Subwoofers: Do They Add Dimension to Music?

[watchnerd]

The Dynaudio 'Ask the Expert' series has turned into a great source of audio education for the lay person.

This episode is on subwoofers. There is a short pitch for their sub line, but overall generally applicable info to any sub:

 

[DonH56]

I did not view the video, sorry, but for decades I have been of the opinion (backed by listening and measuring) that a good sub or two (at least) adds greatly to the music. It fills in the bottom end, something present to some degree in a lot of recordings (piano hammers, plucked strings, drums of course, but also things like the subharmonics generated by multiple players and percussives by singers and many instruments), and greatly reduces distortion in the main speakers by taking the load off (assuming your sub does not add equal or more distortion, natch). I built my own servo sub back in the late 70's/early 80's and was amazed just how much everything improved, with the caveat that my Maggies -- very clean when not overdriven -- distort heavily when driven by large bass signals. Of course, there are plenty of bad subs, though with the HT revolution there are also a great many very good subs that are pretty inexpensive.

FWIWFM - Don

 

[watchnerd]

I did not view the video, sorry, but for decades I have been of the opinion (backed by listening and measuring) that a good sub or two (at least) adds greatly to the music. It fills in the bottom end, something present to some degree in a lot of recordings (piano hammers, plucked strings, drums of course, but also things like the subharmonics generated by multiple players and percussives by singers and many instruments), and greatly reduces distortion in the main speakers by taking the load off (assuming your sub does not add equal or more distortion, natch). I built my own servo sub back in the late 70's/early 80's and was amazed just how much everything improved, with the caveat that my Maggies -- very clean when not overdriven -- distort heavily when driven by large bass signals. Of course, there are plenty of bad subs, though with the HT revolution there are also a great many very good subs that are pretty inexpensive.

FWIWFM - Don

The gist of the video hits a couple of points:

1. BL is the differentiating factor between good / mediocre / bad subs

2. In addition to relieving satellites / mains from producing lows, there are reverberant cues in the low end that make the midrange sound more realistic with a good sub.

3. In a 2.1 system, placement on the edge of the sphere is best, in a multi-sub scenario, placement options are greater, as is the ability to ameliorate room nodes.

4. Tight / fast subs are easier to do with medium sized drivers due to higher BL. Giant woofer + high BL = very big box, very big motor, very expensive.

 

[DonH56]

Thanks, sometimes videos do not work for me (work blocks them, for instance). Here are my comments just to throw them out. This is not something I claim a lot of expertise upon, pretty far afield from my career path.

BL = magnetic flux times path length in a voice coil? Not sure I completely agree with (1) and (4) since there are other ways to skin the cat but as a general rule I could not disagree. This is not my area of expertise by any means. For decades my subs of choice have included servo control, starting with my own design and now commercial products from Rythmik. Having a good driver and supplementing with servo works best for me to provide good (high) output and low distortion. Only so much you can do starting with a lousy driver; the phrase "polishing a turd" comes to mind.

I think (2) is a better, more concise way of my statements about integrating the music (more or less).

On (3), when I had one sub, I found placement between the two L/R speakers worked best for me, partly due to a fairly high crossover and modest slope. This was in a two-channel music system, long before Audyssey and such. Later I designed a higher-order crossover and was able to set it lower and placement became less an issue. My current system utilizes four subs in a mono swarm of sorts, with the front two operating as a single pair and rear another pair, phase/time aligned as two pairs, then I let Dirac Live correct the entire quartet once I had the two pairs in synch.

My 0.000001 cent (microcent) - Don

 

[oivavoi]

Very interesting!

First time I hear about BL in subs. Thanks.

How does this relate to the Q factor? I assumed that was the most accurate way of describing how fast a subwoofer is.

I'm also wondering why there are so few sub manufacturers who provide Q numbers for their subs. One exception is BK electronics, who claim that the affordable XLS200 has a Q factor of less than 0.5. I think that's very fast and tight.

 

[DonH56]

Q relates to bandwidth and damping. Low Q and "fast" do not necessarily go together; high-Q systems that are underdamped typically have faster transient response (attack) but the trade is more ringing after the attack.

