How To......Cardioid Sub Arrays

Hey

With all the talk of Rog’s Delta Array and the general idea of Cardioid sub arrays I thought it would be good to type up a quick tutorial on the basics of the theory of this technique and how to set it up in the real world, partly for all you good folks but partly for myself as well, as I’ve always found it a great test of your own knowledge if you can explain it to other people successfully.

As a quick introduction for people who don’t know what I’m talking about, the idea of a cardioid sub array is simply to set the subs up in a particular way so as to be able to control the direction of the bass output of your speakers (which is usually omni directional). You use the fact that when sound waves meet out of phase they cancel to create a null of energy behind the subs. You minimise the backwash of sound from the subwoofers and help to stop it travelling towards the stage, which makes your life easier when you have vinyl decks or open mics or anything else that is prone to feedback, and if done properly you can also harness this normally rear firing energy and use it our front.

Most of my knowledge comes from a couple of threads on the PSW forums (Pro Sound Web) which i’ll link at the end, and on there you can see nice MAPP diagrams and people who have actually set these up and taken real life measurements so it all definitely works.

Also note that as Rog reminds us in the other thread, this is all just theory, dealing with a “perfect” subwoofer and “perfect” conditions. In real life subwoofers aren’t perfect point sources and they have ports and group delay and lots of other nice real life issues so the calculations below should be used as a guide, and then when you actually set this all up you should play around with things a little to get the best settings for your particular set up.




OK, so the most common way to do create a Cardioid subwoofer array this is what’s known as an End Fire array (although to be honest I don’t know why, don’t really know what’s so End Fire about it LOL ).

There are a couple of basic variations on how to set it up, but the main idea is that you set up two rows of subwoofers, front and back, and you use a combination of time delay, physical distance and phase inversion to create a situation where behind the array the sound from the two rows meets destructively and cancels out, and in front on the audience side they meet in phase and combine.



Preparation

So the first step is to pick a frequency that will be the centre of your cancelation band. As we know every frequency has its own wavelength and as this trick relies on distance and delay is only properly effective across a band of frequencies, centred around whichever one you pick.  If your playing Dub, D&B, Reggae or anything else with big deep synths and lots of really low energy your gonna wanna move this down a bit, and if your playing more kicky stuff like House or Rock then you’ll wanna take that into account. Some report that you may get as much as an octave either side of the centre frequency although I don’t know if it’d effectively cover just as much as that.

So pick a frequency and work out its wavelength and also 1/2 and 1/4 of this, and also worth working out at this point what times these equate to for setting your delays later. 



Physical Set Up

So you have your frequency chosen, the next step is to set up the array. We need to create a path difference of half the wavelength of your chosen frequency, and to get our ideal conditions we will do this half with distance and half with signal delay. This means we have to set up our two rows of speakers so that there is a 1/4 wavelength physical spacing between the sound sources (drivers for front loaded or horn mouth for horns.)  This second row can also be facing forward if you like and have the space.

If your subs aren’t too deep and you can fit them in and still preserve the 1/4  wavelength spacing, you can spin the second row around and have them back to back which’ll save you a little bit of floor space.

For reflex subs I don’t think it’ll make much difference either way but for horns you will have to consider the fact that because if the nature of the horn they will already have some measure of directivity themselves. If you turn the second row around you may get away with using less in that row because the front row will be putting out less sound towards the back and the back row will be putting its full measure of sound backwards to cancel it. Some experimentation with your particular subs would help confirm this.

Note also however that for one of the two processing methods below we also have to reverse the phase of the rear speakers. If you set it up with both sets of speakers facing forwards you have the option of doing this with processing if you like, but if you reverse the back row to save space this effectively does a phase reversal for you so you are limited to just that method.

(edit: after thinking on this i suppose you could also add a processing phase reversal, flipping your backwards facing speakers back into phase again if you really wanted to use method 2. Might be worth trying to see how effective it is in practice.)


However, as long as the distance between the two sound sources is 1/4 wavelength then your good.



Processing and Delay

Now we come to setting the delays. We must add another 1/4 wavelength delay to one of our rows here, which’ll complete the 1/2 wavelength path difference and get our cancelation going, and again here we have two ways we can do this but in this case it WILL affect the sound you get.


Method 1:  Delay the REAR SPEAKERS by 1/4 wavelength for (theoretically) perfect cancelation.

This is the method that includes the phase reversal, so we must do that first and then add a delay time that matches up with our 1/4 wavelength, all to the back row of speakers. So In this case whats happening is the sound coming from the rear of the front row of speakers will take 1/4 of a wavelength to reach the back row. We match that with a quarter wavelength delay on the rear speakers so that they would effectively be in phase normally (just like we would delay a delay tower or whatever to match it up) but because we have also reversed the phase of the back speakers either by processing or physically turning them around, the delays now put them perfectly out of phase which makes them cancel out and you get theoretically perfect cancelation and hence no (or in real life much less) sound behind the array.

