Understanding subwoofers

[DY2]

you mean below 40Hz?

No, I definitely meant above 40 Hz.

In a typical setup, people struggle with localization around 80 Hz. But as you can see in the measurements I just shared, I have extensively tested crossover points from 40 Hz up to 250 Hz.

Because I use 6 subwoofers and have managed to keep harmonic distortion extremely low, I’ve reached a point where the bass remains completely non-directional. In my room, if I can pinpoint the subwoofers' location at any frequency higher than 40 Hz, it simply means my integration isn't perfect yet. It's a strict standard, but the data in the graph above shows why it's possible.

 

[Oddball]

Not sure what is up with your tone. I can post a THD sweep of my sub at 100 dB. My question on the AVR was because I knew you had ART, and an AVR is the only way to get ART. When I saw your system picture I knew routing test tones to different subs would not be straight forward. I hope other people try some expirements themselves and post results. I have no idea why in my room and system it is easy to locate to 50 Hz but I would be surprized if I was the only one.

Sorry have been a bit edgy - my teenage daughter is literally driving me nuts. My apologies for my tone.

I get it. Just have different experience with localization.

 

[levimax]

I get it. Just have different experience with localization.

I make no claims to hear to 50 Hz with music playing, just test tones, I hope the research continues with a focus on what actually happens in "small rooms". I am starting to realize that the goals for sub integration for a MC system and one for a Stereo system can be quite different which is one of the reasons opinions vary so widely. For stereo I treat the mains/sub combination as a single speaker with careful driver placement and since that requires co-location I can chose a crossover determined by the most linear range in the driver overlap region rather than being forced to consider localization effects. This give me a point source transducer with excellent time domain performance and accurately preserve as much stereo information as possible. With MC it is more about smooth FR over a bigger seating area and with multiple subs playing, summing (I know maybe ART doesn't do this in the traditional sense which is why I want to try it) the sub response makes a lot of sense, (otherwise there would be lots of cancelation effects) and the spatial effects are driven more by the multiple speaker channels than any LF spatial cues.

 

[Keith_W]

my teenage daughter is literally driving me nuts.

GOOD!

 

[thcdru2k]

For stereo music, anchoring a subwoofer to its respective speaker helps preserve spatial cues, so co location and higher crossover can make sense. Localization is desirable in the sense that bass is anchored to the speaker location rather than sounding detached.

For the majority of multichannel systems, bass is routed to subwoofers, so uniform response and reduced localization become more important.

A couple approaches I have seen:

1. Single sub / mono sum bass:
Simple but often prone to localization and uneven response due to lack of spatial averaging.

2. Co located / time aligned subs:
Anchor bass to the speakers for stereo image stability and crossover coherence. This can be achieved through physical co location or precise delay and phase alignment, depending on the system. The sub and main sum smoothly in both magnitude and phase through the crossover region and behave like a single source. Can be mono sum or channel-specific.

3. Distributed multi sub:
Minimize localization and improve uniformity across the room through spatial averaging. Bass is typically mono summed in these systems, though alternative routing schemes are possible in principle.

4. Dirac ART / multi speaker control:
Dirac ART focuses on reducing room modes and decay, improving bass clarity and consistency, with reduced localization as a byproduct.

Each approach optimizes a different perceptual outcome such as image anchoring, uniformity, or decay control, so the best method depends on the listening goal.

 

[thcdru2k]

On top of all that with low crossover and channel-specific routing you can reduce localization cues where ITD fluctuations dominate. If subs are not fully correlated you may produce envelopment but that is another thread but still very related to subwoofer integration.

 

[john61ct]

And these approaches can be combined.

Where there are trade-offs e.g. maximum "tightness" vs uniform bass, fixing room modes via mono free placement vs spatial immersion / envelopment

perhaps it is possible to structure different operating modes for the different use cases, different DSP profiles, turning some of the multichannels down or off.

Worst case, in a lower budget context, physically moving enclosures and swapping cables for say dance parties.

Film nights vs critical listening more frequently, in my case.

 

[thcdru2k]

And these approaches can be combined.

Where there are trade-offs e.g. maximum "tightness" vs uniform bass, fixing room modes via mono free placement vs spatial immersion / envelopment

perhaps it is possible to structure different operating modes for the different use cases, different DSP profiles, turning some of the multichannels down or off.

