First time Centre Subs

[Justin Myers]

Hello,

I'm doing a gig next week where I want to try using 4 x centre subs with the 2 a side L/R subs. The 4 centre subs are not the same breed as the L/R subs and The L/R subs will have the mid/hi boxes on top of them.

The L/R subs are all run off 1 x IT8000. 2 of the centre subs are run off 1 x IT8000 and the other 2 run off 1 x IT5000. I was going to run the same mono auxilary sub signal to all amps.

I have never tried using centre subs before. This gig is outdoor and I have the room to put the subs out front. The stage is 36 foot and I have experimented with lots of different Mapp predictions and came up with this.(See Pic) It has 0 delay on the centre subs and the positioning seems to create a reasonable coverage. Well, reasonable compared to the other predictions I tried.

Any word of advice from the more experienced?

Thanks

Justin

 

[Jay Barracato]

If the L/R subs are covering the same frequency band as the center subs you will want to include booth in your model.

 

[Brian Bolly]

Re: First time Centre Subs

Your spacing is a little too wide. Spacing between elements must be less than 240 degrees of the highest frequency being produced in order to get addition on axis and cancellation off axis. If that frequency is 100 Hz: (3.435m x 240/360) = 2.29m. Instead of using 2 blocks of 2 elements each, try using 4 spaced individually in the center plus the left and right. Also consider driving the center pair with one amp, the outer pair with another, and then the L/R with yet another, so you can do a delay arc the array as well.

 

[Justin Myers]

Re: First time Centre Subs

If the L/R subs are covering the same frequency band as the center subs you will want to include booth in your model.

Yep, I have the L/R subs active in the prediction.

 

[Justin Myers]

Re: First time Centre Subs

Your spacing is a little too wide. Spacing between elements must be less than 240 degrees of the highest frequency being produced in order to get addition on axis and cancellation off axis. If that frequency is 100 Hz: (3.435m x 240/360) = 2.29m. Instead of using 2 blocks of 2 elements each, try using 4 spaced individually in the center plus the left and right. Also consider driving the center pair with one amp, the outer pair with another, and then the L/R with yet another, so you can do a delay arc the array as well.

Ok, Here is another pic trying to implement your suggestions. I have the pairs 1ms apart from the inside out.

Less red SPL but definitely more spread with the yellow. Nice to get some more higher spl in the middle but I can't seem to get it without the rest falling apart.

 

[Brian Bolly]

Re: First time Centre Subs

Ok, Here is another pic trying to implement your suggestions. I have the pairs 1ms apart from the inside out.View attachment 4812

Less red SPL but definitely more spread with the yellow. Nice to get some more higher spl in the middle but I can't seem to get it without the rest falling apart.

There's two problems here:

First:

The 4 centre subs are not the same breed as the L/R subs

Your models show (presumably) all subs are the same, when in reality they will not be. So if the subs have different output, different frequency response, etc., all bets are off. Same with the inner 4 that are being driven off 2 different types of amplifiers - if one pair is outrunning the other, this will not hold together. Basically, all subs need to be handicapped to match the lowest common denominator of performance - output, frequency response, delay, etc. And if it's two different types of boxes (i.e. - horn loaded vs front loaded), this will be even more difficult to achieve.


Second, the L/R subs are coupled, whereas the rest are not. Reduce the L/R subs to -6 dB in comparison to the center subs - again, this will likely differ in practice since they are different from the center subs.


You also may try to shorten the delay times to something like 0/0.5/1 ms from center to outer. Also, your spacing from the L/R sub to the outermost sub in the center block looks greater than 2.29m.

 

[Art Welter]

Re: First time Centre Subs

Second, the L/R subs are coupled, whereas the rest are not. Reduce the L/R subs to -6 dB in comparison to the center subs - again, this will likely differ in practice since they are different from the center subs.

Assuming the center subs are within 1/4 wavelength of each other they are coupled.
Coupling gains 3 dB per doubling of cones, assuming they have equal power per cone the L/R coupled subs would be at maximum 3 dB more than uncoupled subs.

