CSA Cardioid Sub Array

So I decided to do a little more research on what i know as a CSA array or Cardioid Sub Array. In the posts on this kind of an array before this one I was mainly interested in getting as much cancelling on the stage side and as much level possible in the audience area.

I will be updating this post but it’s a start. (if you have Mapp on Line installed on your computer here are the ZIP-files from MAPP used to do this post CSA cue2016-3 CSA 2subs cue2016-3 CSA 2subs mics front cue2016-3)

In this post I’ll take a peak at what happens at different mic positions at the back and not at just 1 mic close for getting the maximum cancel. So here we go….

So here’s the start: A front-back-front set-up (3 x 900LFC Meyer Sound)

The front-back-front set-up consists of 3 900LFC with the middle one reverse (shooting back worth). The mic’s are positioned at 1mt 2mt 4mt 8mt and 16mt (so each time the distance is doubled).

The problem with this array: If you phase align the “back sub” with the front subs with the mic at 1mt from the array and then polarity reverse the back sub you’ll get a pretty good cancelation at that mic position but in the far field you get to deal with a level offset between the 2 front subs vs. the 1 back sub.

This means there is a 6dB offset between them so at the 1mt mic position you will see a pretty good cancelation due to the fact that the sound from the front subs has to travel around the array and assuming this is +/- twice the distance compared to the sound travelling from the back sub to the 1mt mic position you’ll end up with both parts of the array having matched level. Has to do with the Inverse Square Law (with each doubling of the distance the SPL level reduces with 6dB).

So with the”back sub” at 1mt and the front subs at 2mt the front subs will have 6dB more level but travelling twice the distance they also loose 6dB.

Within Mapp you’ll see the same thing at the 1mt mic position.

This is the predicted response from the 2 front subs at the 1mt mic position at the back of the array.

This is the predicted response from the back sub at the 1mt mic position at the back of the array.

And a compare between both responses (2 subs front and 1 sub back at the 1mt mic position).

(you might notice the level differences between the first 2 screenshots and ”the compare screen shot”. This has to do with a Gaussian program algorithm implemented within the Mapp program. The screenshots in the first part of this post were done in 2 day’s so i ended up with slightly different levels in the comparing screenshots sorry).

Now let’s look at the 2mt mic position.

This is the predicted response from the 2 front subs at the 2mt mic position at the back of the array.

This is the predicted response from the back sub at the 2mt mic position at the back of the array.

And a compare between both responses (2 subs front and 1 sub back at the 1mt mic position).

This is the predicted response from the 2 front subs at the 4mt mic position at the back of the array.

This is the predicted response from the back sub at the 4mt mic position at the back of the array.

And a compare between both responses (2 subs front red trace and 1 sub back blue trace at the 4mt mic position).

This is the predicted response from the 2 front subs at the 8mt mic position at the back of the array.

This is the predicted response from the back sub at the 8mt mic position at the back of the array.

And a compare between both responses (2 subs front red trace and 1 sub back blue trace at the 8mt mic position).

This is the predicted response from the 2 front subs at the 16mt mic position at the back of the array.

This is the predicted response from the back sub at the 16mt mic position at the back of the array.

And a compare between both responses (2 subs front red trace and 1 sub back blue trace at the 16mt mic position).

So you see the level difference increase between the 2 front subs and the back sub as soon as the distance increases. This means that the cancel you get after aligning this array might look pretty cool at the 1mt mic position but at the 16mt mic position the cancelation is not going to be as deep as the “cool” mic position.

Front subs @ the mic 2 3 4 & 5 positions

Back sub @ the mic 2 3 4 & 5 positions.

Now let’s phase align this array at the 1mt mic position.

1st the 2 front subs solo (red) vs. the back sub solo (brown). The phase trace from the back sub shows it needs to be delayed to match with the solo trace from the front subs. There’s a 120dg phase offset at 100Hz so i have to delay the back sub with 3.33ms (10ms =360dg so 120dg =1/3 of 10ms=3.33ms)

Back sub (brown) delayed with 3.33ms matches with the front subs (red).

