ProSoundWeb Community

Sound Reinforcement - Forums for Live Sound Professionals - Your Displayed Name Must Be Your Real Full Name To Post In The Live Sound Forums => LAB: The Classic Live Audio Board => Topic started by: Robert Lunceford on July 27, 2014, 09:50:18 PM

Title: When is it a line-array?
Post by: Robert Lunceford on July 27, 2014, 09:50:18 PM
I'm sure most of us have seen from one to three line-array "elements" mounted on a pole over a sub, or on a tripod stand as shown in this video.
http://www.youtube.com/watch?v=QZu8deHi6Ug
The DVA series are a very popular choice.
When there are only 2 or 3 "elements" stacked, will it actually perform as a line array?
If not, will it perform more like a point source speaker? How would we realistically expect a stack of 2 to perform?
I looked at the DVA spec sheet which includes the following specs:
Frequency Response [+/- 3dB]: 80 - 19.000 Hz
Directivity: 100x15° Single unit

At what range of the frequency response can we expect the 100X15 degree directivity? I wouldn't expect this directivity for the full frequency range of a single cabinet.

Do any of the professional audio societies have a published standard that defines what a "line-array" is?
I'm wondering how many "elements" you would need before the system actually begins to perform as a line-array.
Title: Re: When is it a line-array?
Post by: TJ (Tom) Cornish on July 27, 2014, 10:00:20 PM
I'm sure most of us have seen from one to three line-array "elements" mounted on a pole over a sub, or on a tripod stand as shown in this video.
http://www.youtube.com/watch?v=QZu8deHi6Ug
The DVA series are a very popular choice.
When there are only 2 or 3 "elements" stacked, will it actually perform as a line array?
If not, will it perform more like a point source speaker? How would we realistically expect a stack of 2 to perform?
I looked at the DVA spec sheet which includes the following specs:
Frequency Response [+/- 3dB]: 80 - 19.000 Hz
Directivity: 100x15° Single unit

At what range of the frequency response can we expect the 100X15 degree directivity? I wouldn't expect this directivity for the full frequency range of a single cabinet.

Do any of the professional audio societies have a published standard that defines what a "line-array" is?
I'm wondering how many "elements" you would need before the system actually begins to perform as a line-array.
What do you mean by "performs as a line array"?
Title: Re: When is it a line-array?
Post by: Tim McCulloch on July 27, 2014, 10:03:32 PM
I'm sure most of us have seen from one to three line-array "elements" mounted on a pole over a sub, or on a tripod stand as shown in this video.
http://www.youtube.com/watch?v=QZu8deHi6Ug
The DVA series are a very popular choice.
When there are only 2 or 3 "elements" stacked, will it actually perform as a line array?
If not, will it perform more like a point source speaker? How would we realistically expect a stack of 2 to perform?
I looked at the DVA spec sheet which includes the following specs:
Frequency Response [+/- 3dB]: 80 - 19.000 Hz
Directivity: 100x15° Single unit

At what range of the frequency response can we expect the 100X15 degree directivity? I wouldn't expect this directivity for the full frequency range of a single cabinet.

Do any of the professional audio societies have a published standard that defines what a "line-array" is?
I'm wondering how many "elements" you would need before the system actually begins to perform as a line-array.

To be a mathematical line source, it needs to be infinitely long.  That presents problems with my truck pack. ;)

The use of 2, 3, or 4 line array elements is a way to make more use of the investment, not a pursuit of audio excellence.

The DBTechnologies T4 is a model I have some experience with.  The 100° horizontal is consistent with itself regardless of how many boxes you line up, and the HF vertical dispersion of a single box is 15° give or take a degree or 2.  What happens in the vertical as you go down in frequency is determined by the length of the line, and so far I don't like the way anything <6ft long sounds, pretty much regardless of brand.

It's not about how many, but how long.  Ask your significant other... /nudge, wink, satire.
Title: Re: When is it a line-array?
Post by: Robert Lunceford on July 27, 2014, 10:11:56 PM
What do you mean by "performs as a line array"?

Hi TJ,
As described in this paper written by Meyer.
http://www.meyersound.com/support/papers/meyer_line_array.pdf
Title: Re: When is it a line-array?
Post by: Jay Barracato on July 27, 2014, 10:25:21 PM
I just spent 4 days at a festival where the stage we were on had 2 baby martin LA boxes on stands per side. This was the first time I gave ever really been happy with that configuration.

Not my system so I have no idea if there was anything really special in the processing but I was able to step up and just do my show and all the eq I really needed was the channel.

Sent from my DROID RAZR HD using Tapatalk

Title: Re: When is it a line-array?
Post by: Tim McCulloch on July 27, 2014, 10:37:11 PM
Hi TJ,
As described in this paper written by Meyer.
http://www.meyersound.com/support/papers/meyer_line_array.pdf

What part of it are you finding most significant?  This was Meyer's response to another brand's marketing claims.

Note, too, that some of the things discussed in the white paper have been addressed by subsequent design advances from several manufacturers (including proprietary designs by that "Global" company that used to be Brothers).

