I often find myself asking the question, how would N of these subs compare to M of those subs in terms of Max achievable SPL. I recognize that loudness alone is not a qualifier or criteria but it plays a very important role in the overall movie and music experience. I was inspired by discussions in Craig's thread where in post #1035 (http://www.avsforum.com/avs-vb/showthread.php?t=768150&page=104&pp=10)I talk about an "ATOMIC SUB". This is common sense and nothing new but often we fail to recognize the efficacy of using multiple subs.
I use the measurements done by our Finnish "philaudioophilist" (did I just create a new word :D ) gentleman Ilkka who has graciously done so many tests and posted them on cyberspace for the audio community. Now straight to the results.
Study the following charts and let me know what you think. I manually picked the data points from Illkka's charts and did the simple extrapolations. The fundamental rule used was, adding another driver adds +3dB to the Max output SPL. I used the Max output level and power compression charts to get the baseline and extrapolate from there. The charts compare the HSU 3.2 to the SVS Plus/2 and for fun, the genelec 7073a. This by no means should be taken to conclude one is better than the other. There are lot more parameters to subwoofer evaluation than just MaxSPL, so take this with a big grain of salt :D, however there is a lot one could take away from the charts if one understand's it's limitations. The best part is, it is derived from the least common denominator and the lcd are real world measurments.
This is just to get people thinking about multiple subs and also I would like to solicit your opinion on the charts and the methodology used for extrapolation.

There could be errors as the numbers were captured by eye. Here is the link to Ilkka's measurments (http://www.hometheatershack.com/forums/subwoofer-tests/).

Enjoy,
-Jai

Do these make any sense at all? I was hoping it will be helpful to people who are wondering if they should buy one sub or more subs especially with questions concerning MaxSPL. If you can find MaxSPL FR sweeps for your subs of interest, you can apply the same techniques. Please let me know if you need more explanations. I was under the impression that the charts would speak for themselves, apparently not.

Actually, this is great information. Many of us have wondered what kind of impact a second sub would make when we consider upgrading. It's one of those deals where we ask ourselves, "Is the $1000 sub really twice the sub a $500 sub is?" It is all subjective to the listener and degree to which they are willing to give their hard earned dollar to home theater.

One thing is for sure. This thread should spark up some interesting conversation.

jmcomp,

Are you sure you have the math right?

You are adding 3dB for each VTF-3MK2, but 6dB for each PB12/2+.

I cannot pull up the graphs posted, But if One VTF-3.2 = 100 dB, then 2 = 106 dB and 4 = 112 dB. You add 6 dB for each doubling of subs.

I cannot pull up the graphs posted, But if One VTF-3.2 = 100 dB, then 2 = 106 dB and 4 = 112 dB. You add 6 dB for each doubling of subs.
jmcomp,

Are you sure you have the math right?

You are adding 3dB for each VTF-3MK2, but 6dB for each PB12/2+.


This is exactly the kind of feedback I was looking for. Thanks for checking. I made the assumption of +3dB based on Brian Florian's article (http://www.hometheaterhifi.com/volume_8_4/subwoofers-12-2001.html).

Under section "What is SPL" this is what he says "Generally if you double the power input or add an additional driver you will increase the output by 3 dB"

Adding a 3.2 sub gives one additional driver and hence +3dB. Adding a Plus/2 gives 2 additional drivers and hence +6dB. Is this assumption correct?
If so, great, if not, I have to correct the charts accordingly.
Thanks,
-Jai

I cannot pull up the graphs posted, But if One VTF-3.2 = 100 dB, then 2 = 106 dB and 4 = 112 dB. You add 6 dB for each doubling of subs.
It looks like he added 3dB per VTF-3MK3 and 6dB for PB12/2Plus because it has dual drivers.

Like you posted, it should be 6dB regardless.

This is exactly the kind of feedback I was looking for. Thanks for checking. I made the assumption of +3dB based on Brian Florian's article (http://www.hometheaterhifi.com/volume_8_4/subwoofers-12-2001.html).

Under section "What is SPL" this is what he says "Generally if you double the power input or add an additional driver you will increase the output by 3 dB"

Adding a 3.2 sub gives one additional driver and hence +3dB. Adding a Plus/2 gives 2 additional drivers and hence +6dB. Is this assumption correct?
If so, great, if not, I have to correct the charts accordingly.
Thanks,
-Jai

That is not correct - you add 6 dB for each doubling of subwoofers. Let's use 0 dB as a baseline for any subwoofer, and assume co-locating them.

1 = 0 dB
2 = 6 dB
4 = 12 dB
8 = 18 dB

The amount of drivers does not matter - if you used the "3 dB per driver" idea ... a single sub using, say, 12 x 6 inch drivers would be 36 dB higher SPL when adding a second sub. That won't happen.

A 6 dB increase would only be realized by two subwoofers in perfect phase with each other and the same distance to the listening position. Otherwise, the increase would be between 3 and 6 dB.

