A Quick Overview of Calsod 3.1

Calsod stands for Computer-Aided Loudspeaker System Optimization and Design and is a product of Audiosoft in Melbourne, Australia. The program can be used to simulate very complex speaker systems. Frequency response, impedance, spatial location, reflections, multiple off-axis observation points, passive and active filters, etc., etc., etc., can all be modeled with great precision and the combined results can be optimized and plotted. It can also handle imported data from all the well known loudspeaker measurement systems. This is the way I use it. Liberty Audiosuite (LAud) exports ASCII files that usually consist of nothing more than three columns of numbers. These normally represent frequency, amplitude, and phase.

The easiest way to describe Calsod's graphics is to show an example. The first figure shows some data from a midrange driver as measured by LAud. The solid line is the raw driver data just as it was imported from LAud. The dashed line is the transfer function of the filter. Combining the filter with the driver data produces the dotted line. This is the predicted output of this driver and filter combination. Last the short-dash, long-dash line represents the target response. In this case the target is a bandpass response from 250 Hz to 2000 Hz with a level of 88 Db. It uses 4th order Linkwitz-Riley filter shapes. In the bottom pane of this figure is the phase data of the items selected for the upper pane.

This next figure shows the filtered data from three drivers, and the combined output. Also shown is the input impedance of the whole system. You can see that there is an impedance rise at 900 Hz and the minimum impedance is 4 ohms at 2500 Hz. You can plot as many parameters as you want, but it gets real hard to read with more than four.

Note the style of line assigned with each curve on the figures. They are assigned in the same order every time. The first parameter that is chosen always gets the solid line and the second parameter chosen always gets the dashed line, etc. This can be a little confusing when the same parameter uses a different style from one plot to the next. I will try to be consistant as I go through the crossover development.

We will start with thecrossover input.