Worst case, in a lower budget context, physically moving enclosures and swapping cables for say dance parties.

Film nights vs critical listening more frequently, in my case.

If I was a a bachelor with unlimited funds I would love to experiment with multi-sub array, weighted decorrelation based on position, and weighted channel-specific routing.

 

[john61ct]

I choose my partners differently from most.

For me, the learning / testing / tinkering process "journey" is the goal, rather than any one destination.

Not interfering too much with my core interests and lifestyle is pretty fundamental.

As for budget, these days I think it's easier than ever to experiment, the main trade-off is time and convenience.

 

[EJ3]

For stereo music, anchoring a subwoofer to its respective speaker helps preserve spatial cues, so co location and higher crossover can make sense. Localization is desirable in the sense that bass is anchored to the speaker location rather than sounding detached.

For the majority of multichannel systems, bass is routed to subwoofers, so uniform response and reduced localization become more important.

A couple approaches I have seen:

1. Single sub / mono sum bass:
Simple but often prone to localization and uneven response due to lack of spatial averaging.

2. Co located / time aligned subs:
Anchor bass to the speakers for stereo image stability and crossover coherence. This can be achieved through physical co location or precise delay and phase alignment, depending on the system. The sub and main sum smoothly in both magnitude and phase through the crossover region and behave like a single source. Can be mono sum or channel-specific.

3. Distributed multi sub:
Minimize localization and improve uniformity across the room through spatial averaging. Bass is typically mono summed in these systems, though alternative routing schemes are possible in principle.

4. Dirac ART / multi speaker control:
Dirac ART focuses on reducing room modes and decay, improving bass clarity and consistency, with reduced localization as a byproduct.

Each approach optimizes a different perceptual outcome such as image anchoring, uniformity, or decay control, so the best method depends on the listening goal.

I'm # 2. Floor firing subs with the mains sitting on top of them.
Stereo 2.2.

 

[Jaxx1138]

If you don't mind me asking, who are you? A Harman researcher hiding in our midst is always exciting! And do you have a link to some publications about the benefits of cardioid bass?

APOLOGIES, for this unintentionally being off topic here.
I am nobody of any note. I worked with Sean Olive, Floyed Toole and Todd Welti for 10 years as a Lab Manager and acoustic engineer for the research department. Everybody is retired or consulting now. I am an audio nut that loves sound reproduction far more than I should. And I enjoy a healthy debate with other audio enthusiast, luminaries and anybody else that loves this stuff. I did run over 400 tests in the double blind test that is in Sean Olives work, Floyd Tooles book and other mentions online and in many publications. I am not as prolific as many of my colleagues.

 

[Jaxx1138]

My case is below. Really happy with it. Subs with ART are going to 100hz and bass localization was never better. Flies in the face of all my previous experiments and stups, but I am not shy to admit it. I was trying my best to get to something that ART does so much better with zero effort.

Don't mind the bed, it is just another convenience feature. I'd take it anytime over the swivel chair that has been resting in the cellar for years.


View attachment 529653

That is actually a really good layout. I would prefer a more symmetrical room but that is something we can't always control. The use of the Gallo loudspeakers in this case is excellent because of their dispersion characteristics. Not bad at all.

 

[MKR]

APOLOGIES, for this unintentionally being off topic here.
I am nobody of any note. I worked with Sean Olive, Floyed Toole and Todd Welti for 10 years as a Lab Manager and acoustic engineer for the research department. Everybody is retired or consulting now. I am an audio nut that loves sound reproduction far more than I should. And I enjoy a healthy debate with other audio enthusiast, luminaries and anybody else that loves this stuff. I did run over 400 tests in the double blind test that is in Sean Olives work, Floyd Tooles book and other mentions online and in many publications. I am not as prolific as many of my colleagues.

Welcome! As with Keith, would love to hear your thoughts on cardioid bass, especially any papers and research results! Personally I am a big believer in the benefits, having a multichannel system with cardioid bass (down to ~40Hz) in my mains, center, and side channels

 

[Jaxx1138]

Welcome! As with Keith, would love to hear your thoughts on cardioid bass, especially any papers and research results! Personally I am a big believer in the benefits, having a multichannel system with cardioid bass (down to ~40Hz) in my mains, center, and side channels

First , Thank You.
Brief answer; cardiod woofer on free standing loudspeakers should be a standard across the industry. Linear dispersion or directivity index from about 10kHz to 100Hz(approximate)would be optimal. This is an opinion and potentially a hypothesis based on the work that I did.