I'd further direct you to some AES papers by Mark Ureda (EV, Altec, JBL) https://www.jblpro.com/ProductAttachments/AES_Ureda_Analysis_of_Line_Arrays.pdf (warning, heavy math lifting) and https://www.jblpro.com/ProductAttachments/MSU_AES_Fall_2001_Monday_Morning.pdf along with http://www.jblpro.com/pub/tour/aes%20may%2001%20ureda%20line%20arrays.pdf and David Scheirman's paper http://www.jblpro.com/pub/tour/scheirmanaes.pdf  You might want to look at Mark Engebretson and Doug Button's work, too.
Title: Re: When is it a line-array?
Post by: Robert Lunceford on July 27, 2014, 10:57:25 PM
To be a mathematical line source, it needs to be infinitely long.  That presents problems with my truck pack. ;)

The use of 2, 3, or 4 line array elements is a way to make more use of the investment, not a pursuit of audio excellence.

The DBTechnologies T4 is a model I have some experience with.  The 100° horizontal is consistent with itself regardless of how many boxes you line up, and the HF vertical dispersion of a single box is 15° give or take a degree or 2.  What happens in the vertical as you go down in frequency is determined by the length of the line, and so far I don't like the way anything <6ft long sounds, pretty much regardless of brand.

It's not about how many, but how long.  Ask your significant other... /nudge, wink, satire.

Thanks Tom,
If the HF vertical dispersion of a single box is 15 degrees, I can see that a single box can be useful as a front fill. But what about 2 boxes on a pole? What practical purpose would you use this deployment for? It seems you would have to be careful how high up they are. If they are above the audience you may not be able to hear a good balance of HF to LF unless they were angled downward.
Title: Re: When is it a line-array?
Post by: Tim McCulloch on July 27, 2014, 11:31:37 PM
Thanks Tom,
If the HF vertical dispersion of a single box is 15 degrees, I can see that a single box can be useful as a front fill. But what about 2 boxes on a pole? What practical purpose would you use this deployment for? It seems you would have to be careful how high up they are. If they are above the audience you may not be able to hear a good balance of HF to LF unless they were angled downward.

Tom will be along shortly.... ;)

How much vertical HF does the application require, and will the HF outrun the low-mids and lows?

Yes, aiming is important, and entire software programs are dedicated to predicting array coverage so the user can optimize coverage as he sees fit... but yeah, "point the speaker system at the listeners" still applies.

BTW, one of the things Meyer talks about that is often overlooked is how much LF is radiated behind the array.  This creates some interesting interactions with certain types of venue architecture (barrel vaulted ceilings and high trim heights...)
Title: Re: When is it a line-array?
Post by: Robert Lunceford on July 28, 2014, 12:03:38 AM
Tom will be along shortly.... ;)

How much vertical HF does the application require, and will the HF outrun the low-mids and lows?

Yes, aiming is important, and entire software programs are dedicated to predicting array coverage so the user can optimize coverage as he sees fit... but yeah, "point the speaker system at the listeners" still applies.

BTW, one of the things Meyer talks about that is often overlooked is how much LF is radiated behind the array.  This creates some interesting interactions with certain types of venue architecture (barrel vaulted ceilings and high trim heights...)
Hey Tim,
My apologies on the typo of your name.
If you had two of these boxes per side with a pair of their companion subs, how large of an area could you reasonably expect to cover?
Title: Re: When is it a line-array?
Post by: Scott Holtzman on July 28, 2014, 12:12:22 AM
Tom will be along shortly.... ;)

How much vertical HF does the application require, and will the HF outrun the low-mids and lows?

Yes, aiming is important, and entire software programs are dedicated to predicting array coverage so the user can optimize coverage as he sees fit... but yeah, "point the speaker system at the listeners" still applies.

BTW, one of the things Meyer talks about that is often overlooked is how much LF is radiated behind the array.  This creates some interesting interactions with certain types of venue architecture (barrel vaulted ceilings and high trim heights...)

Which is a great time to bring something else up.  Where do compact line arrays fit in?  Vue especially has been showing off the al-4, these things are tiny, 5.5" tall and when flown in a large array go in packs of 8 so that would  give you pattern control down to about 400hz. 

Is the compact line array all about beamforming or I am completely missing the point.  In the demo the show them stacked 4 on a sub for a club and on up.

Is this another trend or is some science behind it? 

Title: Re: When is it a line-array?
Post by: Peter Morris on July 28, 2014, 03:09:14 AM
I'm sure most of us have seen from one to three line-array "elements" mounted on a pole over a sub, or on a tripod stand as shown in this video.
http://www.youtube.com/watch?v=QZu8deHi6Ug
The DVA series are a very popular choice.
When there are only 2 or 3 "elements" stacked, will it actually perform as a line array?
If not, will it perform more like a point source speaker? How would we realistically expect a stack of 2 to perform?
I looked at the DVA spec sheet which includes the following specs:
Frequency Response [+/- 3dB]: 80 - 19.000 Hz
Directivity: 100x15° Single unit

At what range of the frequency response can we expect the 100X15 degree directivity? I wouldn't expect this directivity for the full frequency range of a single cabinet.

Do any of the professional audio societies have a published standard that defines what a "line-array" is?
I'm wondering how many "elements" you would need before the system actually begins to perform as a line-array.


