[AJinFLA]

My understanding is that two stereo cardioid subs would have better spatial resolution than the distributed monopole subs, assuming the latter are all receiving the same summed mono signal.   How far that carries over when we have left and right channel signals going to the respective left and right pairs of monopole subs, I don't know.

Hi Duke,

As noted earlier, there is no one size shoe fits all. Ability to discriminate lateralization (low frequency spatial effects) varies quite a bit. Then there is the issue of content and preferences. For the HT crowd or rock/pop music types (the vast majority of audiophiles I've encountered), smooth amplitude via spatially averaged monopoles should be a no brainer. It sounds better than response with peaks and over a wider area to boot.
However, if one's preferences lean more heavily to non-amplified music that contains LF spatial info, then there is no chance of reproducing this in summed mono.
So as always YMMV. Keep in mind that the gradients not only increase discrimination of spatial effects (being fed a non-mono signal), but also have increased clarity due to the lower amount of power resulting in a decrease in decays. This effect can be mimicked with a lossy room as Dr Geddes advocated for (3' of absorber on front wall alone iirc), as this will reduce the far field power to levels similar to gradients.
But as Sean Olive says, why create power only to absorb it, very wasteful in these green times.

cheers,

AJ

[JLM]

AJ,

As I try to read up on gradient/cardioid in-room bass effects it seems that stacking out of phase subs and a delay (perhaps 4 ms, but I understand that is frequency dependent) applied, with one pair per channel is the practical ideal to achieve the effect (for non-amplied music).  True?

And is it not true that the cardio effect is only about 10 dB and that overall it reduces output significantly?

BTW I'm primarily interested in lower frequencies (20 - 60 Hz).

[AJinFLA]

AJ,
As I try to read up on gradient/cardioid in-room bass effects it seems that stacking out of phase subs and a delay (perhaps 4 ms) applied, with one pair per channel is the practical ideal to achieve the effect (for non-amplied music).  True?

Hi JLM,

To an extent, one would achieve a cardioid patter over some bandwidth by doing so, yes. Ideal? Well, maybe . There are several ways to do it (see Kimmo link). One benefit of a "single" enclosure, either flow resistance or multi-source (dipole+monopole, rear delay driver, etc), is that they can be rotated, so the null is directed at different angle, possibly better coupling to modes. This is true for dipoles also.

And is it not true that the cardio effect is only about 10 dB and that overall it reduces output significantly?

Not quite sure what you are referring to. Rear output (nulling)? Far field?

cheers,

AJ

[poseidonsvoice]

Anand, I should be clear that there is no one size shoe that fits all. There are dipoles and cardioids than extend to below 20hz and may prove all one persons needs for bass. The fact is monopoles are simply more efficient in this area and there is no real benefit to gradient output in this region below 40-50hz, or non-mono reproduction.
But a very common misunderstanding is that gradients are less efficient than monopoles. This is false. Above F equal, they are more efficient. So it all depends on what frequency F equal is in the system.
The other issue regards both spatial reproduction and localization. Multiple subs can indeed yield smoother amplitude response, but that is limited to around an upper end of about 80-90hz, due to localization. There are lots of modal issues above this, in typical rooms, up to several hundred Hz, typically around 400hz or so.
Multiple subs are useless between 90-400hz. Most folks then resort to both EQ and/or so called "treatments", but these have there own issues.
Gradient systems can address issues in this region where monopoles by themselves, cannot.
Yes 
It's the best $125 spent in audio, if that is ones interest.

cheers,

AJ

Ah! Thanks for the eye opener. Indeed, my high efficiency Geddes bandpass/passive radiator subs (midbass modules actually) which I love work from about 40 Hz to about 100 Hz and smoothly blend into the mains (no electrical crossover, all acoustic due to the bandpass design, only electrical manipulation I have done is with phase). That being the case, I had to be careful about dealing with SBIR from 100 hz to 250Hz, so positioning of the mains, as well as having very decent room treatments around the corners near the mains and also behind the mains on the side walls made a nice difference. The midbass modules were useless in that bandwidth. This was partially because they are bandwidth limited but also because of what you stated earlier that multiple subs are less useful in that region above 100 Hz.

I will be happy to share my measurements which are a few years old, if Duke doesn't care/mind.

I would like love a demonstration one day with different recordings (that truly do have stereo bass) to illustrate to me the importance of preserving spatialization. I think I will have to read the article you referenced and take a trip down to Tampa. Unfortunately, I do not think I really own any recordings that have 'true stereo bass.'

Thank you once again for the knowledge. In fact, I feel yet another book can be written on all the vagaries of bass reproduction. It is a pity since most audiophiles' "reference" systems have pathetic bass reproduction.


Best,
Anand.

[JohnR]

Duke, thanks for clarifying the definition. Have you considered the way in which response of dipole vs monopole varies with distance from the listening position?

