Originally Posted by GIEGAR
I take it they're seriously flawed? I'm broadly across the issues of a horizontal MTM, but have no idea about line array theory.
Line arrays have two main problems in my experience. The first is that the MF/HF drivers are side by side, putting lobing issues on the horizontal axis rather than the vertical where it is generally less of an issue. The second is, that the taller the panel, the greater the time of flight difference is between the top and bottom drivers versus the central ones. This results in strong comb filtering and is annoyingly audible to me. It makes it sound unfocussed for want of a more technical description. Large panels like electrostatics and Magneplanars are the same.
The MF drivers in this design are quite small about 4" by eyeball. That gives them the area (Sd) of about 52cm^2 so the total area is 1248cm^2 or about the same as a pro 18" driver. The problem is that small drivers have a small Xmax so the volumetric displacement which determines SPL is not going to be the same as a pro driver, so the drivers will be using a larger % of their Xmax than the pro 18 to get the same SPL. More excursion from rest, the more distortion. Also, because a large number of drivers are used in my experience they are cheaper units than designs using a smaller number of drivers in a more conventional way, and the drivers will have to be wired in series/parallel to get the impedance to a reasonable number. Let's hypothesise they are in groups of 3, which are then paralleled; each driver will see the inductance and the Le(x) of the other two, reducing rise rate and increasing distortion.
Even very expensive designs with lots of high quality drivers still have these issues, but to a lesser extent. Many years ago we built a 3 way line array based upon a similar design, just longer and using ribbons for the HF and using a separate enclosure for the LF drivers. Compared to a number of other designs such as a coaxial (compared using the same LF array) we could never seem to get the midrange clarity and imaging of the point sources, through the dynamics were quite good.
It would be interesting to try to work around the TOF differences, by outputting to multiple DACs and amps (1/driver) with small delays to adjust for the TOF differences as much as possible, but that will have to be a project for someone else.
Gotta go shortly. I'm being dragooned to some function. Apparently there is some football match being played tonight, and I'm not a fan.