Setting Subwoofer Phase with REW

[carsknivesbeer]

I am trying to fine tune the phase for a Rhythmic L12 and a near field setup. The subwoofer is behind the listening position. Subwoofer is crossed over at 80hz via the amplifier crossover. Posted is a snapshot of the RTA using Pink PN with a sequence length of 64k, full range. Also included is the SPL and Phase. Is there a link to what this should look like or what I should be aiming for? I am not sure I am doing this right. This setting is 40ms on the subwoofer with a 12db LPF Slope.

 

[ernestcarl]

40ms seems like a huge amount of delay for a nearfield setup -- over 13 meters!

Not sure what's happening in your RTA... and your summed sweep looks sort of okay. Why don't you try looking at the wavelet in the spectrogram view.

 

[carsknivesbeer]

So this is the wavelet with a 40ms second delay then with a 0ms delay. Clearly it should be closer to zero. is there a way to set is more precisely?

 

[Speedskater]

What I would do. Using a cross-over frequency sine wave on narrow band pink noise.
I would reverse the wires to either the main speakers or the sub. Then I would tune for a dip in the response. It's easier to see or measure a dip than a peak. Then return the wires to the correct polarity.

 

[carsknivesbeer]

So Sine wave at 80hz with speaker wires reversed, this is what the RTA looks like at 0ms

 

[bigjacko]

Can you change 1 ms at a time or even smaller? I guess trail and error is the easier way to do.

 

[carsknivesbeer]

Sure but I got it in my head after watching a RMAF video I should work on it and I have REW sitting here so I figured I could learn how to do it without letting my AVR do it for me (which in this case is using a dumb amplifier). The speakers are 2ft from the listening position and the sub is 4.5 feet away from listening position behind so it should be 2ms or so. The sub goes incrementally from 0ms to 135ms. Its more about learning how to use REW and how to properly read measurements.

 

[ernestcarl]

So this is the wavelet with a 40ms second delay then with a 0ms delay. Clearly it should be closer to zero. is there a way to set is more precisely?
View attachment 96967

View attachment 96966

Sorry, wasn’t watching this thread so I did not see the reply. My idea with the wavelet was to see the summed response of both sub and speaker(s) — not separately.

Thus you will be able to visualize the actual offset more easily between the two transducers at your listening position.