There are many benefits that better (read, usually more expensive) subs bring other than just sheer level, so you aren't going to build some expensive sub killer merely by buying a bunch of cheap subs.

A PB-12/+2 is not going to be 6 dB louder than a PB-12 due to cabinet inefficiencies, I believe SVS claims a 4 dB increase or so.

This is exactly the kind of feedback I was looking for. Thanks for checking. I made the assumption of +3dB based on Brian Florian's article (http://www.hometheaterhifi.com/volume_8_4/subwoofers-12-2001.html).

Under section "What is SPL" this is what he says "Generally if you double the power input or add an additional driver you will increase the output by 3 dB"

Adding a 3.2 sub gives one additional driver and hence +3dB. Adding a Plus/2 gives 2 additional drivers and hence +6dB. Is this assumption correct?
If so, great, if not, I have to correct the charts accordingly.
Thanks,
-Jai
Brian is being correct. But when you add another sub, you will double both, amount of drivers AND amp power. The amount of drivers per sub doesn't matter at all (assuming you're adding up identical subs of course).

A 6 dB increase would only be realized by two subwoofers in perfect phase with each other and the same distance to the listening position. Otherwise, the increase would be between 3 and 6 dB.

That's correct.


A PB-12/+2 is not going to be 6 dB louder than a PB-12 due to cabinet inefficiencies, I believe SVS claims a 4 dB increase or so.
That whole louder thing is very silly unless you define the input signal being used.

Thanks for the feedback. The +6dB assumption is looking too aggressive and subwoofer scaling may not be linear due to reasons just raised. Will a +4dB per additional sub due to acoustic coupling be reasonable?

Here is another useful link
http://www.hometheaterforum.com/htf/archive/index.php/t-119865.html

jmcomp124, besides the fact that your chart incorrectly gave only a 3db increase to the HSU rather than the correct 6db for co-located subs, it also is not indicative of current product performance levels because the VTF-3.2 was replaced a couple months ago by the new 3.3 which according to Howard Ferstler's tests has SIGNIFICANTLY higher output then the older version and your graph results would be much different if the 3.3 was included (but to my knowledge there are not yet any ground plane measurements to go by).

But the real major shortcoming of a chart like this is it's based on SPL only and says nothing of sound quality which to many of us is far more important (and tests such as Craigsub's do reveal sound quality differences, as do my own in-room listening sessions between my HSU 3.3 and my SVS PB12-NSD).

Thanks for the feedback. The +6dB assumption is looking too aggressive and subwoofer scaling may not be linear due to reasons just raised. Will a +4dB per additional sub due to acoustic coupling be reasonable?

Here is another useful link
http://www.hometheaterforum.com/htf/archive/index.php/t-119865.html

No, a 6db increase in SPL for dual co-located subs is not an assumption, it is a scientific fact, and many reviewers have measured such in their own tests. And I confirmed this in an email conversation a few months ago with ED Mullen at SVS.

I went with 3.2 because that is what I have data for (Ilkkas). I am not worried about current market situations. This is simply a case study, so look at it that way. I also don't care about which brand performs better.

You raise a good point about SQ.
Can someone shed some light into what effect multiplying subs has on SQ subjectively and THD, impulse response and group delay and other interesting parameters?

For one time, please don't get biased or hurt if one shows better than the other in these charts. These are strictly for learning purposes and streamline the methodology. Each person can then use their own charts. I will forward the excel chart to whoever interested. For the record, I don't have an agenda here to make one look better than the other. I want to understand how the multiply thing works. Some good comments are coming in. Let's keep this going in a positive direction.

I was always under the impression that a doubling of the power adds 3 db, and having two subs effectively doubles the power, so where is this 6 db increase coming from?

No, a 6db increase in SPL for dual co-located subs is not an assumption, it is a scientific fact, and many reviewers have measured such in their own tests. And I confirmed this in an email conversation a few months ago with ED Mullen at SVS.
You are assuming perfect scaling (acoustic coupling) here which is what I am beginning to see is too aggressive. I was assuming the same earlier. I suspect if you take 10 different samples in different rooms and measure the acoustic coupling efficiency it will not always be 6dB. It appears 6dB is the best case. That is why I am asking, is it always the best case (+6dB) or should we go with +4dB that would average out the acoustic coupling efficiency?

I went with 3.2 because that is what I have data for (Ilkkas). I am not worried about current market situations. This is simply a case study, so look at it that way. I also don't care about which brand performs better.

You raise a good point about SQ.
Can someone shed some light into what effect multiplying subs has on SQ subjectively and THD, impulse response and group delay and other interesting parameters?

For one time, please don't get biased or hurt if one shows better than the other in these charts. These are strictly for learning purposes and streamline the methodology. Each person can then use their own charts. I will forward the excel chart to whoever interested. For the record, I don't have an agenda here to look one make better than the other. I want to understand how the multiply thing works. Some good comments are coming in. Let's keep this going in a positive direction.

