Mixing Cabinets and Phase Response.

[IvanMike]

Many threads have engaged in the debate over mixing different cabinets. Apart from subjective opinion I decided to post a thread to be linked in the FAQ to address the more scientific issue regarding mixing different speaker cabinets, namely, phase response. Having been unaware of this phenomenon, I've decided to ask billfitzmaurice to explain the process for us. By asking him the question directly, I'm avoiding any Commercial User Policy issues, hence the odd first post of this thread. Everything outside of this quote box is all Bill.

So... Bill, can you explain phase response and the issues it causes when mixing different cabinets? (Those graphs you included in another thread would be neat too).

Every piece of gear in the chain, from the cables to the speakers, affects the phase of the signal. Speakers affect phase differently at every frequency. This chart shows the phase response of a 1x12:

You can’t hear the shifting phase because there’s only one source, and phase effects are only audible when there are two sources with different phase. As for a second source, here’s a 1x15:

You can see that the phase response is different from the 1x12. That’s not just because the driver sizes are different; you get a different phase response even using the same driver in different box sizes.

Now look at the 1x15 phase response overlaid on the 1x12.

If the two charts were the same there would be no problem, but they aren’t, and that’s wherein the problem lies. Their combined outputs will augment each other at some frequencies, detract from each other at others. That’s why well engineered speaker systems don’t run different drivers in the same bandwidth. They run different drivers in different bandwidths, using crossovers to be sure that multiple drivers always complement each other.

The above charts show phase response of sealed cabs, as it’s easier to see. This chart shows phase response of a ported cab, which is more illustrative of how complicated the interactions between different drivers and different cabs really are.

 

[RickenBoogie]

Interesting. Should put an end to all these weird pairings people keep insisting on putting together, but it won't.

 

[bass4worship]

Nice point, that end my mixing of cabs. Maybe.......

 

[MooseLumps]

No more mixing for me.

 

[Benjamin Strange]

Ok, this is starting to get a bit ridiculous. You experience phasing issues all the time, as you move while you play. Phasing issues between two cabinets will change dependent on your position in the room, movement through the stereo field, etc., etc., etc. There's too much splitting hairs going on here.

I'm all for trying to get the best sound possible, but there are far too many people splitting hairs over such trivialities - and there's no shortage of unscrupulous companies out there who prey on such obsession. Sure, having charts and graphs objectively dissecting your sound can be helpful, but let's not gnash our teeth over it.

Ultimately the 15 drunks listening to your set at the bar don't give one rat's ass about phasing between your mis-matched cabinets - they only care if you rawk.

 

[fdeck]

All of the wild swings in those plots are simply due to the time delay. Try modeling them with zero delay for starters before adding the complicating factor of a constant time delay simply due to distance from the speakers.

Also, the phase response chart has the sawtooth shape because it interprets a swing past 180 degrees as going to -180 degrees. I am concerned that the computation method is making this seem worse than it really is.

 

[Marko 1]

I don’t understand the debate. If one is better, what does it matter how much better.

As for splitting hairs, why not? I think mixing cabs would cause more significant issues than moving about the room.

And I’m not going to judge my sound based on 15 drunks.

And how are companies being unscrupulous here… it’s not like they’re making more money by your mix-matching cabs.

Without being an expert, just some rudimentary sound knowledge, I can figure out that having dissimilar speakers reproducing the same frequencies can’t be as clean as matched speakers.

I wouldn’t even trust my own ears, since the science is undeniable.


Edit to add: It may not make enough of a difference to justify buying new cabinets if it would be costly, that I will agree with.

 

[Benjamin Strange]

I wouldn’t even trust my own ears, since the science is undeniable.

If you don't trust your own ears, why are you making music?

 

[Red Planet]

Interesting, if you say my 2x10 cab mixed with my 15 cab sounds bad it must be. At least on paper anyway.

 

[Marko 1]

Interesting, if you say my 2x10 cab mixed with my 15 cab sounds bad it must be. At least on paper anyway.

Nobody said that, or even anything close to that.


Btw… nice rig.

 

[MIJ-VI]

Every piece of gear in the chain, from the cables to the speakers, affects the phase of the signal. Speakers affect phase differently at every frequency. This chart shows the phase response of a 1x12:

You can’t hear the shifting phase because there’s only one source, and phase effects are only audible when there are two sources with different phase. As for a second source, here’s a 1x15:

You can see that the phase response is different from the 1x12. That’s not just because the driver sizes are different; you get a different phase response even using the same driver in different box sizes.

Now look at the 1x15 phase response overlaid on the 1x12.

If the two charts were the same there would be no problem, but they aren’t, and that’s wherein the problem lies. Their combined outputs will augment each other at some frequencies, detract from each other at others. That’s why well engineered speaker systems don’t run different drivers in the same bandwidth. They run different drivers in different bandwidths, using crossovers to be sure that multiple drivers always complement each other.

The above charts show phase response of sealed cabs, as it’s easier to see. This chart shows phase response of a ported cab, which is more illustrative of how complicated the interactions between different drivers and different cabs really are.

You mean the Earth isn't flat?!

 

[JimmyM]

Ok, this is starting to get a bit ridiculous. You experience phasing issues all the time, as you move while you play. Phasing issues between two cabinets will change dependent on your position in the room, movement through the stereo field, etc., etc., etc. There's too much splitting hairs going on here.

I'm all for trying to get the best sound possible, but there are far too many people splitting hairs over such trivialities - and there's no shortage of unscrupulous companies out there who prey on such obsession. Sure, having charts and graphs objectively dissecting your sound can be helpful, but let's not gnash our teeth over it.

