It's worthwhile to note that all the definitions of overshoot / critical damping etc are referring to a low-pass function (the cone excursion) and not to a high-pass function (the second derivative, acceleration which is proportional to sound pressure). As noted, a high-pass always "overshoots" as its integral must be zero.
Last edited:
True the step response going from zero to one won’t happen that much in music, if ever. However, it’s a good measuring technique to learn how the system responds to a sudden input. Large fast deviations from the long term steady state generally have huge effects on the loudspeaker system and the audio output. Also the system cannot act until the driver’s output settles to its final state. Lots of settling is lots of distortion and ringing because the speaker system doesn’t settle quickly to a zero state. This is easily heard by listening to tonga drums live and then listening to them through a recording and then through a playback system. Listen to live cannon then listen to recordings of cannon. This overshoot in audio reproduction can be heard in much low frequency music. Measuring step response is is a good tool to show just how well a loudspeaker can reproduce clean clear bass and hi fidelity sound.
So Arthur, in your opinion, how can we eradicate the problem with speakers. I am very familiar with sound of cannons, rifles, etc. I have never heard anything that can reproduce the scale or characteristic of this sound. Do we then just give up trying or is there a solution to correcting for it. If there is no solution then what is the point even discussing it?
EDIT: I must say after a cannon shot my ears ring for minutes afterwards.
EDIT2: I watched the Iranian attack on Israel with my headphones on a night ago, and I must admit, the far-off explosions when the munitions hit and my experience with the same was pretty convincing, I will not mistake that sound for anything else ever. Far awy thunderstorms sound very much like it. Do I then accept it that headphones are so much better than loudspeakers.
EDIT: I must say after a cannon shot my ears ring for minutes afterwards.
EDIT2: I watched the Iranian attack on Israel with my headphones on a night ago, and I must admit, the far-off explosions when the munitions hit and my experience with the same was pretty convincing, I will not mistake that sound for anything else ever. Far awy thunderstorms sound very much like it. Do I then accept it that headphones are so much better than loudspeakers.
Last edited:
Arthur, a square wave impulse response consists of a rise, and decay. The rise is purely a high frequency, the decay is also a high frequency the only low frequency content in the pulse is the time between the rise and decay, which I agree could be a low frequency. In a speaker, inertia is the culprit in seeing a slower rise and a longer decay as well as the apparent ringing in the impulse response. Critical damping may be a way to solve this. It would be like putting shock absorbers in a car.
EDIT: I am not absolutely certain but this could be the reason inserting ferro-fluid into a tweeter. Can it be applied to a woofer. I never though of this.
EDIT: I am not absolutely certain but this could be the reason inserting ferro-fluid into a tweeter. Can it be applied to a woofer. I never though of this.
Arthur, now you got me thinking. Amplifiers are not an issue any longer, we can have a 2000 watt amplifier, what if we can dampen a speaker so that it does not have these troublesome characteristics trying to overcome the drop in efficiency caused by adding a fluid damping mechanism in a woofer/speaker. I would agree that there will be some delay in response, but if everything is delayed over the whole spectrum, it does not matter, it would be similar to being some distance away from the transducer only.
Exactly. Some large ported enclosures are still ringing 1/4 second after the signal has ceased, compared with around 15ms or less with a sealed enclosure.
Are we supposed to believe that this will not affect fidelity? To my ear, any system with high energy storage, be it horn loaded/ported/transmission line sounds blurred and imprecise when reproducing drum beats compared with a sealed enclosure. There are those who argue that step response is irrelevant compared with the room's much longer reverb time, but a highly resonant speaker will still be dissipating stored/resonant energy into the listening space when it should not be, topping up the reverberative 'noise', as it were.
It is always slightly comical and ironic to me that a poorly damped Helmholtz resonator (ported box) is used as a sub enclosure, whilst at the same time Herculean efforts are made using thick walls and bracing them to within an inch of their lives to reduce resonances! Waterfall plots and transient decay time concerns seem to be completely disregarded when designing low frequency enclosures, when great import is placed upon them further up the audio spectrum. That's very odd to my way of thinking and fifty years of audio involvement.
