1. I suspect that room sound is treated differently by the brain than speaker sound.
2. Not counting subs, other "passive" drivers generally have their resonance outside their passband. That's proper design, unlike the screwy design for subs.
3. Anybody know whats kind of Q a horn has? Or an ESL? Of course, ESLs are better matched for impedance to air, so hardly worth bothering about their odd resonances.
In practice OP might (still) be able to choose to be as clean and featureless as possible or to relish some mild bass emphasis. As an ESL enthusiast, I aim for clean with obvious implications for enclosure choices.
Ben
2. Not counting subs, other "passive" drivers generally have their resonance outside their passband. That's proper design, unlike the screwy design for subs.
3. Anybody know whats kind of Q a horn has? Or an ESL? Of course, ESLs are better matched for impedance to air, so hardly worth bothering about their odd resonances.
In practice OP might (still) be able to choose to be as clean and featureless as possible or to relish some mild bass emphasis. As an ESL enthusiast, I aim for clean with obvious implications for enclosure choices.
Ben
Last edited:
A horn itself has no Q-factor.
The Q-factor of a system is derived from how its fundamental resonance decays. Q=0.5 means that if the cone sets off at some displacement, it'll return to its rest position with no overshoot.
Q=0.707... means it'll return to it's rest position fastest, with a slight overshoot.
Q>0.707 means it'll ring a little.
Q<0.5 means it'll take longer to get back to its rest position.
The same Q-factor also determines the shape of the rolloff of the system. Its easy to see that a sub with a Qtc=5 will ring at resonance, giving it a big peak in its response. So long as there's no energy storage issues (ie, Qtc<1), you can EQ for pretty much any response shape and it'll be fine.
I suspect ESLs have some resonances in their passband. There's all sorts of interesting standing waves that can show up on a flat membrane, over a very wide frequency range.
What's the problem with going past the resonance of a driver? Its well within the design parameters.
Chris
The Q-factor of a system is derived from how its fundamental resonance decays. Q=0.5 means that if the cone sets off at some displacement, it'll return to its rest position with no overshoot.
Q=0.707... means it'll return to it's rest position fastest, with a slight overshoot.
Q>0.707 means it'll ring a little.
Q<0.5 means it'll take longer to get back to its rest position.
The same Q-factor also determines the shape of the rolloff of the system. Its easy to see that a sub with a Qtc=5 will ring at resonance, giving it a big peak in its response. So long as there's no energy storage issues (ie, Qtc<1), you can EQ for pretty much any response shape and it'll be fine.
I suspect ESLs have some resonances in their passband. There's all sorts of interesting standing waves that can show up on a flat membrane, over a very wide frequency range.
What's the problem with going past the resonance of a driver? Its well within the design parameters.
Chris
Strictly speaking it does - in the case of a horn though, it applies to the directivity characteristics 🙂.
There's no need to guess, there are extensive studies on early reflections, late reflections, room modes, room gain, absorption, diffraction and everything else "room sound" related.
Congrats, you've turned another thread completely OT with your favorite dogma points.
1. Simulators are evil.
2. Resonances are evil.
3. ESLs rock.
Maybe if you try really hard you can slip in something about motional feedback too.
Your words don't match your actions. The subs that you use go completely against the things you describe as important.
Your previous Klipschhorn was riddled with resonances all through the passband and had the driver fs inside the passband.
Your current antique AR sub is not a low qtc design, it uses a TINY box and because of that and the low xmax driver it is necessarily a high distortion design at even moderate spl.
Also, as I've mentioned several times, ESL is not a no compromise design. ESL panels necessarily have serious off axis issues due to the large panel sizes. Some ESLs mitigate this to some extent by having narrow and/or curved panels but it's still a big issue, especially when you use huge 1 meter square panels like you do, the worst of both worlds.
I recently measured a sub that I built with 2 -12" Infinity 1260w drivers. 3/4" MDF, no double baffle. 2.5cuft net internal. I read as much as I could online and decided the use r13 paper backed insulation (I peeled the paper away. My box was 15.5" X 32" X 13" and so before subtracting for drivers and bracing it was 2.8cuft. I multiplied that by 1.5 to arrive at the 4.2lbs of r13 i jammed in there. I measured using Dayton DATS and got a Q of .534. I removed all but just a little bit of batting that I loosely placed in yhe bottom 1/3 of the enclosure and wound up with a Q of .758. I was aiming for .70 but it got dark and removing the drivers so many times to adjust the fill gave me a headache. So...you can definitely change the boxes Q with some R13. I've not tried any other type of stuffing.
Filling both increase the apparent volume (lowering Fr and Q) and add resistive damping (lowering Q). The former effect works best with boxes much smaller than Vas of the driver. The latter works regardless of the Vas/Box ratio.
More like 150-200 Hz in reality.
Mass quantities of fiberglass insulation 😉: https://www.hometheatershack.com/fo...204-new-si-ht-woofer-specs-info-legatos19.jpg
Calculate it like the tiny room it is: https://www2.owenscorning.com/quietzonepro/pdfs/NoiseControlDesignGuide.pdf
GM
Mass quantities of fiberglass insulation 😉: https://www.hometheatershack.com/fo...204-new-si-ht-woofer-specs-info-legatos19.jpg
Calculate it like the tiny room it is: https://www2.owenscorning.com/quietzonepro/pdfs/NoiseControlDesignGuide.pdf
GM
- Status
- Not open for further replies.