What makes a sub musical? (1 Viewer)

[Chu Gai]

A discussion of Group Delay in speakers can be found at http://www.trueaudio.com/post_010.htm for further consideration.

 

[Jack Gilvey]

Indeed. And these same designers apparently attach more import to the results of such tests, and their usefulness in predicting the "sound" of a sub, than many hobbyist/amateurs do. What do they know?



Another good GD discussion:

http://www.trueaudio.com/basslst2.htm

 

[Max F]

OK, i think i'm getting it. It really is very helpful getting advise from folks with as much experience as you'all.

I haven't yet read the most recent links but i did get to the essay on "How a Hole-in-the-Box Works - A Big Dig into Bass Reflex" which was excellent.

Here's a quote that helped alot:
"The phase response, which is sometimes referred to as transient response, will be slightly worse at cutoff in a ported system than in a sealed system. This can result in "slower," or more accurately, "more delayed" sounding bass. But, at the same time, since it will go deeper than a sealed system with the same efficiency and cabinet size, and tends to use a more damped driver, the larger phase error at the tuned frequency may be less noticeable if it can go deep enough. So, subjectively, it could then sound faster."

From this and other peoples observations (and testing) i can assume that a high quality ported system (like SVS) can produce "fast, tight" bass using a better design of the box (i.e., lower Qtc, lowered tuned freq of the port) and a better driver (more damped, lower tuned frequency, and other parameters of a quality driver) than my cheap, but good for the price, Dayton Tiny Mighty.

This of course assumes that the room is adequate for a relatively flat frequency reponse and one that does't produce low freq. sound decay issues.

So now i just need 400-500 bucks for the new 10" SVS sub thats coming out

I would be curious how this sub will compare to the UFW-10 sub for music reproduction - but maybe that just sticking my foot into the poo.

 

[Mark Seaton]

Boy is that discussion a flashback! The good 'ol days of "The DIY BASS List."

The "Tom" referred to at the end of that is Tom Danley, our designer (incidentally, that list-serve is how I met Tom Danley, which lead me to where I am now). Knowing Tom pretty well, his later comment should have had further detail:

Tom has always worked with the TEF measurement system which makes it pretty easy to remove what Dick Heyser termed the "fixed delay." Once this is removed, what remains represents the phase characteristic of the device. The least confusing means to look at this is to also consider group delay, where the goal is to keep group delay as constant across the frequency range as possible. The only problem with group delay is that 1ms at 6kHz is much more significant than 1ms at 60Hz. A view of phase response helps keep this in perspective.

dave alan earlier mentioned keeping the subs the same distance from the listener as the mains in the interest of alignment. While this makes the figuring a little easier, in this setup, the subwoofer is almost always "late" at the crossover. Not always, but most of the time. In those rare cases where things line up, we always find that the main speaker has a significant increase in group delay at it's low end. While the filter type and frequency has a more significant effect, most any of the DSP based filters will add ~= 1ft. of additional delay to the subwoofer.

Remember that with our main speakers the delay/distance settings are helping with the precedence effect and keeping the relative arrival times appropriate to ensure porper imaging. The only reason for delay adjustment on the subwoofer is to adjust the arrival to be proper relative to the mains. The flaw here is that the physical distance does not typically match the required acoustic delay. I generally use the center channel as a reference for the subwoofer, and make adjustments to get everything else reasonably correct as possible (my dad always told me it is better to be reasonably correct than precisely wrong). A nice trick here when you can't take hyper detailed measurements is to use a >1/12th octave RTA and after you have a reasonable integration flip the polartity or shift phase by 180 degrees on the sub or the speaker you are testing (only one of them). Now adjust the delay until you see the greatest notch in the response. The notch should reduce if you go to far or are not far enough. With the greatest dip setting now flip the polarity back to normal, and see what the response is. You may have a slight peak, where you can then use other means to fine tune. This is NOT the same as making real measurements, but it will get you much closer. In your adjustments, know that you will be closest by increasing the subwoofer distance (or reducing the declared distance of the main speaker.

