Bessel vs Critically Damped Enclosure

Does anyone have experience listening to the difference between critically damped (Qtc 0.5) and Bessel (Qtc 0.577) sealed enclosures?

The 'tightness' and 'speed' are noticeably better with lower Qtc enclosures, with systems over Qtc 1.0 not that uncommon, but sound like mud to me.

Will a Qtc 0.577 sound just as tight as Qtc 0.5? If Group Delay is the reason for the muddiness, 0.577 might even sound better.

With a high qts subwoofer, this makes a huge difference in enclosure size. e.g. with the Wavecor SW312WA03, speakerboxlite calculates 236 liters for Critically Damped and 85 liters for Bessel. This is using the 'unbroken in' manufacture numbers, I'd expect qts to drop, and even if VAS goes up, box size comes down. But this is just an example of how big a difference there can be between them for enclosure size.

Obviously extension is also an issue, but I don't want to go into that in this thread.
 
frugal-phile™
Joined 2001
Paid Member
I go for the Bessel alignment if i can afford the room over something closer to Butterworth, By the time you add damping you are dropping the Q to close enuff…

And take your comment about “unbroken in T/S” into context. The factory numbers are taken off of the T/S parameter curves in a more useful spot than what we usually get when we measure them.

dave
 
I go for the Bessel alignment if i can afford the room over something closer to Butterworth, By the time you add damping you are dropping the Q to close enuff…

I meant the measured Q of the built system, not the calculated Q, which can be different.


And take your comment about “unbroken in T/S” into context. The factory numbers are taken off of the T/S parameter curves in a more useful spot than what we usually get when we measure them.

dave

I am not sure what you mean by this. Wavecor specifically says their measurements are before break in. Why they would do this is beyond me, but it is a safe assumption Q will drop, along with Fs and VAS will go up. Usually these offset each other, but not when Q is high in free air and the final system is low Q.

I picked this woofer as an example of something where 0.577 and 0.5 resulted in a dramatically different enclosure. I may or may not actually use it in my build.
 
System Q influences both time domain, which is buried in the reverberation of the room (so it cannot be examined anymore), and frequency domain, which is altered by the room (so a different system Q in the same round still sounds different). So if you compare different system Q's, you are comparing their frequency domain performance.

Lots of things will make far more of a difference than than a Qtc of 0.577 versus a Qtc of 0.5. I was just wondering if anyone reading this had a chance to compare both alignments.
A comparison can be done. Lower Q means less bass. If you happen to have a room which amplifies bass, this bass reduction can be a good thing. If you live in a room which does the opposite, a higher Q is better. Therefore, determining which alignment is 'best' in room X is meaningless for room Y.
 
Last edited:
  • Like
Reactions: 1 users
System Q influences both time domain, which is buried in the reverberation of the room (so it cannot be examined anymore), and frequency domain, which is altered by the room (so a different system Q in the same round still sounds different). So if you compare different system Q's, you are comparing their frequency domain performance.

A comparison can be done. Lower Q means less bass. If you happen to have a room which amplifies bass, this bass reduction can be a good thing. If you live in a room which does the opposite, a higher Q is better. Therefore, determining which alignment is 'best' in room X is meaningless for room Y.

The Q definitely effects how much bass we have "Obviously extension is also an issue, but I don't want to go into that in this thread." We can fix the level of bass with EQ or crossover design. I agree the fix may be different in different rooms.

But if you built a very high Q system (2.0 or higher) by putting a big woofer in a way-to-small small box, it will sound slow and muddy. Some rooms might make this worse, but you cannot find or build a room where it doesn't sound muddy. And there are no rooms where bass sounds too fast, tight and clean so you want a muddy speaker to counteract that.

I've built several systems targeting .707 (max flat response) and ended up high by mistake and got mud. .5 or near also sounds better to me than .707, but this is subjective and might also depend on source material. I have enough experience to say generally I prefer a lower Q, but not enough to say whether .5 would be noticeably better than .577
 
But if you built a very high Q system (2.0 or higher) by putting a big woofer in a way-to-small small box, it will sound slow and muddy. Some rooms might make this worse, but you cannot find or build a room where it doesn't sound muddy. And there are no rooms where bass sounds too fast, tight and clean so you want a muddy speaker to counteract that.
I assume you agree about the time domain performance of the loudspeaker being irrelevant once you put it in a room. Then frequency domain performance is left. So something in the frequency response of a high Q system must sound muddy. That would be the low frequency bump. (Too much) Low frequency sound without accompanying midrange frequency harmonics to add definition indeed sounds muddy.

With regard to a 'too fast room', I suggest thinking in terms of frequency response and impulse response is easier to understand. An certain uneven frequency response results in slow sound. Uneven frequency response + different uneven frequency response can sum to a flat frequency response. Replace the first uneven frequency response by the high Q loudspeaker, the second by a room transfer function from loudspeaker to listener, with a big suck out at some bass frequencies. Those two together result in a flat frequency response, so it must be 'fast'.
 
  • Like
Reactions: 1 user
Subs are supposed to be muddy and slow. Your not supposed to be able to locate them by ear, if you can there putting out too much HF. There supposed to be LPF'd quite steeply which removes the speed and snap ( "tightness" ). Try a very steep LPF at 80hz on music into your full range mains and hear how muddy and not tight the sound is.
 