[Clio09]

I find this discussion interesting as I have always shied away from subs or using electronic crossovers. Living happily with my Spendor monitors or Duke's Jazz Modules in their native form. However, I have always been intrigued by the Swarm and lately I have been using a similar set up to it that was brought to my attention by Roger Modjeski of Music Reference.

Roger builds his own ESL speakers and designs them to cut off at 100 Hz. He supplements them with a pair of woofers in a bi-amped set up. His RM-10 driving the panels and a generic Class A/B solid state amp with his low pass filter built into it. The filter is 4th order Linkwitz Riley but with a passive EQ that goes to 32 Hz or lower if desired. The drivers Roger selected are designed to fit into a small enclosure where they resonate above 100 Hz. It's quite a nice set up.

For me, I went a slightly different route. I have a Beveridge RM-3 using high and low pass filters with the same 4th order L/R configuration and 32 Hz EQ. The RM-3 allows me to use my RM-10 to drive my ESL 57s and a Luxman M-02 amp that has connections for two sets of speakers to power 4 of the bass boxes. I have spread the boxes asymmetrically around the room and have to say the results are quite satisfying. I did some additional re-positioning to dial in the sound a bit more. What I haven't done is reversed the polarity on any of the bass boxes. I really don't feel any excess bass energy so I just left well enough alone. If anything I may change the EQ to push below 30 Hz.

[skvinson]

So, how does EQ figure into the strengths/weaknesses of various sub setups? ... My specific question involves using something like a miniDSP unit as an active crossover. The unit has the capability of EQ. If I EQ a full-range speaker, would that overcome the need for multiple monopole subs to smooth out the frequency? Also, would that provide a smooth bass response in one position, but not do as good a job of smoothing it out room-wide? And, would monopole or dipole subs matter if I am using EQ? ... A related question, with or without EQ, would augmenting to full-range speakers with two subs provide similar benefits to four subs?

Thanks for any input!

[AJinFLA]

Duke will have to pardon me on his forum, but I see no reason not to use your mains full range and use subs to smooth the response and increase output/headroom.
The benefit of the minidsp is it also allows delay (along with EQ, phase, etc), so that the added subs can be placed in various spatial positions that work best for whatever one is hoping to accomplish.
The key of course, mains that are as capable at LF as subs, in extension, power, inaudible levels of distortion, etc.

One of the setups mentioned in the paper are 5 fullrange speakers (no "subs").

cheers,

AJ

[skvinson]

Thanks AJ. Duke, any thoughts on this?

[Duke]

Hi skvinson,

A combination of DSP and multiple bass sources is probably better than either one alone.   Imo DSP alone is not as effective as multiple subs alone for smoothing out the modal region over a large listening area, but DSP alone can do a very good job in a small listening area. 

Personally I'd prefer only having DSP in the signal path to the subs, and not in the signal path to the mains, but that's just my personal prejudice. 

[skvinson]

Thanks Duke! ... Makes sense.

[Tyson]

Hi skvinson,

A combination of DSP and multiple bass sources is probably better than either one alone.   Imo DSP alone is not as effective as multiple subs alone for smoothing out the modal region over a large listening area, but DSP alone can do a very good job in a small listening area. 

Personally I'd prefer only having DSP in the signal path to the subs, and not in the signal path to the mains, but that's just my personal prejudice. 

I agree - multiple subs are great at evening out room modes, especially the ones created by the front/rear walls and the side walls.  Floor/ceiling tends to NOT be addressed by multiple subs, and DSP is a good option to knock that mode down. 

[rajacat]

I agree - multiple subs are great at evening out room modes, especially the ones created by the front/rear walls and the side walls.  Floor/ceiling tends to NOT be addressed by multiple subs, and DSP is a good option to knock that mode down.

Would  a few ceiling mounted and/or wall mounted subs take care of the floor/ceiling mode issue?

[goskers]

I agree - multiple subs are great at evening out room modes, especially the ones created by the front/rear walls and the side walls.  Floor/ceiling tends to NOT be addressed by multiple subs, and DSP is a good option to knock that mode down.

Why can't subs be mounted above the room centerline? 

Yes, aesthetics is a concern for some but if you are already doing multi-sub then you should do it right.  Your room is 3D so you should think about trying to energize as many room modes as possible throughout the space.  Keeping all subs on the ground is not optimum as Geddes has shown.

[poseidonsvoice]

Why can't subs be mounted above the room centerline? 

Yes, aesthetics is a concern for some but if you are already doing multi-sub then you should do it right.  Your room is 3D so you should think about trying to energize as many room modes as possible throughout the space.  Keeping all subs on the ground is not optimum as Geddes has shown.

One of my subs is about 1 foot off the floor relative to the other 2 subs. And I have a 4th one to add soon, as a result of a project that I recently collaborated with Duke...and yes, I wish one of my other subs was hanging a bit from the ceiling. That being said, the only frequency response area that I am using DSP is right around the 1st mode (i.e. below the modal region) to boost 20Hz which is expected given the lossy nature of all small rooms.

Best,
Anand.