Here is the basic theory: https://en.wikipedia.org/wiki/Q_factor

And this article is probably easier to follow with pictures showing the response: http://www.electronics-tutorials.ws/filter/second-order-filters.html
Note Q is related to the inverse of the damping factor; an overdamped system has lower Q, under-damped higher Q.

Hopefully others have links to articles with better lay presentations...

There is also this thread here: https://audiosciencereview.com/foru...-damping-factor-and-why-is-it-important.1257/

I suspect the difficulty in explaining Q and the trades involved to most folk are part of the reason manufacturers rarely publish their data. Some subs, like my Rythmiks, do include some form of damping (Q) control.

HTH - Don

 

[oivavoi]

Q relates to bandwidth and damping. Low Q and "fast" do not necessarily go together; high-Q systems that are underdamped typically have faster transient response (attack) but the trade is more ringing after the attack.

Here is the basic theory: https://en.wikipedia.org/wiki/Q_factor

And this article is probably easier to follow with pictures showing the response: http://www.electronics-tutorials.ws/filter/second-order-filters.html
Note Q is related to the inverse of the damping factor; an overdamped system has lower Q, under-damped higher Q.

Hopefully others have links to articles with better lay presentations...

There is also this thread here: https://audiosciencereview.com/foru...-damping-factor-and-why-is-it-important.1257/

I suspect the difficulty in explaining Q and the trades involved to most folk are part of the reason manufacturers rarely publish their data. Some subs, like my Rythmiks, do include some form of damping (Q) control.

HTH - Don

Thanks Don! I need to read up on this more, I think. Yes, this is complicated material.

But let's keep it simple: What is the best practical advice for getting "fast" bass from subwoofers? I had kind of assumed this was partly an audiophile myth, but it seems the Dynaudio engineer thinks there's something to it. I assume a big part is managing room modes and bass overhang in the room. But beyond that?

 

[svart-hvitt]

Thanks Don! I need to read up on this more, I think. Yes, this is complicated material.

But let's keep it simple: What is the best practical advice for getting "fast" bass from subwoofers? I had kind of assumed this was partly an audiophile myth, but it seems the Dynaudio engineer thinks there's something to it. I assume a big part is managing room modes and bass overhang in the room. But beyond that?

Scientists (at least the social «scientists») like metaphors. So when it comes to subwoofers and agility (which is «fast»), I think of the mouse and the elephant. The mouse is much more agile and faster. So a mouse subwoofer must be your choice if «fast» is your preference.

Irony aside; I think this argument is widespread in audiophile communities. On the other hand, I think they define «mouse» as 8 inches or bigger. So they apply a rule of thumb (8 or bigger) on top of the mouse-elephant-theory. Which means bigger than 12 is too big, I reckon...

 

[Brad]

Any idea what number would be considered large BL and small BL?

 

[Cosmik]

I presume that if all the drivers in the system are active, phase-corrected and time-aligned with DSP then the notion of 'fast' and 'slow' is rendered moot - the system is behaving as it should: a single neutral driver. But maybe that would be too easy..!

The only drawback (for some applications) is the latency involved.

 

[watchnerd]

I presume that if all the drivers in the system are active, phase-corrected and time-aligned with DSP then the notion of 'fast' and 'slow' is rendered moot - the system is behaving as it should: a single neutral driver. But maybe that would be too easy..!

The only drawback (for some applications) is the latency involved.

If "fast" is described as the ability of the amp to control the woofer better, this is an electromechanical issue. I don't see how DSP helps that.

 

[Fitzcaraldo215]

Thanks Don! I need to read up on this more, I think. Yes, this is complicated material.

But let's keep it simple: What is the best practical advice for getting "fast" bass from subwoofers? I had kind of assumed this was partly an audiophile myth, but it seems the Dynaudio engineer thinks there's something to it. I assume a big part is managing room modes and bass overhang in the room. But beyond that?

I think fast or slow woofers are largely an audiophile myth spawned by speculation by creative, subjective magazine reviewers. I think they were indeed hearing room modes and bass overhang because they were oblivious to the details of room acoustics and how to measure and fix it.