Dude, chill out - my "feelings" are not hurt by your chart, and I never accused you of having an agenda - you seemed to be asking for honest input regarding the usefulness of your chart and I gave mine. I have no ties to any audio company and have never met anyone in the industry - I am simply a 41-year-old Northwest-based professional photographer by trade who has enjoyed as a hobby (to the great aggravation of my wife ;) ) dabbling in budget audiophile products for the last 15 years as a pasttime, and in order to avoid endless forum speculation I went ahead and purchased both an SVS and an HSU sub so I could evaluate them in my room with my measurements and my ears :) .

Cheers Jai,
Ross

I was always under the impression that a doubling of the power adds 3 db, and having two subs effectively doubles the power, so where is this 6 db increase coming from?
2x power and 2x cone area = +6 dB

You are assuming perfect scaling (acoustic coupling) here which is what I am beginning to see is too aggressive. I was assuming the same earlier. I suspect if you take 10 different samples in different rooms and measure the acoustic coupling efficiency it will not always be 6dB. It appears 6dB is the best case. That is why I am asking, is it always the best case (+6dB) or should we go with +4dB that would average out the acoustic coupling efficiency?
IF the subwoofers are co-located (closer than 1/4 of the length of the shortest sound wave they have to produce, for example around 1 meter for 80 Hz), there will most definitely be +6 dB. But co-locating four PB12-Plus/2's might a bit difficult...not impossible though. :D

You are assuming perfect scaling (acoustic coupling) here which is what I am beginning to see is too aggressive. I was assuming the same earlier. I suspect if you take 10 different samples in different rooms and measure the acoustic coupling efficiency it will not always be 6dB. It appears 6dB is the best case. That is why I am asking, is it always the best case (+6dB) or should we go with +4dB that would average out the acoustic coupling efficiency?

All I can tell you is that many reviewers have measured a full 6db increase with dual co-located subs, Ed at SVS told me I would achieve a 6db increase with dual co-located PB12-NSDs, and in fact two weeks ago when I did the unthinkable and stacked my PB12NSD on top of my VTF3.3 (and adjusted the phase properly) I too measured a 6db increase. Now the sound quality was not as articulate with those two different subs mixed, but crap I have never experienced WOTW like that before! :D Unfortunately the look did not pass the WAF :p

All I can tell you is that many reviewers have measured a full 6db increase with dual co-located subs, Ed at SVS told me I would achieve a 6db increase with dual co-located PB12-NSDs, and in fact two weeks ago when I did the unthinkable and stacked my PB12NSD on top of my VTF3.3 (and adjusted the phase properly) I too measured a 6db increase. Now the sound quality was not as articulate with the those two different subs mixed, but crap I have never experienced WOTW like that before! :D Unfortunately the look did not pass the WAF :p
That is good to know. I will wait for some more feedback before I decide if I should use +6dB. Thanks for your inputs.

That is good to know. I will wait for some more feedback before I decide if I should use +6dB. Thanks for your inputs.

Jai, sorry to come accros earlier as dumping on your efforts here, I do apologize. I can totally see where such a chart would be useful for someone contemplating their various options for increasing sub headroom, but it sure would be awesome if we had ground plane measurements for many more models :) I do like data and numbers myself, but I also like to balance it with lots of listening to determine SQ.

Just an idea, you could drop both SVS and HSU a quick email to confirm the validity of the 6db increase with dual co-located subs.

The difference in dB between any two sound pressure measurements is:

dB = 20 * log (Px/Py) (where Px and Py represent the two pressure levels)

If you double the sound pressure, then Px/Py = 2. Solving for the dB increase = 6.0206 dB. So a doubling of the sound pressure equates to a 6 dB increase in sound pressure.

Remember, you will only see a 6 dB increase when the two sound pressure sources are co-located and are identical in frequency, phase, and individual sound pressure.

The difference in dB between any two sound pressure measurements is:

dB = 20 * log (Px/Py) (where Px and Py represent the two pressure levels)

If you double the sound pressure, then Px/Py = 2. Solving for the dB increase = 6.0206 dB. So a doubling of the sound pressure equates to a 6 dB increase in sound pressure.

Remember, you will only see a 6 dB increase when the two sound pressure sources are co-located and are identical in frequency, phase, and individual sound pressure.

