Question on accuracy of PCD and speaker workshop

[mintos]

This is a general question on how accurate PCD and speaker workshop is in modeling crossover and driver responses. I tried googling but turned up nothing. Anyone did any measurements after implementation of a crossover and found the results vastly different?

Also, I was wondering how important driver modeling driver offset is. I can't figure out what to put for driver offset execpt 0,0,0 is at the listening position. Can anyone help? Thanks

[---k---]

CJD uses Speaker Workshop. His results are good. The Khans measured at DIY Iowa last year and were probably as close to the model as can be measured in room with a gated response. The model and the actual measurements are posted near the end on the Khanspire thread.

I think the math is well enough understood that both PCD and Speaker Workshop are accurate enough. The benifits of Sound Easy and LSPCad are usability (though some would debate that they are useable ).

Unfortunately, someone else is going to have to answer your driver offset question.

[Saurav]

There's a SW manual floating around somewhere. I use SW. I measure all the drivers on the baffle, with the mic at a fixed spot. Since SW measures phase, this means you leave the driver offsets at 0 for all the drivers, because the relative distance has already been accounted for in the phase measurements. If you measure the drivers separately, you'll have to enter the distance into SW.

Get a hold of the manual, it goes into a lot of detail about how to use SW, how to do the initial setup and calibration, and so on. It helped me a lot with understanding what was going on.

I've found the SW measured and calculated response to be pretty close, within a fraction of a dB. Since I didn't measure the caps/coils/resistors, I'd expect a difference given component tolerances, DCR of the coils, etc.