 

[Justin Myers]

Re: First time Centre Subs

There's two problems here:

First:

Your models show (presumably) all subs are the same, when in reality they will not be. So if the subs have different output, different frequency response, etc., all bets are off. Same with the inner 4 that are being driven off 2 different types of amplifiers - if one pair is outrunning the other, this will not hold together. Basically, all subs need to be handicapped to match the lowest common denominator of performance - output, frequency response, delay, etc. And if it's two different types of boxes (i.e. - horn loaded vs front loaded), this will be even more difficult to achieve.


Second, the L/R subs are coupled, whereas the rest are not. Reduce the L/R subs to -6 dB in comparison to the center subs - again, this will likely differ in practice since they are different from the center subs.


You also may try to shorten the delay times to something like 0/0.5/1 ms from center to outer. Also, your spacing from the L/R sub to the outermost sub in the center block looks greater than 2.29m.

Yeah, I know what your saying with the different subs but I can't change it unfortunately.
As for the different amps on the centre subs, they have the exact same response, phase and input sensitivity and I shouldn't need to drive any harder than what the smaller amp is capable of.
Sorry, but I miss understood with the 2.29m. I thought you meant I had to bring all the center subs within 2.29m of each other.

Spacing between elements must be less than 240 degrees of the highest frequency being produced in order to get addition on axis and cancellation off axis. .

Assuming the center subs are within 1/4 wavelength of each other they are coupled.

Ok, I'm bit confused on these 2 comments. Are you talking about different things? I always thought it was a within quarter wave for coupling.

 

[Per Søvik]

Re: First time Centre Subs

Ok, I'm bit confused on these 2 comments. Are you talking about different things? I always thought it was a within quarter wave for coupling.

You need to be under 2/3 wavelength to get rid of the side lobe, it is gone completely when you are at 1/2 wavelength.
At 1/4 wavelength you are getting stronger addition on-axis, but you are starting to loose directivity

 

[Rasmus Rosenberg]

Re: First time Centre Subs

You need to be under 2/3 wavelength to get rid of the side lobe, it is gone completely when you are at 1/2 wavelength.
At 1/4 wavelength you are getting stronger addition on-axis, but you are starting to loose directivity

Hey Per and Brian
What puzzel me is that you describe on-axis addition as a loose of directivity, why is it not a gain of directivity?
Very interesting topic by the way.

mvh
R

 

[Per Søvik]

Re: First time Centre Subs

Hey Per and Brian
What puzzel me is that you describe on-axis addition as a loose of directivity, why is it not a gain of directivity?
Very interesting topic by the way.

mvh
R

Sorry if I'm confusing you, but when you get under 1/4 wavelength, you may gain some efficiency, so on-axis might show a small gain because of this.
Off-axis you are loosing cancellation, at quarter wavelength, maximum cancellation is 3dB, so it is pretty omni-directional.
On-axis doesn't really change that much, except for a minute change in the angle between the sources, untill the sources are close enough together to interact in a way that changes the acoustic resistance the units experience.

Edit: Too confusing, here are some illustrations:
If you look closely, you'll see that on-axis is unchanged, the changes in directivity is due to off-axis cancellation

1/4 wavelength
1/2 wavelength
2/3 wavelength
1/1 wavelength

 

[Rasmus Rosenberg]

Re: First time Centre Subs

Thanks Per,
Great Pic's! makes it much easier to understand and break the language barrier i was lost in

 

[Kip Conner]

Re: First time Centre Subs

I suppose that with a enough processing you could make it work just like the model. Do you have access and have comfortability with measuring ALL the subs individually? You could conceivably make all the subs react the same way with individual processing and then use a master processor to create your sub array. It's not something you can do in the field real easily though.

My current subwoofer set-up is to 2 per side in a cardioid pattern. Even after doing all the math for spacing I started the process by measuring all the cabinets (even though they were identical) to make sure the frequency and phase response was the same. I think that we sometimes take for granted that because all the components of the system are the same the will produce the same Freq & Phase response. Never hurts to analyze the gear from time to time to be certain or other wise modeling will be off.

 

[Brad Weber]

Re: First time Centre Subs

Maybe I missed it but why are you doing this and what are you hoping to achieve by it? Is this just to experiment or are you trying to get some specific result or advantage from it? It seems that what you do might be very dependent on the result you are hoping to obtain.

