Two Subs time alignment question ...

[FidelioX]

Hi guys,
I'm new here and new about REW.
After watching some videos I've done my first test to time align my two JL Audio subs.
Both subs have default settings....so polarity to zero, phase to zero and xover at the highest.

My receiver is a Denon X4700 and I own a MiniDSp 2x4 HD.
Latest beta REW build with a UMIK1 mic.

My SUB1 is at 3,5 meters from the measuring point.
SUB2 is near, it's 2,80 mt. from the measuring point.

So my starting idea was to apply a delay to the SUB2.

I've measured both Subs and after some simulations via REW's Time alignment tool I find out that the best positive summation seems to be with Sub2 in "inverted" mode and with a delay -21 (!!!)
So i've setted the Sub2 inverted (via minidsp) and applied a +21 ms to the delay of Sub1 .... I've done a real measurement with both sub actived and the prediction seems to be good ...

So the real measurement seems to confirm the rew's prediction tool but .... I'm not sure if it's correct to apply a delay to the farhest sub (my Sub1) from the measurement point ...????

Thank you so much for any feedback

 

[sm52]

You are doing everything right. What I wouldn't do is not use inversion. I would try to play with delays instead. But many consider polarity reversal to be normal. And of course, the last word is yours. Compare before and after.

 

[FidelioX]

Hi Thank you for your feedback.
The first thing tried yesterday was not inverting the second sub and add some delay using the REW's prediction tool.
It seems that the result was worse respect inverting it and subtract delay.

Today I try again to not invert it and add delay. I will do a comparison.

But, in general, ... Do you can confirm that when one sub is inverted .... it's not an error to add delay to the farhest sub ?

 

[sm52]

Yes. It could be. Low-frequency waves indoors behave in ways we don't expect. If you want someone to try to create delays based on your measurements, publish the mdat files here.

 

[FidelioX]

Hi guys, I dont saved yesterday test.
Attached a measurements just made now....
I don't know why ....maybe a different mic position but the situation is different now....
I'n not able to find with Rew's tool ... a better alignment respect both decalibrated subs playing together ...

Very very strange ...let mw know If you are able to find a better solution....

 

[sm52]

The measurements are weird. You can try to apply a delay of 7.2 ms to Sub1. And it is better to make 3 measurements from three points located at a distance of 1 m from each other.

Second option, apply 29.9 ms to Sub2.

 

[John Mulcahy]

The delay figures for the 2nd and 3rd measurements are wrong, try measuring with 5.20.14.

 

[teyhyrh4r]

What I wouldn't do is not use inversion.

sometimes it is necessary not for the FR, but in order to get the sub close to zero on excess group delay graph. IMO there should be a excess group delay graph in the alignment tool, cause sometimes you are aligned, but a cycle to late/early

 

[John Mulcahy]

a cycle to late/early

There is an impulse alignment mode in the 5.20.14 early access build which avoids that potential issue.