@Tenson
The physics part of my heart left a little tear of joy with that explanation!
Little additional side note, this does assume all system will perform on a linear fashion.
As soon things start to become less or non-linear, things don't line up so well anymore.
Since there is not one single system in the universe that is perfectly linear, that is nothing to be amazed about.
But it's important to keep in mind when one wants to compensate certain things.
Some things can be very easily compensated for, because they behave on a very linear fashion through the range of interest, other things do that a lot less.
So it's important to know and understand where, when and how these systems will become non-linear*.
Ported systems can be tricky in this for two reasons;
1 - the loudspeaker doesn't behave as function of cone excursion + other additional losses
2 - the port doesn't behave linear as function of air flow/velocity + other additional losses
The higher the order of a system, the more variables there are at play, higher errors and (more important) uncertainties can be expected.
* Non-linear means how well the system behaves and follows the ideal minimum-phase behavior.
Not just only "distortion" (THD or IMD), although never being used that way in audio, all these other forms ARE actually a type of distortion as well.
The physics part of my heart left a little tear of joy with that explanation!
Little additional side note, this does assume all system will perform on a linear fashion.
As soon things start to become less or non-linear, things don't line up so well anymore.
Since there is not one single system in the universe that is perfectly linear, that is nothing to be amazed about.
But it's important to keep in mind when one wants to compensate certain things.
Some things can be very easily compensated for, because they behave on a very linear fashion through the range of interest, other things do that a lot less.
So it's important to know and understand where, when and how these systems will become non-linear*.
Ported systems can be tricky in this for two reasons;
1 - the loudspeaker doesn't behave as function of cone excursion + other additional losses
2 - the port doesn't behave linear as function of air flow/velocity + other additional losses
The higher the order of a system, the more variables there are at play, higher errors and (more important) uncertainties can be expected.
* Non-linear means how well the system behaves and follows the ideal minimum-phase behavior.
Not just only "distortion" (THD or IMD), although never being used that way in audio, all these other forms ARE actually a type of distortion as well.
A sealed cabinet just means it can't have holes in it....because holes make all kinds of noise that is clearly audible. A small hole, like from a screw that missed, makes a lot of noise. I have found this on my own builds, and I have helped others fix noisy boxes by patching little holes.
So we are confusing "air tight" with "air permeable". MDF is absolutely air permeable, and most wood finishes are aslo slightly air permeable. And the drivers themselves may be slightly air permeable as well.
Some people may not realize that on CNC machines, plain MDF sheets are used as part of a vacuum hold down system. You skim off the outer surfaces of the MDF, and then lay it on top of a vacuum table. The MDF acts like a diffuser, spreading out the pull of the vacuum pump across the entire sheet. The vacuum pulls air through the sheet, holding material down securely enough that it can be machined with no other clamping besides the vacuum.
So MDF can be very air permeable. And nobody is going to build a pressure vessel out of wood, mainly because wood is air permeable. But you don't need a speaker cabinet to be a pressure vessel. You just need to stop air flowing in and out and resonating.
Here is a demonstration to support this - https://www.diyaudio.com/community/...fix-resonance-or-ringing-to-advantage.352796/
Do you mean an impulse response? I don't think that would show much we don't see more clearly with the graphics provided. What did you want to see?
Any change from zero to a non-zero velocity includes a transient. Have a look at the spectral display here.
Last edited:
I stole a page out of Genelec's book when I build this port. They have their "Laminar Spiral Enclosure", and I made a radius port:
Granted it is a subwoofer, so resonance is a simpler problem to solve; just use a steep filter. But I am getting very low distortion as long as I keep the crossover around 100hz. (Not the best photo to show it, but the best I can do on short notice!)
That's basically just what a Q-factor is.
You either go fast but overshoot or go slow without overshoot.
Pick one of the two.
Damn law of conservation of misery.
I'm curious what software that second image is? Is it CAD or a simulation system?
No, I am fairly certain that my tweeter gaskets pass a small amount of air. It is a slow leak. I am not sure about the mids and woofers, since those get screwed down tight and the gasket gets compressed. But with tweeters I don't screw them down tight, I screw them down only until they are flush. There is usually significant gasket thickness left in the joint (1 or even 2 mm). In addition, most tweeters are supplied with a gasket made of open cell foam, which is air permeable.
We shouldn't pass by the fact that in a room with a fairy stable temperature, the difference in pressure is around ±0.1-0.2% or so.
In practice I think even less.
Since this is directly proportional to the DC cone excursion, this is also how much we could expect how much the cone would move.
Assume that the enclosure and the cone is a 100% seal.
Which they are not at all.
Getting a 100% seal with the gasket and the baffle is already not possible.
When we talk about the attack of kick-drum sound in BR vs. sealed bass (subwoofer) difference comes mostly from transient behaviour and phase match of fundamental to harmonics. This is where a low-tuned BR comes close to sealed/direct radiator.
https://output.com/blog/get-perfect-kick-drum
https://www.credland.net/kick-drum-theory.html
The first sealed woofer I made was HT sub with two 15" woofers. At that time I didn't have measurement equipment or dsp. The realism of drum or acousic bass was astonishing when flat response starts from 20Hz. Really good integration to mains with dsp (acoustic nearfield LR2 at 100Hz, with necessary delay for mains) made it even better.
https://output.com/blog/get-perfect-kick-drum
https://www.credland.net/kick-drum-theory.html
The first sealed woofer I made was HT sub with two 15" woofers. At that time I didn't have measurement equipment or dsp. The realism of drum or acousic bass was astonishing when flat response starts from 20Hz. Really good integration to mains with dsp (acoustic nearfield LR2 at 100Hz, with necessary delay for mains) made it even better.
Last edited:
You can give a BR the same response as a sealed box if you want with some EQ'ing and vice versa.
So talking in terms of BR or sealed is not really that meaningful.
It's the order of the system as well as Q that matters.
Obviously assuming that both systems don't have any additional issues etc.
Oh I see where the miscommunication started. I never advocated that a small hole was necessary, and I don't believe the steady state pressure in the box will ever be different than atmospheric (different enough to matter).
I was agreeing with @AllenB 's statement to @diyuser2010 assertion that a sealed speaker box could be made so finely that it would become a leak-free pressure vessel. Allen and I disagree with diyuser2010, we don't think a wood box could ever be so tight, and basic ideal gas law shows that even in an air tight box, it is not an issue with normal barometric variations.... but if he feels strongly about the issue, he should drill a small hole in his box so he can sleep at night.
Regarding cone offset due to pressure differences inside a sealed enclosure with respect to ambient pressure - entirely minimal over time, I suspect - I would bet good money that almost every ported enclosure ever built would have considerably greater net cone offset due to port rectification. I have zero data to support this, but always pleased to be proved wrong.