Measuring far field vs. near field + bafflestep

[igy137]

Hi,

I'm a bit lost during measurements, please help.
I did some far field ground plane measurements and also near field measurements.

I tried to align the two graphs, applying calculated baffle step to near field.
They actually match quite well up to 2-3khz, ignoring some dips (like ~1050Hz) in the near field graphs, which seems to be at box width or so. I guess that might do for crossover design.

However, the phase is not that clear. If I adjust delay to match somewhere, it soon goes different.

The far-field measurement was gated to 4ms or so, so probably this is the reason of mismatch below 250hz? .. in this case which one should I use for crossover sim?

Also, for the mid, levels do not line up nicely, either range 100-700Hz is ok or range 700-3k. Could this be due to baffle step simulation problem or is this more a measurement issue?

Sorry if all this is too stupid

[cjd]

If you are merging multiple data collection points you MUST work in a minimum phase model, which means getting exact z-axis information and entering that into your simulation software.

The alternative is to use gated measurements where you measure every driver without moving anything at all in the setup, so the phase is exactly right. Doing this means you need a clear picture of exactly where your data starts to fall apart as far as frequency response, and a low crossover point in a 3-way can make it really tough to get usable data.

C

[Face]

The alternative is to use gated measurements where you measure every driver without moving anything at all in the setup, so the phase is exactly right. Doing this means you need a clear picture of exactly where your data starts to fall apart as far as frequency response, and a low crossover point in a 3-way can make it really tough to get usable data.

C

I'd like to emphasize the last part with this method. IMO, using this method if the woofer's crossover point is at, or above baffle step it's not as bad. If you have some bass traps or insulation, lay them down on the floor between the speaker and mic as it will help remove floor bounce from your readings.

[igy137]

Thanks guys, I go back and check that Z-axis to be sure.
I guess the 2nd method would not really work for me, as I plan to cross somewhere around 250-300 Hz should be below baffle step with my 32cm wide baffle.

[cjd]

How are you getting to a minimum phase model? (What software?)