Intergrating dual subs with mains

[John24sj]

Hi everyone,

Using REW + UMIK-1 and miniDSP 2x4HD I got the subs time alligned and EQed to this response:

Now I want to properly time align them to the mains and pick the best crossover as the one that audyssey provided is utter rubbish. I was following this video for guidance:

But my main concern is that it appears a 120Hz crossover works best with delays of 38.7ms on left sub and 24.20ms on right sub (picture below is Subs + Mains). I would like to avoid 120Hz as I don't want to localise the subs in the room.

Now I am just wondering if measuring the subs only at each crossover and then L+R only at each crossover again and later in REW using "Align phase slopes at cursor" at each crossover point is the best way to work out delays and best crossovers?

I saw other people measuring with L and R individually instead of L+R, thus I am slightly lost. Because when I do that, I get a positive delay meaning I would need to advance the subwoofers and not add a delay in miniDSP:

Just in case I attached my .mdat files: Subs + Mains alignment with Crossover.mdat is measuring L+R and Subs + L and R crossovers is measuring L and R individually.

Any help greatly appreciated

 

[jtalden]

Observations about the second file:
Two of your measurements; L-60 and R-110 appear to be timed incorrectly relative to the others. Possibly there is an occasional problem with the acoustic timing. If this cannot be corrected for any future work, you could take each measurement 3 times to confirm the timing agreement of the 3 impulses and then delete 2 of them.

All the L and R measurements have a huge SPL rise between 30-50 Hz. It rises to the level of the average upper bass SPL even though the XO appears to have something approaching the proper influence at the XO frequencies. The SWs response there is thus impacted significantly by the 2 mains. The XO filter should have suppressed this range very significantly, The cause of this issue should be determined and addressed. If necessary a PEQ filter could be used to suppress this L, R channel peak by at least 10 dB.

The group delay and spectrogram of the SWs response suggests the timing between the 2 SWs is very offset. It is sometime possible to achieve a much smoother SPL response doing that, but it spreads out the duration of bass range and results in greater group delay and phase rotation than is normally recommended. The relative timing of the 2 SWs is normally set close to value needed to account for the difference in their distance from the LP.

I did a timing analysis on the 80 Hz XO data anyway to determine my recommended XO timing. I would reverse the 2 SWs polarity and also reduce their delay by 8.5 ms. The 1st chart below shows the resulting L+R (mono) response with this change compared to the current settings.

The left and right mains phase is opposed to each other in much the measured range. There is no clear evidence that the polarity is actually switched between them so it is likely this is just due to the room influence. The 2nd chart below is a comparison of the SPL difference between the 2 channels (using my XO timing changes). It may be prudent to confirm the proper polarity of the L and R using full-range, near field measurements. If it is just the the room influence as expected, then some adjustment to the L and R speaker positioning may be helpful.

In general:
I don't know if addressing any, or all, of the above observations will significantly impact the sound quality. The SPL is not significantly improved with just the timing changes. I would still try to sort out some of the other issues noted to see if the sound is improved. The setup would at least be closer to more conventional settings.

New XO 80 Hz settings (green) vs current settings for mono signals.

Left vs Right channel SPL.

 

[John24sj]

Thank you for such a great and detailed responce. Tbh I am rather lost for what I could as I am righting with my room I guess.
Ths is my current layout:

Tbh I am not sure what I can really do I clearly got some awful room modes in the room, but at the frequency range where normal acoustic dapening would need to be way too large to help