Thanks for chiming in Ed. Looking back on my notes when I experimented with the PB12NSD and VTF3.3 co-located (stacked), on pink noise I was getting closer to a 5db increase, and on individual test tones I achieved 6db on some frequencies but not all, as would be expected since these were not identical subs. But despite it being an imperfect combo it still did enough right to blow me away - I have never felt the impact on the Lightning and Pods Emerging scenes of WOTW so viscerally, and what really stood out was how effortless the sound was. Peter Marcks mentioned that distortion should be cut in half for dual co-located units which must have contributed to the sense of ease and power I heard. Now I can only imagine how two identical brand & model subs would be, either co-located PB12NSD's or co-located VTF3.3's :eek: . In 5 months I will be getting my own dedicated HT room in a new house so two subs will become a permanent reality for me, the WAF be damned :D

Ed, I've had a vision stuck in my head ever since I read the CES report of your soon-to-come new 13" Ultra...two of them co-located in a 2700 c.f. room in the 10hz tune...I am assuming this combo would bring about a few nail pops and plaster cracks in a new home construction :D

Thanks for chiming in Ed. Looking back on my notes when I experimented with the PB12NSD and VTF3.3 co-located (stacked), on pink noise I was getting closer to a 5db increase, and on individual test tones I achieved 6db on some frequencies but not all, as would be expected since these were not identical subs. But despite it being an imperfect combo it still did enough right to blow me away - I have never felt the impact on the Lightning and Pods Emerging scenes of WOTW so viscerally, and what really stood out was how effortless the sound was. Peter Marcks mentioned that distortion should be cut in half for dual co-located units which must have contributed to the sense of ease and power I heard. Now I can only imagine how two identical brand & model subs would be, either co-located PB12NSD's or co-located VTF3.3's :eek: . In 5 months I will be getting my own dedicated HT room in a new house so two subs will become a permanent reality for me, the WAF be damned :D

Right, the phase and frequency responses are not identical, so you won't see exactly a 6 dB increase at all frequencies with these two subs colocated, but it will be close.

While distortion is considerably lower with duals, what you are reacting to is a lack of dynamic compression where the subwoofers are operating within their clean/uncompressed limits more of the time - even during the toughest passages on WOTW.

You could probably overload the pair if you tried hard enough. :) The goal really isn't "unlimited headroom" per se - it's sufficient headroom for the given application - i.e., room size, distance from the listening position, preferred playback level, etc.

Some people would be happy with a single subwoofer in your room, and others would need four of them - tastes in bass vary greatly.

The difference in dB between any two sound pressure measurements is:

dB = 20 * log (Px/Py) (where Px and Py represent the two pressure levels)

If you double the sound pressure, then Px/Py = 2. Solving for the dB increase = 6.0206 dB. So a doubling of the sound pressure equates to a 6 dB increase in sound pressure.

Remember, you will only see a 6 dB increase when the two sound pressure sources are co-located and are identical in frequency, phase, and individual sound pressure.
Ed,
People like me admire you for these kind of posts.
-Jai

Jai, sorry to come accros earlier as dumping on your efforts here, I do apologize. I can totally see where such a chart would be useful for someone contemplating their various options for increasing sub headroom, but it sure would be awesome if we had ground plane measurements for many more models :) I do like data and numbers myself, but I also like to balance it with lots of listening to determine SQ.

Just an idea, you could drop both SVS and HSU a quick email to confirm the validity of the 6db increase with dual co-located subs.
No problem :). I am glad you understand my intentions.

No problem :). I am glad you understand my intentions.

Jai, have you tried dual co-located subs in your room? I'm curious to read comments of what others have noticed regarding sound quality, dynamics, etc. when moving from one sub to two co-located units. From previous reading I gather it can be the Holy Grail for movie LFE reproduction :) I want to experiment more with my two subs (pending temporary permit approval from spouse :p ) but somehow it seems 'sinful' to me to marry an SVS with an HSU :D ;)

From extensive discussions, my takeaway is don't try to mix and match but go with the a sub that performs well across the LF spectrum and then add multiples. The reason that marrying an SVS to a HSU may not be a good idea is because of their difference in FR along with other differences.
Don't be unequally yoked :)

I will be correcting the +3dB and +6dB error shortly. Thanks to cschang, craigsub and others who caught this right away. The intent of posting and asking for feedback is to address such miscalculations and wrong conceptions.

From extensive discussions, my takeaway is don't try to mix and match but go with the a sub that performs well across the LF spectrum and then add multiples. The reason that marrying an SVS to a HSU may not be a good idea is because of their difference in FR along with other differences.
Don't be unequally yoked :)

Ha, love that NT reference ("Don't be unequally yoked") :)

You are totally right and I knew that mixing the two different brand/model subs would not be a good long-term solution, but it was enough of a very fun experiment to give me a taste of the dynamic headroom benefits, and 2- 12" drivers and 3 - 4" ports sure did move some serious air :D

Ha, love that NT reference ("Don't be unequally yoked") :)

You are totally right and I knew that mixing the two different brand/model subs would not be a good long-term solution, but it was enough of a very fun experiment to give me a taste of the dynamic headroom benefits, and 2- 12" drivers and 3 - 4" ports sure did move some serious air :D

As an experiment, you might try plugging a port on the Hsu (or installing the external port extensions); the FR and phase response will then be closer to that of the PB12-NSD (which is tuned to ~18 Hz). While overall SPL peaks might decrease a bit, you might notice a more coherent/cohesive sound at the very deepest frequencies. Can't hurt to try.....experimenting is cool.

