Line arrays and inter-element angles.

[Phil Graham]

Re: Line arrays and inter-element angles.

Helge, et al.,

I could write a huge tome on this topic, which is really our industry fumbling about in the world of antenna theory. Here are some brief points that may give you some things to think about:

1. The lobing and/or narrowing behavior along, above, and below a curved, finite arc is a function of the amplitude and phase at each point along the arc. These effects are purely physics, and exist in the presence of a perfect arced source, before we introduce real drivers/waveguides.
2. The far-field response of a finite, perfect arced source is not flat along the listening plane we would normally define for our audience area.
3. The behavior of a box in an array is a function of its radiation environment. The vertical directivity and overall axial response of a box in the middle of an array is different than a box at the top or the bottom.
4. The behavior of the off axis lobing is influenced by the curvature of an arc. Curvature changes the arc shape, total group delay of each element at a given listening point, and radiation environment around a given box.
5. Once wavelengths get long enough, the effect of array curvature does very little, as the delays induced are small versus wavelength.
6. The 3dB per distance thing that gets tossed around is completely, fundamentally incorrect for how we use vertical sources. It is merely a mathematical curiousity that falls out of playing with an infinite array, and thus the observer completely within the array length.

Practical implications for system techs using vertical arrays:

1. Generally, increasing inter-box curvature helps distribute lobes and reduce the intensity of any individual one.
2. Effects of arc sources typically require box aim points to be higher up the back wall than on would choose based on a waveguide's nominal coverage angle.
3. The overall coverage of the array is essentially never simply the included angle of the array curvature. Furthermore, it is frequency dependent as function of array length.
4. While a deep and challenging topic, shading is not your enemy. Certain, well designed shading methodologies can greatly reduce side lobe effects. See, for example, Don Keele's CBT array.
5. The "virtual source location" of the various frequency ranges in the array tends to move from the top of the array towards the middle as the frequency gets lower. A lot of advanced system tech techniques are about managing this effect, and bringing these in line.
6. The primary down lobe from the straightest section at the top of the array can land inside the audience coverage zone, depending on geometry.
7. The dominant unwanted down lobes of a typical line array usually falls between 1 and 3kHz.
8. Even if the bottom boxes of the array are providing reasonably uniform high frequency coverage up near the stage, downlobes from the array can render very uneven response in the upper midrange near the stage.

There's much more on this topic, but this is a workable start.

An important PS: Curved, finite arrays of all types, regardless of brand/orientation/configuration exhibit behaviors dictated by physics. Horizontal arrays of point source boxes show their same incarnations of the effects above. The real question is whether it makes sense to segment an audience vertically, or to segment it horizontally, and that is room by room dependent.

 

[Max Warasila]

Re: Line arrays and inter-element angles.

Kinda/Sorta off-topic, but in terms of point sources being pointed at the last audience member: is it worth aiming the coverage down a bit in a auditorium like room (think floor slanted at about 3-6 degrees from front to back) where there is a back wall that is extremely reflective about, lets say, 5 feet from the last listener? The actual room I'm thinking of has the speakers aimed about 8 feet over even the farthest back listener, and because of this, actually tends to hit the ceiling almost more than it does the floor, as well as generate crazy reflections off the back wall.

Not that I can change it; just want to know if there is some hidden secret explaining that aiming.

P.S. Great thread to read. Thanks for all of those who posted so far. Solid information and experience going around.

 

[Jason Lavoie]

Re: Line arrays and inter-element angles.

Kinda/Sorta off-topic, but in terms of point sources being pointed at the last audience member: is it worth aiming the coverage down a bit in a auditorium like room (think floor slanted at about 3-6 degrees from front to back) where there is a back wall that is extremely reflective about, lets say, 5 feet from the last listener? The actual room I'm thinking of has the speakers aimed about 8 feet over even the farthest back listener, and because of this, actually tends to hit the ceiling almost more than it does the floor, as well as generate crazy reflections off the back wall.

Not that I can change it; just want to know if there is some hidden secret explaining that aiming.

