[Chris L]

Hey all

I know a few bits and pieces about live sound, but am looking to improve my knowledge of system design.
This is for a hypothetical situation, where the spread of bass needs to be controlled. For example, keeping as much bass as possible inside a club or just giving more control over sound dispersion. Now I know that some manufacturers do make cardiod subs, but all that I've heard good reviews for are a bit expensive.

So:
Is it possible to create a cardioid response if a few subs are arranged in the right way and one or more have their phase reversed?
Does the type of sub matter (ie front/back loaded, vented etc...)?
Are there any other ways of controlling bass dispersion?
A search of the BR revealed a little info about manufacturered cardioid subs, but nothing I could find for these q's.

Thanks
Chris

Edit: sp.

This post has been edited by Chris L: 18 June 2008 - 07:32 PM

[JackPott]

This is very possible and indeed can work quite well.

I can't remember the specifics (ie the maths) but I think you can use either a front and rear row, with delay so they cancel behind OR a row of them along the front of the stage properly spaced so they follow line law.

I don't think it matters much which sort of sub you use.

Sorry I can't remember the exact details, I've never done it myself

[RustyBrooks]

[Theres more than one way to skin a cat, some subs out, some of the nexo stuff, and d&B j subs, are cardiod by design, and use complictaed controller settings to time and phase the signal in such a way to kill the sound radiating out the back of the boxes, this helps to throw more energy forward, and more imprtantly make the area behind the subs (the stage) less noisy, with less spill into open mics, though this can somtimes mean that you have to add more sub to the stage with monitoring, as the bands may miss the ooommmppppfffff of the main pa subs! The cardiod pattern just k=looks after directivity of the subs, not o much dispersion, as the pattern drops to 240 degress instead of 360.... to make the subs more precise, it is possible to array subs horizontally and then, useing some clever maths, or more so now, a spreadsheet, to work out the coverage you want from the subs, and then delay each sub section of the array... this works really well in arenas and stadiums, but theatres, where you can't get the subs accross the front of te stage, it is of less use...

For the stackk them up accross the front of the stage anitcs, almost any sub can be used, as long as each column is on its own delay feed, whereas for cardiod, you need specific subs, and amps, or masses of time which isn't really availible.

Read the L-accoustics and d&b manuals for proper information.... no just jaded roadie words.

This post has been edited by RustyBrooks: 18 June 2008 - 09:10 PM

[Chris L]

Cool, so it's possible, just requires a fair bit of effort...
The hypothetical situation I have in mind is an install for a live music venue where there are neighbours in fairly close vicinity as this happens quite often, so there would be time to set up a complex system properly.

Jackpott, would the cancellation from using a front and a rear row with delay not be frequency dependant?

Does anyone know enough physics/maths to explain the basics behind how to achieve this, or is there any good (and free) software out there that I could fiddle around with?

Looking through the L'acoustics and the D&B manuals, I now know that at least 1 speaker need to face backward, presumably with its phase reversed?

Thanks
Chris

This post has been edited by Chris L: 18 June 2008 - 10:02 PM

[Rob_Beech]

Yes, it is frequency dependant.

You'll notice a change in the response at other related frequencies aswell, which you may be able to use to your advantage, or may cause problems.

No, one speaker does NOT need to face backwards in all cases. D&B's CSA mode has 2 speakers facing forwards and 1 facing back. There are more complicated alterations to the signal than just polarity and delay but the amps have this setting built in so its very straight forward to setup.

A simple cardioid pattern does not need one box facing backwards. It needs to be spaced correctly, inverted polarity, and delayed. The spacing and delay times are the frequency dependant parts.

BR User Simon West has studied different sub arrays in alot of detail over the past 12 months, if you're lucky enough to catch him on a not so busy day he may be willing to explain some of it to you.

[JackPott]

Indeed.
It will be centred around a frequency, but as Rob says, with some jiggling this is hopefully complinmentary to its nearby frequencies.

If you have access to any of the manufacturers prediction tools (Meyer's MAPP, Martin Viewpoint etc) you should be able to set up a simple model and play around with phase delay and distance and see the results instantly.

