Measured phase in SE

[augerpro]

I have quick question for the 50 lb. brains here. When doing SPL measurements in SE I reference the baffle as 0 cm distance. So when performing FFT all excess phase is removed and what's left is the phase difference due to AC offset. This way it's included in the measured phase and crossover CAD has an accurate phase to work with.

I have two problems though. Why would I use the "include distance in phase" in the HBT then if I've already taken care of it? I tried it and with tiny values I could adjust HBT phase to match the measured phase. But i can do the same thing just adjusting LP stop and rolloff too so I'm not sure why I want to mess witht the phase this way when it has already been taken care of when I did the FFT?

Also when measuring the nearfield, do I place the mic very close (like .25") to the cone or just at the plane of the baffle?

Anyway thanks everyone for the help so far! :T My first diy project is moving along steadily now. I should have it finished by the end of Feb. It's a 5.1 Bose buster type of thing using HiVi woofers. I'll post the measurements if anyones curious as to the difference between the B and M series. I also have 2 MCM woofers and a Goldwood I was considering for a cheap first project in LA's. I'll post those too.

Brandon

[Mark K]

Hi Brandon,

First, the easy one. A true SPL measurement is very close to the diaphragm. So 0.25" or less, in general, would be a true nearfield. It can be instructive to look at multiple nearfields, and spl at 1" 6" 1ft etc. However, for the purpose of splicing a nearfield to the farfield, 0,25" is fine.

It's been a while since I used SE. I usually fall back on LspCAD though I have SE.

If you do a measurement at 1m the phase response will be the natural driver's phase behavior, called the minimum phase response, plus the distance to the mic, which is the distance from the acoustic center, kind of a mythical entity...let's call it the driver's G spot, to the mic. It's often hard to know exactly where this is.

So there are a couple ways of getting around this. You can measure exactly at, or close to the listening point, and then use exact measured FR and phase, to cook up your xover.

Or, you can measure the drivers FR/phase, and "subtract out" the distance to the baffle. I assume this is what you mean when you say "reference to 0." Once you've subtracted the baffle difference, what's left in the phase response is minimum phase plus the phase change from G spot, uh, AC offset.

Now, "FFT'ing" doesn't remove all the phase. That's somewhat of a misleading statement. You remove an arbitrary amount of phase by deciding where to put the start marker, and then perform and FFT.

Whether you want to "include distance" depends on whether you want to generate a simulated FR at a certain point in space, or whether you're trying to generate a minimum phase curve. Or, min phase curve plus AC, as you've chosen to reference it.

When designing though, you do want the final FR curve to include distance, so that you get the modeled FR curve at the listening point.

hope that helps!

[augerpro]

Thanks Mark. I think I see what you're saying. I'd never really thought of using a simulated FR at the listening point. Interesting. I'll have to check this out in SE. I figured as long as phase of each driver is represented accurately for a given distance I'd be fine.

You'd think by my age elusive G spots wouldn't cause me so much trouble, but alas...