ouch, that's a really bad answer
basically ignore/discount everything before " A very small sealed box with a very large driver moving at huge excursion is the worst case"
(and most of the rest since cone breakup is the bigger cause of "ripples")
in loudspeaker electro-mechanical-acoustic modeling a "spring" is lossless - real air springs aren't perfect, just better than anything else we can make
the fact that a reasonable volume air spring is so much better in terms of loss and linearity than a dynamic loudspeaker surround/spider is the reason boxes are used - beyond just blocking the back wave
basically ignore/discount everything before " A very small sealed box with a very large driver moving at huge excursion is the worst case"
(and most of the rest since cone breakup is the bigger cause of "ripples")
in loudspeaker electro-mechanical-acoustic modeling a "spring" is lossless - real air springs aren't perfect, just better than anything else we can make
the fact that a reasonable volume air spring is so much better in terms of loss and linearity than a dynamic loudspeaker surround/spider is the reason boxes are used - beyond just blocking the back wave
Completely wrong. But I'm sure j.a.g. will write 1000+ words trying to defend it, maybe even run some hornresp sims as a bonus.
In subwoofers? Cone breakup causes ripple in subwoofers?
The pic he posted and asked about showed a graph that had a top end of 200 hz and the ripples were about +/- 1/2 db across the whole bandwidth.
Resonances and diffraction could be possible causes of ripple but the certainly not cone breakup below 200 hz. The more important point is that the +/- 0.5 db ripple in the pic isn't caused by any of those, it's just what an unsmoothed measurement often looks like.
Just looking at a few examples of good speakers measured on a Klippel you can see that the good speakers don't have a problem with suspension linearity until you get highish excursions. And I've certainly never heard of anyone complaining that an IB or even OB had linearity problems at less than rated xmax that was likely due to the suspension that should be corrected with a tighter air spring. This idea is ridiculous. No one would ever use IB or OB if this was an issue.
Loss can be measured by looking at efficiency or even frequency response, linearity can be measured by looking at THD. IB vs small sealed box will have efficiency at different frequencies but the IB won't have less overall efficiency than a small box. And I certainly doubt you will get less THD in a small sealed box either, in fact if the box is very small and the displacement is very large you will get MORE THD, not less.
Perhaps I could be wrong about springs and losses, that's fine.
But you can't argue that the IB takes MUCH less power at lower frequencies to reach the same excursion and provide the same spl compared to a small sealed box.
Air spring non linearity is a valid concern if the box is too small and the displacement is too large.
Those are my two points and I can't see a need to defend them as they are both true and I've showed plenty Hornresp graphs of IB vs small sealed already.
Thanks all,
As I reread the thread, I can't see that anyone ever really suggested using an overly small box (for whatever reason). Not sure what made the idea surface....
And I'm left thinking, what can the motivation be for deviating significantly from the tried and true sealed volume that gives maximally flat Qtc=.707?
(Which is 100L for the bms 18" I'll be using.)
In hornresp, I looked at response, displacement, and driver electrical watts for 100L vs larger all the way to IB. It seems to me there is really nothing to gain by going much larger than 100L. It appears there's more to lose...
So, I guess it's 100L or bust 🙂
As I reread the thread, I can't see that anyone ever really suggested using an overly small box (for whatever reason). Not sure what made the idea surface....
And I'm left thinking, what can the motivation be for deviating significantly from the tried and true sealed volume that gives maximally flat Qtc=.707?
(Which is 100L for the bms 18" I'll be using.)
In hornresp, I looked at response, displacement, and driver electrical watts for 100L vs larger all the way to IB. It seems to me there is really nothing to gain by going much larger than 100L. It appears there's more to lose...
So, I guess it's 100L or bust 🙂
1. nothing beneficial about being flat; when done and installed you want a rising low end ("house curve") because most folks like the sound better*
2. nothing beneficial about "critical damping" as a goal; bet nobody would say so except that it represents a compromise choice they happen to like
You are right, there is some size growth limit that represents a sensible trade-off of performance and domestic aesthetics. Maybe some would say your choice of 3 cu feet is on the small side.
B.
*I don't know why "flat" is touted as the goal. While always a matter of taste and according to your music genre collection and without there being any real physical basis that a person who understands psychoacoustics would ever endorse, an "objective-like" decision (that folks on this forum would twig to) would be based on a Fletcher-Munson curve
Last edited:
I was in the process of modifying a big reflex/vented speaker that was far too loud for domestic use.
I had lots of sensitivity and max SPL to spare.
I was advised to look at using a Bessel style roll-off rather than Butterworth for home use.
I massaged the roll-off curve using WINisd as my prediction tool to give a roll-off that looked more Bessel.
The final result is plenty loud enough. It goes very low. It does not need massive power. and it sounds very natural on all spoken voice and with "natural" sound effects on movies.
It was the best advice I have been given on this Forum.
That would require a bigger sealed box than "flat" Butterworth.
Why don't reduce the volume pot ?
(is there something sounding "excessive" or "unnatural" with this alignment ?)
If you are digitally equalized the Bessel is the best alignment, if not butterworth seems adequate.
Overdamped alignements "eat" the lows resulting of an truncated "unnatural" response.
overdamped is equivalent to low Q
Bessel is lower Q than Butterworth.
Underdamped is higher Q. That is what gives the bass hump that sounds unnatural when reproducing real acoustic sounds.
Butterworth plus the bass reinforcement from a room often results in a higher Q bass humped response.
A big reflex with extended bass response is not the same as
Bessel is lower Q than Butterworth.
Underdamped is higher Q. That is what gives the bass hump that sounds unnatural when reproducing real acoustic sounds.
Butterworth plus the bass reinforcement from a room often results in a higher Q bass humped response.
A big reflex with extended bass response is not the same as
and from my experience does not result in
Sorry again 🙄
I was talking about the underdamped (High Q) alignment.
A bessel (low Q) gives more headroom for equalization than the Butterworth one (higher Q) IMHO.
A big reflex with extended bass response always appears as a enigma to me because of the resonator ringing in a peculiar room always appear as a unpredictable phenomenon in my POV.
It is beneficial loudspeaker frequency response to be flat, because when you put that loudspeaker in the room, room gain will rise the low end. Oh, and you need a flat response because female vocal should sound as a female, not as a male.
If needed, Flecher-Munson curve is attainable with pre-amp "loudness" control (or bass + treble). There is no passive loudspeaker which can do a dynamically variable Fletcher-Munson curve (different shape for different SPL).
Last edited:
- Status
- Not open for further replies.