Although the T4 is a line array box, two or three boxes is not a line array, its just an array with a narrow and adjustable vertical HF pattern.

In the case of the T4s, they have 2 / 3 box settings and work fine … and you can add boxes until they behave as a “real” line array and the pattern control is extend to the low-mid frequencies. The longer the array the lower the frequency of the pattern control.

FWIW the T8’s sound much better,  go much harder and integrate well with the T12s.
Title: Re: When is it a line-array?
Post by: TJ (Tom) Cornish on July 28, 2014, 07:22:28 AM
Hi TJ,
As described in this paper written by Meyer.
http://www.meyersound.com/support/papers/meyer_line_array.pdf
I ask, because even really long line arrays don't truly have what some folks claim - loss of only 3dB over double the distance distance compared to 6dB/double the distance of a "point source" - at least not intrinsically.  The advantage of a vertical array is that elements can be combined to throw more sound at the back of the coverage area and gradually less towards the front which to some degree equalizes the volume, however even really long arrays lose pattern control at low frequencies. This means that at some point you will lose the ability to produce equal tonality over distance if you are trying to chase equal volume, as the array only "works" down to a certain frequency, which is determined by the array length.

There are some ways around this, with caveats, as nothing is free.  JBL's 4886 system can be hung with the mid-sub 4883 elements in cardioid mode, which greatly improves LF pattern control, but using a different mechanism than line-array element superposition.

If your question is "Are 'dash' arrays useful and/or better than a traditional point source speaker?", that's a different thing, and the answer depends on what you're trying to do.  I have recently gotten to do some larger events with my 4886/4883 gear.  Saturday night we had 2 4883 over 7 4886, and it sounded really nice for our application from about 20' to 200'+.  Other than a couple Danley boxes like the Jericho and Genesis horns, I don't know of too many point sources that could do what we needed to do, especially with our hang weight of only 400lbs.

One application that is more questionable is for a permanent install.  It becomes harder to make the case for a relatively expensive "Swiss Army" system like most small format vertical arrays, when you can go to the catalog and buy the right point source boxes for the room and probably get better sound for less money.

Title: Re: When is it a line-array?
Post by: Ivan Beaver on July 28, 2014, 07:50:58 AM
I would argue that most people that talk about "line arrays" simply have no idea what makes up a line array-what the pattern control really means or have any idea how long (NOT the number of elements) they have to be to even begin to perform "line array magic" (ie 3dB/doubling of distance) or have a clue that even "point sources" can provide equal SPL (no losses) from the front seat to the back.

So why is the 3dB/doubling really important? And what happens to cause that "apparent" loss?  Hint- it is because of DESTRUCTIVE INTERFERENCE.  It does not make the far seats louder-the cancellation causes the near seats to be not as loud. 

But at a price of all sorts of "finger lobes" (you know-the things the marketing depts erase from the measured data) that can cause erratic response at different seats.

They look at the "pattern" number and assume it means that is the pattern over a wide range of freq. but yet if you look at the measured polar pattern (or do a little math yourself), you will see that often the stated pattern is only for the top octave or 2-not even getting down into the "important freq" for intelligibility. 

But that does not stop the marketing engines from churning out false information.
Title: Re: When is it a line-array?
Post by: Kevin McDonough on July 28, 2014, 08:52:08 AM
I would argue that most people that talk about "line arrays" simply have no idea what makes up a line array-what the pattern control really means or have any idea how long (NOT the number of elements) they have to be to even begin to perform "line array magic" (ie 3dB/doubling of distance) or have a clue that even "point sources" can provide equal SPL (no losses) from the front seat to the back.


hey

yeah what most people don't get is that there are two things happening at the same time to control the vertical pattern, one for the HF and one for the LF.

For the HF its simply controlled by the horn/waveguide as it would be for any other box. Yes they can be a bit fancy and complicatedly designed to turn the circular mouth of the HF driver into a rectangular exit running from the bottom to the top of the cab (and so join with the next cab) and try and keep a consistent path length etc, but at the end of the day, the waveguide's control of 5, 10 or 15 degrees is what allows you to aim the box.

however at the low/mid end, its the opposite. Especially with a reflex mounted driver, the coverage pattern will be the same as a point source. You stick a 12" driver in the front of a box and is doesn't matter whether its a "line array" box or a point source, its still a 12" driver on the front of a box and will output sound the same way. Its the vertical stacking of the boxes, and as Ivan says the constructive and destructive interference of the sound from these drivers when they're stacked within 1/4 wavenength of each other (which is largely what manufacturers aim for with line arrays), that begin to create pattern control just in the same way that a long line of subs also begins to aim the sound in an ever narrowing alley.

The extent to which this happens is down to the length of the line against the wavelength your talking about. So as the line gets longer it is able to maintain pattern control to a lower and lower set of frequencies. 

A small line array of only two or three boxes on a pole will have line array type pattern control, however it'll just be at relatively high frequency and will be increasingly more like a point source as you move down.

And note again, as Ivan says the length of the line overall, not the number of boxes. Whether its made up of twelve 6" boxes or six 12" boxes its the overall length that counts.  (well, technically the smaller boxes will have smaller drivers with slightly different characteristics and tonality than the bigger ones, and the individual drivers will beam differently and have their own individual pattern control, but BROADLY speaking its the length of the line overall).


