Soundstage Pull and Disparate Phase Measurements

[mrlawng9]

Despite careful amplitude matching, I continue to have a soundstage pull left using D&D 8Cs. My hypothesis is that the walls of my townhouse have disparate materials in them.

I've noticed that the left and right speaker's phase measurements don't align. But, I'm not sure I'm taking the measurements correctly. The LP is ~2.5 meters from each speaker. Below are the plots using an FDW of 8 cycles and setting t to 0 for both.

Right is Red and Left is Blue:

I'd be very appreciative of the forum's help in:

• Validating or invalidating the methodology I used to generate these phase measurements. Please let me know if there is a more accurate method
• Providing feedback on whether or not these disparate measurements could be the culprit in pulling the stage left?
• Providing solutions I can try if any exist
  • For example, could RePhase be employed to align phase between left and right speaker?

Thanks!

 

[ernestcarl]

Likely it’s the asymmetry in the room. You could try switching the left and right speaker positions to see if the same happens. Also try 3 cycles FDW to remove any remaining room induced phase rotations.

 

[mrlawng9]

It's definitely the room and not the speakers nor electronics. I've gone to extensive lengths to vet all that, including recently getting my hearing checked.

Viewing phase at 3 cycles FDW did not change the measurement.

Would you think this disparity is audible?

And if so, are there DSP options to fix it? I should add that I have an audiolense XO license, which unfortunately, has not fixed the soundstage pull - though it sounds excellent in all other aspects. I was curious about the phase EQ in RePhase as being a possible solution? Unfortunately, the sound is pretty unlistenable as it's very difficult to hear anything coming from the right side of the stage.

Thanks!

 

[ernestcarl]

Would you think this disparity is audible?

Yes. I would expect there to be so…. as well as a cancellation in some of the bass frequencies with mono content — you will see this with a L+R sweep, say, 0-500Hz.

But it’s difficult for me to understand what you exactly mean by “soundstage pull” and how it relates to the phase difference in the bass… meaning that I presume it’s not bass being the main audible issue for you, but something else altogether.

 

[ernestcarl]

I was curious about the phase EQ in RePhase as being a possible solution?

I have “aligned” the phase response difference induced by the room between left and right speakers before, but nothing as big of a difference as seen in yours. So I’m not so sure… It would help if you could upload the mdat of a left and right sweep. But I have to ask if this really is the only possible physical arrangement for your speakers? Because I’d probably start with changing the placement or positioning first.

 

[mrlawng9]

Good question, I can clarify. Vocals and dialogue that are meant to be centered sound like they are coming from the left side of the soundstage. About halfway between center and the left speaker.

 

[ernestcarl]

Good question, I can clarify. Vocals and dialogue that are meant to be centered sound like they are coming from the left side of the soundstage. About halfway between center and the left speaker.

Normally, this would be caused by any timing and amplitude level difference between speakers — moreso above 300Hz, rather than in the bass itself.

 

[mrlawng9]

I have “aligned” the phase response difference induced by the room between left and right speakers before, but nothing as big of a difference as seen in yours. So I’m not so sure… It would help if you could upload the mdat of a left and right sweep. But I have to ask if this really is the only possible physical arrangement for your speakers? Because I’d probably start with changing the placement or positioning first.

I’ve explored all options short of moving! mdat attached.

 

[mrlawng9]

Normally, this would be caused by any timing and amplitude level difference between speakers — moreso above 300Hz, rather than in the bass itself.

Agree. The above mdat is the 8Cs prior to equalizing them, so the raw in-room response. My eq and timing efforts have bene successful but have not fixed the soundstage pull. And yes, there is bass wonkiness. It doesn't always sound like it's coming from in front of me, it can sound like it's coming from my left.

My hypothesis is that the materials behind my walls vary significantly, potentially with one or two steel beams in play.

 

[ernestcarl]

I’ve explored all options short of moving! mdat attached.

You will have to compress it in a zip format to attach in a post. Alternatively, an external download link would work fine as well.