Ok. Based on feedback I did the +3dB to +6dB corrections for the respective models. Ilkka just wrote me and he thought the genelec was not a good candidate here due to limiters, so I removed it. The new charts have 2 sets of results for very poor coupling and ideal coupling.
Again, I am reminding people, this is not for you to infer which is better, it is only an exercise so you can do your own homework while thinking of multiplying subs.
The CALL TO ACTION is for people to provide more data on subs like the way Ilkka does. I am going to try and learn from him sometime in the near future about how to go about the measurment methodology.

Please treat this as preliminary data and kindly forgive mistakes and do bring them to my notice as and when you find them.
Thanks,
-Jai

nice stuff jm, keep working at it - i bet the data will be useful in some way :) Also - is there any way to make the graphs bigger?

As an experiment, you might try plugging a port on the Hsu (or installing the external port extensions); the FR and phase response will then be closer to that of the PB12-NSD (which is tuned to ~18 Hz). While overall SPL peaks might decrease a bit, you might notice a more coherent/cohesive sound at the very deepest frequencies. Can't hurt to try.....experimenting is cool.

Thanks for the feedback Ed. That's actually how I ran the HSU (max extension) the glorious day I was playing with the two stacked, which from what you noted is probably why it sounded overall so dang impressive! You are right, experimenting is cool, and I think more is in order :)

Ok. Based on feedback I did the +3dB to +6dB corrections for the respective models. Ilkka just wrote me and he thought the genelec was not a good candidate here due to limiters, so I removed it. The new charts have 2 sets of results for very poor coupling and ideal coupling.
Again, I am reminding people, this is not for you to infer which is better, it is only an exercise so you can do your own homework while thinking of multiplying subs.
The CALL TO ACTION is for people to provide more data on subs like the way Ilkka does. I am going to try and learn from him sometime in the near future about how to go about the measurment methodology.

Please treat this as preliminary data and kindly forgive mistakes and do bring them to my notice as and when you find them.
Thanks,
-Jai

Jai, you get a 6db increase when you double the number of co-located subs, so +6db when going from 1 to 2, and another +6db when going from 2 to 4, but only +3db when going from 3 to 4.

Ed/craigsub/cschang, correct me if I'm wrong...

Ok. Based on feedback I did the +3dB to +6dB corrections for the respective models. Ilkka just wrote me and he thought the genelec was not a good candidate here due to limiters, so I removed it. The new charts have 2 sets of results for very poor coupling and ideal coupling.
Again, I am reminding people, this is not for you to infer which is better, it is only an exercise so you can do your own homework while thinking of multiplying subs.
The CALL TO ACTION is for people to provide more data on subs like the way Ilkka does. I am going to try and learn from him sometime in the near future about how to go about the measurment methodology.

Please treat this as preliminary data and kindly forgive mistakes and do bring them to my notice as and when you find them.
Thanks,
-Jai

Assuming each additional sub adds 3 dB is not correct as the number of subs grows. I know you are assuming incomplete acoustic coupling, but even then the math is not correct because you are dealing with log10 functions and they are not additive like that.

For the purposes of your comparisons, I would advise assuming complete acoustic coupling and using the following:

1 sub = baseline
2 subs = + 6 dB
3 subs = +9.5 dB
4 subs = +12 dB
5 subs = +14 dB
6 subs = +15.5 dB
7 subs = +17 dB
8 subs = +18 dB

Ah, so it is logarithmic, I should have know this. Thanks a lot Ed.
Another correction will follow this evening.
-Jai

I updated the charts to reflect Ed's corrections using a log scale. In short
subwoofer scaling efficiency follows a log scale.

For those of you who are curious, there are 2 components to the gain.

Total gain = 10 log (Number of woofers) + 10 log (Impedance of single woofer / Impedance of woofers when wired)

The above total gain is basically "Mutual coupling gain + Electrical gain".

The numbers Ed provided are based on the above formula. Thanks for the useful PMs that helped clarify the above. I should say there are some very nice folks around here who correct in a very gentle manner and who prefer that their names not be mentioned. I appreciate their help and publicly want to thank them for their contribution. Amazing what simple math like this can do.

I think the data posted here now is pretty solid and I have a higher degree of confidence in it.

Now Craig is going to test some multiple subs and prove these projections. Right Craig :p :) ;)
-Jai

I updated the charts to reflect Ed's corrections using a log scale. In short
subwoofer scaling efficiency follows a log scale.

For those of you who are curious, there are 2 components to the gain.

Total gain = 10 log (Number of woofers) + 10 log (Impedance of single woofer / Impedance of woofers when wired)

The above total gain is basically "Mutual coupling gain + Electrical gain".

The numbers Ed provided are based on the above formula. Thanks for the useful PMs that helped clarify the above. I should say there are some very nice folks around here who correct in a very gentle manner and who prefer that their names not be mentioned. I appreciate their help and publicly want to thank them for their contribution. Amazing what simple math like this can do.

