Because I have a lot of free time these days, I started testing few multi-way speakers I have here, sitting around, unfinished.
One in particular came to my attention: a project of 2-way using some nice pro audio 15'' woofer that happens to climb as high as 4khz while having some very manageable specs to go down to the first octave.
Nothing to win the gold medal of Hi-Fi Olympics, mind you. Just a project for fun, to pass time while I'm isolated at my place. 🙄
So, that particular project reminded me of something very important. Maybe one of the oldest problems in audiophilia, especially for us DIYers.
How, supposedly, the box is the least worst of the solutions
We all know there is basically 3 types of boxes:
The sealed one is the simplest, obviously. The ported one is the efficient one, while the PR is the ''technological evolution'' of the ported one, so to speak.
So far, nothing too complicated to understand. That's the ''box world'' even though many exotic configurations exists which are essentially variations on the theme.
But here comes the Open Baffle. And, just to be very clear, that's not a thread about how magical are OB's... They're not.
On one side you want to avoid the acoustic short circuit of an open baffle, a ''naked driver'' that most of low-frequencies abilities are stripped off, but on the other hand you ''bury'' the problem in a box... and that generates other problems. Problems ''solved'' mostly by stuffing absorbent materials into said box.
Compromises. Compromises everywhere. Insert a popular meme here.
Problem is, there is a general misconception about an idea in which the box itself -if poorly built- will generate resonances, noises and bottomline what can be called distortion, from the source (music) to our ears.
In fact, the vast majority of boxes, even built by beginners, are probably built in a way that they cannot generate resonances that will be audible, especially not at low/moderate SPL.
What's left then? LEAKS.
Ports, obviously, but also the driver itself. The 15 inches pro audio driver I was talking about earlier. Event though I use it in a sealed box, the internal resonance builds up and it leaks badly through the (big) cone. Of course, I can stuff stuff inside. But it's not enough. Unless I'm looking for other acoustical problems somewhere down the road...
Which do you think, of 25mm thick plywood V.S. 0.5mm paper will block the most sound energy? The plywood, obviously. The big cone here is a massive leak. Only the highest frequencies are blocked.
And, bottomline, it's distortion. Sometimes not too uncomfortable, but distortion nonetheless. The resonance build-up of the box which leaks through the membrane is audible enough to pollute the final sound rendering, sometimes in a very bad way. Like it's the case with my huge 15 PA driver.
The resonances are sound energy that bounces internally, trapped, on the (very) rigid enclosure's walls, and that generates reverberation (echoes) until that energy dies or leak somehow.
Almost impossible to correct on EQ. You can correct amplitude and frequency on a EQ but not reverberation time (RT), unless you try to dim the frequencie(s) where the resonance build-up is at the worst, but then you cheat on the signal, which is not good. Better to solve the problem at the source. Instead of putting out a fire, it is better to avoid having one from the start ...
These resonances are well-known with the ported boxes. That's another leak, ports. The resonances build-up leaks through the port(s) and even though there is many designs to avoid partially the unwanted effects, it's still a leak. At the perceptive level, the ''ported box reputation'' is that the bass is ''slow'', which is mostly explained by the reverberations, of the higher frequencies, that are building up and then leaked through the port, which cause a ''blurr'' in the perception. Less ''clean'', also a common way to describe it. And that is exactly what is it: less clean because polluted by distortion: unwanted sound energy on the original path from the source.
The human ear is unprecise regarding amplitude and frequencies, but very precise in the time domain. We can detect very small RT and that affect our perception a lot.
So here it is: I wonder if a fellow DIY enthusiast or any lab or company ever did a study, a test about the following:
To enclose a small full range speaker IN a larger box.
First, a completely closed box, sealed like a can. Then measure what comes out using sweeps or pink noise.
The result will be the resonance of the box itself, its walls and possible sound energy leaks, within limits of what the small fullrange can generate, in both frequencies and SPL.
Then, the same process but with a box, still of the sealed type, loaded with an inactive (bigger) driver. Ideally with a burnt/locked voice coil so it won't generate sound waves by itself...
The idea of course would be to measure what comes out via the membrane. So we can have a good idea of the leak compared with that the benchmark (the walls of the box, from the first test).
Any idea? Any comments?
🙂
One in particular came to my attention: a project of 2-way using some nice pro audio 15'' woofer that happens to climb as high as 4khz while having some very manageable specs to go down to the first octave.