So are short arrays worthwhile?  Yes, its another tool in the toolbox, they have good points and bad points.

The fact that you only have one line of boxes (each side) horizontally helps with interference between boxes. However you're stuck with a specific dispersion horizontally you can't just add a smaller number of narrow coverage boxes if you need narrow dispersion.

They have flyware built in (usually) so if you have points or truss to hang them from you can fairly simply. But if you don't they can be a little harder to groundstack if they don't come with a pole mount.

They will probably be made up of lots of smaller drivers at the polemount/two or three box size, and while they'll be more sensitive they'll handle less power and you may loose a lot of the low mid balls you'd get from just having a single 12" or 15" point source box, and so your subs may have to play much higher to meet them or you may need a dedicated low mid/upper bass box as well as subs.


However at the end of the day though, one of the main ways they will be different will be perception, and that unfortunately can be key.  While we would all like to make our decisions based on pure sound quality and what is the absolute best tool for the job sound wise, often there are many other factors.

And if you're gonna get a whole load of extra work because something that looks like a line array 'sounds' better to promoters, band members and venue owners then its hard to ignore that ROI.



k







Title: Re: When is it a line-array?
Post by: Brian Jojade on July 28, 2014, 11:23:58 AM
The very simple answer is, it's a line array when the marketing department deems it so.  Simple as that.
Title: Re: When is it a line-array?
Post by: Mark Cadwallader on July 28, 2014, 02:43:22 PM
The very simple answer is, it's a line array when the marketing department deems it so.  Simple as that.

Oh, you mean a "lie arrary"!  Mark C.
Title: Re: When is it a line-array?
Post by: Scott Holtzman on July 28, 2014, 04:44:41 PM
I am quite sure that the experts are tired of repeating this information but there is so much FUD concerning arrays that these conversations (especially with the same participants) have tremendous value to the community.

I touched on the Vue AL-4 and others mentioned compact and ultra-compact have a place.  Vue claims that they can alter the phase to each driver within a narrow range of frequencies and be able to control the pattern via software and physical alignment.

Is this an accurate claim?  If you went into one of these null zones would the energy still be there just canceled out?

Title: Re: When is it a line-array?
Post by: Kevin McDonough on July 28, 2014, 05:01:37 PM
I am quite sure that the experts are tired of repeating this information but there is so much FUD concerning arrays that these conversations (especially with the same participants) have tremendous value to the community.

I touched on the Vue AL-4 and others mentioned compact and ultra-compact have a place.  Vue claims that they can alter the phase to each driver within a narrow range of frequencies and be able to control the pattern via software and physical alignment.

Is this an accurate claim?  If you went into one of these null zones would the energy still be there just canceled out?


yes, its accurate.  Just as I said above the basics of line array directionality are essentially the same as that of subs. More drivers radiating closely together in a line will, thanks to cancellations, will narrow the pattern. The longer the line the narrower the beam will become, and the lower it'll hold to.

And just like subs, by delaying cabinets you can make the waves meet in or out of phase at certain points and control the pattern.  Its very common to, for example, progressively delay the outer cabinets when you have a long line of subs across the front of the stage to counter the narrowing effect and widen the pattern again.

And essentially its the same process, a little harder to program/calculate/predict because of the smaller wavelengths and the aiming/j shape of the array and because you generally have more drivers and a much wider bandwidth to try and control, but essentially the same overall idea.


Martin and EAW are already doing it to great effect with MLA and Anya respectively. And their advantage is that you're not trying to do all the work, the trial and error of manually changing delay times and seeing what effect it has, change again and test, change again and test......  With their software you just tell it the coverage you want, where you want it to start and end and where/if you want 'holes' in the coverage to miss balconies and things, and then the software does all the hard work to calculate what processing all of the drivers need to get it done.


k



Title: Re: When is it a line-array?
Post by: Jay Barracato on July 28, 2014, 05:05:08 PM


Is this an accurate claim?  If you went into one of these null zones would the energy still be there just canceled out?

I am not sure what you mean by this as it does not seem to tally with how energy is transferred in a wave form.

When two waves overlap with destructive interference at a particular point that doesn't cancel the energy that is being transferred through that point.
Title: Re: When is it a line-array?
Post by: Scott Holtzman on July 28, 2014, 05:52:41 PM
I am not sure what you mean by this as it does not seem to tally with how energy is transferred in a wave form.

When two waves overlap with destructive interference at a particular point that doesn't cancel the energy that is being transferred through that point.

That is correct, basics if you connect two speakers out of phase and reproduce a sine wave and stand right in the deepest null with a SPL meter you will see the "energy" in that space still exists.

So even with the very granular pattern tuning these arrays are capable of it is not efficient as you need the original power and the power to cancel the unwanted coverage.

I just wanted to make sure I had the full picture.

