What makes a sub musical? (1 Viewer)

[Max F]

Someone mentioned that a good "musical" sub makes all the difference in listening to concert DVDs.

So what makes a good musical sub?

I'm trying to figure that out. I figured, so far, that a sealed sub has, overall, lower group delay (i.e., tighter bass). I noticed that my Dayton 10" gets pretty fat when i turn it up. I really notice it when it comes to bass drums - the sound lingers too long - THuummp instead of THUMP!. Thats what it sound like even though the frequency is flat (relatively for my room) from 25-100hz. I turn it down for a more natural sound (to my ears).

Can a ported sub be a "musical" sub? Kindof a loaded question, isn't it:b SVS guys feel free can chime in as i may buy a SVS sub in the future (as a ported musical sub).

 

[Mark Seaton]

There are many, many intertwined factors that go into what we consider to be natural sounding. Frequency response is largely the dominant factor. Other factors include, but are not limited to dynamic changes in the response (see Keith Yates's subwoofer report in Ultimate AV), the spectrum of the distortion created, and most importantly, how the sub is integrated with your main speakers.

That last factor is largely ignored in home theaters, and is most certainly intertwined with frequency response. It is not just the response of your subwoofer which matters, but also how it integrates with your LCR. I do in fact believe that group delay plays some part, yet I think it matters most through the top end of the subwoofer and where it mates with the main speaker. If I were to measure most subwoofers in most home theaters, I know we would find that in most cases, the sub is "late to the party" relative to the center channel or R/L. We can adjust for this with the distance settings for our subwoofers. Note that this affects frequency response as well since a time delay results in an equivalent phase shift at some frequency. In reality you can adjust the time to make the phase match at or through the crossover range, without much need for a phase knob (but it can be used to tweak further.

 

[Max F]

I think this is what i'm getting at:

Subwoofer group delay (document found on adireaudio.com's website which appears to be down now). For the vented (ported) sub the group delay increases as the freq. decreases. The group delay of the example vented (ported) subwoofer peaks at 28ms at 15hz. Would a 28ms delay be audible? Would this explain why a bass drum doesn't sound right when i turn up my sub (to a flat freq. response to 25 hz)?

The group delay for the sealed sub (in the above document) peaks at 6 ms, but, overall does not extend as deeply as the ported sub.

 

[ScottCHI]

i've read that a "musical" sub can reproduce with accuracy the frequencies that are well above the crossover setting; more than an octave above. for example, even with the usual 80Hz crossover and it's corresponding slope (is it 6 or 12dB/octave? can't remember), a sub still has to reproduce a significant amount of output an octave above the crossover, at 160Hz. and even at 320Hz, 2 octaves from the crossover in this example, there can still be significant output that contributes to the sub's "musicality".

 

[DavidLW]

Speed is the other quality that "muscial" subs have. Most "muscial" subs are built with 8" or 10" drivers. Not too often with 12" and rarely with 15" drivers. The less the mass the faster the drvier can move. If more output is needed then they use multiple 8" or 10" drivers per enclosure. High End "musical" subs are nearly always servo-control. That is they have a built in sensor on the voice coil that senses the movement of the cone and compares it to the amps output. It corrects for distortion by altering the amp output. The movement of the cone is non-linear in that the wattage it takes to move the cone increases as the cone is away from the middle of the voice coil. This is due to the elastic nature of the driver's suspension. The servo-control sensors corrects for this by increasing the oputput when needed. It also knows when the cone is about to bottom out or over extend. When the sensors senses this they immediately throttle back the amp. This does two things 1) it prevents the distortion caused by the cone hitting bottom or leaving the voice coil and 2) the amp can provide more wattage immediately without the fear of having the cone distorting because it reached it's limit of travel. This in effect makes the driver respond quicker to the music and thus "speeds up" the driver. You may lose the very bottom end but muscially it's more important to keep up with the music. Besides the only really, really low notes in music is from an organ or synthesizer. Nearly everything else is above 30hz.

High End "Musical" subs also tends to be expensive. My Entecs (out of business) are High end, servo-control subs with a built in 175 watt amp that cost nearly $3000.00 a pair (with electronic Xover). They are great with my Martin Logan CLS2z but a pair of $400.00 Dayton or SVS would be better for HT than these puppies. In fact a home built sub with a 100w plate amp would probably be more impressive than most "musical" subs. "Musical" subs are designed to bend in with the music. You want them to disappear and not distract you from the music. Which is opposite of what you want from a HT sub. With HT the lower and longer the bass goes, the better. It doesn't matter that the bass no longer blends in with what's on the screen. But with music, the bass must naturally blend in with the rest of the music or else it doesn't sound right. With "musical" subs it more about quality than quanity. The more expensive M&K and Velodyne are servo-control. These companies were making subs for music way before the HT craze hit. Martin Logan also makes very "musical" subs. These would be my first choice if my Entecs ever bites the dust. That being said Hsu makes some pretty good "musical" subs that aren't servo but doesn't cost an arm and a leg either.