In front of the array, the sound from the rear speakers travels forward which takes 1/4 wavelength of time, and is also delayed in processing by 1/4 wavelength so it is half a wavelength behind overall and would be out of phase, but the phase reversal puts it back into phase and it combines to add its energy to the front and hence the audience. 

Note however that this changes the nature of the bass sound slightly, because although the waves are combining together in phase out front they are actually half a wavelength behind and then inverted back into phase again so they are not perfect mirror images of each other and there is a subtle but noticeable sound difference. This is reported by actual users as an ever so slight “rounding” off of the sound, a very slight loss in kick or attack, but only really noticeable to a trained ear who has the ability to A/B the two sounds, for the most part it is regarded as unnoticeable to the public.

However if you would like to make sure that you have perfect sound out front then you can use...



Method 2: Delay the FRONT SPEAKERS by 1/4 wavelength for (theoretically) perfect sound out front.

OK so in this case we have NO phase reversal. What happens is now the reverse of what happened in method one. The sound travelling from the rear row takes 1/4 wavelengths of time to reach the front row, and as we have now delayed this front row by a matching 1/4 wavelength they meet exactly in perfect phase and they combine and all our sound out front is  normal, just like delaying speakers or towers to match up. From the perspective of behind the speakers the sound from the front speakers is delayed 1/4 already in processing, and then takes another 1/4 wavelength to travel back to the back ones so when it finally gets there it is 1/2 a wavelength behind and is out of phase. Not that like in the front of method 1 although it is out of phase it is not a perfect mirror image but is half a cycle out, and so will cancel very well but not as good as in method 1.

So method 1 to have maximum cancelation sacrificing a shade of sound out front, and Method 2 to have the best sound out front but loose a little of the cancelation.


Ok think I have that all right although feel free to throw in any comments or tell me i'm all wrong  Tongue


the link that i mentioned to the PSW thread is....


http://srforums.prosoundweb.com/index.php/mv/msg/3954/0/0/0/


where they have lots of pretty pictures to look at and graphs to see Tongue



kev


Edited by kevinmcdonough - 18 April 2009 at 1:43pm

For sub-bass control i'm not convinced there is any advantage over just pointing all the subs forward & forming a wide horizontal array, in most circumstances.  IE try to induce as much diffraction directivity as possible & then the radiation pattern tends to be cardioid shaped anyway.

Edited by roborg - 17 April 2009 at 10:54pm

jamwa wrote: or if you use DB you can just point a thrid of your subs 180 degrees the other way run these on seperate amps and switch the CSA mode on - simples!!!!

Well... I do have and use Q-Subs and S2 cabs. The CSA method does work to some extend, but you need more gear ( cabs and amps ) to end up with less SPL than straightforward non-CSA. I tried it several times in various venues from very small to rather big and from my experience it's a small benefit at best at the cost of more gear. The feeling creeps on me that it's a marketting ploy to sell more gear because they ( not only D&B ) convinced you bass splill at the rear is THE BIG problem. But is rear spill really that big a problem ? I admid it can be in some cases, but an environmental problem primarly in open air gigs. Every time I used it, musicians start complaining they need more bass in the monitors... . In small venues CSA is just plain useless. More cabs and amps where you don't have the place for it in the first place and definately a visual problem with one or more cabs facing the public with it's rear... try to sell that to your client... good luck ! Bass spill is slithly less behind the cabs, but nobody talks about what gets thrown in the air or in the ceiling. You have 50% more gear making noise and some of it interacts with the first 100%, but not all of it. In bigger setups I always get better result from stacking and arraying multiple sub and low cabinets and play with fysical positions, (sometimes) frequency overlap and delays between cabs or arrays of cabs. My main goal is allways to get the most possible SPL to the audience and as little as possible EVERYWHERE else, not only at the rear. The dimension and situation of a venue usually dictates placement and I start from there.
For me, CSA is just one of more out-of-the-box solution that 'can' sometimes partly fix a problem or part of a problem. Fine for folks that don't have the knowledge or time to figure things out for themselves in various situations and setups, but I prefer the 'manual' way. Far better result with the same gear instead of limited benefit at the cost of more gear.


Edited by Peter Jan - 21 June 2009 at 2:20am

This thread...

http://srforums.prosoundweb.com/index.php/t/44778/7154/

and back to Dave`s blog for his take on it...

http://ratsound.com/blognav.htm

Highly informative and some intriguing questions?

.p.

Hi 

I can sort it out for you ;) Will PM.