Acoustic Prediction

K

Kip Conner

Junior
Mar 13, 2011
370
0
0
Athens, GA
For all you guys using prediction software I need to clear up something. After reading some of Bennett's articles I ran across an iPad app that does acoustic predictions. It's only does Omnidirectional Subs, but that's all I need it to do.

I was able to get the exact cardioid shape that I want from a block of four subwoofers by placing them a quarter wavelength apart on the grid, reversing the polarity and adding 2 ms of delay. I'm assuming that it's 2ms since the only parameter you can change is whether or not you want the field in feet or meters.

So, here's my question... if I were to take this prediction and put it into a real world scenario... Is the 2ms of delay assuming that the phase has already been aligned/corrected so the additions are going to be perfect? Or would I just add the 2ms and go with it?

550988_3866751942206_399817057_n.jpg
 
Re: Acoustic Prediction

Kip Conner said:
For all you guys using prediction software I need to clear up something. After reading some of Bennett's articles I ran across an iPad app that does acoustic predictions. It's only does Omnidirectional Subs, but that's all I need it to do.

I was able to get the exact cardioid shape that I want from a block of four subwoofers by placing them a quarter wavelength apart on the grid, reversing the polarity and adding 2 ms of delay. I'm assuming that it's 2ms since the only parameter you can change is whether or not you want the field in feet or meters.

So, here's my question... if I were to take this prediction and put it into a real world scenario... Is the 2ms of delay assuming that the phase has already been aligned/corrected so the additions are going to be perfect? Or would I just add the 2ms and go with it?

View attachment 4480
Click to expand...
What freq are you looking at? Remember that the delay (and physical distance) will get larger as you go lower in freq.
 
Re: Acoustic Prediction

Bennett Prescott said:
Kip,

The 2ms of delay is what aligns the phase!
Click to expand...

Yes, since polarity and phase are (mostly) independent. The time aligns the phase to 0 degrees, then swapping polarity puts them 180 degrees out of phase - at all frequencies. If the time delay weren't used, it would be 180 out of phase, but only at one frequency.
 
Re: Acoustic Prediction

Caleb Dick said:
Yes, since polarity and phase are (mostly) independent. The time aligns the phase to 0 degrees, then swapping polarity puts them 180 degrees out of phase - at all frequencies. If the time delay weren't used, it would be 180 out of phase, but only at one frequency.
Click to expand...

Which is how the array works out front... it's 360° out of phase (read: in phase) over a limited frequency band.
 
Re: Acoustic Prediction

So I would assume that the absolute best way- even after you have done your modeling, would be to use the Phase display in Smaart to do your alignment. Thus letting smaart test the entire signal path and account for the latency in the processor??
 
Re: Acoustic Prediction

Kip Conner said:
So I would assume that the absolute best way- even after you have done your modeling, would be to use the Phase display in Smaart to do your alignment. Thus letting smaart test the entire signal path and account for the latency in the processor??
Click to expand...

I was say the "best" way is to use 2 mics-1 in front and 1 in back of the array.

Adjust the delay to get "the best" freq response and level out front-while getting the most rejection to the rear.

If you have a live 2 channel measurement system you can look at both mics at the same time and see what happens.

1 mic alone is not going to really tell you much-except in that location.

Note that it will probably be a compromise in one or the other-you have to determine which is best for the particular situation.

While you can get good results-it will not be "perfect".
 
Re: Acoustic Prediction

Hi Kip:

Measurement will always trump modeling.

Modeling is an extension to the art of educated guessing. Measurement is an extension to our perception of reality.

The former used correctly will reduce the time required to do the latter correctly.

Each Thanksgiving my Dad puts about 30 bottles of wine on the table and our family scores each on a pad of paper and the winners end up becoming Christmas gifts. My Dad and I are the only ones good at windowing the data such that reliable interpretation of the results are obtained. I can usually tell if a wine is worth tasting with my nose, thus avoid bad data and its potential to skew the entire measurement session. The remaining scores from my family are generally discarded. We do this every year and they still don't get the concept of windowing. They'd be better off reading wine reviews, i.e. modeling.

---

Start looking at the group delay of each passband during each step (almost said phase!) of your alignment session. There's gold in those traces. :)
 
Re: Acoustic Prediction

Kip,

I always set up a directional array by measuring in the direction where it is most clearly defined. For me, for a cardioid array, that is behind the array a reasonable distance. Line up the phase traces with delay and then invert polarity, boom, you're done. If you've set your driver spacing well and are not too close to boundaries everything should be good.
 
Top Bottom