Subs with reversed polarity (surprise!)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
It is an Article of Faith that polarities must be preserved. And for dipoles, that the phase annihilation diagram for woofer frequencies in the textbook is an Eternal Truth.

On the other side, anybody with experience fooling with phase and even with polarity and certainly with dipoles, knows that the sound hits walls and is in a multiple of ways shaken from its phase locksteps.

So I figured there's a real simple way for a person with a Behringer DCX2496 - the all-singing all-dancing DSP - to explore the notion with about 6 button presses of the DSP box.

The subs are in the front corners of the room. L is a 5-cu ft sealed box and the R a 17-foot labyrinth. Both with some EQ. A mic was placed mid-way between the subs and near my listening chair. The REW sweep is 10-500 Hz (OK, 10 Hz is pushing it a bit with my 1980 JVC receiver amp, but you can see for yourself what the mic picks up).

The first illustration shows the (1) L and (2) R traces when run separately. Many similarities are reflections of the room (literally, of course).

The second illustration shows the (1) R sub trace and the (2) R sub with polarity inverted trace. Many readers will notice a certain overlap.

The third illustration shows results when (1) both subs are running and when the (2) R sub is inverted. Needless to say, the reversed trace is the lower one. But not annihilated. Far from it.

The fourth illustration repeats the inverted trace shown in the third illustration. Funny, it looks so much better than the non-inverted trace. Maybe I'll raise the volume control and keep that. I bet just the thought of somebody doing such a thing would cause indigestion among many members here.

B.
 

Attachments

  • L plus R.jpg
    L plus R.jpg
    63 KB · Views: 415
  • R plus R invert.jpg
    R plus R invert.jpg
    57.8 KB · Views: 415
  • both plus both w invert.jpg
    both plus both w invert.jpg
    68.9 KB · Views: 422
  • L and R invert.jpg
    L and R invert.jpg
    57 KB · Views: 417
Last edited:
Are you saying that the right sub response changes that much, from just reversing its polarity? Is the flatter response the one with correct absolute polarity, or reversed?

+1 nigelwright7557.

As the second illustration shows, there is zero change in the FR of the R sub when its polarity is reversed.

If the orthogonal polarity led to a flatter response, I wouldn't bother posting it, eh. Pretty much accidental, but the reverse connection (which does show some of the expected cancellation) looks flatter. Accidental, I am sure. But I set up my mic stand and took the first runs without moving it. So I didn't fish for a good result.

B.
 

ICG

Disabled Account
Joined 2007
An inverse function would suggest the speakers would just cancel each other out.
But in a real room the speakers will be a distance apart and there will be phase shifts etc.

You are right, that will probably fix the phase in one place, in others it will likely get worse. It's much better to find better positions for the subs. Or build a double bass array.
 
The third illustration shows results when (1) both subs are running and when the (2) R sub is inverted. Needless to say, the reversed trace is the lower one. But not annihilated. Far from it.

The fourth illustration repeats the inverted trace shown in the third illustration. Funny, it looks so much better than the non-inverted trace. Maybe I'll raise the volume control and keep that.
If you measure the response is your listening area, eg within 3' of your favourite chair, you will find the 'in-phase' subs give MUCH more consistent performance over a larger area than your 'out-of-phase' subs.

You may also like to listen to a central (mono) voice comparing in & out of phase if your 'subs' go up to 100Hz or more.
 
If you measure the response is your listening area, eg within 3' of your favourite chair, you will find the 'in-phase' subs give MUCH more consistent performance over a larger area than your 'out-of-phase' subs.
Is that a question or an assertion of faith?

You may also like to listen to a central (mono) voice comparing in & out of phase if your 'subs' go up to 100Hz or more.
Huh?

Is that a question or an assertion?

First, the measurements show there's only a bit of "correct" phase because there's only a bit of interference. Second, there's no content down there to hear. Third, there's no sense of localization down there. Fourth, any issue would not be better anywhere except up close to one of the subs in which case, the polarity reversal wouldn't be evident either way.

B.
 
Absolutely yes. As would anybody who isn't running a dance establishment. Wouldn't you?

Not that I was 100% serious about doing so in the OP.

B.
You might also paint a thick layer of honey over speaker cones, both inside and outside surfaces.

It will definitely dampen resonances, kill all cone cry, put all peaks and throughs on their knees and give you smooth and sweet sound across the passband.