 

[MKR]

First , Thank You.
Brief answer; cardiod woofer on free standing loudspeakers should be a standard across the industry. Linear dispersion or directivity index from about 10kHz to 100Hz(approximate)would be optimal. This is an opinion and potentially a hypothesis based on the work that I did.

Preach it brother!

But I would contend cardioid down to ~35-40Hz actually provides maximal benefit

 

[Jaxx1138]

It does appear that a lot more research is needed on bass frequencies in rooms. David Griesingers research into the perceptual attributes of low bass in rooms did not yield anything conclusive but does raise more questions(stereo bass). There are questions as to what can be perceived as a source of sound below 80Hz in room less than 1000 square feet. With sine tones it is very difficult (this also happens at very high frequencies). By the use of techniques used to eliminate modes and nulls used currently , it may change the way we perceive sound in a room. It obviously changes the overall timbre of the sound (as it should). Floyds book tells us that that the activation of most of the modes(the use of multiple subs/distributed mode) and then EQing helps to better balance the overall timbre especially for a larger listening area. Based on my experience, cardiod woofer patterns from 350Hz to 80Hz on any free standing loudspeakers offers a significant improvement in bass quality and articulation. One consideration is that the overall transfer function/target curve will most likely change somewhat as we are artificially changing the perceived room boundary effects. The Shroeder frequency may be shifted down with this application. We need evidence to support any of these conjectures. Your experience may vary.

 

[Jaxx1138]

Preach it brother!

But I would contend cardioid down to ~35-40Hz actually provides maximal benefit

You are most likely correct based on my experience.

 

[levimax]

For stereo music, anchoring a subwoofer to its respective speaker helps preserve spatial cues, so co location and higher crossover can make sense. Localization is desirable in the sense that bass is anchored to the speaker location rather than sounding detached.

There are a few other advantages to this approach:

1. co-location and crossing over higher and keeping tight 1/4 wavelength spacing puts the crossover point in a more linear range of both drivers allowing the use of steep complementary Linear Phase crossovers which effectively cancel out pre-ringing and greatly improve LF time domain performance (which I contend it quite audible but I know the science is not settled).

2. By crossing a powerful sub with plenty of headroom at higher than normal frequency it is often times possible to wisely fill dips with DSP in addition to knocking down peaks. This allows for frequency response smoothness like distributed subs with just two point sources and all the advantages that come with it.

 

[levimax]

It does appear that a lot more research is needed on bass frequencies in rooms. David Griesingers research into the perceptual attributes of low bass in rooms did not yield anything conclusive but does raise more questions(stereo bass). There are questions as to what can be perceived as a source of sound below 80Hz in room less than 1000 square feet. With sine tones it is very difficult (this also happens at very high frequencies). By the use of techniques used to eliminate modes and nulls used currently , it may change the way we perceive sound in a room. It obviously changes the overall timbre of the sound (as it should). Floyds book tells us that that the activation of most of the modes(the use of multiple subs/distributed mode) and then EQing helps to better balance the overall timbre especially for a larger listening area. Based on my experience, cardiod woofer patterns from 350Hz to 80Hz on any free standing loudspeakers offers a significant improvement in bass quality and articulation. One consideration is that the overall transfer function/target curve will most likely change somewhat as we are artificially changing the perceived room boundary effects. The Shroeder frequency may be shifted down with this application. We need evidence to support any of these conjectures. Your experience may vary.

Do the techniques to generate cardiod bass affect time domain behavior at all?

 

[Jaxx1138]

Do the techniques to generate cardiod bass affect time domain behavior at all?

They do but in a good way if implemented correctly. Your destructive interference is used where much of the boundary effects cause indirect sound to cause cancellations with the direct sound. By utilizing the back wave of your woofer, you cancel as much as 6-15dB of this sound. Thusly creating a more coherent direct sound (in the 400Hz to 80Hz region) . This region is where many colorations occur and one of the main areas that room correction are attempts to fix but often cannot because it's an acoustic issue not a loudspeaker problem. In short, it aligns more of the bass to be in phase much better.