I think the data posted here now is pretty solid and I have a higher degree of confidence in it.

Now Craig is going to test some multiple subs and prove these projections. Right Craig :p :) ;)
-Jai

Why wouldn't anyone want to be publicly thanked for helping out? that's odd.

Anyhow...

Wait - why does the number of drivers have anything to do with total spl?

You are implying that, say, 1 dual driver sub = X, and 2 dual driver subs = 2.5X (or something like that.

It makes no sense to me..it doesn't matter if the sub has 4x 8" woofers or 1x 12"...the UNIT can only make a certain max spl (per distortion limit yadda yadda).....can you elaborate? I'm sure i'm not the only one confused....

LOL, This was the exact same reason he didn't want his name to be mentioned. He didn't want to be answering questions and he said if I post it I have to deal with it, so here I am :). I verified the forumula and it works.
Here is what I recommend before I try to explain it. Use an excel sheet or just a notepad.
The formula for impedance when wired in parallel is 1/z = 1/z1 + 1/z2.
Now just plug in the number of woofers and choose an impedance value of say 4 ohms for the woofer. Give it a shot and if it does not work then I will come back and try to explain.

The numbers Ed provided are based on the above formula.

Actually, they aren't. That doesn't mean your formula doesn't work, though. My numbers are based on the original forumula I provided which is simply a ratio of sound pressures created by adding subs. I'll annotate the equation and the numbers for clarity:

The difference in dB between any two sound pressure measurements is:

dB = 20 * log (Px/Py) (where Px and Py represent the two pressure levels)


Px = 1 sub = baseline (Px/Py = 1.0)
Px = 2 subs = + 6 dB (Px/Py = 2.0) First doubling of sound pressure.
Px = 3 subs = +9.5 dB (Px/Py = 3.0)
Px = 4 subs = +12 dB (Px/Py = 4.0) Second doubling of sound pressure.
Px = 5 subs = +14 dB (Px/Py = 5.0)
Px = 6 subs = +15.5 dB (Px/Py = 6.0)
Px = 7 subs = +17 dB (Px/Py = 7.0)
Px = 8 subs = +18 dB (Px/Py = 8.0) Third doubling of sound pressure.

So you can see, every time the sound pressure is doubled (2, 4, 8 subs), the dB increases 6 dB. In order to double the sound pressure again, you would need to go from 8 to 16 subs. Naturally the biggest "bang for the buck" is going from 1 to 2 subs. After that, it gets expensive very quickly. :)

Actually, they aren't. That doesn't mean your formula doesn't work, though. My numbers are based on the original forumula I provided which is simply a ratio of sound pressures created by adding subs. I'll annotate the equation and the numbers for clarity:

The difference in dB between any two sound pressure measurements is:

dB = 20 * log (Px/Py) (where Px and Py represent the two pressure levels)


Px = 1 sub = baseline (Px/Py = 1.0)
Px = 2 subs = + 6 dB (Px/Py = 2.0) First doubling of sound pressure.
Px = 3 subs = +9.5 dB (Px/Py = 3.0)
Px = 4 subs = +12 dB (Px/Py = 4.0) Second doubling of sound pressure.
Px = 5 subs = +14 dB (Px/Py = 5.0)
Px = 6 subs = +15.5 dB (Px/Py = 6.0)
Px = 7 subs = +17 dB (Px/Py = 7.0)
Px = 8 subs = +18 dB (Px/Py = 8.0) Third doubling of sound pressure.

So you can see, every time the sound pressure is doubled (2, 4, 8 subs), the dB increases 6 dB. In order to double the sound pressure again, you would need to go from 8 to 16 subs. Naturally the biggest "bang for the buck" is going from 1 to 2 subs. After that, it gets expensive very quickly. :)
Though you used a different forumla, the formula I posted solves to be the same. I just had to apply some simple log rules and simplify the equation and voila, I saw how our results matched.

So this is how it works.
Let's represent the first formula as
Total Gain Tg = 10 log (N) + 10 log (xi / xw)

where, N is the # of woofers and xi and xw are the impedances of a single woofer and the effective impedance when wired. The above equation covers the generic case of variable impedances and also circuits that are not just wired in parallel. Since we are multiplying the same subwoofer, all impedances remain the same and all the woofers are in parallel. This causes xw to become xi/N. Substitute this in the formula above, xi and xi cancel out. This answers why the impedance does not matter in our case. Now we end up with teh simplified formula 10 log (N) + 10 log (N) which is 10 log N^2 or "20 Log N" which matches your formula.
Mystery solved :cool:

Is something becoming clear here about gaining SPL?

1 sub to 2 subs - Best "BANG" for the buck
2 subs to 3 subs - More "BAng" for the buck
3 subs to 4 subs - A little more "bang"
4 subs to 5 subs - Stop wasting your money and think of a new "ATOMIC SUB", unless you want to start smoothing out the response.

