Stereo Bass using subwoofers

[bmc0]

True as with the wavelengths involved the enclosure can be considered a point source but that property is nearly useless for all practical purposes anywhere outside an anechoic chamber.

Sure, but the measurements I was referencing were done with a Klippel NFS (if I'm not horribly mistaken), so are effectively anechoic.

 

[levimax]

Well, if the person who holds the data can do this, do a linear fit on the phase of the system in the range of interest. THEN subtract out that linear fit. What remains is the non--delay part. The linear fit part is pure delay. See the thread I just posted in general audio discussions for the terms, etc. That can be minimum phase or maximum phase, or some of each, but with the time delay sorted out.

I have attached the .mdat file from my SB-3000 taken with REW and a Earthworks M23 using loopback timing taken nearfield. Since it agrees fairly closely with Erin's Kipple measurements and is repeatable I have some confidence that it is accurate. I don't really know how to subtract out things but if someone is interested have at it and I would be interested in the results.

 

[OCA]

I have attached the .mdat file from my SB-3000 taken with REW and a Earthworks M23 using loopback timing taken nearfield. Since it agrees fairly closely with Erin's Kipple measurements and is repeatable I have some confidence that it is accurate. I don't really know how to subtract out things but if someone is interested have at it and I would be interested in the results.

The response is not minimum phase according to your measurements, see excess group delay compared to minimum phase version even after IR delay is removed:

It seems to be caused by a HP filter at around 17.5Hz and a LP at around 200Hz as @j_j predicted:

 

[bmc0]

I have attached the .mdat file from my SB-3000 taken with REW and a Earthworks M23 using loopback timing taken nearfield.

Magnitude, measured phase, and excess phase (approximately compensated for time delay):

Measured, minimum phase, and excess group delays:

It seems to be caused by a HP filter at around 17.5Hz and a LP at around 200Hz

If the filters are minimum phase, they do not contribute to the excess phase.

 

[OCA]

If the filters are minimum phase, they do not contribute to the excess phase.

True, maybe they are not minimum phase or maybe it's some boundary reflection leaked by suboptimal anechoic conditions. Maybe @j_j can explain better but the response is not minimum phase. By its very definition an initial move towards the opposite direction before the impulse is non-minimum phase behavior:

 

[bmc0]

By its very definition an initial move towards the opposite direction before the impulse is non-minimum phase behavior

If you want an accurate minimum phase impulse response from a measurement, you usually need to carefully set HF and LF tail frequencies and slopes. Otherwise strange things can happen either due to noise or assumptions made by the software about system behavior outside the measurement range. Here's what I get for the minimum phase impulse response (along with the measured impulse, which was shown inverted in your plot):

 

[OCA]

If you want an accurate minimum phase impulse response from a measurement, you usually need to carefully set HF and LF tail frequencies and slopes. Otherwise strange things can happen either due to noise or assumptions made by the software about system behavior outside the measurement range. Here's what I get for the minimum phase impulse response (along with the measured impulse, which was shown inverted in your plot):
View attachment 433160

That seems a bit forced to look min phase and peaks in the opposite direction but more importantly does not explain the excess group delay!

 

[bmc0]

That seems a bit forced to look min phase and peaks in the opposite direction

Both impulses are positive polarity in my plot. Yours shows the measured impulse as inverted while the minimum phase impulse is not.

 

[OCA]

Both impulses are positive polarity in my plot. Yours shows the measured impulse as inverted while the minimum phase impulse is not.

The mic is inverting it during measurement acc. to this experiment:

 

[thewas]

Shouldn't we consider though the total system of loudspeaker plus room (gain)?

In this case it is possible to get a flatish response down to DC as shown here:

Analytical Analysis - Room Gain

-Introduction- The term 'room gain' is often seen used when the topic of subwoofers and the low frequency sound reproduction comes up. As I see many explanations that are lacking or incorrect, I thought I might try and clear up the topic once and for all. In all rooms there are so-called...

www.audiosciencereview.com

 

[Blockader]

Shouldn't we consider though the total system of loudspeaker plus room (gain)?

In this case it is possible to get a flatish response down to DC as shown here:

View attachment 433193

Analytical Analysis - Room Gain

-Introduction- The term 'room gain' is often seen used when the topic of subwoofers and the low frequency sound reproduction comes up. As I see many explanations that are lacking or incorrect, I thought I might try and clear up the topic once and for all. In all rooms there are so-called...

www.audiosciencereview.com

This assumes pressure chamber conditions where displacement equals to pressure. Only a perfectly sealed room can satisfy such conditions.