Title: Re: When is it a line-array?
Post by: mark lonow on July 28, 2014, 07:18:22 PM

thanx kevin nice posts

too some of the others...
how come line arrays need to be 6 foot?
but  its ok for point source/ trap arrays  to have no 250hz  directivity?
another random stab....
and its ok for trap arrays to be non 1/4 spacing?

its all about $$$$$$ and directivity in audio is high dollar
a dash is just another compromise in a imperfect world on a budget
pick ur evil





Title: Re: When is it a line-array?
Post by: Mac Kerr on July 28, 2014, 07:26:10 PM
how come line arrays need to be 6 foot?
but  its ok for point source/ trap arrays  to have no 250hz  directivity?
another random stab....
and its ok for trap arrays to be non 1/4 spacing?

Point source can have directivity at low frequencies if they're big enough, just like line arrays. The length of the line determines the low frequency corner for pattern control. Six feet gets it down to a useful number.

Many line array elements also rely on the positive summation of the low frequency drivers to have enough low end, so short lines may be low end deficient.

Many point source speakers do not sound good used in multiples.

Mac
Title: Re: When is it a line-array?
Post by: Peter Morris on July 28, 2014, 07:51:48 PM
I am quite sure that the experts are tired of repeating this information but there is so much FUD concerning arrays that these conversations (especially with the same participants) have tremendous value to the community.

I touched on the Vue AL-4 and others mentioned compact and ultra-compact have a place.  Vue claims that they can alter the phase to each driver within a narrow range of frequencies and be able to control the pattern via software and physical alignment.

Is this an accurate claim?  If you went into one of these null zones would the energy still be there just canceled out?

http://firmaker.afmg.eu/

http://www.youtube.com/watch?v=PA2fKdbeDT8
Title: Re: When is it a line-array?
Post by: mark lonow on July 28, 2014, 08:43:11 PM
thanx mac
if u cant do 6 feet of line array(for what ever reason)
u probably  cant do 6 feet of point source ither
or 6 feet   trap array
or pay for cardioid low mid

or a extra low box+amp 75-150hz or what ever

performance cost what it costs








Title: Re: When is it a line-array?
Post by: Lee Buckalew on July 28, 2014, 11:28:22 PM

Martin and EAW are already doing it to great effect with MLA and Anya respectively. And their advantage is that you're not trying to do all the work, the trial and error of manually changing delay times and seeing what effect it has, change again and test, change again and test......  With their software you just tell it the coverage you want, where you want it to start and end and where/if you want 'holes' in the coverage to miss balconies and things, and then the software does all the hard work to calculate what processing all of the drivers need to get it done.


k

These are two very different systems.  Anya is beam steering but, as far as I am aware it uses no FIR filters.  MLA on the other hand is all FIR filters and does not beam steer at all.  It creates a phase and frequency coherent summation at the audience plane and it creates cancelation in the non audience and the hard avoid areas.  It does not steer sound away from those areas but merely reduces the level.

Lee
Title: Re: When is it a line-array?
Post by: Scott Holtzman on July 28, 2014, 11:52:13 PM
http://firmaker.afmg.eu/

http://www.youtube.com/watch?v=PA2fKdbeDT8

Thanks Peter great stuff.  Need to spend more time on Firmaker site.

Title: Re: When is it a line-array?
Post by: Peter Morris on July 29, 2014, 02:06:15 AM
These are two very different systems.  Anya is beam steering but, as far as I am aware it uses no FIR filters.  MLA on the other hand is all FIR filters and does not beam steer at all.  It creates a phase and frequency coherent summation at the audience plane and it creates cancelation in the non audience and the hard avoid areas.   It does not steer sound away from those areas but merely reduces the level.

Lee

Exactly - while the MLA is not a beam steered array, and you could even argue its not a line array,  it will however steer, just a little  ;).

I would be surprised if Anya’s processing didn’t use some FIR filters.  While EAW does not mention FIR filters they do say it’s "focused", and AFAIK that requires some FIR magic.


http://www.youtube.com/watch?v=th1VqW2dD4w
http://www.youtube.com/watch?v=e8OgtU29M8k
http://www.youtube.com/watch?v=GTEK3v2oMDk
Title: Re: When is it a line-array?
Post by: Kevin McDonough on July 29, 2014, 04:19:06 AM
Exactly - while the MLA is not a beam steered array, and you could even argue its not a line array,  it will however steer, just a little  ;).

I would be surprised if Anya’s processing didn’t use some FIR filters.  While EAW does not mention FIR filters they do say it’s "focused", and AFAIK that requires some FIR magic.


http://www.youtube.com/watch?v=th1VqW2dD4w
http://www.youtube.com/watch?v=e8OgtU29M8k
http://www.youtube.com/watch?v=GTEK3v2oMDk

Hey

Yeah I've read all Martin's papers etc on MLA, and am aware of the technical jargon they use. I'd still argue that it was beam steering though, even if they use a slightly different method to get there.

There are many very successful reports of them having hard avoid areas at say 50 or 75 metres from the stage, then suddenly using the processing to extend it out to 100m. Or suddenly adding a notch in the coverage to miss an area, changing the start of the coverage to add in or miss out a pit area.

Sounds like beam steering to me...  ;)

Lol, no seriously, they way their software works (which I'm sure you know) is by repeated testing of processor settings. What the human would have to do only the computer can do it far faster. It uses a very accurate model of the cab's dispersion both as a single cab and how it behaves in lines, and plots thousands of different graphs of coverage. By trial and error it works out which one is closest to what you have told it is your ideal coverage pattern for this job.