Ultimately the 15 drunks listening to your set at the bar don't give one rat's ass about phasing between your mis-matched cabinets - they only care if you rawk.

-I- care, and that's enough. Besides, I have way more drunks than 15 at my gigs. If you're happy with your mixed rig, great. It's never once worked out for me, although I did have a mixed rig a couple years ago that sort of worked well but was greatly improved with matching cabs. Never knew why until I got on here, but in retrospect it makes perfect sense because the phasing was what I was hearing screwing everything up.

 

[lomo]

I don’t understand the debate. If one is better, what does it matter how much better.

As for splitting hairs, why not? I think mixing cabs would cause more significant issues than moving about the room.
And I’m not going to judge my sound based on 15 drunks.
And how are companies being unscrupulous here… it’s not like they’re making more money by your mix-matching cabs.
Without being an expert, just some rudimentary sound knowledge, I can figure out that having dissimilar speakers reproducing the same frequencies can’t be as clean as matched speakers.
Edit to add: It may not make enough of a difference to justify buying new cabinets if it would be costly, that I will agree with.

This. If perfect is the enemy of good, then good may also be the enemy of better. I know my 310/210 stack sounds good, but now I'm going to test it against the 310 alone (but at equal volume) in a large room, by listening carefully to a broad spectrum riff (say low B up a couple of octaves) from many places in the room. It seems to me that even if box size plays a lesser role than driver size (which I'm not sure about), even a bit of phase incoherence can be a bad thing. As posted above, it may not be a enough reason to redo a rig, but it certainly seems like a logical reason not to design a rig from scratch with differing drivers and no properly designed crossover.

 

[IvanMike]

I did pick up the following tidbit from one source:

"Even with identical woofers, if there's a path length difference to your ears there can be degradation of clarity. It's a slam of just about every bass cabinet that uses more than one woofer, but that was my reluctant conclusion after conducting blind listening tests of various array geometries. In many cases multiple woofers are necessary to get sufficient loudness, but there is usually a tradeoff."

 

[IvanMike]

It's interesting to note that the ear/brain can't tell the difference between a sound where the fundamentals and partials are all in phase, and one where the phases vary quite a bit. Heard seperately the speakers could sound very close, together you could have issues.

Now, I'll play Devil's advocate and say of course you should trust your ears before anything else (i like my mixed setup better than two of either cab i own). But, the science is worth knowing, and helps explain some combinations I've tried which sounded plain weird or aweful.

 

[OriginalCrash]

I did pick up the following tidbit from one source:

"Even with identical woofers, if there's a path length difference to your ears there can be degradation of clarity. It's a slam of just about every bass cabinet that uses more than one woofer, but that was my reluctant conclusion after conducting blind listening tests of various array geometries. In many cases multiple woofers are necessary to get sufficient loudness, but there is usually a tradeoff."

At the same time, though, doesn't having multiple drivers also smooth out some of the peaks and valleys in frequency response that a single driver cab may have...?

ETA: I'm not saying one thing or another... I suppose I'm just A) checking to make sure I have my facts straight and B) stressing the point that it's a trade off.

 

[greenboy]

As is so often the case with bassists, concern for what they themselves hear in their little spot they usually stick to seems to be all that matters to them.

But in the larger world, like the one that also includes the room, audience and even maybe <gasp> bandmates, the phase issues of poor porting, poor driver spacing, mixing drivers/enclosures with sufficiently different phase response, is what can separate a muddy unintelligible sludge from a coherent tone.

Myself, when playing without PA support or depending on backline only for bandmates stage mix (or just having a big jam at a house party), consider it a blessing when the tone doesn't mud out (or turn to tinny garble). Any design or cab-mixing issues that mess with this, I try to avoid.

That said, some cab combinations that sound good to their users in their fixed position also don't have that different of a phase response anyway. The reason those combinations might mud up in a room has more to do with dispersion itself, and not phase mismatching specifically.

 

[billfitzmaurice]

I did pick up the following tidbit from one source:

"Even with identical woofers, if there's a path length difference to your ears there can be degradation of clarity. It's a slam of just about every bass cabinet that uses more than one woofer, but that was my reluctant conclusion after conducting blind listening tests of various array geometries. In many cases multiple woofers are necessary to get sufficient loudness, but there is usually a tradeoff."

There's a trade off with every speaker. The ideal speaker would be a single element with zero moving mass, infinitely narrow and infinitely high. It doesn't exist. Electrostatic speakers with those goals in mind, Quad ESL for example, are acknowledged to be some of the best sounding speakers in existence.
As one goes further and further from that ideal speaker, for whatever reason, more and more compromises are necessary. That's all well and good if one is aware of those compromises and how to make the best of them. A well designed speaker system will consider phase response as one of those compromises. A poorly designed system will ignore it. That doesn't mean that one might not find a system assembled by the 'throw of the dice' method to sound perfectly acceptable, you're bound to throw a 7 or 11 every once in a while. But you're just as likely to roll craps.

 

[MrTAteMyBalls]

so, to clarify, if you were to use a crossover and run each cab with a different set of frequencies you would not have the phasing issues???? i dont plan to mix different cabs, but it's still good info to have.

 

[billfitzmaurice]

so, to clarify, if you were to use a crossover and run each cab with a different set of frequencies you would not have the phasing issues???? i dont plan to mix different cabs, but it's still good info to have.

+1. Phase is of no consequence if the system has only one phase response, it's when you have different phase responses competing with each other that things get muddy. If each driver set operates within its own frequency bandwidth there's no problem.