You will also lose 10/20dB average sensitivity but who cares?

Except night club dancers or car subwoofer competition hacks, of course.

Just keep the ants away :p
 
If you measure the response in your listening area, eg within 3' of your favourite chair, you will find the 'in-phase' subs give MUCH more consistent performance over a larger area than your 'out-of-phase' subs.
Is that a question or an assertion of faith?
Actually its measured results. (Also some early FEA work culminating in full PAFEC simulations when Patrick Macey developed his Acoustic Boundary Element. The FEA stuff in the 1990s was some of the earliest validated FEA/BEA of response in real rooms.)

The FEA stuff is still difficult these days but you can easily measure the results with your REW setup.
First, the measurements show there's only a bit of "correct" phase because there's only a bit of interference. Second, there's no content down there to hear. Third, there's no sense of localization down there. Fourth, any issue would not be better anywhere except up close to one of the subs in which case, the polarity reversal wouldn't be evident either way.
I can pontificate on the theoretical implications of out-of-phase speakers at LF and localisation too but its MUCH simpler for you to listen to the effect on a mono signal. If you don't find out-of-phase mono disturbing at frequencies above 100Hz .. you have exceptional hearing facility and should ignore anything I say. :)

The aural effects I am referring to will be most obvious/unpleasant at your listening position but I gather your music sources have nothing near or below 100Hz :eek: This must be true as you say you hear little or no difference when polarity is inverted even though it leads to 10-20dB loss of level below 150Hz.
 
Last edited:
My post was commenting on the DATA* that you posted. Other peoples data (including mine) would be of absolutely no relevance to my post.

You are suggesting that losing up to 20dB of output due to phase cancellations is better than using EQ to bring down a frequency to the same SPL. Your amps / subs are going to have to be working around 20dB more to produce the same output as the 'in phase' setup. So if the 'in phase' setup is doing 100dB at 8W then your 'out of phase EQ solution' will be running at ~500W.

Please tell me how all that excursion / power consumption is going to improve distortion / headroom?

I'm not trying to troll you Ben, I just cannot understand your reasoning on this one.

Rob.
 
My post was commenting on the DATA* that you posted. Other peoples data (including mine) would be of absolutely no relevance to my post.

You are suggesting that losing up to 20dB of output due to phase cancellations is better than using EQ to bring down a frequency to the same SPL. Your amps / subs are going to have to be working around 20dB more to produce the same output as the 'in phase' setup. So if the 'in phase' setup is doing 100dB at 8W then your 'out of phase EQ solution' will be running at ~500W.

Please tell me how all that excursion / power consumption is going to improve distortion / headroom?

I'm not trying to troll you Ben, I just cannot understand your reasoning on this one.

Rob.
Your data on the consequences of polarity reversal in subs would be an excellent addition to my own. What is keeping you from posting? Even with no DSP, takes 30 seconds to reverse the leads to one of your subs, eh. Please post these two curves in your room. Don't be shy. We're all friendly here.

My point is that ONE accidentally chosen mic location hardly tells the story about other locations, such as my chair.

Over-Interpreting the accidental location and offering your well-intentioned advice seems to miss that point. And that is exactly what i prove in the other thread. The FR for my chair is about as nice as I can make it with my poor powers and yet it bears just minor resemblance to the accidental location.

More directly to your reply above, YES, power is cheap but SQ is dear. That's a trade-off that anybody but a dance establishment would make. Wouldn't you?*

B.
* just to rest your mind, the chair FR is pretty good and I am certain that nobody would want to see the inverted polarity subwoofer curve which has to be a lot worse. But, the inverted curve does kind of look better in the accidental location. Gosh, I hope somebody will insist I post the inverted polarity chair curve.
 
Last edited:
You really need other people to post graphs to prove that running 2 subs out of phase loses all the bass ? Really ? Your point in post numbers 1 to 3 seem to suggest that you think out of phase subs seem flatter and that is a good thing.

Suddenly we now have an 'accidental' mic placement and that you were really talking about listening at your chair.

"More directly to your reply above, YES, power is cheap but SQ is dear. That's a trade-off that anybody but a dance establishment would make. Wouldn't you?*"

I was saying the opposite ? You're cancellation eq thing would be burning 500W against my 'in phase EQ' solution burning 8W. I was assuming the 8W solution with much less power / excursion would win the SQ trade off.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.