One acid test of fast transient decay in a sub is double pedal bass drum beats, such as Metallica et al; a sealed enclosure will reproduce this as the staccato machine gun-like sound which it is, whilst a poorly damped system sounds - to my ear - like a blurred, indistinct, and extended fart...
Finally, Q0.5 sealed enclosures need not be huge - my 18" subs are 55l.
You can match a low-boost shelf filter to a critically damped woofer and you’ll get the superior pulse response AND fatter bass.
The only price you pay is amplifier power and stress on the driver.
For that matter, you can use any Q driver+box you want and flatten it with DSP and the net result will be the same.
I 100% agree with the general sentiment that steep filters ring and it’s better to avoid them where possible.
And I 100% agree with @Arthur Jackson that you can easily hear the difference.
The only price you pay is amplifier power and stress on the driver.
For that matter, you can use any Q driver+box you want and flatten it with DSP and the net result will be the same.
I 100% agree with the general sentiment that steep filters ring and it’s better to avoid them where possible.
And I 100% agree with @Arthur Jackson that you can easily hear the difference.
The practical tradeoff is transient response vs dynamic range. This is a fundamental unavoidable dilemma. The steep slope systems (horns and reflex) favor dynamics and the sealed favors step response.
Which compromise is more objectionable is purely a matter of taste.
Which compromise is more objectionable is purely a matter of taste.
Last edited:
....... or the effort you are willing to make. If you have a losed box with a Q slightly above 0.5 and fb slightly higher than you actually want then you don't need that much of gain below the physical fs and you can even reduce the power slightly in this area when using EQ. If you want closed box AND dynamics then there is no other choice than enough Vd.
Regards
Charles
A few personal takes on the subject of damped bass.
As a couple of folks stated above, the system high-pass is the issue. So we need to look at the measured acoustic high pass achieved, with all electrical filters in place. The acoustic roll-off, or same thing said another way,.... the order of the acoustic high-pass is ALL that matters with regard to dampening, ime/imo.
Type of box, sealed, reflex, horn ...gives us a starting natural low-end roll-off order...sealed is assumed 2nd, reflex 4th, and horn depends on type.
My understanding is a front-loaded horn is a sealed (with a horn attached), so 2nd order for it too. (Other horns, dunno.)
On top on the natural low end roll-off, we have to increase the order of the roll-off if an electrical high-pass is needed.
Maybe not needed for sealed, but an essential for reflex, adding 2 orders of addition roll-off give or take.
On top of the natural roll-off, and any additional electrical high-pass, we have to increase the order of the roll-off further due to mechanical limits of excursion.
Sealed without an electrical high-pass runs into this; its suspension puts the brakes on excursion (hopefully!)
If low-end shelving or a Linkwitz Transform is in play, the natural 2nd order roll-off of sealed will be pushed into a corner up against the suspension limitation and increase the high-pass order sharply. Or if the suspension isn't safe to be so stressed, and an electrical hpf is added, order of course increases, but perhaps less abruptly.
If voltage limiting is in place, particularly frequency dependent to limit over-excursions, it also acts to increase the order of the roll-off.
Bottom line imo. It all comes down to the systems acoustic high-pass response.
The flatter the final acoustic roll-off, the better the step response and probably the better the sound.
Enough sealed capacity to play as low and as loud as desired without running into any type excursion limit, is holy grail of sub bass imo. Much harder to achieve than realized imho.
As far as qtc and dampening........i see qtc as simply fine tuning the degree of natural low-end roll-off, and knee created. For instance 0.5 qtc has a flatter roll-off that 0.7. Above 0.7 a knee starts forming.
Qtc though, is a bit of twiddly-dee twiddly dum, in comparison to the bigger picture issues effecting the total acoustic high-pass order, imo.