Nothing to win the gold medal of Hi-Fi Olympics, mind you. Just a project for fun, to pass time while I'm isolated at my place. 🙄
So, that particular project reminded me of something very important. Maybe one of the oldest problems in audiophilia, especially for us DIYers.
How, supposedly, the box is the least worst of the solutions
We all know there is basically 3 types of boxes:
- Ported
- Sealed
- Passive radiator
The sealed one is the simplest, obviously. The ported one is the efficient one, while the PR is the ''technological evolution'' of the ported one, so to speak.
So far, nothing too complicated to understand. That's the ''box world'' even though many exotic configurations exists which are essentially variations on the theme.
But here comes the Open Baffle. And, just to be very clear, that's not a thread about how magical are OB's... They're not.
On one side you want to avoid the acoustic short circuit of an open baffle, a ''naked driver'' that most of low-frequencies abilities are stripped off, but on the other hand you ''bury'' the problem in a box... and that generates other problems. Problems ''solved'' mostly by stuffing absorbent materials into said box.
Compromises. Compromises everywhere. Insert a popular meme here.
Problem is, there is a general misconception about an idea in which the box itself -if poorly built- will generate resonances, noises and bottomline what can be called distortion, from the source (music) to our ears.
In fact, the vast majority of boxes, even built by beginners, are probably built in a way that they cannot generate resonances that will be audible, especially not at low/moderate SPL.
What's left then? LEAKS.
Ports, obviously, but also the driver itself. The 15 inches pro audio driver I was talking about earlier. Event though I use it in a sealed box, the internal resonance builds up and it leaks badly through the (big) cone. Of course, I can stuff stuff inside. But it's not enough. Unless I'm looking for other acoustical problems somewhere down the road...
Which do you think, of 25mm thick plywood V.S. 0.5mm paper will block the most sound energy? The plywood, obviously. The big cone here is a massive leak. Only the highest frequencies are blocked.
And, bottomline, it's distortion. Sometimes not too uncomfortable, but distortion nonetheless. The resonance build-up of the box which leaks through the membrane is audible enough to pollute the final sound rendering, sometimes in a very bad way. Like it's the case with my huge 15 PA driver.
The resonances are sound energy that bounces internally, trapped, on the (very) rigid enclosure's walls, and that generates reverberation (echoes) until that energy dies or leak somehow.
Almost impossible to correct on EQ. You can correct amplitude and frequency on a EQ but not reverberation time (RT), unless you try to dim the frequencie(s) where the resonance build-up is at the worst, but then you cheat on the signal, which is not good. Better to solve the problem at the source. Instead of putting out a fire, it is better to avoid having one from the start ...
These resonances are well-known with the ported boxes. That's another leak, ports. The resonances build-up leaks through the port(s) and even though there is many designs to avoid partially the unwanted effects, it's still a leak. At the perceptive level, the ''ported box reputation'' is that the bass is ''slow'', which is mostly explained by the reverberations, of the higher frequencies, that are building up and then leaked through the port, which cause a ''blurr'' in the perception. Less ''clean'', also a common way to describe it. And that is exactly what is it: less clean because polluted by distortion: unwanted sound energy on the original path from the source.
The human ear is unprecise regarding amplitude and frequencies, but very precise in the time domain. We can detect very small RT and that affect our perception a lot.
So here it is: I wonder if a fellow DIY enthusiast or any lab or company ever did a study, a test about the following:
To enclose a small full range speaker IN a larger box.
First, a completely closed box, sealed like a can. Then measure what comes out using sweeps or pink noise.
The result will be the resonance of the box itself, its walls and possible sound energy leaks, within limits of what the small fullrange can generate, in both frequencies and SPL.
Then, the same process but with a box, still of the sealed type, loaded with an inactive (bigger) driver. Ideally with a burnt/locked voice coil so it won't generate sound waves by itself...
The idea of course would be to measure what comes out via the membrane. So we can have a good idea of the leak compared with that the benchmark (the walls of the box, from the first test).
Any idea? Any comments?
🙂
Passive Radiator article on Wikipedia english:
They mention the vent turbulence -which I think is barely audible at low/moderate volume- especially with correctly designed ports. And even at high SPL it's probably drown into... the high SPL.
So vent pipe resonances.
I think it's rather the whole enclosure's resonances, vent pipe included.
Builded-up by multiple parallel walls, bouncing to a point that internal RT climbs to an audible (and annoying) level. Mainly because of the enclosure itself and maybe worsen by the vent pipe, poorly designed or not.