I assume you agree about the time domain performance of the loudspeaker being irrelevant once you put it in a room. Then frequency domain performance is left. So something in the frequency response of a high Q system must sound muddy. That would be the low frequency bump. (Too much) Low frequency sound without accompanying midrange frequency harmonics to add definition indeed sounds muddy.

With regard to a 'too fast room', I suggest thinking in terms of frequency response and impulse response is easier to understand. An certain uneven frequency response results in slow sound. Uneven frequency response + different uneven frequency response can sum to a flat frequency response. Replace the first uneven frequency response by the high Q loudspeaker, the second by a room transfer function from loudspeaker to listener, with a big suck out at some bass frequencies. Those two together result in a flat frequency response, so it must be 'fast'.

I disagree. I know from experience 2 subs can have the exact same frequency response, and 1 sounds tight and fast, the other sounds slow and muddy, both in the same room.

Suppose we have a sound track with gunshot, which is dead silence followed by 6 waves of equal amplitude at 31 Hz followed by dead silence. A perfect speaker will reproduce that perfectly. Suppose we have a speaker where the first wave is low amplitude, the 2nd is a little higher but still to low, then the 3rd, 4th, 5th and 6th are correct, but then we have 3 extra waves of decreasing amplitude. We cannot fix this with EQ, Linkwitz transforms, or crossovers.

Our 2nd speaker is correct in a mathematical sense in the frequency domain - the only frequency is 31 Hz. It may or may not be correct as you are thinking of frequency domain, as the amplitude is wrong.

In a time domain, it is wrong.

Is how a muddy speaker behaves? I don't know, but it seems logical from the description of overshooting do to being under-damped. I do know what I hear isn't what I want, and it cannot be fixed by EQ.
 
Subs are supposed to be muddy and slow. Your not supposed to be able to locate them by ear, if you can there putting out too much HF. There supposed to be LPF'd quite steeply which removes the speed and snap ( "tightness" ). Try a very steep LPF at 80hz on music into your full range mains and hear how muddy and not tight the sound is.

This is why we shouldn't use imprecise terms like 'muddy' and 'slow'. Underdamped is probably the correct work, but I am not sure.

As we go below 80Hz, sound gets less localizable, this is not what I am talking about, and both a 'fast' sub and a 'muddy' sub are equally less localizable at the same frequencies.

Try a very steep LPF at 80hz on music into a cheap, small sub and compare it to the same music and slope on a well made expensive one. On a well made sub, a large drum sounds like a drum (even without the overtones). On a cheap small sub, at best it sounds like a thump.
 
frugal-phile™
Joined 2001
Paid Member
I meant the measured Q of the built system, not the calculated Q, which can be different.

Only by the amount caused by driver-to-driver variance. And with a sealed alignment and a ! this low not a whole lot of difference unless your woofer is junk.

Why they would do this is beyond me

That is strange. How is one to have a set of T/S parameters to work with as the ones we typically measure are different than the ones measured by the facvtory kit. A manufacturer will usually take a sample of drivers, break them in and measure them for the data sheet.

Not doing that is a shortcoming.

as an example of something where 0.577 and 0.5 resulted in a dramatically different enclosure.

That is usually the case independent on the driver you choose.

dave
 
Suppose we have a sound track with gunshot, which is dead silence followed by 6 waves of equal amplitude at 31 Hz followed by dead silence. A perfect speaker will reproduce that perfectly. Suppose we have a speaker where the first wave is low amplitude, the 2nd is a little higher but still to low, then the 3rd, 4th, 5th and 6th are correct, but then we have 3 extra waves of decreasing amplitude. We cannot fix this with EQ, Linkwitz transforms, or crossovers.

Our 2nd speaker is correct in a mathematical sense in the frequency domain - the only frequency is 31 Hz. It may or may not be correct as you are thinking of frequency domain, as the amplitude is wrong.

Do the math - a tone burst of 6 cycles if 31 Hz has frequency content on either side of 31 Hz. Unless your speaker is flat D.C. to daylight there will be distortion of the amplitude and relative phase between the cycles. Yes, that means there is content below 31 Hz. And above, obviously. Much of that content will get fed to the mains not the sub.

I've found that acoustic guitar sounds more realistic on speakers flat anechoically to 20 Hz than on something that rolls off sooner. And the fundamental tone is 82 Hz. Even if all you have is a 40 Hz capable sub it's still better than no sub with just an 82 Hz fundamental. You can definitely tell when the sub is switched off- even before the drums and bass kick in.
 
Hi All,

FYI:

A Tone-burst .wav file containing 6 Cycles of 31 Hz.
Duration time= ~0.19354839 sec and an Off-time of 1 sec at a Level of -6(dBFS) making it possible to test a Sub and or the Main Speakers in a Small Room where RT-60 Reverberation time is equal or less than the Off-duration.
wave-burst_31Hz_-6dBFS_x5.zip ===> wave-burst_31Hz_-6dBFS_x5.wav

b:)
 

Attachments

  • wave-burst_31Hz_-6dBFS_x5.zip
    514.1 KB · Views: 65
  • Like
Reactions: 1 user