How it achieves this coverage would be open to debate. While yes Martin does say phase coherence at the audience etc, it's pretty hard to get the kind of changes in coverage the system is capable of just by creating a phase relationship at a single point in space. Mostly because cancellation is very frequency and wavelength dependent and that single point you were cancelling at would only hold true for a fairly narrow band of frequencies, the other frequencies with different wavelengths would still be full volume at that point and so it wouldn't be very 'cancelled'.

I feel as if there would have to be a significant amount of what you would think of as traditional beam steering in there too, deliberately creating 'lobes' in the coverage for given areas to cancel the sound out.

But technical details aside, as I said both cabinets use arrays of multiple, independently processed and amplified drivers, to be able to achieve changes in coverage using software and processing alone. Even if there are some technical differences they're both still beam steering to me.

K
Title: Re: When is it a line-array?
Post by: Lee Buckalew on July 29, 2014, 06:27:17 AM
Exactly - while the MLA is not a beam steered array, and you could even argue its not a line array,  it will however steer, just a little  ;).

I would be surprised if Anya’s processing didn’t use some FIR filters.  While EAW does not mention FIR filters they do say it’s "focused", and AFAIK that requires some FIR magic.


http://www.youtube.com/watch?v=th1VqW2dD4w
http://www.youtube.com/watch?v=e8OgtU29M8k
http://www.youtube.com/watch?v=GTEK3v2oMDk

It does not steer at all.  It covers the entire area that is available based upon the array inter cabinet angles.  From there the top and bottom apertures, as they call it, are affected by either being available to be heard or by being canceled.  Same goes for exclusion zones.  This is all budgeted into the processing ahead of time so that there is enough DSP processing available to allow for the required cancellation or addition.  There is no steering occurring.  When steering is done every time you change where a beam is pointed you change the frequency response adjacent to that beam.  From a physics standpoint what MLA does is very different from beam steering.

Lee
Title: Re: When is it a line-array?
Post by: Lee Buckalew on July 29, 2014, 07:19:43 AM
Hey

There are many very successful reports of them having hard avoid areas at say 50 or 75 metres from the stage, then suddenly using the processing to extend it out to 100m. Or suddenly adding a notch in the coverage to miss an area, changing the start of the coverage to add in or miss out a pit area.

Sounds like beam steering to me...  ;)
But it's not.  Since each driver can only produce a given amount of energy MLA pre calculates how that coverage will happen for the entire potential coverage of the array based upon the inter cabinet angles and it calculates these inter cabinet angles based upon what your anticipated start and stop coverage needs will be and where you have indicated you will need audience vs. non-audience vs. exclusion areas.  This coverage already includes the areas that you indicate as "suddenly using processing to extend...".  The change in processing is not steering, it is making the area that is already being covered able to be heard.  No change in energy inside the already existing coverage area occurs, they just make the energy that is already covering the "new" area able to be heard.  They do not redirect the energy.
In a beam steered example, since there is only a given amount of energy available from each driver, when you process the beams differently you redirect where that energy is focused.  This takes energy away from areas that were being covered creating a change in the frequency response, SPL, and phase interaction both inside and outside of the "new" coverage area.


Lol, no seriously, they way their software works (which I'm sure you know) is by repeated testing of processor settings. What the human would have to do only the computer can do it far faster. It uses a very accurate model of the cab's dispersion both as a single cab and how it behaves in lines, and plots thousands of different graphs of coverage. By trial and error it works out which one is closest to what you have told it is your ideal coverage pattern for this job.

It is hundreds of thousands of calculations.  Something along the lines of 180,000 per virtual measurement position.  In 1/32 octave resolution for frequency response.  Compare that to EASE resolution.
It does not work out the one solution as there is not a single available solution.  Each time you run a new calculation a different solution may be obtained to create the result that you want because there is a range of combinations that can create the solution to the problems presented.
The important point is that the entire potential coverage area, including the possible "aperture" areas at the top and bottom of coverage and any exclusion zones, have been calculated into the processing ends and are being covered with acoustic energy all of the time.  Sometimes that energy is audible such as when you open up a top or bottom aperture or choose not to exclude a balcony face and sometimes it is inaudible such as when you don't extend an aperture or you exclude the balcony face.  The energy coverage is not changing, only canceling or not in given areas. 


How it achieves this coverage would be open to debate. While yes Martin does say phase coherence at the audience etc, it's pretty hard to get the kind of changes in coverage the system is capable of just by creating a phase relationship at a single point in space. Mostly because cancellation is very frequency and wavelength dependent and that single point you were cancelling at would only hold true for a fairly narrow band of frequencies, the other frequencies with different wavelengths would still be full volume at that point and so it wouldn't be very 'cancelled'.

MLA creates cancellation on purpose across the entire band and it does not change the focus of the energy outside of the exclusion area to achieve this.  Beam steering refocuses energy in order to create a shift in the coverage pattern.  It actively takes energy away from certain areas in order to focus it elsewhere.  It does not create a continuous and even coverage of energy and only change the ability to hear or not hear certain areas of that energy it instead actively changes the focus of the energy.  This in turn changes the sound throughout the coverage area as it changes all of the acoustical interactions.