Ok, so lets throw out a hypothetical.
Let's say I have a fat wallet and a theater to fill with bass with a volume of 5000+ ft^3. Should I:
A) Run out and buy the biggest and baddest "fathom 113"
B) Order about 4 or 5 HSU VTF 2.3s(shipping would be a killer, maybe a package deal though)
C) Order 3 HSU VTF 3.3s
D) Order a Bose Acoustamass system and kick myself in the groin.

This is rediculous of course and we would only be concerned with max SPL levels. I tried to keep the costs near the benchmark $2500. Any thoughts?

graniteguy,
Recommending a choice in this thread maybe a bad idea as it will get very controversial. Please start a different thread and point to this one and ask for recommendations. I will PM you my thougths.
-Jai

Jai!

I'm liking your thinking. I'm in the midst of studying horn loading as it relates to acoustical impedance matching for "typical" listening environments. You see, there is a "window" represented by aggregate driver surface emitting area "Sd" that is somewhat commensurate with the exit mouth of an optimum flare bass horn. Bass horn mouth effect is further scaled by the local acoustical loading, such as 4p, 2p, wall & floor, or corner loading. For corner loading (ie. Klipsch corner horns) the tuning factor (and thus cross sectional area) is DIVIDED BY 8. What that means, is that there is a corner frequency walking towards you as you increase the number of your "atomic subs" in corners (I refer to this as a SUBWOOFER ARRAY :) ). Room loading will then cause a DECREASE in required excursion above that frequency, and an INCREASE in peak available SPL for certain sub/amp arrangements over your idealized 6dB per doubling once you cross that threshold at a given frequency. I see "room gain" bandied about, but not as much predictive analysis as I'd like. The seminal work is half a century old or more, and I'm cracking the books in this direction. It will also change transfer function compensation EQ requirements for the better. Keep up the good work.

:)

Martin

Doubling the number of subwoofers while holding the power constant results in a +3db gain. Double the subs *and* the power is +6db gain.

My previous comments were completely hypothetical. I do not have a fat wallet or a theater room that big. The purpose was to gather opinion and grasp an understanding whether it is better to have one high performing/high dollar sub or multiple mid level/mid dollar subs.
I don't know which is correct. But there are people out there who would like the ultimate performance but are not willing to jump in with both feet at once. Instead, they might take comfort in knowing that they can add to a mid level sub in the future and obtain high level results.

My previous comments were completely hypothetical. I do not have a fat wallet or a theater room that big. The purpose was to gather opinion and grasp an understanding whether it is better to have one high performing/high dollar sub or multiple mid level/mid dollar subs.
I don't know which is correct. But there are people out there who would like the ultimate performance but are not willing to jump in with both feet at once. Instead, they might take comfort in knowing that they can add to a mid level sub in the future and obtain high level results.
Sorry. I didn't want to come across as discouraging. Welcome to the forum, I didn't notice your post count. There have been very many debates and arguments here with sub comparisons, so I was trying to avoid that from happening. What you ask is a hard question. One high performing/high dollar sub or multiple mid level/dollar subs? In my mind, it is best not to mix and match subs. Opinions vary though. If you are thinking of upgrading in the future, buy a good sub with a good design and buy more of the same later. If you don't want to keep upgrading, buy the best you can afford now. I thought I wouldn't upgrade a couple of years ago, but the bug bit me again, just to let you know.
Good luck,
-Jai

JAI you told me to look at these.....

I want to look at SINGLE sub vs SINGLE SUB

not dual vs trio or 4's.....that doesnt help me....I'm trying here to get more information but this isnt helping me out

I was answering your question "...........is a 2nd one really going to make that 18hz sound BETTER? "

When you start pushing your sub to the limits, it begins to compress and then you hear port chuffing, bottoming out etc. You are not interested in the trio, but you are in the dual right. By looking at Max SPL sweeps and the compression charts, you can figure out where the 10-NSD stands. Find out from Ed, how much better the 12 would perform. Add the delta to the 10-NSD chart line. Add another +6dB for second sub gain. By comparing a few data points with that data and working backwords you can have some idea if indeed adding a second sub will help with the chuffing. You will have to play 1.5x (6dB) louder to reach those limitations with a 2nd sub.

It might be informative to include the single sub, and double sub lines....just for the sake of completeness...

Add the delta to the 10-NSD chart line Jai you are a help that is no doubt....but that "add the delta" line is above my knowledge base at the moment.... :D

I do understand that adding a 2nd sub will give me 6db more which is a TON....

But why have 2 subs when there might be one sub that can do it for me and even do it better ESPECIALLY LOWER like 14-16hz........namly the 3.3Turbo or the HO Turbo or maybe even the new Ultra coming out after we get some reviews on it....

Add the delta to the 10-NSD chart line Jai you are a help that is no doubt....but that "add the delta" line is above my knowledge base at the moment.... :D

I do understand that adding a 2nd sub will give me 6db more which is a TON....

