also, I'm interested in having a seperate direct term, so thus the fudge. In any case don't forget that this is all a fudge if/since the absorption is non 0! THe model a fudge anyway, and probably in other ways too if you go through the derivations with a fine tooth comb.

I thought that I went through this with you before. Since the series does contain the direct field, adding it again is wrong. You can subtract out the direct field as a free space Green's function if you want, BUT then you must modify the remaining series to account for the part that you have subtracted. Morse shows how this is done for a membrane, but the exact same procedure works for the three dimensional acoustics problem and the remaining series is the same as Morse shows for the membrane.

It is true that a larger sphere as a source converges faster, but this is not so simple to do. Basically what happens is that the source is accurate beyond the radius of a 1/4 wavelength for your highest terms. Closer than that and the series solution is wrong. The bigger problem is the fact that your solution is real and thus cannot account for energy flow. If you look at the pressure gradient anywhere in the room, you will not find any net energy flow from the source or into the walls from absortion. In the interior of the room this is not a big factor, but near the source, where ALL the energy comes from, it is.

I did a paper in the early 80's showing this. A real solution will not find any energy flow from the source or into a sink. However a complex solution will. You will see energy flow from the source, but for a given limit of the series terms it only starts to appear at about 1/4 wavelength out. You will also see energy flow into a sink. We had a small patch, like an open window at one end of the room and the source at the other. The complex modes and eigenvalues were calculated with FEA. the energy could be seen to flow from the source, circulate arround the room as a complex pattern of standing and traveling waves and then exit the room.

It was this complex simulation that proved to me that source directivity at LFs in a room is a misnomer. The sound energy does NOT flow directly from the source to the sink but travels in a very complex, circuitous and indirect way. No directivity was ever evident at all.

Just for the record, I was doing room FEA calculations back in the late 70's. I had to wrote the code at that time as none existed to do acoustics. ANYSIS later adapted my techniques for use in their code. They used to credit this to me in their manuals. In the 80's I was doing complex modes and energy flow in rooms with FEA (NASTRAN). I'm not a novice at this.

[Patrick Bateman]

I'm really happy with the use of multiple subs. I started with three, and ramped it up to eight, mostly because my mains are so efficient. I am personally using a bunch of DIY subs, but it would be a lot less hassle to just go and buy a bunch of them.

If anyone reading this thread wants to give this a whirl, and do it cheaply, this might be an opportunity to do so.

Newegg has a $250 Polk subwoofer marked down to $100 at the moment. At that price, it's hard to go wrong. Try it out for a few weeks, and if you don't like the sound, go sell them on Craigslist. I wouldn't be surprised if you could RE-sell them for more than $100 each. They are currently $200/each at Crutchfield, and $150/each at buy.com. It's hard to find a plate amp for under $100, much less a finished sub.



Here's the manual:
http://www.polkaudio.com/downloads/m..._12_Manual.pdf

Here's the sub:
Polk PSW10

The promo code to get it for $100 is EMCLNLP28.

[gedlee]

I recommended that sub recently. A few of those would do the trick, and very cheaply.

[cap'n todd]

Originally posted by gedlee


I thought that I went through this with you before. Since the series does contain the direct field, adding it again is wrong. You can subtract out the direct field as a free space Green's function if you want, BUT then you must modify the remaining series to account for the part that you have subtracted. Morse shows how this is done for a membrane, but the exact same procedure works for the three dimensional acoustics problem and the remaining series is the same as Morse shows for the membrane.

Yes, I thought we agreed that the series does contain the direct sound field IF CALCULATED TO A VERY HIGH DEGREE. Since I'm only using the first few terms, it should not have much of the direct field in it, but still reasonalby accurate for modes (also I generally dont have source and receiver that close together). So then adding in a direct term would be appropriate. Of course its not mathematically absolutely correct, but unless your telling me that it's a significant error, I'm happy. Remember, in my work I often need to simulate a large number of rooms/configurations in one go. In other words the code needs to be fast, so a good approximation is necessary. For example I did a study of Bonello criteria where I simulated hundreds of different room dimensions, each with hundreds of receiver locations.

I hope you're not quibling about a theoretical mathematical issue which has no significant effect. I already agreed long ago that it's an approximation