Sub/Main Phase Alignment Questions

[Steell]

Hey everyone, I'm fairly new to REW and acoustic measurements in general. I'm trying to integrate my two subwoofers with my mains. I've selected my crossover filter and am now trying to align the phase at the crossover. Specifically, I'm trying to follow the advice from Rythmik Audio here. Some questions:

1. When taking phase measurements, should I use an acoustic timing reference? I read in the REW manual that they aren't useful when measuring subwoofers, due to the low frequencies.
2. Should I measure the phase of the mains in stereo? When measured individually from the listening position, the phases do not perfectly align (about 30 degrees apart according to the measurements)
3. If the answer to (2) is no, should I phase align the subwoofers to each other first? Or should I individually align them with their respective main?

Thanks, and forgive me if this is pretty basic, I'm trying to learn!

 

[jtalden]

Info on your specific equipment and room setup is really needed to understand the situation better. There is a wide range of limitations/capabilities depending on the equipment used. I will make some assumptions and comment anyway in case general direction is all that is needed.

Assuming

1. A manual stereo setup.
2. You prefer a stereo SW setup even though mono is normally recommended.
3. LP and speaker positioning is
   
   established and is relatively symmetrical in a small room.​
4. XO setting has been selected and applied with a MiniDSP or similar.
5. Speaker distances/timing is now based on distance measurements.
6. We now want to see/adjust the phase tracking using REW measurements,

I suggest 4 measurements:

Setup

• Mic at LP
• Acoustic Timing - Active.
• Use the same reference channel for all 4 measurements.
  [If this is a problem, let us know your room and equipment setup.]
  
• The XO needs to be active for all 4 measurements​
• Sweep full range for all measurements (as 20 - 20k Hz)

Measure

1. Left main
2. Right Main
3. Left SW
4. Right SW

The rest of the timing/phase analysis can be done within REW. It's easy to combine the 2 mains and/or 2 SW measurements in REW into a single L + R to see mono interaction of the 2 channels. Those can be used for the timing adjustment if it's appropriate.

 

[Matthew J Poes]

Hey everyone, I'm fairly new to REW and acoustic measurements in general. I'm trying to integrate my two subwoofers with my mains. I've selected my crossover filter and am now trying to align the phase at the crossover. Specifically, I'm trying to follow the advice from Rythmik Audio here. Some questions:

1. When taking phase measurements, should I use an acoustic timing reference? I read in the REW manual that they aren't useful when measuring subwoofers, due to the low frequencies.
2. Should I measure the phase of the mains in stereo? When measured individually from the listening position, the phases do not perfectly align (about 30 degrees apart according to the measurements)
3. If the answer to (2) is no, should I phase align the subwoofers to each other first? Or should I individually align them with their respective main?

Thanks, and forgive me if this is pretty basic, I'm trying to learn!

How you measure will effect how accurate the relative phase measurement is here. However a more important point to make is that precise phase alignment is unimportant. The length of a period at low frequencies is so long that our gear won’t detect small phase variation (or large phase variation for that matter). What matters (and should be the focus) is amplitude response integration. You should adjust phase such that he amplitude response is the smoothest it can be.

The only phase related anomaly I’ve been looking at as of late is group delay. I use wavelets, which I’ve posted on before, to help visualize the relative group delay change. Assuming properly taken measurements, if I don’t see a sudden change of peak energy time in the wavelet, and instead see a smooth steady change at LF’s around the crossover frequency, then things are good.

Under the RT60 graph, you might also look at TS. Same thing, if you don’t see a sudden sharp increase in the TS value at the crossover region, then assume you are good. If you do, then you have too much added delay at LF’s.