But why have 2 subs when there might be one sub that can do it for me and even do it better ESPECIALLY LOWER like 14-16hz........namly the 3.3Turbo or the HO Turbo or maybe even the new Ultra coming out after we get some reviews on it....
Adding a 2nd sub is your easiest upgrade path and it sure to give you a lot of headroom. If you want to go lower, then of course, change your "atomic sub" :).
The 3.3 Turbo or HO turbo or 16-46 PC+, all are very good choices. If you are ok with going through the process of selling, then go for it. How would a 3.3 Turbo or HO turbo or 16-46 PC+ perform at 30Hz and above compared to dual 12-NSD is a question you need to ask and unfortunately we don't have a baseline to play with. I wish we get more data on the 3.3 Turbo and HO Turbo soon.
Food for thought ...

bgillyjcu,
From the charts, the HSU sure looks better for the group delay and impulse response. Are those audible? I don't know. Please don't make a decision based on SPL alone. I am sure you already know that, but wanted to make sure.

Arrays are nothing new. Grateful Dead's Wall of Sound is famous.

30+ years ago
http://www.terrapin-station.de/grafisches/pics/wallofsound.jpg
http://dozin.com/wallofsound/index.html

The Wall of Sound consisted of 89 300-watt solid-state and three 350-watt vacuum-tube amplifiers generating a total of 26,400 watts RMS of audio power. This systems projected high quality playback at six hundred feet with an acceptable sound projected for a quarter mile.

http://en.wikipedia.org/wiki/Grateful_Dead

26.4kw of power is nothing by today's standards. hehe
I have 12 proamps waiting for me to start a home audio line
array project, over 60kw. :)

Hey Thylantyr!

Let's start a WATTS war! :cool: Seriously, I like using the Crown K2 for subwoofer duty. No fans! I'd like to see more of your array idea based on 60,000 watts.

:D

Martin

Adding a 2nd sub is your easiest upgrade path and it sure to give you a lot of headroom. If you want to go lower, then of course, change your "atomic sub" :).
The 3.3 Turbo or HO turbo or 16-46 PC+, all are very good choices. If you are ok with going through the process of selling, then go for it. How would a 3.3 Turbo or HO turbo or 16-46 PC+ perform at 30Hz and above compared to dual 12-NSD is a question you need to ask and unfortunately we don't have a baseline to play with. I wish we get more data on the 3.3 Turbo and HO Turbo soon.
Food for thought ...


You are right about all this, and no single SPL chart will make my mind up.....

My theory is that any system I have I'll probably add an MBM down the road and place it close to my LP for that extra SLAM....then let the main sub do the 50hz and below duty...

I'd only sell my PB-12NSD if I get a great offer. Then I'd probably go for the 3.3+turbo, HO+turbo, 16-46+ or maybe even the new Ultra. Those are probably the front runners in my book right now since I do care alot about the sub 20hz frequency response because I love that subsonic shake, rare as it may be, I still love it :D

The 13.5inch woofer will make a dramatic difference compared to the 12's on the market. The JL shows proof just how much an inch and a half can make. We'll just have to see how the SVS design compares to it and the other 12's still on the market. I just wish the new Ultra was out or we had a date for presale...

Did someone say line array ? Here are some samples of my brother's handiwork ... This system is absolutely fantastic ...

High output 2 x 10 inch drivers with 115 dB output per 27 pound enclosure. 16 stacked will deliver 139 dB per side for your outdoor theater...

http://www.a-lineacoustics.com/Images/LAfacingLeftcc.jpg

Then add these subwoofers, with dual 18's tuned to 32 Hz (a slight mod required at the factory) - with a solid 28 Hz floor...

http://www.a-lineacoustics.com/Images/2Subscc.jpg

3 of those per side, and we have 139 dB in a GP environment from 28 Hz and up.

I'm oggling those pics, Craig. What kinda tweet is in those line array modules? I've gotten to tear apart some SLS recently to discover a proprietary ribbon and a pair of PHL audio 8".

:)

Martin

the more subs the better... ;]

Did someone say line array ? Here are some samples of my brother's handiwork ... This system is absolutely fantastic ...

High output 2 x 10 inch drivers with 115 dB output per 27 pound enclosure. 16 stacked will deliver 139 dB per side for your outdoor theater...

http://www.a-lineacoustics.com/Images/LAfacingLeftcc.jpg

Then add these subwoofers, with dual 18's tuned to 32 Hz (a slight mod required at the factory) - with a solid 28 Hz floor...

http://www.a-lineacoustics.com/Images/2Subscc.jpg

3 of those per side, and we have 139 dB in a GP environment from 28 Hz and up.


WHHAAAAT?

WHAAHHHHTTT?

NO REALLY I CAN'T HEAR A WORD YOU'RE SAYING BECAUSE I HAVE SERIOUS DEBILITATING HEARING DAMAGE, BUT WASN'T THAT CONCERT AWESOME?