Another Measurement Techniques Question

[dsrviola]

Let's say that one is taking measurements of the raw drivers installed in their intended enclosure. If one were to take off axis measurements at this time, would the crossover simulations be accurate for on and off axis measurements, or are other considerations necessary?

[cjd]

short answer: usually, but it depends.

If you do all all your work with raw data (one mic position for measuring all drivers at a good farfield distance for proper summing, proper gating for bass, etc...) then it can yield accurate off-axis simulations if you use the off-axis measurements (you don't mention what tool you're using for sims)

C

[Jed]

Let's say that one is taking measurements of the raw drivers installed in their intended enclosure. If one were to take off axis measurements at this time, would the crossover simulations be accurate for on and off axis measurements, or are other considerations necessary?

The answer is... maybe. For example, while driver offset of say a midrange or tweeter can yield a flatter on axis response during the design phase, sometimes the off axis response can be a bit unpredictable on the horizontal plane because of the different distance of the tweeter's center to baffle edge from the left or right of the baffle relative to the mic location. So, measuring the off axis response at varying degrees is a good reality check after modeling the polar response with minimum phase based measurements or diffraction simulators (the Edge simulator seems accurate). If you aren't using minimum phase frequency response measurements with the acoustic offsets of each driver determined accurately, forget about the accuracy of the off axis modeler in most software programs. This is especially noted in the vertical plane since using a fixed mic measuring technique accounts for actual phase and magnitude at a fixed point in space, but not the driver's orientation on a baffle, and how the center to center spacing influences the power response at the listening position (which programs like LSPcad can predict accurately with the correct data). An important consideration in my book (I use LSPcad).