I feel as if there would have to be a significant amount of what you would think of as traditional beam steering in there too, deliberately creating 'lobes' in the coverage for given areas to cancel the sound out.

But technical details aside, as I said both cabinets use arrays of multiple, independently processed and amplified drivers, to be able to achieve changes in coverage using software and processing alone. Even if there are some technical differences they're both still beam steering to me.

K

When MLA first came out I looked at it simply as another version of beam steering just with much higher resolution.  As I studied further what was going on I realized that it was actually a wholly different process occurring as there are significant differences in the physics applied. 

Both jet engines and propellers are used on aircraft.  Aircraft with either can fly.  If we applied the same physics to both we would not be able to understand how each works so differently from the other.  We don't just say that both help to make the plane fly so they are both the same.  They are significantly different from each other.

Lee
Title: Re: When is it a line-array?
Post by: Peter Morris on July 29, 2014, 08:32:55 AM
It does not steer at all.  It covers the entire area that is available based upon the array inter cabinet angles.  From there the top and bottom apertures, as they call it, are affected by either being available to be heard or by being canceled.  Same goes for exclusion zones.  This is all budgeted into the processing ahead of time so that there is enough DSP processing available to allow for the required cancellation or addition.  There is no steering occurring.  When steering is done every time you change where a beam is pointed you change the frequency response adjacent to that beam.  From a physics standpoint what MLA does is very different from beam steering.

Lee

Hi Lee,

I may have used the term "steer" too loosely and we maybe talking about the same thing - as you know, for a given coverage area you calculate the inter box angles and FIR coefficients for each of drivers and up loaded them to the boxes.... about 15 minutes of number crunching on an i7.   Without rehanging the array you can change the coverage area slightly by recalculating new FIR coefficients and up loading them again...

Title: Re: When is it a line-array?
Post by: Peter Morris on July 29, 2014, 08:33:40 AM
Kevin,

I agree with Lee, its not a beam steered array, its exactly as Lee has described above.

It was not done by trial and error but complex mathematics. It appears to use a fuzzy logic type of optimization programme to manipulate the complex vectors that describe the phase, amplitude and frequency response of each speaker; maximizing their summation on the listening plan by determining the required EQ, delay, amplitude of each speaker within the array, taking into account the different angles and height of each speaker. 

When I saw what Ambrose had done I was in awe!

Peter           
Title: Re: When is it a line-array?
Post by: Lee Buckalew on July 29, 2014, 11:42:56 AM
Hi Lee,

I may have used the term "steer" too loosely and we maybe talking about the same thing - as you know, for a given coverage area you calculate the inter box angles and FIR coefficients for each of drivers and up loaded them to the boxes.... about 15 minutes of number crunching on an i7.   Without rehanging the array you can change the coverage area slightly by recalculating new FIR coefficients and up loading them again...

Peter,
I am not trying to be picky but precise.
To me, understanding these types of differences goes a long toward helping me to understand the complex interactions that are going on.

I too am in awe of what Ambrose understands in his brain and then translates into filter sets, algorithms, and computer interfaces.

Lee
Title: Re: When is it a line-array?
Post by: Brandon Wright on July 29, 2014, 12:14:54 PM
It does not steer at all.  It covers the entire area that is available based upon the array inter cabinet angles.  From there the top and bottom apertures, as they call it, are affected by either being available to be heard or by being canceled.  Same goes for exclusion zones.  This is all budgeted into the processing ahead of time so that there is enough DSP processing available to allow for the required cancellation or addition.  There is no steering occurring.  When steering is done every time you change where a beam is pointed you change the frequency response adjacent to that beam.  From a physics standpoint what MLA does is very different from beam steering.

Lee

And how exactly do you think Anya achieves its "beam steering? ???
Title: Re: When is it a line-array?
Post by: Lee Buckalew on July 29, 2014, 10:36:41 PM
And how exactly do you think Anya achieves its "beam steering? ???

Precisely that, through beam steering.  It is a flat hung array that only steers.  It lacks phase coherence and demonstrates this because it does not hold up well over distance or in wind.  It is very flexible in its ability to adjust to varying coverage needs from a single hang, more flexible than similar Renkus systems.  Anya is a very useful tool but Anya and MLA are significantly different approaches to the physics of audio reproduction.

Forgive me if I am seeming short.  This week is very long as I am working with the marvelous Brad Ricks and George Georgallis from JBL tuning a large stadium install.  Long tired days seem to make my answers more clipped than I intend.

Lee
Title: Re: When is it a line-array?
Post by: Peter Morris on July 30, 2014, 07:48:43 PM
Peter,
I am not trying to be picky but precise.
To me, understanding these types of differences goes a long toward helping me to understand the complex interactions that are going on.

I too am in awe of what Ambrose understands in his brain and then translates into filter sets, algorithms, and computer interfaces.

Lee
.... its actually a pleasure to talk to some who has an understanding  of what these guys have done  :)
Title: Re: When is it a line-array?
Post by: Tom Danley on August 01, 2014, 01:59:45 PM

I'm wondering how many "elements" you would need before the system actually begins to perform as a line-array.