 

[Chu Gai]

Well, 40 hZ is 40 hZ whether its being reproduced with 8" or with 12". The driver must still vibrate at the same speed. However with a smaller surface area, it must by necessity travel further to achieve the same SPL which is why you'll need more small drivers simply to equal the surface area of the larger one. Further, smaller drivers just don't have the xmax so they're quite limited as to how loud they can play.

Whenever I see the word musical subs, it sends up flags and alarms in my head. Generally it means it doesn't go down very far, is incapable of clean, undistorted, SPL's, and I can be sure of a lot of bullsh*t from whomever is marketing it. As far as blending goes, that's pretty much a function of setting the crossovers and slopes appropriately and that holds true regardless of what kind of sub you choose. You'll often hear terms like a tight sub which is really another way of saying the sub is incapable of reproducing the lower frequencies at the levels they were intended to be. IOW, the lower end has been truncated. Another thing you generally won't see with musical subs are comprehensive measurements indicating their performance capabilities. I think we've all heard enough stories about how fast their cars are. Of course, if you never put them on the track then you can always rely on the stories instead of the performance.

 

[Edward J M]

Transient response is a function of upper frequency response. If the subwoofer exhibits a flat frequency response to say 150 Hz, then it has adequate transient response for almost any HT or music application. The low pass filter in the pre/pro bass management circuit is typically the limiting factor in a subwoofer's transient response, not the driver itself.

The vented alignment will typically exhibit a higher GD than a sealed alignment (however the addition of a HPF in the sealed sub can add GD). But that point may simply be academic if the differences cannot be perceived by the listener; it really boils down to where this occurs in the bandwidth.

The audibility of group delay is not the question; it IS audible. The question is at what threshold and frequencies does it become audible. According to speaker and TrueRTA designer John Murphy (of True Audio) one of the most widely referenced reports of the audibility of group delay is:

Blauert, J. and Laws, P "Group Delay Distortions in Electroacoustical Systems"
Journal of the Acoustical Society of America
Volume 63, Number 5, pp. 1478-1483 (May 1978)

Blauert and Laws report approximately the following thresholds for audibility:

Frequency Threshold of Audibility
8 kHz 2 msec
4 kHz 1.5 msec
2 kHz 1 msec
1 kHz 2 msec
500 Hz 3.2 msec

According to John, the data available on group delay audibility shows a threshold of audibility of 3.2 ms at 500 Hz and increasing about half again for every octave of falling frequency. John extrapolates and finds 6.4 ms at 200 Hz, 12.8ms at 100 Hz, 25ms at 50 Hz and around 50 ms at 20 Hz. Admittedly this is an extrapolation, but John feels it should not be in gross error. This would suggest that both the deep tune vented and closed alignments are both well below the approximate audibility thresholds of 25-50 ms in the 50-20 Hz bandwidth.

Talk with Tom Vodhanel (of SV subwoofer fame) and he'll point out the major problem with trying to determine the audibility of ONE factor (like GD) is isolating that variable and holding all others (like FR, THD, IMD, etc.) constant.

Study the writing of people like Mark Seaton, Keith Yates, Tom Nousaine and Tom Vodhanel and you'll find a common theme about their views on subwoofer design and performance - focus on the big five: flat frequency response, deep extension, low distortion, high bandwidth linearity, and high dynamic compression limits - as these factors affect our subjective perceptions of subwoofer performance to a much greater extent than other factors like the difference between 8 ms and 20 ms of GD at 20 Hz or the difference between a 2nd order and a 4th order impulse response. This is not to diminish the importance of GD and IR, but rather to merely rank them in the overall subwoofer performance food chain.

According to Floyd Toole (of Harman International): "It turns out that, within very generous tolerances, humans are insensitive to phase shifts. Under carefully contrived circumstances, special signals auditioned in anechoic conditions, or through headphones, people have heard slight differences. However, even these limited results have failed to provide clear evidence of a 'preference' for a lack of phase shift. When auditioned in real rooms, these differences disappear..."

