This whole topic is likely to set the teeth of engineer-purists on edge. And that is because the theory is inherently heteroskadastic.
There seems to be no or little hope of manipulating room responses feasibly and effectively using architectural means, at least in the domestic music-room context. Likewise, and somewhat inexplicably, experience with signal processing and filtering have not been favourable.
Which means that using multiple subwoofers is the method of choice to tame room influences - as taught by Toole's chapter and recently advocated by Geddes too.
Due to the necessities of research, tests of multiple subs are generally systematic and coherent. But the underlying message that I get is that the more heterogeneous ("mix-and-match") are the woofer locations, suspension types, resonances, and other factors, the better off you are in combating the pesky room influences.
I'm delighted by the results with my highly heterogeneous mixture of a corner horn and a midwall giant IB.
My impression is that you can add almost any woofer into the mix and through the magic of log-addition, even woofers with not-so-low cut-offs will improve the room.
BTW, as I have firmly believed for the last 40 years, it is largely a waste of effort to have more than one subwoofer signal. This belief is, of course, inherent in the multiple sub approach, because it is all but inconceivable to have any of the subs playing different signals.
There seems to be no or little hope of manipulating room responses feasibly and effectively using architectural means, at least in the domestic music-room context. Likewise, and somewhat inexplicably, experience with signal processing and filtering have not been favourable.
Which means that using multiple subwoofers is the method of choice to tame room influences - as taught by Toole's chapter and recently advocated by Geddes too.
Due to the necessities of research, tests of multiple subs are generally systematic and coherent. But the underlying message that I get is that the more heterogeneous ("mix-and-match") are the woofer locations, suspension types, resonances, and other factors, the better off you are in combating the pesky room influences.
I'm delighted by the results with my highly heterogeneous mixture of a corner horn and a midwall giant IB.
My impression is that you can add almost any woofer into the mix and through the magic of log-addition, even woofers with not-so-low cut-offs will improve the room.
BTW, as I have firmly believed for the last 40 years, it is largely a waste of effort to have more than one subwoofer signal. This belief is, of course, inherent in the multiple sub approach, because it is all but inconceivable to have any of the subs playing different signals.
Notwithstanding of course, frequency independent, continuously variable phase as you would no doubt have noticed. If I was doing this for a living I'd get on to a self contained product that had this feature as it would save so much time.
I'm referring to the times that a sub enhances the lower response but upsets the higher bass frequencies and when I invert the drive the upper bass pulls into line etc.
I might mention that I'm experimenting with helper woofers, which I'm liking, but they restrict the locations for some of the subs.
I might mention that I'm experimenting with helper woofers, which I'm liking, but they restrict the locations for some of the subs.
Perhaps it is that "etc" that fogs my understanding of what you might possibly be talking about. Could your comment be related to what I - equally obscurely - referred to as "magic of log-addition"?
I am sure the excellent subtlety of your thought would be admired even if written so that simple minds could understand what you are saying as well.
Sorry Ben, I'll come at it from another angle.
When I have set up subs, I've found it a little frustrating at times when placing a woofer that helps some frequencies but makes others worse. I find myself faced with the possibility of crossing over lower to avoid the problem region and maybe having to build another sub to make up for it.
The problem I'm talking about is phase. This may even show up when I change the low pass crossover only to find that the frequencies around the crossover no longer blend anymore. I assume that this is a factor in your theory on heterogenous woofers as I know that if you have two different subs that have otherwise been set up the same you can swap their location (to exactly where each other used to be) and the overall response will change.
By the way, I looked up 'heteroskadastic'. Great word
When I have set up subs, I've found it a little frustrating at times when placing a woofer that helps some frequencies but makes others worse. I find myself faced with the possibility of crossing over lower to avoid the problem region and maybe having to build another sub to make up for it.
The problem I'm talking about is phase. This may even show up when I change the low pass crossover only to find that the frequencies around the crossover no longer blend anymore. I assume that this is a factor in your theory on heterogenous woofers as I know that if you have two different subs that have otherwise been set up the same you can swap their location (to exactly where each other used to be) and the overall response will change.
By the way, I looked up 'heteroskadastic'. Great word
Last edited:
Yeah, used the word just 'cause it sounded spiffy (and is roughly appropriate).
Now, "phase" is one of those concepts that can be neatly handled by quantitative analysis. So it has a prominence in acoustic textbooks.
