What's the best / easiest way to measure frequency response using MLS full range?

[fjhuerta]

Hi all!

I'm planning to build a 3 way system. Previously, I've only been concerned with two way systems, and adding the required baffle step correction was as simple as checking the transfer function plot and making sure it went 6 dB's down.

Now, with a three way crossover, I know things are not going to be that simple anymore.

The way I measure drivers is - I place a microphone on the tweeter axis, feed the MLS signal to each driver, and save the results. I tried aligning the microphone with each driver and then correcting frequency response at the tweeter axis by playing around with the driver size and position in LSPCad, but I found that LSPCad could never simulate off axis response properly.

So, now, I need to do a full range MLS capture, and I don't know how to do the bass part. Since I usually use a 6ms gate, I can't capture the full bass response.

How can I do such a thing? I know using a longer gate will only add floor reflections. Should I do a 500-20 KHz capture and then splice a nearfield measurement? If so, how can I take into consideration off-axis response?

[ThomasW]

Yep bass needs to be done separately since it needs a longer gate time than higher frequencies.

For bass there's no such thing as off axis.

The best way to do off axis is put the speaker on a rotating stand, aka turntable or lazy-susan.

[fjhuerta]

Thanks Thomas!

I guess the question is - can it be done using a longer gate, or must I do a 1M measurement at a shorter gate, then a nearfield mesurement, and then splice them together?... and then using the KHF tool to find out the baffle step correction needed?

If so, I'd assume off-axis response isn't that important.

[augerpro]

Do a short gate farfield. Then a long gate nearfield, I've been using 200ms and it seems accurate. I'm not sure what software you're using but I think you would add BS to the nearfield and then merge. You don't want to add BS to the farfield.