P.S. Great thread to read. Thanks for all of those who posted so far. Solid information and experience going around.

Because as someone mentioned previously, if you aim the speakers really high the floor coverage (ignoring reflections from the ceiling and walls) starts to look really good on paper.
This is why modeling is useful, but it takes a brain to read between the lines and know what is not being shown.

Jason

 

[Ivan Beaver]

Re: Line arrays and inter-element angles.

Because as someone mentioned previously, if you aim the speakers really high the floor coverage (ignoring reflections from the ceiling and walls) starts to look really good on paper.
This is why modeling is useful, but it takes a brain to read between the lines and know what is not being shown.

Jason

+1

Just because a models says it "works", does not mean that it does. You HAVE you use some common sense about "the other things" that will be happening (reflections etc) and how practical that is in the real world.

When doing any model-you should always ask yourself "Does it make sense"? If it doesn't, then you need to look a bit closer.

 

[Ivan Beaver]

Re: Line arrays and inter-element angles.

Kinda/Sorta off-topic, but in terms of point sources being pointed at the last audience member: is it worth aiming the coverage down a bit in a auditorium like room (think floor slanted at about 3-6 degrees from front to back) where there is a back wall that is extremely reflective about, lets say, 5 feet from the last listener? The actual room I'm thinking of has the speakers aimed about 8 feet over even the farthest back listener, and because of this, actually tends to hit the ceiling almost more than it does the floor, as well as generate crazy reflections off the back wall.

Not that I can change it; just want to know if there is some hidden secret explaining that aiming.

P.S. Great thread to read. Thanks for all of those who posted so far. Solid information and experience going around.

That is simply "part of the game" and tradeoffs you choose between. How far can you "push" one thing without affecting something else to much?

How much is to much? Hard to say. Often it depends on a lot of other factors that exist in reality-but do not show up in the model.

In EVERY situation you HAVE to use your own experience and knowledge to decide which way to go.

Here is a good example. How do get get nice even coverage in just about any room (in a model)? Simple-make every surface hard and reflective and point the loudspeaker away from the audience and into a upper corner. Turn off all interference etc. You will have a very even coverage on the audience area.

But does it make sense? NO-but the "model says" it does.

 

[Max Warasila]

Re: Line arrays and inter-element angles.

That is simply "part of the game" and tradeoffs you choose between. How far can you "push" one thing without affecting something else to much?

How much is to much? Hard to say. Often it depends on a lot of other factors that exist in reality-but do not show up in the model.

In EVERY situation you HAVE to use your own experience and knowledge to decide which way to go.

Here is a good example. How do get get nice even coverage in just about any room (in a model)? Simple-make every surface hard and reflective and point the loudspeaker away from the audience and into a upper corner. Turn off all interference etc. You will have a very even coverage on the audience area.

But does it make sense? NO-but the "model says" it does.

Thanks to both of you; I thought it was something like that.

But Ivan, that example… my ears ache just imagining the effects. Talk about cruel and unusual. I'll have to remember that next time I talk to our system installer.

 

[Bennett Prescott]

Re: Line arrays and inter-element angles.

A topic near and dear to my heart.

Personally, I prefer a smooth coverage transition, building an array more like a spiral or exponential array as you personally prefer. I like to think in terms of using half the array for the back half of the audience, the next 1/4 for the nearer 1/4 of the audience, etc. This tends to line up relatively well with doubling boxes for doubling throw. I have never, ever, found any application where using the zero degree angle was useful. Even for very long throw audio (800'+) I found it causes more lobing problems than it helped get energy out there, by the time you're at 300' or so you have to give up on a lot of HF anyway and the environment becomes a serious issue. At that kind of distance 1° is plenty tight. All line arrays enclosures are designed with a certain wavefront curvature, taking it into account will produce the smoothest results.

