I found an online calculator of box resonances. It’s easy to use, simply fill the speaker cabinet dimension and it will return results.
Picture is the result of my speaker’s dimension inputted.
Unfortunately, I don’t understand what they mean. Could anyone please explain the results for me?
Also, according to the results, which frequencies could cause the mid-bass boom phenomenon?
Link to the calculator: https://web.archive.org/web/20230603010703/http://www.mh-audio.nl/Calculators/standingwaveinbox.html
Picture is the result of my speaker’s dimension inputted.
Unfortunately, I don’t understand what they mean. Could anyone please explain the results for me?
Also, according to the results, which frequencies could cause the mid-bass boom phenomenon?
Link to the calculator: https://web.archive.org/web/20230603010703/http://www.mh-audio.nl/Calculators/standingwaveinbox.html
It's calculating the fundamental standing wave frequencies for each internal dimension (mode 1) and the harmonic series based on multiples of each one (mode 2, 3, 4, etc.)
So for depth, the dimension supports reinforcement at 477.8 Hz, 2 x 477.8 = 955.6, 3 x 477.8 = 1433.4, etc.
It can be helpful to view them on a dot plot on a single line to see how much they are bunched up.
So for depth, the dimension supports reinforcement at 477.8 Hz, 2 x 477.8 = 955.6, 3 x 477.8 = 1433.4, etc.
It can be helpful to view them on a dot plot on a single line to see how much they are bunched up.
So, I should focus on mode 1, right?
And if I’m going to account them into the crossover design, in order to compensate for these resonances, such as using underlap crossover point technique, I could use the mode 1 as the center frequency of the gap created by underlapping crossover points?
But it still be confusing because which side of mode 1 should be picked; depth (477.8Hz), width (593.1Hz), or height (143.3Hz)?
And if I’m going to account them into the crossover design, in order to compensate for these resonances, such as using underlap crossover point technique, I could use the mode 1 as the center frequency of the gap created by underlapping crossover points?
But it still be confusing because which side of mode 1 should be picked; depth (477.8Hz), width (593.1Hz), or height (143.3Hz)?
Last edited:
Mode 1 is the lowest frequency for each axial mode, so if you can avoid exciting those at all by crossing over lower than them, you are in decent shape.
There are other modes (tangential and oblique), but they tend to be lower in intensity so aren't focused on as much.
https://sengpielaudio.com/calculator-roommodes.htm
I don't understand exactly what you mean on the crossover topic. If you want to cross over low enough to avoid the axial modes completely, that makes sense.
From a theoretical design standpoint, it's optimal to distribute the various modes so they are evenly spaced and not overlapping. Practically speaking, some are high enough in frequency that they can be easily absorbed, so it's pretty common to make some choices based on that and aesthetics. That also gets you back to really only needing to focus on the lowest few modes in many cases.
There are other modes (tangential and oblique), but they tend to be lower in intensity so aren't focused on as much.
https://sengpielaudio.com/calculator-roommodes.htm
I don't understand exactly what you mean on the crossover topic. If you want to cross over low enough to avoid the axial modes completely, that makes sense.
From a theoretical design standpoint, it's optimal to distribute the various modes so they are evenly spaced and not overlapping. Practically speaking, some are high enough in frequency that they can be easily absorbed, so it's pretty common to make some choices based on that and aesthetics. That also gets you back to really only needing to focus on the lowest few modes in many cases.
The claim the a cube subwoofer is the worst shape is silly as the wavelengths are so long modal frequencies don't really get played to excite standing waves.
Is this the suggestion of leaving a gap of one octave between each crossover point in a multiple driver speaker system?
I suggest that crossover underlap is not practical as a method to "notch out" fundamental resonance modes in an enclosure.
With low order filters at least, an underlap of one octave can still give a surprisingly smooth overall response.
Besides, as you said, which fundamental mode would you attempt to eliminate - depth, width or height?
With low order filters at least, an underlap of one octave can still give a surprisingly smooth overall response.
Besides, as you said, which fundamental mode would you attempt to eliminate - depth, width or height?
Neither do I, what I see from your post isn't complete.
Firstly, you need to consider where in the box the source is placed, this makes a difference to the modes in both distribution and level.
Secondly, it doesn't tell you what's happening. For example: Consider one wall 1.7m from a source where all the energy goes, reflects then recombines. You'd have cancellations at 50Hz, 150Hz and 250Hz. You'd have an increase below 25Hz, at 100Hz and at 200Hz.
Furthermore, this is usually managed (reduced) by damping material. Probably a good thing considering the delays added by reflected energy.
Seems just a old basic audio rule of thumb.
You dont wanna use a completely square box.
Since all equal sized panels will resonate more.
More typical box is rectangle in shape.
Of course the longest panel is then your concern.
All leads to the usual = adding bracing.
With the calculator it basically shows the longest
panel resonates at 144 Hz
This is a low frequency and where there is a lot
of energy with a speaker.
The trick with any brace is it raises the resonant frequency
higher.
All panels resonate no matter what.
We just want it high enough to where their isn't enough
energy made by the speaker to be a audible rattle.
Relation to crossover not even a concern.
If you add a brace this calculator wont tell you
what it is raised to. As noted the smaller panels
are already above or near 500 Hz
Meaning high enough to never ever be heard or rattle.