Sum of both measured at the back shows +/- 6db addition at the back. This is not what we want so i have to polarity reverse the back sub to get cancelation.

By reversing polarity on the back sub the sum of all goes from 114.5dB to about 96dB (at some frequencies even deeper cancels 83 to 87dB).

This means a +/- 18.5dB reduction Knipogende emoticon (you can see the peaks settling around the 96dB line)

But this is at the 1mt mic position let’s take a look at the 2/4/8 and 16mt mic positions.

2mt mic position: 10dB reduction so still a pretty nice result but less reduction ((you can see the peaks settling around the +/- 98-100dB line).

4mt mic position: +/- 8dB of reduction

8mt mic position: +/- 7dB reduction

16mt mic position: +/- 6 to 7dB reduction

So watch out with putting such an array in to action because what might give you nice results in the near field can be misleading. The far field mic position shows less reduction and might cause problems regarding environmental laws (the fine on exceeding the levels on a open air thing in Amsterdam can run up to about 10000 euros trust me on that since i nearly got fined a couple of times Bedroefde emoticon ).

So this is the prediction field from MAPP at the 1/3 octave point of a graphic eq.

20Hz

25Hz

31.5Hz

40Hz

50Hz

63Hz

80Hz

100Hz

125Hz

Now let’s look at a CSA array of 2 subs.

Here’s the start.

Front sub @mic 1 1mt

Back sub @mic 1 1mt

Compare of front vs. the back sub @mic 1 1mt (so a +/- 6dB more on the back sub)

Front sub @mic 2 2mt

Back sub @mic 2 2mt

Compare of front vs. the back sub @mic 2 2mt (so a +/- 4dB more on the back sub)

Front sub @mic 3 4mt

Back sub @mic 3 4mt

Compare of front vs. the back sub @mic 3 4mt (so a +/- 3.5dB more on the back sub)

Front sub @mic 4 8mt

Back sub @mic 4 8mt

Compare of front vs. the back sub @mic 4 8mt (so a +/- 2.5dB more on the back sub

Front sub @mic 5 16mt

Back sub @mic 5 16mt

Compare of front vs. the back sub @mic 5 16mt (so a +/- 2dB more on the back sub)

And now the phase alignment. I moved this set-up a little to get the centre of the array in-line with the mic’s (so just checking).

Front Sub solo

Front & Back sub solo

Front & Back sub 3ms delay on back sub to phase align with front sub.

Front & Back sub 3ms delay+polarity reverse on back sub.

SUM (mic1-1mt from array) Front & Back sub 3ms delay+polarity reverse on back sub (solo front sub level 109.4dB goes to +/- 107dB so roughly 2dB reduction).

SUM (mic2-2mt from array) Front & Back sub 3ms delay+polarity reverse on back sub (solo front sub level 104.9dB goes to +/- 99dB so roughly 6dB reduction).

SUM (mic3-4mt from array) Front & Back sub 3ms delay+polarity reverse on back sub (solo front sub level 99.4dB goes to +/- 90dB so roughly 9dB reduction).

SUM (mic4-8mt from array) Front & Back sub 3ms delay+polarity reverse on back sub (solo front sub level 94.1dB goes to +/- 84dB so roughly 10dB reduction).

SUM (mic5-16mt from array) Front & Back sub 3ms delay+polarity reverse on back sub (solo front sub level 88.4dB goes to +/-76dB so roughly 12dB reduction)

Most of the time a –6dB is implemented on the back sub in an CSA array of 2 subs but by doing a –6dB on the back sub the reduction in the far field will be less. Also you run the risk of the cardioid effect disappearing. If this set-up is run at maximum level the the limiters on the front sub will be activated while the back sub still has 6dB to go before it’s limiters are activated.

If you implement a CSA of 2 subs you have the choice. Maximum reduction on stage or maximum reduction in the FAR field. Take your pick Knipogende emoticon

There’s also another challenge. If you do not implement the –6dB on the back sub it will chance the response on the audience side of the array.

So here are the predictions at 1/2/4/8 &16mt.