Taking measurements from an Open Baffle

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello. can anyone please care to tell me how should I measure an Open baffle loudspeaker with a D'Apollito configuration that is 7 feet tall? I will use a miniDSP and REW, but I need to know how do I take the measurements of the drivers if I have 2 woofers, 2 midbass and a tweeter on each channel.
 
If you are trying to decide on a crossover starting point. :) 3' from the tweeter, on your listening axis to get the "far-field" response. This will be your guiding light in everything else you do.

Then the big question is, are you going to use 5 channels/speaker? If so, then it may be worth while to measure each driver separately.

Use a volt meter to ensure your test signals are 2.83V (1W into 8 Ohms) so that you get the correct absolute levels.

Then see how bad it is. :) Chances are you'll want to resort to near field for the lower mid and upper woofer responses. Use the far-field data to guide your levels.

Best,


Erik
 
Oh, wait, that above assumed you would be doing measurement -> simulation -> crossover design.

If you want to stick with REW and miniDSP, ignore my comments on near-field, just know you're going ot have to literally read between the lines. :) If you have problems seeing the right response move your mic around. Different distances will have different interference patterns so you may find it more convenient to see the responses somewhere else.
 
Hello again and thanks for replying... I Just built the definitive version of my open baffle project. The baffle measures 72" by 24" and has 2 Alpha 15A drivers, 2 PRV 8" midbass drivers and a Selenium D220Ti driver in a Pyle PH12s Constant Directivity horn, a rectangular one with a 90x60 dispersion pattern, per channel... Just a big "D'Apollito" vertical LCR for each channel.The speakers will be tri amplified by a Multichannel amp and the brain is the miniDSP OpenDRC-8. So, my question is.... which will be the appropiate procedure and order to measure and calibrate the speakers? Do I need to near-field test each driver first? How do I take the measurement of the 2 woofers which are mounted one top and one bottom and parallel wired? The midbass units are also separated by the horn and wired in parallel too....Basically a 3 way system with 5 drivers per channel...
 
It's kind of a, just how do you want to proceed kind of question. Far-field is king, but it's also damn messy. We do near field to get rid of the position dependent interference patterns which throw off everything. If this was a two way project with a crossover around 800Hz or higher, no problem. Far-field almost equals near field. In fact, that may be the approach you take for tweeter and midrange.

The problem is when you get below that range, below round 500 Hz or so and are trying to figure out the right crossover slopes and phases, the comb filtering from the room and baffle makes it very hard to get the "right" answer so we revert to near-field. When you do that, use your far-field measurements to guide your absolute levels, but near field to decide on slope/phase, etc. You will also need the far-field measurements to figure out the acoustic distances. If you are using a sloped-back baffle, such that the top and bottom MW drivers are at different distances from the listener, then you DO need to treat them separately.

There's another thread here I started asking how to simulate DSP crossovers I think you will find informative.

Best,


Erik

If all you have is a miniDSP and measurement mic, then that's what you should use.
 
Last edited:
Hi juliopr,

You may find a solution here.

measure a dipole

Peter

This is how it should be done. Quasi-anechoic, not messy and misleading ungated in-room responses. I really don't understand why people that already have access to the software and equipment to do it right insist on those types of measurements. It's so useless I don't even respond anymore when people ask why there speaker measures like it does.
 
I looked at that link, there isn't much there. It basically just says splice in the measured top with a simulated bottom, yeah?

No, the lower frequencies are measured near field but then summed with a phase inverted and delayed version of itself. Above that it is measured far field with a gate. The closest you can get to anechoic measurements.
 
No, the lower frequencies are measured near field but then summed with a phase inverted and delayed version of itself. Above that it is measured far field with a gate. The closest you can get to anechoic measurements.
Nearfield is typically an inch. Dipole behaviour will not appear in such measurement, appropriate distance depends on baffle width, usually 2 feet (60cm) is enough and gives satisfactory result with say 20-40ms gating. But measuring a total multi-way louspeaker in a small room is pain-in-the-ashbox.
subwoofer/main matching, x-o slope matching... been there, done that!
 
Last edited:
Nearfield is typically an inch. Dipole behaviour will not appear in such measurement, appropriate distance depends on baffle width, usually 2 feet (60cm) is enough and gives satisfactory result with say 20-40ms gating. But measuring a total multi-way louspeaker in a small room is pain-in-the-ashbox.
subwoofer/main matching, x-o slope matching... been there, done that!

Reread what I wrote
 
Last edited:
Administrator
Joined 2004
Paid Member
OK, I looked at the paper, no big whoop. John gives a nice way of simulating the open baffle cancellation. Simple and effective. But he goes on to say:
"The method outlined here is designed for drivers intended for use as midranges. For woofers, where the useful response is typically below the on axis dipole peak, only the near field data need be measured and the appropriate sum performed with the SPL scaled to appropriate distance. "

The method does not help us with woofer measurements, which a difficult part of any in-room speaker measurement.
 
I don't think you understand what John is saying. He's saying there is no need for the far field measurement that is spliced to the low frequency near field measurement since the entire passband of the woofer exists in classic dipole-only mode.

If you have problems with getting good measurements perhaps it is because you are not attempting proper quasi-anechoic methods? Which was my original point, and is indeed a "big whoop" when ya'll are always fighting with messy in-room measurements.
 
Administrator
Joined 2004
Paid Member
I think the method is smart and good for midrange and up, as it lets you calculate what the baffle will do. But it isn't valid for woofers (and John says so) so you still are going to have problems with in room measurements.

FWIW, quasi-anechoic measurements of an Open Baffle bass aren't going to tell you the full story. You have to take the rear radiation and the wall behind the speaker into account.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.