Toole, Floyd E., "The Acoustics and Psychoacoustics of Loudspeakers and Rooms - The Stereo Past and the Multichannel Future," 109th AES Conv., Los Angeles, Sept 2000.

It's ironic, some of the biggest proponents of sealed subwoofers (who are always the first to point out group delay) are the same people who typically own vented loudspeakers. The vented loudspeaker will exhibit its group delay much higher in the audible frequency spectrum than will any well designed vented subwoofer. And yet you never hear owners of vented loudspeakers complaining of "time smear" or "slow mid-bass" despite the fact that the typical vented loudspeaker has a resonator tuning point in the 40-50 Hz bandwidth.

Regards,

Ed

 

[Jack Gilvey]

Really?

 

[Max F]

Excellent answer Ed, thank you!

You state that GD using the numbers we are talking about for a ported design (say 17 ms at 30 hz or 25 ms at 25 hz, for example) at low frequencies are not noticeable even though this is based on extrapolation (as you point out). I'm not sure how this was tested, but lets say you use a test sound like an instrument that people are familiar with (e.g. kick drum). Wouldn't you suspect that the human ear would be much more sensitive to GD for those familiar sounds? I think the human ear is dead on for instruments we have heard all of our lives. Contrast this with fake explosions on the movie screen in which GD is no problem. Honestly, i don't know if GD is a big factor or not when comparing subs, just throwing out the thought. Seems like a reasonable argument.

I don't have a very good sub like a SVS so lets take my little Dayton 10" as an example. So if my cheap Dayton doesn't sound right reproducing the bass (kick) drum and its not due to group delay (which is dependent on the box), what parameters would explain this phenomena assuming that i have a relatively flat frequency response in combination with my front speakers (i.e., the flat frequency response sound too fat to me).

 

[Edward J M]

As Mark stated above, there are many factors that could contribute to your subjective perception that the Dayton 10" doesn't produce the sound of a kick drum properly.

You state that when you turn it up loud, it gets "fat". The sub could be creating audible distortion and compression artifacts, and the sound signature of the subwoofer will definitely change when this occurs.

The frequency response of the sub might also change under a high dynamic load as opposed to lower levels. This will also be audible.

The energy decay in the room itself might be a factor. As you turn it louder, the decay signature will also become louder and more noticeable. What you think might be subwoofer "overhang" could just be audible decay products in the room. People often notice an improvement in the bass quality after room treatment. Check out the article in Secrets where John E. Johnson treats the Secrets sound lab. He reports a substantial difference in bass sound quality before/after.

So while you are definitely hearing "something" wrong with the sound of the Dayton on a kick drum at louder volumes, it is almost assuredly not the result of group delay, even if the Dayton is equipped with a high pass filter at say 25-30 Hz (which would cause a phase shift and hence an increase in GD). The fundamental of the kick drum, and its associated harmonics (which are probably as much or MORE responsible for the sound signature than the fundamental itself) occur too high in the bandwidth for GD to ever be a genuine consideration.

Regards,

Ed

 

[Brian L]

Didn't Tom Nousane once say that if a woofer were fast it would be a tweeter?

BGL

 

[Philip Hamm]

I don't know about the numbers and theories much. I have been playing the electric bass in bands for 25 years (since junior high!). I know a musical sub when I hear it, because the bass sounds like a bass guitar through an amp. My vintage Crown powered dual 15" infinite baffle sub may not meet many of the technical criteria defined above, and it was very inexpensive in materials, but it sure as hell is the most musical sounding sub I've ever heard.

 

[dave alan]

Since no one listens to only a subwoofer, the satellite speakers and the transition between them and the sub is at the top of the list (agreeing with Mark Seaton).

Using whatever crossover scheme a person uses in his/her set up, it's easy enough to plot the mains alone, then the sub alone and calculate the slope of each, in-room. Few people do that. If the high passed speaker's slope doesn't match the low passed speaker's slope, the two halves won't sum to unity across the crossover region.

It doesn't matter what order of alignment the satellites are if they have anechoic response to 1 octave below the crossover point, because at the crossover point the group delay differences are not a factor.

The Toole statement that Edward quoted above may have been taken out of context. If not, I'm certainly confused by it. I have never met a multi-way speaker designer who believed that phase shifts in the crossover region of any 2 of the speakers are of little concern, or allowed for generous tolerances.

Distance disparity between the sub and sats can be reconciled using a delay, but delay always causes phase shift. I've consistently found that equidistant placement is the surest way to time align sub and sats, without the price of additional phase problems.