I've been reading a great book, The Information, about Claude Shannon. Information transmission, in some contexts, is related to the statistical characteristics of a language or code. In the same sense, the influence of phase on music reproduction may well depend on the actual characteristics of music. And "music" now-a-days, maybe be divided into separate species of traditional mic-pickup music versus music that is synthesized and/or highly processed.
Is there enough phase-coherent information in mic-pickup music to make phase matter? Or, do even pedal organ notes wander all over the place, phase-wise, and make any model based on phase, such as your discussion, less relevant?
Obviously, if you test with sine waves, your music room will have hot and cold spots, as you point out arising from the familiar principles of acoustics. But with actual music, does it have any practical importance, given the "wandering about"?
Now, "phase" is one of those concepts that can be neatly handled by quantitative analysis. So it has a prominence in acoustic textbooks.
I've been reading a great book, The Information, about Claude Shannon. Information transmission, in some contexts, is related to the statistical characteristics of a language or code. In the same sense, the influence of phase on music reproduction may well depend on the actual characteristics of music. And "music" now-a-days, maybe be divided into separate species of traditional mic-pickup music versus music that is synthesized and/or highly processed.
Is there enough phase-coherent information in mic-pickup music to make phase matter? Or, do even pedal organ notes wander all over the place, phase-wise, and make any model based on phase, such as your discussion, less relevant?
Obviously, if you test with sine waves, your music room will have hot and cold spots, as you point out arising from the familiar principles of acoustics. But with actual music, does it have any practical importance, given the "wandering about"?
The way you are referring to phase, I agree it is of little consequence. I am talking about the relative timing of sound coming from more than one source. Similar to reversing the polarity of a tweeter and experiencing a hole around the crossover.
When adding a sub to the mix, reversing the polarity of the sub tends to produce the opposite effect. For example if adding the sub filled one hole and created another, when you switch the polarity it would then make a larger hole and create a peak (in that order)
Reversing polarity takes the phase to 180 degrees from 0 (360) degrees. Continuously variable phase as I am suggesting would allow anywhere in between as well as different amounts across the spectrum.
When adding a sub to the mix, reversing the polarity of the sub tends to produce the opposite effect. For example if adding the sub filled one hole and created another, when you switch the polarity it would then make a larger hole and create a peak (in that order
)Reversing polarity takes the phase to 180 degrees from 0 (360) degrees. Continuously variable phase as I am suggesting would allow anywhere in between as well as different amounts across the spectrum.
Being a good boy, I used to check polarity of speakers. Yup, with conventional speaker set-up geometry makes a difference for woofers.
Let us talk about two heterogeneous subwoofers.
Even with sine waves, you get different results with different frequencies (and so does moving your head or mic a few inches). On music, beyond my ability to tell which polarity makes a difference with my subs.
So, in my view, phase can't matter much in practice with mic-recorded music.
And I am skeptical about lab studies where using trick stimuli, folks can sense phasing a wee bit. It proves that only with trick stimuli folks can....
Let us talk about two heterogeneous subwoofers.
Even with sine waves, you get different results with different frequencies (and so does moving your head or mic a few inches). On music, beyond my ability to tell which polarity makes a difference with my subs.
So, in my view, phase can't matter much in practice with mic-recorded music.
And I am skeptical about lab studies where using trick stimuli, folks can sense phasing a wee bit. It proves that only with trick stimuli folks can....
If you were to reverse the polarity of all of your speakers at once, it may not be noticeable (or if it is, it may not be due to the phase itself). If, on the other hand, you take measurements before and after you can make a good +/-5dB response blow out to e.g. +/-15dB by reversing only the sub polarity.
Maybe when you measure with mics and/or with a standard speaker set-up and an announcer on a recording saying "My voice is now OUT OF PHASE" with the whole left speaker system out of phase, but is it detectable when you listen to music and reverse just subs (AND test blind!!!).
What do you think?
Footnote: elsewhere I suggested it might be a very valuable tool to have some kind of stimulus with the stochastic properties of music for physical testing (ghost of Claude Shannon again?)
Last edited:
It is known complex (music) signals have different behaviour from single (clean) sine wave tones. Therefore it is a little strange to assume that different genres/types of music have different outcomes on the same loudspeaker system.
If you analyse the signals of both, even pure electronic music compared to fully acoustical music, the information contains enough complexity to conclude they are all far from single sine waves test tones or phase-coherent signals if you prefer. All music, except for some extreme experimental music maybe, can therefore be classified as complex signals and all suffer from the same phase issues, if any, that are part of a loudspeaker system.
I can’t deny certain individual tracks can 'disguise' phase issues better then others but I would not go so far trying to connect certain music genres/styles with specific electric or acoustic problems.