It also seems like you aren't really looking at a single center array of subs but rather at four subs distributed in some fashion over a distance to either side of the centerline along with two box arrays under each main. That makes it sems like you may have some specific goal in mind in terms of the resulting directivity, but it is not clear what that goal may be.

FWIW, full coupling occurs with no separation while 1/4 wavelength separation relates to the nominal 3dB increase for two random sources of equal level. That's why the "less than or under 1/4 wavelength" and "up to 3dB" caveats relating to coupling. Also, keep in mind that the coupling and interaction is acoustic and since crossovers are not hard wall electronic or acoustic filters, the subwoofers will likely have output above the crossover point, and mains below the crossover point, whose acoustic interaction may be relevant. This is even more of a factor if you run the subs hotter than the mains as that can shift up the effective acoustic crossover and thus raise the highest frequencies relevant to the subs.

 

[Bennett Prescott]

Re: First time Centre Subs

You could conceivably make all the subs react the same way with individual processing and then use a master processor to create your sub array.

Until one of them approaches limit, either in the amp or in the box, and changes character... at which point your array will become non-linear in a frequency and level specific way. This can affect pattern, and certainly affect how it sounds. I like arrays made of identical components, or if that is not possible for one reason or another separate arrays doing different things (like flown subs of one type and ground of another, the ground subs being aux fed).

 

[Justin Myers]

Re: First time Centre Subs

Thanks for all the replies.

Great pics Per. They helped heaps. I see what Brian and Art were trying to explain to me.

I suppose that with a enough processing you could make it work just like the model. Do you have access and have comfortability with measuring ALL the subs individually? You could conceivably make all the subs react the same way with individual processing and then use a master processor to create your sub array. It's not something you can do in the field real easily though.

I'll try that when I get a chance Kip. I can run Smaart no probs and all pairs of subs will have their own processing since everything are Itechs HDs and Itechs.

Maybe I missed it but why are you doing this and what are you hoping to achieve by it? Is this just to experiment or are you trying to get some specific result or advantage from it? It seems that what you do might be very dependent on the result you are hoping to obtain.

It also seems like you aren't really looking at a single center array of subs but rather at four subs distributed in some fashion over a distance to either side of the centerline along with two box arrays under each main. That makes it sems like you may have some specific goal in mind in terms of the resulting directivity, but it is not clear what that goal may be.

Gday Brad. What I am aiming to acheive is a more even spread of Subs in the audience area. Trying to get away from the power alley effect from having L/R subs.

 

[Bennett Prescott]

Re: First time Centre Subs

What I am aiming to acheive is a more even spread of Subs in the audience area. Trying to get away from the power alley effect from having L/R subs.

Justin,

Just be aware that while you can create a sub array with very even coverage, your mains will still exhibit power alley to the same extent as they cover the same frequencies... especially in the 70-80Hz kick drum range, where a typical crossover point is. So now you have two different coverage patterns in the same frequency band from three different sources, which really messes up your seemingly perfect subwoofer array and can create a lot of frustration since you were expecting it to be fixed. It also smears time arrivals and makes it more difficult to get good impulse response. If there was a perfect answer, we'd all be doing that!

 

[Ivan Beaver]

Re: First time Centre Subs

Justin,

Just be aware that while you can create a sub array with very even coverage, your mains will still exhibit power alley to the same extent as they cover the same frequencies... especially in the 70-80Hz kick drum range, where a typical crossover point is. So now you have two different coverage patterns in the same frequency band from three different sources, which really messes up your seemingly perfect subwoofer array and can create a lot of frustration since you were expecting it to be fixed. It also smears time arrivals and makes it more difficult to get good impulse response. If there was a perfect answer, we'd all be doing that!

+100+
Yet another good example of how easy it is to look at the "simple" part of the problem.

As with many things speaker related-if all you look at is one part of the whole problem-then most likely you are missing many other huge chunks. It is the TOTAL picture that needs to be looked at.

You never hear anybody talk about interference patterns when it comes to the left/right stacks/hangs etc. They interfere just as much.

That is why in an install situation-my first "go to" is a center source-unless some other directive is given.

And what "fixes" a problem at one freq may very well cause other problems at different freq. So you HAVE to look at the "big picture". But is so much easier to look at simple numbers and answers and "assume" (wrongly) that all other freq will behave exactly the same.