It's been pointed out many times that a sealed sub generally has a SS filter to prevent extreme excursion down low, when, in fact, the opposite is true. When a sealed sub design employs an L/T circuit, it allows for a smaller box, which prevents over excursion. OTOH, if a ported sub is designed with a very low tune to keep GD low in the music bandwidth, a SS filter is always necessary. What happens to the ported sub's GD when a SS filter is applied?

The system 'Q' is another important factor that should be included in the discussion. TV has said that if a person thinks a sub is 'fast, tight, musical', etc., that it's probably because of the sub's roll off characteristic. I agree with that, in that system damping is critical to subjective impressions.

When the discussion is headed by the word 'musical', it's assumed that music listening is the priority. If we also assume normal listening levels (we certainly aren't talking about peaks of 114 dB at the LP) and proper execution of the 2 alignments of subwoofers, we can assume similar THD and compression numbers as well.

Let's then say that we could afford to have Mark Seaton show up and adjust time arrival, volume levels, slopes and relative phase.

What's left to describe the difference you might perceive in the low frequencies of MC music playback? Would there be no difference?

Group delay certainly has to be considered, studies and extrapolations thereof aside. We probably can say with a high degree of certainty that the majority of people who think a sealed sub is tighter (or whatever) haven't, by some mass miracle, gotten all of the other points mentioned above right.

Another thought comes from Linkwitz:

"Most likely it is not the start up transient response of the speaker, because by the time we hear the signal, it has undergone many reflections and the waveform fidelity has been lost. It could well be, though, the slow decay of transients due to energy storage in resonant mechanical and acoustical structures of the speaker which we recognize as typical for a loudspeaker and missing in the corresponding live event."

I hope we don't get into which sub is louder during playback of Blackhawk Down or some similar soundtrack, but stay on the subject of music reproduction.

I like threads that chip away at this mystery, and don't have any bias for or against any alignment of sub. Nor do I have any difinitie answer, other than the logical one that I've proposed many times, that is, get the music part as right as possible (redirected bass) and then get the subsonic effects right (.1). Keep the 2 systems discrete and optimized for their respective job and don't expect any one system will do both as well as each can do it's part.

 

[Tom Vodhanel]

>>>It's been pointed out many times that a sealed sub generally has a SS filter to prevent extreme excursion down low, when, in fact, the opposite is true. When a sealed sub design employs an L/T circuit, it allows for a smaller box, which prevents over excursion.>OTOH, if a ported sub is designed with a very low tune to keep GD low in the music bandwidth, a SS filter is always necessary

 

[Martin Rendall]

Geez. You guys keep on going. You're making me start to think that the "compromise" sub I just bought was actually the better choice!

Martin.

 

[Edward J M]

I think the question that needs to be answered is which has a greater magnitude and longevity - the energy decay signature of the room itself, or the impulse response of the subwoofer. IOW, are differences in impulse response audible under normal listening conditions?

 

[Manuel Delaflor]

... what makes a sub musical??


Size of the driver? NO

Its group delay? NO

If it is ported or sealed? NO

All those factors are (relatively) irrelevant. The single most important and easily discernible (by our own ears) is:

THE ROOM

Once we have a good room there is another incredible important factor:

THE CORRECT LOCATION

And then:

ITS INTEGRATION WITH THE MAINS, in phase and frequency.


Thats about it.

 

[Max F]

Here's the answer that i was hoping for (although it may be wrong):

Your Dayton Tiny Mighty is an not a very good sub for music. It's designed to reach a relatively low frequency (25hz) using a modest size driver (10 inch). The ported design of the box was meant to help achieve this, but this gives a high Qtc resulting in audible group delay (i.e,. not accurate and tight).

Ported subs in general have higher Qtc and faster fall-off than sealed subs resulting in higher group delay numbers. This does not mean that a ported sub cannot be "accurate and tight" subs as long as these values are not too high. The document from Adire Audio mentions a good rule of thumb of less than 25 ms at 20 hz. So maybe the SVS and other quality ported subs have designs that effectively deal with GD? We're still talking music here.

Heres the document (just add www): ***.adireaudio.com/Files/TechPapers/GroupDelay.pdf

 

[Max F]

Manuel,

Based on that answer, my Dayton Tiny Mighty should have no problem sounding just as musical as say UFW-10 subwoofer, as long as the room is good.

I know that can't be right.

 

[Brian L]

Ed,

I have seen pictures of you in action out in the back yard...test gear at hand, sub at the ready.

I bet YOU can answer that question for us!

BGL