Help on Ease Cluster - Cluster Calculation Module
Why the Cluster routine looks wrong sometimes.
A) Cluster A is the Project with the four original Loudspeakers
S1 with hor=30, ver=-25 and S1* with hor=-30, ver=-25
S2 with hor=90, ver=-25 and S2* with hor=-90, ver=-25
used speaker model: cemh64-3.spk
look:
B) Cluster B is the Project with the 4 Loudspeakers replaced by a cluster
with an aiming axis ver=-47.1° hor=0° (average)
cluster name: clustera.spk
picture: clpro_b.gif
C) Cluster C is the Project with the 4 Loudspeakers replaced by a cluster
with an aiming axis ver=0° hor=0° (default)
cluster name: clusterc.spk
picture: clpro_c.gif
D) Cluster D is the Project with the 4 Loudspeakers replaced by a cluster
with an aiming axis ver=-25° hor=0°
cluster name: clusterd.spk
picture: clpro_d.gif
Explanation:
1. The angle of the resultant Loudspeaker (if the originals are replaced)
is the same as the chosen aiming axis. Look at:
clpro_b.gif in case B)
clpro_c.gif in case C)
clpro_d.gif in case D)
2. All Direct-SPL-1000Hz-InterferenceSum Maps are the same! Look at:
spl_a.gif in case A)
spl_b.gif in case B)
spl_c.gif in case C)
spl_d.gif in case D)
Legend:
Just the case A) seems to be a bit more detailed because the cluster
in B), C), D) is averaging the balloon data.
3. If, after replacement, the map looks the same, but the new Loudspeakers of B), C), D) have different angles, then the balloons of those Loudspeakers must be different too. Obviously a balloon must look that much upward the Loudspeaker looks downward. So:
4. The balloons of B), C), D) are of course different, because different
coordinate systems (aiming axes) have been chosen. (Frequency Band = 2 kHz) Look at:
clustera.gif in case B)
clusterc.gif in case C)
clusterd.gif in case D)
5. Final question - why is the average axis of those 4 loudspeakers at -47.1° and
not at -25°?
This is simple math. The average axis is not just the plain
average of all axes, but it is the axis where the biggest attenuation
is to be expected. Therefore all speaker balloons are assumed to be spheres,
otherwise the cluster computation itself has to be done to find the
axis with the biggest attenuation.
So the axis is chosen that way that the distance to all other 4 individual
axes is a minimum. And that is NOT the case for ver=-25°.
6. Limitations of application:
The EASE cluster routine is just the far field approximation of a loudspeaker array. When you use a cluster in computations please make sure that the distance from the array to the point of interest is much bigger than the average distance between two components of the cluster. Otherwise the far field assumption is not right.
In such a case we recommend to ungroup the cluster and to use the single devices for simulation. The cluster creation procedure also allows you to treat a big set of loudspeakers like a single loudspeaker. This is not only very useful for computations like mentioned above, but also for moving and turning several loudspeakers all together.
For bigger arrangements it has also proved to be sensible to use bigger bandwidths for calculations, e.g. 9..15.