FSP-LA

DUe to Heathrow's naming of the other topic, I've decided top make a little thread about a possible FSP-Line array cab.

Have fun talking about flying hardware and planar waves![img]smileys/smiley1.gif[/img]

Here's a few transferred posts:

Tony ASS:

I assume that because of what I read in the first thread that this cab would have to be flown.

Just to point out that the companies that do line array systems can
spend more time on the metal work than the cabs. Array cabs have to
work as a system and not necessarily as an individual cab. The ability
to come up with the flying rig is the thing that would make this a
reality.

jsg:
<div>I've seen some examples of line array cabs being used in very
short lines to good effect. My local club in Cambridge has 3 of the
small DB cabs per side, in a short vertical line. And Fabric in London
has 2 of the Martin large-format line array cabs per side and it sounds
unbelievably good.</div>
<div></div>
<div>This is an example of practical experimentation winning over the theory, so I guess youmight approve of that.</div>
<div></div>
<div>Anyway,the thinking is if you can get good results with as few as
2 line array cabs per side, then they don't need to be flown - just
stacked on top of the bass with props to get them at the right angle.
So something like this might be practical for FSP, and yet comparable
insome waysto one of the world's leading clubs.

Tony again:
Deadbeat, The MX system was calculated with the help of Mick Nash, who
was MAN Flying Systems. He told me how the balance should be in free
fall so that it would behave itself when flown and kelped back to the
shape reqiured. The transformer system that it flew on, was the most
elaborate system ever designed. All that changed when line array hit
us. He never got involved in that, and neither did I.
<div>All the flying components were engineered superbly, Mick and I
were toolmakers together in a former life, so the ideas used to roll.</div>
</div>Wrighty:

From my limited knowledge of line array theory:

At the end
of the day, you'll get noise out of a short line array and in the size
of venue it's likely to be used, it'll probably sound fine. The drivers
will all couple well, because they're configured as designed. However,
you won't get proper pattern control at lower frequencies, since the
line has to be a given length to achieve this. I can't find the figure
at the moment - it's either 1 wavelength or 1/2 a wavelength. If I
find/remember it, I'll edit this post!

Me:

Do we need a FSP-LINE ARRAY?

Sounds like an interesting project. I wouldn't build one personally, but still a great exercise.

Norty: (@ Wrighty)
ts not pattern control, its the ability to function as a line
rather than a point source (e.g. -3db per doubling rather than -6db per
doubling)
<div></div>
<div>Was looking at the D&B J system at the weekend and was with
the system tech who was showing me the software. They have
compensation for different length hangs at the frequency where it
becomes line source.

AJW:
The compression driver really is the clue to designing a correctly
operating LA. Its no use just bunging in any old waveguide just "to
do". It needs to play pretty low and also be capable of very high
SPL's. Obviously dispersion (usually around 120 x 10 degrees) is also
of paramount importance.

IMO
one of the reasons that they can sound so good is purely down to the
fact that the upper bass/mid drivers are very often pretty small ( 5" -
10") and also of excellent quality. Putting in any design of band pass
horn that also does lower mid would probably seriously compromise
overall quality.

The hardware design is actually pretty simple
and with the availability of inexpensive Laser profiles can be made to
look very professional.

Tony
</div>

I have been thinking about building one for a few months now. It would be a fairly simple 2 x 8" or 10" units per box with suitable Hi-end driver plus waveguide.
As I would only ever envisage using 4 units per side I probably would not achieve proper LA lobing down to the sub crossover point but I think it would still be a good exercise.

At the moment I am stuck between trying to design and build my own waveguide (not too difficult) or buying a suitable off the shelf unit, probably BMS as they are the devil I know.

The flying hardware design and testing would present no problems. By incorporating the hardware into "Genie" lifts the whole site assembly phase could be turned into a simple 1 man job.
I envisage twin 18" subs sitting on the bottom frame of the Genie to aid stability.

More soon.

Tony

Hi Tony,

Would one cross the two small drivers differently (like a 2.5 way design)?

Deadbeat wrote: Hi Tony,

Would one cross the two small drivers differently (like a 2.5 way design)?

No I would keep it to the tried and tested design of having the waveguide in the centre of the Bass-mid drivers and running them in parallel.

By doing it this way and crossing to the Hi unit at a pretty low value i.e. no higher than 1k then the units will couple together almost perfectly.

Having a good top end that will run happily from this low figure to at least 16k and have a sufficiently high output is really what separates the men from the boys.

Tony

Good. I know a few major manufacturers do that in their LAs with a passive xover...can't remember which but it is used, I guess to make up for the weediness of their comps.

I think two 10" and 1 or two comps will be great.

BMS 4510 seems to be a good one.

For waveguides, I think DDS' DSLA series is interesting.