Most of these require registration but you might know someone with this, or you might have your own software which can do it.

[andy_s]

I saw an interesting "live" (ie not modelled in a computer) demonstration once by John Taylor from D&B, using 3 stacked QSubs and a turntable thingy, he showed us the marked difference between the 3 boxes pointing in the same direction, and the cardioid arrangement with the middle one pointing backwards.

This post has been edited by andy_s: 19 June 2008 - 12:44 PM

[simonwest]

Someone Rang.

Theres several ways to 'beam steer' Low frequency energy, heres a few..but not limited to:

'One in front one behind'
Delay a rear positioned bass loudspeaker speaker (say 1m behind) , invert polarity...and away you go, although you may find that there are lobes that appear in the negative half at around the 135degrees and 225degrees...which funnily enough makes the beam look 'cardioid', with good processing you can make it look more like a hypercardioid, which is good or bad...depends what you want to achieve.


'Inverted loudspeaker', look up D&B's CSA stack.
Rotate an element, Edit the Phase, EQ, and few other bits and bobs, rather complicated. requires FIR filters, so if your good with C/C++ and have the appropriate sound cards / DSP units to process through this can be achieved..but rather time consuming.

'End fire array / Logarithmic End fire array / Log end fire array with phase shift '
Use progressive delay, where by the very rear box is 0delay and the most front box is the last delay. The diffraction causes destructive interferance in rear projection and good summation in forward, I managed to get a minus 30dB drop in all the negative 180degrees areas (90 through 270). Down side is, you require ALOT of boxes, processing, amps and space..which makes this slightly infavourable to crew and alot of events. This has by far the best results sonically (IMO), and for the event licences. Although, I feel very sorry for the people in the front row, if you dont limit right, the forward summation can be very scary and ear ripping. So limit properly to protect the punters too!

Theres a fair bit of science behind all this aswell, My advice would be to read Bob McCarthys book, its green and called Sound Systems: Design and Optimization..very good, highly recommended..not just for sub stuff. Also look up Glen Ballou, Handbook for sound engineers, and if your lucky, Harry Olson's Acoustical Engineering. Reprint. Philadelphia, Professional Audio Journals. Page ,535

If you need specific stuff, Please feel free to PM me. Apart from that, theres stuff on prosoundweb and previous posts on here. Its very difficult to put it all in here, my dissertation was around 100pages long (Sorry Simon Lewis), but the maths is actually not scary for the basic cardioid stuff.. just space the subs (preferably as close together as possible, but usually you wont get closer than 80cm from the acoustic centre due to sub size, to be even better, space at a frequency in which you want forward summation to be 1/4 wavelength such as for 100Hz, 344 / 100 = 3.44 / 4 = 0.86cm distance for factor 1.4 (3dB) increase at 100Hz). Then go (assuming you've done spacing for 1/4 wavelength at given freq:

331.4 + (0.606 * temp in celcius) = C (speed of sound),

1/C = Speed of sound per Meter (1/344 = 2.91)

0.86 (distance) * 2.91 = 2.5ms delay on read speaker, and invert rear polarity.

its that easy

Play around on Meyersounds Mapp online, start by using CQ1's..I know they are not a sub, but because their prediction is a 'true omni' (unlike the HP700 as thats modelled for its actual predicted performance) its low frequency representation is a good starting point, before you play with the differences of 'omni resoponse' as frequency varies.... also note, just because theory teaches 'as you go up in frequency, the omni projection becomes narrower', does not mean to say 125Hz cannot be more omni that say 50Hz, its a matter of box construction, driver positioning and so forth. this means you may have to 'tune' the rear boxes / inverted boxes to get a tailored response, particularly to sort out unwanted boom / bass spill onto the stage.