 

[Max F]

OK, i just read the latest links to some great discussions.

The coolest thing i figured out is that the lower the tuning frequency of a ported sub the lower the group delay at musically important low freq. (e.g., 30-50 hz). That means those big badass SVS subs that are tuned at

 

[Edward J M]

Mark:

I know we have briefly discussed this subject in PM, but (for the benefit of others) in Tom Danley's and your experience, are the differences in impulse response (energy decay) between the sealed and vented subwoofer audible under normal listening conditions? How does the Servo-Drive system or the B-DEAP horn compare in impulse response to a more conventional subwoofer alignment (sealed/vented/PR).

Also, I would imagine the IB alignment would exhibit a worse impulse response than the 2nd order sealed, because there is no enclosure air spring to help damp the woofer? IOW, how well can a woofer damp itself without the help of the enclosure? Can the lack of the enclosure air spring be compensated for by adjusting the T/S parameters and/or suspension design, etc.? TIA.

Regards,

Ed

 

[Mark Seaton]

Let's address these in reverse order and in two separate posts:

The point to understand here is that the air spring in a sealed box is modeled as being in parallel with that of the driver itself. Basically they are combine to one lumped value, not two independent values. This means that there is no difference in a sealed box with a driver-air combined spring constant of X and an infinite baffle design where the driver's own spring constant is X. In a real system there is some variation of this constant with position, just as there is variation of motor strength. This is related to the Cms/Kms curves we see plotted in the DUMAX reports. An air spring is highly linear to a point of significant pressure change where the volume displaced is more than 5-10% of the total enclosure volume. In earlier Acoustic Suspension designs distortion was significantly reduced in-box as the sizes were such that the more linear air-spring largely dominated the spring force.

Short answer? In an IB, the linearity of the suspension is more important than in an acoustic suspension system. Arguments for different cases can be made for and against either system, it all depends on how they are used and at what operating levels.

Finishing up, remember that the restoring force from a driver or the confined air in a box is a spring constant. Thinking in terms of "damping something out" will mislead you. It is a matter of having the right amount of damping to compliment all of the other factors, not a matter of absolute damping. This is a common misconception in the audio community.

 

[DavidLW]

I know F=MA is a way over simplification of how the dynamics of a electro-magnetic driver works. I was just trying to convey in a simple way how servo-control can alter the moving dynamics of the driver by controling the force.

But...........

What's wrong with my logic here. I hook up two identical self powered subwoofers to the same source. Now it doesn't matter what type of enclosure, size of driver, Q, or power of the amp. The only requirement is that they are identical and do not have any servo-control circuitry. So that room acoustic plays no part in the result I place them outdoors. I position the subs so that I can see the drivers of both sub at the same time. I play a test cd so that a 20hz signal is sent to the subs. Now I turn up the amp of each sub so that each sub is playing at exactly the same volume. I look at the both drivers and they should be moving in and out in unison. Another words, both driver should be at their center point at the same time. And both driver should extend out and go in from center at the same time and the same distance. Let's say for convience sake the driver goes in or extend out from center 1/8". Another words it's total excursion is 1/4". I now turn up the volume of one of the sub so that it's noticablly and measurablly louder not subjectively louder. Now correct me if I'm wrong but the loudness of a driver is based on how much air it moves. The more air it moves the louder it sounds. Right? So there's only two ways for a driver to move more air. You got to either increase the square area of the driver or increase it's excursion. I take a tape measure and measure the diameter of the driver of the softer playing sub and then I measure the diameter of the louder sub driver. Sure enough they're still the same size. This means that the louder sub must be exhibiting more excursion. If I measure the excursion it should be more than the softer playing sub. Right? Now, again for conveince sake let's say that the louder sub is now moving in and out from center 1/4". Another words twice the distance of the softer playing sub. Now it gets tricky and once again correct me if I'm wrong. If I step back and look at the drivers of both subs they should still be moving in and out in unison. Right? Another words when the softer driver is at center so is the louder sub. When the softer sub goes in 1/8" the louder sub has to go in 1/4" to remain in unison. And when the softer sub travels back to center and then extend out by 1/8", the louder sub travels back to center and has to extend out by 1/4". So the louder sub driver is traveling twice the distance of the softer sub driver in the same amount of time. This means that the louder driver is traveling faster than the softer driver. Right? So how is this NOT related to F=MA. And like you said, by increasing the mass of the louder driver I can make it the same volume as the softer driver. And again correct me if there's a flaw in my logic. When both drivers are again at the same volume they will once again have the same excursion because they must be moving the same amount of air, even though one driver has more mass but also more force.