 

[Doodski]

You will have to compress it in a zip format to attach in a post. Alternatively, an external download link would work fine as well.

Whatever happened to using IRC and C2C file transfers. sigh*

 

[ernestcarl]

Whatever happened to using IRC and C2C file transfers. sigh*

Gone the way of the Dodo it would seem... now have a gazillion online accounts for everything.

 

[mrlawng9]

Rookie mistake!

Here's a link to the mdat.

And, thank you for your time on this - I appreciate it.

 

[Doodski]

Gone the way of the Dodo it would seem... now have a gazillion online accounts for everything.

C2C really does solve a lotta issues though.

 

[ernestcarl]

The phase actually looks less worse than I thought. Auto adjusting the IR delay and FDW 7 cycles already makes it clearer:

These room induced kinks in the phase is best left alone, IMO:

Theoretically, the unequalized sum shouldn't even be all that bad (FDW 5):

However, there are notable differences in the time domain as seen in the spectral views esp. when zoomed-in:

Notice the very strong HF reflection around 0.5ms of the left channel -- better find out what's causing it and minimize/eliminate the issue -- it could potentially shift/bias sound localization.

Also check out that blob of early reflected energy (4-6 ms) between 1kHz & 300Hz in the left channel which is not present in the right.

Early reflection spike of the left channel can be seen in the ETC as well:

Fourier spectrogram view mode, B-H 7, 50 ms window, 50 dB scale

The peak energy time appears more intact in the left channel despite lower amplitude level in some frequencies.

quck edit: dB scaling was actually slightly off so changed to match both exactly (previous pasted the wrong one)

A bit more delay and late "time smearing" coming from your right channel -- to be clear: I would not say that it's bad, I've seen far worse from conventional speakers.


Now, I'm not an acoustician, but it seems obvious to me that even the Dutch 8C can use some help (acoustic treatment) to improve the LR balance further beyond what EQ can "fix" when placed in a difficult space.

 

[mrlawng9]

Wow. This is great analysis and greatly appreciated. It’s a bit beyond my REW pay grade but gives me a lot to investigate and learn - which is why I’m on this forum in the first place. That said, I get that you have diagnosed via the measurement that this is a difficult room and that you have been able to identify stage-smearing issues / reflections. Thank you.

 

[ernestcarl]

Notice the very strong HF reflection around 0.5ms of the left channel

Actually, it might actually make the very HF sound a little weird or possibly even "phasy"?

Wow. This is great analysis and greatly appreciated. It’s a bit beyond my REW pay grade but gives me a lot to investigate and learn - which is why I’m on this forum in the first place. That said, I get that you have diagnosed via the measurement that this is a difficult room and that you have been able to identify stage-smearing issues / reflections. Thank you.

Your welcome. If find that there is always something to learn new here.

 

[Daverz]

Acourate has an algorithm to even up phase differences between channels he calls Inter-Channel Phase Adjustment (ICPA). Mitcho demos this here:

I'd probably try to fix it first by rearranging things or arranging existing furniture. Might be expensive to fix with commercial room treatments (diffusors?).

 

[amirm]

Good question, I can clarify. Vocals and dialogue that are meant to be centered sound like they are coming from the left side of the soundstage. About halfway between center and the left speaker.

Do you know for sure that the music is not mixed that way?

What happens if you just play left or just right (of the same channel in music)? Does the image come from the same position relative to said speaker?

 

[ernestcarl]

Acourate has an algorithm to even up phase differences between channels he calls Inter-Channel Phase Adjustment (ICPA). Mitcho demos this here:

I'd probably try to fix it first by rearranging things or arranging existing furniture. Might be expensive to fix with commercial room treatments (diffusors?).

Well, the phase response difference isn't that big at all, I believe. Even the timing looks quite good on second glance -- better than most I've seen. But there is still a noticeable difference in the early reflected spectral energy between channels which may account for the vocal shift/imbalance.