Instead of playing with online calculators.
probably dwell into the learning curve of crossover design
software. You will more easily observe the behavior
of 1st order crossovers. And quickly abandon them
Aside from a few applications. Mainly 2 way
with 90 degree mounted woofer, which relieves the need for higher order
filter since the woofer is off axis
You dont wanna use a completely square box.
Since all equal sized panels will resonate more.
More typical box is rectangle in shape.
Of course the longest panel is then your concern.
All leads to the usual = adding bracing.
With the calculator it basically shows the longest
panel resonates at 144 Hz
This is a low frequency and where there is a lot
of energy with a speaker.
The trick with any brace is it raises the resonant frequency
higher.
All panels resonate no matter what.
We just want it high enough to where their isn't enough
energy made by the speaker to be a audible rattle.
Relation to crossover not even a concern.
If you add a brace this calculator wont tell you
what it is raised to. As noted the smaller panels
are already above or near 500 Hz
Meaning high enough to never ever be heard or rattle.
Instead of playing with online calculators.
probably dwell into the learning curve of crossover design
software. You will more easily observe the behavior
of 1st order crossovers. And quickly abandon them
Aside from a few applications. Mainly 2 way
with 90 degree mounted woofer, which relieves the need for higher order
filter since the woofer is off axis
It's calculating internal eigenmodes / standing waves raher than panel resonances. Still, the basic principles are similar -if you're running a sealed or vented box based purely on Helmholtz assumptions (uniform internal air particle density & no standing waves), then ignoring everything else, an efficient design is one that hits the target alignment with the minimum of acoustic damping. So avoiding equidistance between panels, & preferably using one of the various acoustic ratios to distribute those modes on the three dimensions is usually a good plan, when it's possible / practical.
You should not do anything with that information, ignore it.
First, if the speaker is closed, it will not matter all that much. Second, if the speaker is ported the resonance leakage to the outside world is very complex. It is affected by the woofer placement as well as port placement. Additionally other objects such as the woofer magnet or bracing beams and damping materials will change the result. You need a deep understanding to optimise this result and IMO making multiple prototypes or adjustable porotypes before a final build is the best way because even advanced simulation easily give false information that lead you in the wrong direction.
Example: Red - Far Field on axis. Blue - Woofer Near Field. Green - Port Near Field.
Visual of height mode at about 350Hz
First, if the speaker is closed, it will not matter all that much. Second, if the speaker is ported the resonance leakage to the outside world is very complex. It is affected by the woofer placement as well as port placement. Additionally other objects such as the woofer magnet or bracing beams and damping materials will change the result. You need a deep understanding to optimise this result and IMO making multiple prototypes or adjustable porotypes before a final build is the best way because even advanced simulation easily give false information that lead you in the wrong direction.
Example: Red - Far Field on axis. Blue - Woofer Near Field. Green - Port Near Field.
Visual of height mode at about 350Hz
Last edited:
In outright terms I'd agree; unfortunately, councils of perfection often tend to run up against what many are able to achieve for a variety of 'complex' (pun intended) reasons. Since many do not have access to, and / or the interest in FEM / BEM etc. modelling, nor necessarily have the time or funds to produce multiple or easily adjustable prototypes, logically anything that can be done to reduce the potential for issues is a step in the right direction. Basic acoustic ratios are far from the panacea they are often made out to be, but in those circumstances, they can be better than nothing at all. Optimising driver & vent tap locations as far as is practical relative to a basic predicted longitudinal mode is the next step, as we do all the time with QW design. Beyond that, then you're into more refined processes such as you describe, when (or if) time / funds / inclination allows.
Last edited:
Yeah, you are right something like HornResp can give a good indication of the matter.
I do not think the online calc the OP shared has much value though, and I would not expect any of it to change the crossover design. So in that sense the info should be ignored.
P.s. @presscot these frequencies are easy to calculate. It is just the speed of sound (342m/s) / by the box diamension ( example 0.35m) / 2.
342/0.35 = 977
977 / 2 = 488 hz
This is the lowest frequency resonance that can fit directly between those panels. There will also be a series of multiples x2, x4 etc.. these are the 2 dimensional modes. However as shown by the graphic I posted of the speaker internals, you can see the strength of this mode is different at the top, middle and bottom on the cabinet, where the middle is less strong.
I do not think the online calc the OP shared has much value though, and I would not expect any of it to change the crossover design. So in that sense the info should be ignored.
P.s. @presscot these frequencies are easy to calculate. It is just the speed of sound (342m/s) / by the box diamension ( example 0.35m) / 2.
342/0.35 = 977
977 / 2 = 488 hz
This is the lowest frequency resonance that can fit directly between those panels. There will also be a series of multiples x2, x4 etc.. these are the 2 dimensional modes. However as shown by the graphic I posted of the speaker internals, you can see the strength of this mode is different at the top, middle and bottom on the cabinet, where the middle is less strong.
Last edited:
No, I'm not a fan of adjusting filters to try to compensate for problems elsewhere (e.g. panel resonances) either. Some do it & power to them of course, but it always seemed to me that you'd be better off fixing the problem itself rather than trying sticking-plaster methods.