Hi,
Date: Sun, 20 Mar 2011 19:56:40 +0100
From: J?rn Nettingsmeier <netti...@stackingdwarves.net>
On 03/20/2011 07:25 PM, Dave Hunt wrote:
The whole PA world has gone rather line-array mad since the idea was
re-introduced and re-engineered by L'Acoustique and Intellivox (Duran
Audio) some 10-12 years ago.
mostly for two reasons:
* they are cheaper than designs with delay towers
* with their better directivity, you can get reasonably good sound out
of mediocre or outright bad-sounding halls, where stacks would be very
hard to get right (since you can effectively reduce room excitation or
even steer around problematic walls altogether)
my subjective view is that you can also scale them higher than stacked
systems before they start to sound really funny.
and of course they are a lot easier and safer to rig, and to get
really
high up in the air, which is totally uneconomical for stacks.
Basically I tried to cover all that and more with the phrase "well
engineered, powerful, and good sounding", though that obviously
doesn't cover all your points.
I would hardly describe any of the line array systems as "cheap",
though rigging time is reduced, coverage is predictable, and the need
for secondary delay systems is reduced but not always eliminated.
"Better directivity" is really only true in the vertical sense. Most
PA boxes have had 90 degree horizontal dispersion (or narrower) for
some time. Of course this is only really so at mid to high frequencies.
The height does help with more even coverage, as you increase the
distance from a speaker system to the nearest listener. This also
makes it rather more comfortable for the nearest listener. It has
always been hard to make a system sound good for the rows very near
the stage.
As someone who used to do quite a lot of corporate work I did notice
that when VDOSC (L'Acoustic) came along, people were so impressed by
the sound (and the theory and engineering, including the flying) that
staging designers were suddenly almost forced to design around good
PA positions, with the result that sound engineers could finally get
what they had been asking for for years, but rarely achieve (though
there were always good designers who considered more than just the
visual).
Arraying is a bit of an art. A friend, who is far more technical than
I am, once said that if you keep on stacking up drivers you reach a
point where all the undesirable cancellations and reinforcements
become so numerous that they average out. Seems somehow believable.
I have noted that several manufacturers offer directional sub-
woofers, and even read the papers, but not come across them myself.
Big venues are almost bound to be lumpy at LF due to nodes and anti-
nodes.
I too, like J?rn, am curious to know at what point does ambisonics
break
due to the non-simultaneity of arrival of sound from widely spaced
sources. Obviously lower order components will suffer most from this,
why do you think that?
Because higher order components will come from fewer loudspeakers and
those will be in the right direction. Of course they will be even
later in their time of arrival for those listeners furthest from
these speakers in comparison with the lower order signals,
particularly W which is at the same level in all speakers.
With a widely spaced array, the "sweet spot" will be correspondingly
larger,
that's not really true. the sweet spot in the strict ambisonic
sense is
a function of order and wavelength only. but intuitively, i think the
same. my rule of thumb to the question of how many tickets can be sold
without falling from grace is "two thirds of the array diameter at
third
order", since it does indeed seem that large arrays are more
lenient in
this respect.
even under lab conditions, the effect is obvious: fons once gave me a
"virtual ambisonic" demo with point sources rendered on the sala
bianca
wfs system (using material i was intimately familiar with), and the
result was clearly improved by moving the virtual sources well behind
the physical wfs speakers.
i guess the explanation is that we get closer to the plane wave ideal.
I think that it comes down to how near the listener gets to an
individual speaker(system) or small subset of them, as this will tend
to be loudest and their sound arrive first, leading to them being
perceived as the source of the sound. Thus W and first-order signals,
where with rE and rV decodes have substantial signals of polarity
opposite those of the speakers where the sound is supposed to be
coming from, are particularly problematic.
As you know, even good mono coverage can be difficult to achieve, and
using delay systems is a compromise, as it is impossible to get the
delay correct for more than one point. Stereo generally only works
for a line equidistant from the left/right speakers. A LCR system,
with a wide horizontal dispersion centre system, can often help, as
noticed by cinema sound system designers.
Dolby 5.1 (etc.) generally work as far as they go. The main LCR
systems are behind the screen, with the audience normally some
distance in front of it. The 'surround' system is usually a fairly
large number of small speakers spread around, so that none is
particularly loud and no-one is particularly close to an individual
speaker. Of course the surround is very diffuse, with no real
directionality.
and thus a bigger audience can be within it, but the ultimate
constraints will be size of venue and environmental noise concerns.
that's another thing worth exploring: the funktion one guys have
reported that ambi rigs have an advantage in this respect, because the
ratio of useful loudness inside to leaked emission outside of the
array
is better than with stereo (or maybe even conventional four-point)
playback systems. i haven't experienced this yet and can't think of a
hypothesis to explain it, but i'd like to talk to the f1 guys at some
point about this observation.
The only explanation I can offer is that with ambisonic systems,
particularly lower order, all the speakers contribute, so that each
component doesn't have to be driven so hard.
Ciao,
Dave
_______________________________________________
Sursound mailing list
Sursound@music.vt.edu
https://mail.music.vt.edu/mailman/listinfo/sursound
