Acoustics of Non-Box Shape Rooms

Most published and internet commentaries on room acoustics related to loudspeaker applications (or mis-applications)deal with "the room" as an enclosed area (ex: 12x16x8). But most modern housing doesn't offer us rooms with such regular dimensions. Ceilings can be of wildly varying heights, and there are usually big chunks of open wall on one or more sides. Often these wall gaps open into adjacent rooms that are as big as or bigger than the intended primary listening room. These conditions raise questions about matching speakers and speaker placements with the room. For example, in an L-shaped enclosed area, is "the room" the vertical or horizontal segment of the L that you wish to consider as "the room," or do you need to consider both the horizontal and vertical segements of the L together as a whole? What tools are available to help us understand the acoustic workings of such a space?
If anyone knows of any books, articles, or other resources that deal with real-world spaces, please let me know. Everything I've ver seen just deals with simplistic cubic volumes -- not the irregular combinations of such volumes that we really live with. Or if you happen to be an expert in acoustic engineering, your personal observations would be just as welcome.
Thanks -- Brett
You need to consider all the continguous and, perhaps, adjacent spaces/barriers to know the "room." You might take a look at CARA, a CAD/Acoustics modelling program which can accept almost any configuration.

FWIW, I have an L-shaped room and found a bass hotspot in the portion not used for equipment/listening.
Kal beat me to the punch. With that in mind, has Kal or anyone else used the newer "updated" version of CARA that uses "English" measurements rather than the metric equivalents? I haven't kept up to date with CARA, but do they have more loudspeaker configurations availabe in terms of radiation patterns, etc ??? Sean
This doesn't really pertain to home audio but in sound reinforcement we have mutiple speaker enclosures or speaker clusters aimed for coverage for the majority of listeners. For the other areas we used fills, basicly high/mid speakers placed at sides, staight down (front rows coverage)on the main clusters or under balconies. By using delays to the fills we get better intelligibility.
The modals for non-parallel walls can be found using finite analysis (which is beyound the scope of this book I looked at last night, I have six different books pertaining to this field so the title is at home). There was, however, a footnote to this and somebody at Philips in the Netherlands studied and published this (again this is at home).

Lots of variables and details. Gross room dimensions, ceiling shapes, adjacent spaces -- size and shape and dimensions of doorways, wall construction material (concrete or wallboard), etc. For complex spaces, calculations get complicated.

The size and shape of the space relates to bass resonances and to midrange and treble reflections. For an existing room, I would run a bass sweep with a mic at the listening position(s) to identify resonances, and experiment with bass traps or equalizer. First reflections for midrange and treble are easily spotted by moving a mirror along the walls to identify the reflection points, then apply an absorbant panel or material. While these are not a complete solution, room calculations that may be difficult to perform theoretically can be dealt with using relatively standard procedures.