Edited by: Deadbeat

Why not try this for a waveguide:

www.vtcproaudio.com/pdf/VTC_flyer_Web.pdf

Seems like it should be easy to build just using a router and several layers of wood. Figuring out the dimensions to get the required behaviour would be harder, though.

I reakon it should be possible to use just a single compression driver per cabinet and still get the required wave shape.

Edi: typos
Edited by: jsg

I've been thinking about using line array techniques to produce a mid-top for music venues requiring a wide horizonal but narrow vertical dispersion.

Think something like a single dVDosc cabinet on a dVSub but all in one box.

I would really like to have a vertically arrayable cabinet so that you can just add extra boxes to get more SPL but without adding extra horizontal dispersion.

I forgot, I should have linked Steve Burton's webpage (lots of LA info there, very helpful).

Here's 'thoughts on Line Arrays', an overview of many concepts in line arrays, it also goes over many concepts used in the commercial cabs.
www.burton-manor.co.uk/Audio/LAthoughts.htm

The thing about line array is that everyone goes on about line sources and cylindrical wavefronts, and then everyone else rambles on about how it's not a line source if iits finite in length or curved or finite in width or has non-omni horiz coverage. They bang on and on and on and it's boring.




Conventional arrays are often described as a coherant curved 3D wave source (or as a point source, which is a special case of the above). They are also often described as a bunch of independant sources each covering its own area. You don't need fancy modelling SW to know that both of these are in fact utter garbage. Faced with reality, conventional array advocates start muttering something like "Its the dynamics man. Yeah, man. The dynamics. Yeah" before passing out.

The fact is that line arrays in practice really do sound better than conventional arrays in practice. Some have had weak midbass, but that's not hard to correct.

The *real* reason isvery simple: line arrays provide a coherant source because there are no gaps in the source and it has a smooth shape. That's it. That's *all* there is to it.

Executive summary: No gaps and smooth.

No gaps horizontally because of the single hang (gap between left and right is too big to matter, and the gap between dual low-mid/low is small enough forthem to act as a single source).No gaps vertially because every channel including the tweets is stretched and phase aligned to simulate a long thin source from the top ot the hang to the bottom with only the cabinet woodwork as an interruption. Smooth because the curvature and intensity are kept roughly constant along the array (smoothness is only important in the vertical dimension).

No gaps and smooth in the physical radiating plane means no gaps and smooth in a histogram of path lengths to your ear. That means the impulse response has no gaps and it is a smooth function.

Now the story splits into 2 parts but each is really, REALLY simple.

1. If you're out of the coverage area (too high or too low), this smooth impluse response become areasonably good low pass filter. For big line arrays, it filters out the whole passband. So up-spill rejection is MUCH better than a conventional array.

2. If you're in the coverage area, there's a point on the radiating surface that's pointing right at you. That's the same as saying you're on the perpendicular at that point. It's also the nearest point on convex or flat line arrays.

Since the rate of change of listener distance with respect to position on the radiating surface is a local zero at this point (think geometry) the density of the path length histogram peaks here (at the shortest distance). And the impulse response peaks too, at the lowest delay. A *single* peak in the impulse response, at the *front* of the impulse response is what gives us clean sound.

And that's *IT*. Absolutely NO NEED to sweat whether you have a true line source. No need to cry buckets over astigmatism and whether a cylindrical wavefront (or a sort-of-partially cylindrical wavefront) is a legitimate way to propogate sound, whatever we mean by "legitimate". Short line arrays, horizontal line arrays, highly curved line arraysare all good if you understand the *real reason* for the benefits. The rest of them are howling at the moon. Let 'em.

Two thread sort of flowing parallel, aren't they...ceharden's and this.

Here's some fresh Tom Danley for you guys, commenting on line arrays, specifically replying to this thread.

Hi,

Everyone is entitled to there own opinion, regardless of if it was
formed as a result of measurements or if its "just a feeling".

In this case, how line sources work is well known, the conditions
needed for sources to add coherently are well known also and the "self
interference" most line source systems have is obvious when you lay one
out horizontally.

The fact is to the degree a line source has a reduced SPL fall off vs
distance compared to a point source, it also stretches out the time
occupied by a single impulse of energy via the same geometry.



The advantage they have over the previous approach is that while the
old horizontal arrays had self interference in both Horizontal and
Vertical planes, the line arrays are mostly confined to interference in
the vertical plane which is much harder to hear (unless you lay one on
its side where it is easy to walk the pattern).

The other advantage they have (for mfr's) is that since the desired
effect of reduced SPL fall off is produced by "nearfield" cancellation,
it takes more drivers, more amplifiers, more processing to reach a
given SPL at a given distance that "IF" one had a real point source at
the same SPL.


Best,


Tom