 

[thewas]

This assumes pressure chamber conditions where displacement equals to pressure. Only a perfectly sealed room can satisfy such conditions.

Real rooms though have and show an intermediate behaviour so its quite possible to get linear bass response very low with real and not insane sized loudspeakers there.

 

[Blockader]

Real rooms though have and show an intermediate behaviour so its quite possible to get linear bass response very low with real and not insane sized loudspeakers there.

Yes, Geddes even suggests that a moderately sized room with only one open door can come very close to ideal pressure chamber conditions. That's a wild assumption, however I think it is worth to investigate.

 

[bmc0]

The mic is inverting it during measurement acc. to this experiment:

You misunderstand what is shown in that video. First, the microphone is not to blame—it accurately converts pressure to voltage (with the correct sign!) in the frequency range of interest. Second, only the acoustic transfer function (i.e. the sound pressure) is relevant to this discussion.

The cone motion being out of phase with the pressure above the system resonance frequency seems curious at first glance, but indeed follows directly from the physics involved, as explained in this thread. Sound pressure in free space is proportional to the diaphragm acceleration, which is 180° out of phase with the displacement. Do note as well that the displacement is not in phase with the input voltage except near DC.

 

[j_j]

This assumes pressure chamber conditions where displacement equals to pressure. Only a perfectly sealed room can satisfy such conditions.

Well, the modulated fan subwoofers can do DC quite well in a mostly sealed room, but in general we are not talking about those. In such cases, the volume of the room and its transition to pressure mode must be known, or measured, and will change with door, window, etc, position.

In general there could be a plan to match woofer transition rolloff to room volume, but I'm not aware of many people considering this.

 

[OCA]

You misunderstand what is shown in that video. First, the microphone is not to blame—it accurately converts pressure to voltage (with the correct sign!) in the frequency range of interest. Second, only the acoustic transfer function (i.e. the sound pressure) is relevant to this discussion.

The cone motion being out of phase with the pressure above the system resonance frequency seems curious at first glance, but indeed follows directly from the physics involved, as explained in this thread. Sound pressure in free space is proportional to the diaphragm acceleration, which is 180° out of phase with the displacement. Do note as well that the displacement is not in phase with the input voltage except near DC.

"Relevant to discussion" would be an explanation of why you think shared sub measurement's response is minimum phase because it's NOT and throwing a bunch of technical gibberish will not change that. FWIW, sound pressure is proportional to velocity not acceleration.

 

[bmc0]

"Relevant to discussion" would be an explanation of why you think shared sub measurement's response is minimum phase because it's NOT

I never stated that it was minimum phase nor am I claiming that it is. In fact, I'm the one who (unintentionally) started this multi-page argument about minimum phase, excess phase, etc. by pointing out that it didn't look minimum phase in post #615.

 

[j_j]

"Relevant to discussion" would be an explanation of why you think shared sub measurement's response is minimum phase because it's NOT and throwing a bunch of technical gibberish will not change that. FWIW, sound pressure is proportional to velocity not acceleration.

Um, sound pressure is proportional to f^2 * volume velocity. There's 2 accelerations involved, cone displacement and microphone displacement, reacting to pressure, so from end to end, it's f^2 * displacement. That's why speakers can be flat, after all, the second-order lowpass above resonance works to provide flat reproduction.

 

[OCA]

I never stated that it was minimum phase nor am I claiming that it is. In fact, I'm the one who (unintentionally) started this multi-page argument about minimum phase, excess phase, etc. by pointing out that it didn't look minimum phase in post #615.

Cool then. But that deems your time shifted graph below wrong because it attempts to show a min phase response:

while my inverted response below shows clear non-min phase behaviour:

And I'd be happy to hear your opinion (and others) on why it's not minimum phase although it's a point source for the wavelengths in focus. Is it the filters, is it the boundary reflections during the measurement...

 

[bmc0]

But that deems your time shifted graph below wrong because it attempts to show a min phase response:

I don't understand what you're accusing me of. My plot shows the original measured impulse with the original polarity (red trace), along with the minimum phase version (gray trace).

And I'd be happy to hear your opinion (and others) on why it's not minimum phase

As far I I can tell, it basically has to be the digital filters used; I don't think there's another explanation. Exactly why the filters aren't minimum phase in this case, I don't know.