Another issue is that neither line arrays nor humans are point sources, but it is easy to read the modeling software like we are. Just because your array can produce a really sharp cutoff and put a beam of energy right where you want it doesn't mean it will be experienced well by the time it gets there. If the difference between a 6' person and a 5'4" person can mess up your model it's time to reconsider your goals. High hung arrays are really great for the far seats, but the front section of the audience starts to be really off axis. Just like it's hard to get a sub array to cover a very wide room, it's hard to get a line array to have consistent coverage down close. Using more angle on your array IMHO reduces this problem because you don't get audience essentially directly underneath a 20' long broadside array. I like to lean on fills as well, but not flown ones. I generally try and cover 80% of the audience with the arrays, and get the front 20% with some other method.

YMMV, take what I say with a grain of salt since it's been at least a year since I've flown anything at this point.

 

[Jay Barracato]

Re: Line arrays and inter-element angles.

A topic near and dear to my heart.

it's been at least a year since I've flown anything at this point.

It has sounded to me like you have been doing a lot of flying recently...

 

[Helge A. Bentsen]

Re: Line arrays and inter-element angles.

I did a small aircraft hangar this weekend, 180 degrees of coverage with JBL 4886. I designed that array using some ideas from this thread, I went for even sonic signature at various distances and not even SPL everywhere. Gradual curvature on the arrays starting at 2, ending at 8 degrees in the main hangs. Sounded very smooth, it look like designing for even sound, not even SPL is easier and translates better across the room.

Agree on using fills. Now, if I could stop people from moving them because they are aimed "wrong" (not at the mix position) or tell me that they are off(I run fills as an extension of the main system, I really like them if you barely notice them until you turn them off).

Thanks for the input

 

[Phil Graham]

Re: Line arrays and inter-element angles.

...I went for even sonic signature at various distances and not even SPL everywhere...

This is the most reasonable goal, in my opinion. It is important to remember that the most even coverage volume front to back is a consequence of the highest level of intra-array cancellation in the nearfield. A finite line source, when far enough in the far field, behaves as a point source. Near the array, the evenness in amplitude is due to the increasingly incoherent summation of the array elements. This is a useful effect, but it has limitations on how far you can rely on it.

Also, if you work too hard in getting the maximum horsepower at the back of the space, you'll find that the down lobes are causing serious issues at points within the primary audience area.

Ideally you can configure the array to scuttle the down lobes in a region where the audience is not present, Not usually possible, but an interesting target goal to aim for.

 

[Jay Barracato]

This is the most reasonable goal, in my opinion. It is important to remember that the most even coverage volume front to back is a consequence of the highest level of intra-array cancellation in the nearfield. A finite line source, when far enough in the far field, behaves as a point source. Near the array, the evenness in amplitude is due to the increasingly incoherent summation of the array elements. This is a useful effect, but it has limitations on how far you can rely on it.

Also, if you work too hard in getting the maximum horsepower at the back of the space, you'll find that the down lobes are causing serious issues at points within the primary audience area.

Ideally you can configure the array to scuttle the down lobes in a region where the audience is not present, Not usually possible, but an interesting target goal to aim for.

I agree with even tone over even volume. Either patrons are free to move to the volume they prefer or they are not and soon adjust to what is at their position.

Not to throw out a monkey holding a wrench into the gears but I wonder about the effect of low ceilings especially over the stage on a systems lobes.

Sent from my DROID RAZR HD

 

[Brandon G Romanowski]

Re: Line arrays and inter-element angles.

I don't think I know what even tone over even volume means ...

 

[Brad Harris]

Re: Line arrays and inter-element angles.

We have several line array 'styled' cabinets in our inventory (some pushing 20+ years old in design to the latest crop from various manufactures).

A starting point I have with most of them (the oldest is tight or not rigging ..... talk about having 2 completely different hangs in 1!) is to do the old VDosc way ... have each box sight line hit every linear X distance on the floor. i.e. each box has 30' between where the next box hits, and then massaging the array from there.

The newer prediction programs have some nice polars that are a very nice guide on how to 'shape' the arrays sound output in SPL and Tonality, and especially with their predictive microphones, you can see what MAY happen at a certain point in a plane.

I try for equal tonality for most shows (large corporate) and sometimes, well you have to cater for the 'Code 18' de jour.....