If there's no flaw in my logic then the only way that F=MA does not play any part in the reproduction of sound with a driver is if I did the same experiment in the vacuum of space.

edit typo to read ".....does NOT play...." in last statement.

 

[Mark Seaton]

Part 2:

I wouldn't say that I am certain, more like a strong suspicion, that there is a difference when the phase response through the operating band is correct. A direct radiator operates in "acceleration mode" which is why we need 4x the displacement to maintain SPL at 1/2 the frequency. An ideal, efficient horn operates in a velocity mode where 1/2 the frequency will require 2x the excursion to maintain the same SPL. Real horns we build only follow this behavior over limited bandwidths, but the trend and reduced excursion requirements are most certainly there. Also evident is the 90 degree phase shift between the two modes of operation. A conventional subwoofer has a general -90 degree phase shift through the passband (range of flat response). An efficient horn can be seen to have either a zero or 180 degree phase angle through its passband. Accordingly, the direct radiator will have an increasing group delay with lowering frequency (90 deg at 50Hz is twice as long a time as 90 deg at 100Hz). In a horn with toleably flat response we see a more constant group delay with lowering frequency, down to the frequency where the loading is lost and/or the response falls off.

What most forget is that the room greatly affects the shape of the low frequency response we hear. I personally suspect that after the impact of the upper frequency response as observed at the listening position, we do identify the "shape" of the low frequency roll off. As some of the links show, a sharper corner with a steep roll off results in more group delay and phase shift. An overdamped (shallow roll-off) or a more gradual roll-off results in less group delay. A lower corner frequency pushes the onset of this delay lower which makes a better case for critical ranges. With a very low tuned reflex design or a sealed system with EQ you can achieve very gradual roll-off shapes to very low frequencies, especially in well confined rooms.

For those still following along, remember that observation without knowledge of the conditions opens us up to coincidence rather than conclusions. Most people have never looked at the response they are listening to, nor have they looked at the spectral content of their source material.

 

[MingL]

The very sad fact is that many people don't understand that music is made up of many spectral components. And some are even worse that they refuse to understand the basics of what they are hearing.

 

[Mark Seaton]

In your original post you used your description of F=MA to attribute factors of "speed" to different drivers. The absolute value of Acceleration is tied to volume and nothing else. We only care that the acceleration doubles when we ask it to, and the value is large enough to produce the SPL we require.

To add some points of reference, your example of 2x the excursion from the louder driver would be observed as a 6dB increase in SPL. Now consider this one...

Using your own example, let's take one of your woofers moving 1/4" p-p delivering some output X. Now let's bring in another woofer (let's say they are both sealed to keep things simple). This woofer has exactly 1/2 the surface area of your woofer. To produce the same SPL at the same frequency, this woofer must move twice as far, or 1/2" p-p. Since they both cross the static zero position at the same time, we see that the smaller woofer has to move twice as far to produce the same level. Larger drivers have to move fractions the distance of a small driver, which increases the likelihood that acceleration will in fact remain constant through it's operating range, and will continue to increase as we want it to when asked.