Hi All
I had intended to reply sooner but we have had a death in the family which has occupied much of my time.                 
Acoustically speaking, a line source is “a single source” which is acoustically “very large” in one dimension and small in the other.   To be an acoustic line source (reaching the theoretical -3 dB per doubling of distance) it is assumed to be straight and infinitely long but like optics, “infinity long” isn’t from here to the edge of the universe and back etc  but rather, infinity here (where the behavior essentially stops changing as a function of length) is normally about 40 or 50 wavelengths or longer.   
Using the word “line array” is a way around the acoustic requirement for that acoustical condition of a continuous source.     Here one instead has an array made of individual radiations where each source radiates with it’s individual directivity and as a result creates some overall directivity via Huygens wave front construction but as Ivan mentions the radiation going up and down and elsewhere while present is ignored. 
This interference (not coherent addition) also means it takes many more sources to reach a given SPL at a distance as much of the energy is lost in the self canceling process and also much energy projected outside the desired direction.

The problem of changing behavior vs frequency when the line length is less than infinitely long is reduced by bending the array which acoustically makes it an astigmatic point source instead and so if one measures one of these directly on axis, the spl essentially rarely falls off at the vaunted “half as fast” as a point source.   While we think in terms of frequency response, time enters into the picture musically as not everything is a continuous sine wave and a single impulse fed to such a system, produces a train of arrivals reflecting the different path length to each source from the listener position. This condition can be corrected with DSP but only in one location by applying appropriate delays etc.

 As Ivan mentioned, the array of individual sources also produce “finger lobes” as the configuration of individual radiations still produces an interference pattern, exhibited by a pattern of lobes and nulls above and below the aimed axis and comb filtering.  When one makes a high resolution measurement or plays pink noise and move around you can see and hear all that..     
It is the self interference or interference pattern and changing directivity vs frequency that cause the line arrays to be strongly affected by a crosswind and to have a very limited useable working distance, which humorously is the exact opposite of the marketing which asserts they have “longer throw” than a point source.  Many of the stadium systems using real point sources “throw” and sound fine at 700 or 800 feet or more.

The “point source” term itself is another name that is normally used incorrectly in loudspeaker marketing and sales, a point source (acoustically speaking) radiates from a single point in space (and ideally time if you’re a fan of Heysers work) and so even a simple 15 and a horn is no longer a real or acoustic point source but an array of two point interfering sources.     
One can add individual sources and have them add coherently but the sources must be less than about ¼ wavelength apart at the highest frequency of interest (as one finds with multiple subwoofers).
The up side of the tiny line array is that they are mostly not a line array acoustically, they have less self interference and so one can achieve a more constant sound spectrum vs location more like a point source.   
Many have observed that the smaller the number of boxes one uses, generally the better it sounds, one can have a “perfect box” but when you stack 16 of them up, what do you have then?   
 
At the far opposite end of this effect, a real acoustic point source will sound essentially the same at any distance (minus some hf loss due to air absorption, only the spl falls with distance) and lacking an interference pattern, also sounds essentially the same even in a strong cross wind or walking the pattern with pink noise.     
If one also has constant directivity and a high front to back sound ratio (easy to do with a single large horn),  if one can place the source at the right height, then one can have a near constant SPL and spectrum from right under the source to the farthest point AND radiate much less energy in the wrong directions than an array which is great when you have a room or stadium to contend with.  The down sides for MFR’s is this is very hard to achieve acoustically and it requires far less cabinets, drivers, amplifiers and dsp for a given audience SPL.

In some of the stadiums where real constant directivity point sources are used in, one finds the SPL varies only + - 3 or 4 dB from under the system out to 700 or 800 feet and subjectively it sounds the same in every seat and produces only one arrival in time which is a big plus musically speaking.
One can steer either type of an array as well, before the line source craze and before I found a way to have drivers combine coherently over a wide bandwidth, I had a patent on the approach which is  popular now for making a steerable array’s.  Examine the Fig’s for this old expired patent.

https://www.google.com/patents/US4845759?dq=tom+danley&hl=en&sa=X&ei=nsvbU5X8JJGqyASyy4LIAg&sqi=2&pjf=1&ved=0CDgQ6AEwBA

The down side here is the more one angles the radiation relative to the source locations, the greater the sound energy projected outside of the desired pattern becomes and it remains difficult to make all the sources combine coherently and not radiate as individual sources..
Best,
Tom Danley
Danley Sound Labs
Title: Re: When is it a line-array?
Post by: Tommy Peel on August 01, 2014, 02:23:10 PM


Hi All
...
Best,
Tom Danley
Danley Sound Labs

Great information as always. That's one of the easiest to understand descriptions of line arrays vs point source I've read.

My condolences for the loss in your family.

Have a great weekend,
Tommy



Sent from my Moto X (XT1053) using Tapatalk

Title: Re: When is it a line-array?
Post by: Ivan Beaver on August 04, 2014, 09:09:42 PM
Here is an interesting article that puts a lot of things in perspective

http://education.lenardaudio.com/en/07_horns_3.html