With many of the arrays out there, the more boxes (and chunks of signal processing ... now thats a whole 'nother debate) you can cram into before Budget, Producers, Video and truss boarders get in your way, the better, just for control of the array, and not necessarily SPL.



BRad
* Code 18: The problem is 18" behind the console ......

 

[Helge A. Bentsen]

Re: Line arrays and inter-element angles.

Not to throw out a monkey holding a wrench into the gears but I wonder about the effect of low ceilings especially over the stage on a systems lobes.

Sent from my DROID RAZR HD

Could be. I´ve noticed that sometimes changing the fly height changes the tonal characteristics more than your computer model suggest. Especially in the start of the mid-field coverage area. If I fire up the array and hear some strange things in rooms with low ceilings, I try to change the height on one of the sides and compare them before doing anything else. This can change your array from sounding weird to usable in some rooms. Physics suggest that where you place your speakers can have a great effect on their sound.

There´s probably people here with a lot more knowledge about this who could either confirm or rubbish this theory

 

[Bennett Prescott]

Re: Line arrays and inter-element angles.

I don't think I know what even tone over even volume means ...

Brandon,

Some tools aim for most even SPL over the audience area. That is to say, if 400Hz is loudest at the front and 8kHz loudest at the back that is fine with them as long as the average to create the most even SPL. Personally I don't care how even the SPL is if every audience member is getting a different experience. I am always willing to give up overall energy in order to get the most even frequency response (tone) at every seat.

 

[Max Warasila]

Re: Line arrays and inter-element angles.

I agree with even tone over even volume. Either patrons are free to move to the volume they prefer or they are not and soon adjust to what is at their position.

Not to mention that it seems like people will complain about a guitarist's "tone" more than whether or not they are hearing something at 100 dB compared to a person 20 feet away from them hearing the same thing at 105 dB.

 

[Phil Graham]

Re: Line arrays and inter-element angles.

I don't think I know what even tone over even volume means ...

Brandon,

In addition to Bennett's excellent reply, the crux of the issue is this:

At high frequencies you can make a much more narrow array beam, and or steep array level gradient than at low and mid frequencies. The shorter the array, the more severe this difference in behavior is. The directivity balloon at lower frequencies, even with long lines, is "squishy" and broader than the high frequency array directivity. Because of this, if you want to tailor the array for consistent magnitude response throughout the room, you're going to have to accept a gradient in level from front to back. This gradient will be dictated by the relative level variations at each point along the listening plane as dictated by the overall directivity (or lack thereof) of the array at the bottom end of its coverage frequencies.

 

[Brandon G Romanowski]

Re: Line arrays and inter-element angles.

If we had to accept a level gradient down to our low frequency we would never do better than 6dB per doubling of distance. My aim is typically to keep a fairly flat response at a fairly consistent spl over a fairly broad range of frequencies. When the word tone comes into play I usually consider low lows and high highs......

 

[Tim McCulloch]

Re: Line arrays and inter-element angles.

If we had to accept a level gradient down to our low frequency we would never do better than 6dB per doubling of distance. My aim is typically to keep a fairly flat response at a fairly consistent spl over a fairly broad range of frequencies. When the word tone comes into play I usually consider low lows and high highs......

Yeah, I want the best of both, too. If the budget and trim height are there, I'll have a longer line.

For me, I notice the lack of 80-200Hz from the main hangs toward the back of the room if the line is too short. In these cases I'll re-pin the array to get more even magnitude response at the expense of total SPL at the back, and amplitude-shade the HF down front.

 

[Steve Ferreira]

Re: Line arrays and inter-element angles.

Great topic guys. Getting some great info in here.

I'm currently plotting a room with 4 prediction softwares from different manufacturers as we are looking to buy a new line array. They all try to get to the same result, but they all have a different path of getting there. As Tim described earlier, LAC II told me to go from a box with a splay of 4 with the next 3 boxes at 10. This just didn't seem right to me.

I typically aim for 1K as my frequency of choice for my plots and try to get that even in the room from front to back. 100% of our work is corporate with video rolls, but we do also get bands for gala dinners.