Test CD for positioning speakers and subwoofer


Can anyone recommend a good test CD to help position stereo speakers and a subwoofer? Whenever I fix one problem (weak center image, poor integration of the sub, uneven bass), another problem pops up.
tantra
I use the Rives audio test CD
http://www.amusicdirect.com/products/detail.asp?sku=ARIVESCD2

Great test cd, calibrated to use with the Radio Shack SPL meter.
Ditto on the Rives CD - grand shtuff. Though it does become a love/hate thing, you know? I'd gone through some other discs & done the (trivial, in retrospect) calibration to the RS meter by hand, etc. Then I broke down and got Rives and got a bit more thorough. I also became acutely aware of nodes and stuff like that. Wound up turning the whole room 90 degrees, researching treatments (and doing entry-level stuff), and in the end it was the nodes and the math that kinda led me to just throw up my hands, kick back, and just listen. I did learn a lot with the Rives, though.

You do have to guard against becoming way too anal about it all - I mean, it's not like I've outlined the exact spot for my butt in the "A" listening position or anything like that...
Anyone have the Rives Audio test CD for sale?
I will even setlle for a burned CD-r copy.
I worked out a procedure to locate speakers (nearly full range floor standers) with a Stereophile Test CD and RS SPL meter.  My meter is placed on a tripod at ear height at the listening seat position.  I record response throughout the 200 to 40 Hz range (after setting level for 80 dB at 1K Hz) for each position, moving speakers forward in 2" increments from the front wall.  I then add up the deviations from the 80 dB reference and compare totals for each speaker position.  This is to locate the smoothest bass response, not necessarily the deepest range.  I've found this quite satisfactory.

Regarding the Rives Test CD, I believe their supposed correction for RS meter readings assumes too much precision from the RS meter.  I doubt very much if at that price, RS could offer a meter with such consistency from one to another that an absolute correction is possible.

And so I consider any readings I make with my RS meter to represent relative data rather than absolute.