I always liked Peter Aczel's quip that if woofers were fast, they would be tweeters.

 

[DonH56]

My "audiophile" definition of "fast subwoofer" stems from decades ago when I was puzzled how something that cuts off over 50~100 Hz or so could be considered "fast". I did speak with a few reviewers at the time (mainly Audio and TAS). My conclusion at the time, and I believe valid today, is that ringing is more impactful than the initial attack. In other words, the "fast" nomenclature applied to a sub that did not "muddy" the bass with an underdamped (or underdampened, depending on how you were weaned) response. That was fairly easy to measure and to hear and is one reason my initial sub design was servo-controlled. So, a sub that exhibits excessive ringing sounds "slow". But, of course, "excessive" is subjective, and some folk like the sound when the sub fils in the space between notes or drum licks.

FWIWFM - Don

 

[Cosmik]

If "fast" is described as the ability of the amp to control the woofer better, this is an electromechanical issue. I don't see how DSP helps that.

I think it is more a question of the woofer's (and attendant crossover filter's) inherent phase shift effectively smearing the transient out over time. This is a result of its electromechanical, physical characteristics, and the amp can do nothing about it beyond behaving properly i.e. duplicating the input signal with gain. In order to correct the smear, the signal itself needs to be changed. DSP, with latency, based on measurements of the driver's output, has the ability to pull the smear into the ideal shape. Time alignment delays can then align the impulse with the other drivers (which should also be corrected with DSP) to give the ideal, neutral impulse. This is just the same thing as the ideal step response being discussed in another thread.

The contrast between audiophiles' acceptance of non-DSP'ed drivers and the credence given to MQA is just one of many paradoxes in the world of audio!

 

[Wombat]

Way back when I studied Control Systems at university the term 'fast' was used to describe the rise-time of the output response of an underdamped system.

I don't see a common definition of the term 'fast' as used by the subjective opiners. The audiophile language is full of terms mal-appropriated from the practical sciences and this makes discussion moot.

 

[dallasjustice]

If you compared an impulse of ANY subwoofer crossed over at say 80hz to ANY tweeter impulse, it is clear that there is no such thing as a “fast” subwoofer.

IMO, a “fast” subwoofer could only mean a sub woofer with very little group delay.

The best descriptor I can think of to describe proper subwoofer integration would be invisible. If someone walks into the room and listens to a wide variety of music for an extended period of time without suspecting a subwoofer was playing, the integration is good.

Here is a good song to test whether subs are properly integrated:

 

[Brad]

Yes, I agree when the sub(s) don’t draw attention to themselves, the question of fast or slow is moot.

 

[watchnerd]

I think it is more a question of the woofer's (and attendant crossover filter's) inherent phase shift effectively smearing the transient out over time. This is a result of its electromechanical, physical characteristics, and the amp can do nothing about it beyond behaving properly i.e. duplicating the input signal with gain. In order to correct the smear, the signal itself needs to be changed. DSP, with latency, based on measurements of the driver's output, has the ability to pull the smear into the ideal shape. Time alignment delays can then align the impulse with the other drivers (which should also be corrected with DSP) to give the ideal, neutral impulse. This is just the same thing as the ideal step response being discussed in another thread.

The contrast between audiophiles' acceptance of non-DSP'ed drivers and the credence given to MQA is just one of many paradoxes in the world of audio!

Up to a point...if the driver is exhibiting ringing / overhang, it's not going to duplicate the electrical signal.

 

[Cosmik]

Up to a point...if the driver is exhibiting ringing / overhang, it's not going to duplicate the electrical signal.

Well, yes. Hopefully it's behaving itself in that regard. Another reason for it to be active, direct coupled to the amp.

 

[Fitzcaraldo215]

Up to a point...if the driver is exhibiting ringing / overhang, it's not going to duplicate the electrical signal.

Agreed, but I doubt that is a big problem with most decent subs these days. Yes, some low budget, cheapies might have something of a problem, also including limited LF extension and dynamics.

The big and persistent problem is the room and modal non-linearities, ringing/overhang, as measurements will easily show in most any room anywhere. DSP and multiple subs to the rescue!