As a couple of folks stated above, the system high-pass is the issue. So we need to look at the measured acoustic high pass achieved, with all electrical filters in place. The acoustic roll-off, or same thing said another way,.... the order of the acoustic high-pass is ALL that matters with regard to dampening, ime/imo.
Type of box, sealed, reflex, horn ...gives us a starting natural low-end roll-off order...sealed is assumed 2nd, reflex 4th, and horn depends on type.
My understanding is a front-loaded horn is a sealed (with a horn attached), so 2nd order for it too. (Other horns, dunno.)
On top on the natural low end roll-off, we have to increase the order of the roll-off if an electrical high-pass is needed.
Maybe not needed for sealed, but an essential for reflex, adding 2 orders of addition roll-off give or take.
On top of the natural roll-off, and any additional electrical high-pass, we have to increase the order of the roll-off further due to mechanical limits of excursion.
Sealed without an electrical high-pass runs into this; its suspension puts the brakes on excursion (hopefully!)
If low-end shelving or a Linkwitz Transform is in play, the natural 2nd order roll-off of sealed will be pushed into a corner up against the suspension limitation and increase the high-pass order sharply. Or if the suspension isn't safe to be so stressed, and an electrical hpf is added, order of course increases, but perhaps less abruptly.
If voltage limiting is in place, particularly frequency dependent to limit over-excursions, it also acts to increase the order of the roll-off.
Bottom line imo. It all comes down to the systems acoustic high-pass response.
The flatter the final acoustic roll-off, the better the step response and probably the better the sound.
Enough sealed capacity to play as low and as loud as desired without running into any type excursion limit, is holy grail of sub bass imo. Much harder to achieve than realized imho.
As far as qtc and dampening........i see qtc as simply fine tuning the degree of natural low-end roll-off, and knee created. For instance 0.5 qtc has a flatter roll-off that 0.7. Above 0.7 a knee starts forming.
Qtc though, is a bit of twiddly-dee twiddly dum, in comparison to the bigger picture issues effecting the total acoustic high-pass order, imo.
Do you have any evidence of those numbers?
Here's step response of a BW 12 dB/oct at 20Hz, to simulated a sealed sub.
I'll let you call where step has fully settled. 15ms or less seems overly optimistic.
(Note the 10ms delayed started to move the pulse start off t=0 for visibility, that needs to be subtracted from the scale to judge settling)
Here's a BW 36 dB/oct, to simulate a 4th order ported along with a 2nd order hpf in place.
1/4 second seems quite pessimistic.
1/8 second is closer to the truth actually. But shockingly long when we only look at the bare number !
Regards
Charles
Step response measurements from DataBass.
Thanks for the simulations - do you have any real world measurements?
I totally agree. It is entirely possible with modern computers and amplifiers to produce very accurate low distortion bass using sensors and feedback. I don’t know why it isn’t being done everywhere.
People are using room correction with high power amps and small high excursion woofers to get more distortion than audio output in the lowest octave. They think it’s bass but really it’s noise. Just take a walk through Best Buy.
Any delay or phase anomalies can be removed from the signal. I don’t know why computer controlled woofers aren’t everywhere.
Regarding the fluid damping in woofers, I don’t think it’s done because woofers have more travel and turn the fluid into foam rendering it useless.
6 x 18" & 4 x 21". Total linear displacement volume about three and a half US gallons...
Very welcome.
The only sub measurements that I have that are worthy of assessing step response, those made ground plane outdoors, were all transfer functions with Smaart which doesn't save an impulse response when taking transfers. So no, dammit.
Here's a just-made nasty indoor measurement of a sub currently in use, a large dual 18" reflex tuned to 25Hz.
Mic 4" from sub, in-between drivers and port, mic about 3 ft off ground. Like said, pretty nasty.
Impulse first, step second. Not sure how much / if anything, can be gleaned from indoors.
Step.