Last edited:
When adding a sub to the mix, reversing the polarity of the sub tends to produce the opposite effect. For example if adding the sub filled one hole and created another, when you switch the polarity it would then make a larger hole and create a peak (in that order
Sometimes, it's worse. The timing of the bass is off. Sounds like slightly out of step with the rest of the music.
For HT, subs are easier to integrate as they are for effects. For music, it's more demanding.
Some music made by some synthesizers (and perhaps DSPs) must have, I a priori assume, substantial content resembling an audio oscillator. If so, MAYBE for that music, phasing heterogeneous subwoofers might matter. That is the only reason I wanted to define synthetic music out of discussion.
Otherwise, agree with your valuable post.
I’ll try to explain by an example. If you take Drum & Bass as electronic music genre you can describe the main signal, which is often the bass line, as a waveform from an audio oscillator. Since this bass line is part of a total signal the outcome is best described as complex. The unstable character of the total amplitude will confirm that. With unstable I do not mean the percussive signals within the music. If the signal was purely based on clean sine waves like in a measuring setup and could therefore be defined as not complex the amplitude should be stable and therefore predictable. So in my view even Drum & Bass can be included in the discussion.
I need a little explanation about the interpretation of your “phasing heterogeneous subwoofer”. Forgive me since English is not my native language but "phasing" is a word which is used when phase shifts over time, if I’m not mistaken. The word "Heterogeneous" in physics means more then one phase, if I’m correct. That's where I find trouble with interpreting because when we talk about multiple subwoofer setups their phase differences do not shift over time (except for subs that have the ability to ‘walk’ maybe...).
I need a little explanation about the interpretation of your “phasing heterogeneous subwoofer”. Forgive me since English is not my native language but "phasing" is a word which is used when phase shifts over time, if I’m not mistaken. The word "Heterogeneous" in physics means more then one phase, if I’m correct. That's where I find trouble with interpreting because when we talk about multiple subwoofer setups their phase differences do not shift over time (except for subs that have the ability to ‘walk’ maybe...).
Last edited:
I don't know if anyone disagrees with Djim: hidden beneath a forest of Fourier transforms are pure sine waves and on the other hand (b) "natural" sound is complex. So I would conclude it is not important to know how well sine waves are reproduced in a room... unless you listen to your audio oscillator a lot or synthesized sounds.
I am especially interested in that conclusion when you install a mix-and-match subwoofer system to combat room influences. The mix-and-match theory is that you can best combat room influences by adding a second and more subwoofers as different from the first subwoofer as you can in such parameters as room location, resonances, peak, and so on. Ordinarily, I find adding drivers ("log addition") does acts to boost sound in frequency low points rather just average the loud and soft parts of each speaker's compass.
I am especially interested in that conclusion when you install a mix-and-match subwoofer system to combat room influences. The mix-and-match theory is that you can best combat room influences by adding a second and more subwoofers as different from the first subwoofer as you can in such parameters as room location, resonances, peak, and so on. Ordinarily, I find adding drivers ("log addition") does acts to boost sound in frequency low points rather just average the loud and soft parts of each speaker's compass.
It's been done before commercially
Here's a speaker that uses 2 woofers of different sizes, from years back
I tried to find a picture i remembered seeing of the internals in a review to post, but couldn't locate it
Wilson Audio Specialties X-1/Grand SLAMM
But then goes on to state
Only $65,000 in 1994 ! Think i'll order a pair for each room
Here's a speaker that uses 2 woofers of different sizes, from years back
I tried to find a picture i remembered seeing of the internals in a review to post, but couldn't locate it
Wilson Audio Specialties X-1/Grand SLAMM
But then goes on to state
Only $65,000 in 1994 ! Think i'll order a pair for each room
That's just like a Karlson which traffics in multiple resonances and sounds like a cello.
While I am not sure about the sense of sharing a port, it certainly does makes sense to have multiple bass-reflex boxes with multiple (different) resonances.
Indeed, all BR boxes have at least three output bumps (at resonance and above and below) and you can design the size of each of them (I think the famous Altec Voice of the Theater emphasized the upper one).
So mix-and-match has a history.
While I am not sure about the sense of sharing a port, it certainly does makes sense to have multiple bass-reflex boxes with multiple (different) resonances.
Indeed, all BR boxes have at least three output bumps (at resonance and above and below) and you can design the size of each of them (I think the famous Altec Voice of the Theater emphasized the upper one).
So mix-and-match has a history.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.