Is it possible to create a cardioid response if a few subs are arranged in the right way and one or more have their phase reversed? Yes

Does the type of sub matter (ie front/back loaded, vented etc...)? Technically no, but easier, depends on your processing and what your trying to achieve

Are there any other ways of controlling bass dispersion? Yes, Bessel array, End fire array, Beam steering... (1/t, 2/t etc and steering the audio into an area desired)

HTH

Simon

This post has been edited by simonwest: 19 June 2008 - 11:09 PM

[Chris L]

Many thanks for that, it's given me plenty to ponder over
I'll have a look at a couple of those books if I can and a search or 2 on the internet. I'm not afraid of maths, in fact I quite enjoy it, so I'd love to see it all working in action properly
Will be in contact if I get stuck

Thanks
Chris

[shytower]

Chris,
I own a sound company in St Louis. I have six JBL VRX 918 subs. I arranged the subs (three subs a side) to cut down on the low end coming off the back of the cabs by facing four subs forward and the two outside subs backwards with the phase reversed. IT WORKS! Try it.
John

Moderation: Edited to remove needless quoting--twice!--of original post.

[mackerr]

shytower, on 30 May 2009, 10:57 AM, said:

I own a sound company in St Louis. I have six JBL VRX 918 subs. I arranged the subs (three subs a side) to cut down on the low end coming off the back of the cabs by facing four subs forward and the two outside subs backwards with the phase reversed. IT WORKS! Try it.

There is more involved than just turning around a speaker and reversing its polarity. It may work better if the rear facing speaker is not reverse polarity, and just delayed. There is a long informative thread on steering low frequencies here. There are a lot of posts, but it is worth the effort to read it.

Mac

[shytower]

mackerr, on 31 May 2009, 3:56 AM, said:

shytower, on 30 May 2009, 10:57 AM, said:

I own a sound company in St Louis. I have six JBL VRX 918 subs. I arranged the subs (three subs a side) to cut down on the low end coming off the back of the cabs by facing four subs forward and the two outside subs backwards with the phase reversed. IT WORKS! Try it.

There is more involved than just turning around a speaker and reversing its polarity. It may work better if the rear facing speaker is not reverse polarity, and just delayed. There is a long informative thread on steering low frequencies here. There are a lot of posts, but it is worth the effort to read it.

Mac

Do what I said Mac. It works. I'm not the only one doing it. I have a good friend who is an Harman Group rep. He is doing it with vertec system in Branson. Use your ear, not the crap you read on the posts about delays. I've read it all too. In the real word on gigs I don't have the space to set my subs up in a cardioid pattern with delays. I set my subs up side by side, three a side with the outside subs backwards with the polarity reversed. This makes the low end drop out behind the sound system and force all the low end forward. If you own a sound company try it. If you are just a tech, you are SOL unless you can talk someone into trying this. I know it sounds crazy but it with drop your jaw the first time you do this. I've been in this business since 1978. Reversing the polarity does cool things. Bob Heil told me he use to reverse the polarity on the monitors for the "The Who" back in the 70's. That was the only way he cound get them so loud without feedback.

[Rob_Beech]

The point made is that there is more than one way to do this. And different ways have different effects on the sound and dispersion characteristics of the LF.

Shytower, I'm sure it works in the environment you use it in and it's incredibly useful that a simple arrangement works well. It will not however achieve the results the engineer (or whoever else) in every setting. The reason being, the results they want can differ. Some forms of array create better rejections in the centre points between 90 degrees and 180 and between 180 and 270 where 180 degrees has a lower rejection. Other types of array allow very good rejection at 180 degrees but a very non linear pattern of rejection between 90 and 180 and the points between 180 and 270. Depending on what you want to achieve will depend on what form of array you create. Some need much more processing than others. Some, as you state need nothing more than a polarity reversal.


Rob

[Bobbsy]

What Rob said.

I'm sure Shytower's recommendation works well--with his speakers and in his situation. However, it's not a universal solution that'll work with every cabinet in every space. For an example of what can go wrong, I suggest the first post in the link mackerr provided should be required reading.

Done without care, cardioid subs can cause far more problems than they solve.

Bob

[dbuckley]

shytower, on 31 May 2009, 8:40 PM, said:

If you are just a tech, you are SOL unless you can talk someone into trying this.

Thats the first time I've heard Mac described as "just a tech". I think most of us would gladly donate a body part or two to be "Just a tech" with the experience, resume, and credibility of Mac Kerr...