 

[ScottCHI]

huh? that doesn't make too much sense. could you restate that? F=MA in space, too, btw.

 

[Sam A]

davidLW-
"So the louder sub driver is traveling twice the distance of the softer sub driver in the same amount of time. This means that the louder driver is traveling faster than the softer driver. Right? So how is this NOT related to F=MA. And like you said, by increasing the mass of the louder driver I can make it the same volume as the softer driver. And again correct me if there's a flaw in my logic. When both drivers are again at the same volume they will once again have the same excursion because they must be moving the same amount of air, even though one driver has more mass but also more force."

once you change the mass of one driver, the subs are no longer equal to each other. one is more efficient than the other [the one with more mass] so yes if you added enough mass to the driverfor it to be less eficient enough for its volume to match the other sub, then you are correct.

"If there's no flaw in my logic then the only way that F=MA does play any part in the reproduction of sound with a driver is if I did the same experiment in the vacuum of space" um, first off, THERE WOULDNT BE ANY SOUND if this experiment was conducted in a vacuum, and SCOTTCHI was right, F=MA doesnt just apply to earth's surface, its applied everywhere. so with those two facts, F=MA is still a factor in the A vs B sub experiment above. the new higher massed driver [A] is using more force [mass was increased] than the original driver even though they are playing at the same volume.

 

[Manuel Delaflor]

IMO, in this discussions the car analogy works well. Suppose you have a small cheap domestic that happens to be very lightweight. The analogy runs when it is assumed that the small car will always accelerate and stop faster than a heavier car (assuming both can do 20MPH) because of their mass.

Of course thats wrong.

 

[Martin Rendall]

Manuel Delaflor,

Does the cheap domestic car have a killer aftermarket sound system with a huge subwoofer? If so, what's the Q of that subwoofer, and is it vented or sealed?

Martin.

 

[frank manrique]

Wowee! What a thread!...

For those of you who have considered this one to be a waste of time and energy...well, just take a second, careful look at the exchange between Ed Muller and Mark Seton; simply brilliant!

That sort of thing sure causes me to feel awfully humble spite my practical experiences, let me tell you... :b

-THTS

"...hi, my name is Frank...and am an SVS bassaholic..."

 

[frank manrique]

quote:
__________________________________________________ _______

They used to take cage birds down in mines to warn of impending danger frank!
__________________________________________________ _______

Yea...but you think the French scientists would have done so too back when they were developing those infrasonic weapons of innards mass destruction!

By the way...did check on that poor feathery critter and he/she seems OK,if still a bit shaken by the experience...at least for now anyway!

Best regards...

-THTS

"...hi, my name is Frank...and am an SVS bassaholic..."

 

[DavidLW]

OOPS- my mistake. I meant to type "...does NOT play...."

I'd edited it. :b

Of course I knew that there would be no sound produced in the vacuum of space. Therefore the law of F=MA (which stills applies) would have no relation to the sound produce because there would be no sound produced.

The confusion seems to stem from some of us thinking that the "speed" of the sub is actually refering to how fast the woofer driver is moving when most us (including me) are refering to how fast the sub respond to signal changes.

 

[Edward J M]

Great posts, Mark. Thanks so much for taking the time to share your knowledge and expertise. So if the woofer in a vented enclosure can be just as effectively damped as a woofer in a sealed enclosure, is the difference in impulse response demonstrated by these two alignments strictly attributable to the Helmholtz resonator in the vented subwoofer? Is it the stored energy in the resonator that takes more time to decay, and not the woofer itself?

Your "Part 2" post (while very informative) seemed to speak more to phase anomalies and GD than to impulse response. Are the two issues necessarily related? TIA.

Regards,

Ed

 

[dave alan]

Yes, IMO, the 2 issues are necessarily related.

The delaying of the decay process, whether by mechanical storing/late release or group delay, has the same audible result. When the 2 combine the effect is compounded.