They mention the vent turbulence -which I think is barely audible at low/moderate volume- especially with correctly designed ports. And even at high SPL it's probably drown into... the high SPL.
So vent pipe resonances.
I think it's rather the whole enclosure's resonances, vent pipe included.
Builded-up by multiple parallel walls, bouncing to a point that internal RT climbs to an audible (and annoying) level. Mainly because of the enclosure itself and maybe worsen by the vent pipe, poorly designed or not.
I'd go Open Baffle with it. But I"m like that. Pair that 15 with a fullrange anywhere from 4" up to 8" and with the proper crossover you'd have a heck of a good speaker and not even have to build a box. 😀
Another interesting wikipedia article:
Bass reflex - Wikipedia
Transient response is, on paper, a key element. The sealed box is a clear winner on that aspect among enclosure types.
But I believe the ''unwanted midrange frequencies reflected from within the box'' mentionned above is far more impactful (audible) on the final sonic result of a sound system. Both leaked by the port and the driver's cone.
Also, ''midrange frequencies'' it's pretty vague. Low-mid, mid, high-mid? Could be very wide. From basically 200hz to 5000hz, maybe even wider. All frequencies that we are very sensitive.
Bass reflex - Wikipedia
Transient response is, on paper, a key element. The sealed box is a clear winner on that aspect among enclosure types.
But I believe the ''unwanted midrange frequencies reflected from within the box'' mentionned above is far more impactful (audible) on the final sonic result of a sound system. Both leaked by the port and the driver's cone.
Also, ''midrange frequencies'' it's pretty vague. Low-mid, mid, high-mid? Could be very wide. From basically 200hz to 5000hz, maybe even wider. All frequencies that we are very sensitive.
Hi Pano 🙂
Few years back I made some tests with OBs, using the big brother of my 15PR400, the mighty 18FH500. Was absolutely fantastic.
Only problem (as expected, even though that driver performed better than most) was the lack of low frequency response. No EQ could overcome the acoustic short circuit, and I didnt want to end with baffles as large as my room 😱
So unless you can live without bass (I sure cannot) Open Baffle is not a solution, only a potential partial one.
Also, the room acoustic becomes trickier, as explained here:
Are open baffle speakers a flawed technology? - HomeTheaterHifi.com
You left out tapered TL as an option. I have found that it can turn a ported speaker into something very special with deep extension AND articulation and low group delay.
In that case, I tend to agree.
But Wikipedia is open for edition by everyone 😉
Also, Wikipedia as ranked quality articles. Over 30,000 articles are categorized as ''good'' and are extremely trustworthy, maybe even the best source of information available on a specific subject.
Wikipedia:Good articles - Wikipedia
Speaking of which, I'd be curious to know if many members of DIY audio have created or edited wiki articles regarding audio?
anyway, that's another topic. 🙂
As I see it, it's part of the ''exotic'' ported variation mentioned in my post #1.
I don't see how the resonances problem would be fixed with a tapered TL or similar.
In fact, on paper the only way to avoid such resonances would be an IB where the back is basically... outdoor. You replace a window by a baffle.
Wife not happy with the idea.
Hi Jon,
I once build up 4 different subs next to each other in my living room:
one closed (7inch)
one smaller Bass reflex (6inch)
one medium Bass reflex (8inch)
one open baffle (2x 12 inch)
All crossed around 140 Hz.
I compared them quite some time (over a few days), also used some measurements and active EQs to flatten the responce and to extent the lower range if necessary.
In my opinion the biggest difference was, how they interacted with the room.
The open baffle acted different than all the other speakers since it has a narrow diffraction.
The bass was really "coming out of nowhere" which sounded somehow fascinating, but was lacking of power.
The closed sub interacted a bit more with room, but was quite sensitive regarding wall distance.
Not so boomy, but for my opinion still a little bit weak. somehow the fun was missing. But sure, it was very "clear".
I actually liked the 6inch ported bass the best. it was interacting quite a lot with room, but it gave power to the music.
The 8inch ported caused to many problems in my living room, since it is quit big and has a strong room mode at 32Hz with was stimulated to much by the low bass port tuning.
So, that was only MY experience in MY listening room.
But maybe its still somehow interesting.
Edit:
for the open baffle calculation and positioning, I just this spreadsheet:
OB guide/book
I once build up 4 different subs next to each other in my living room:
one closed (7inch)
one smaller Bass reflex (6inch)
one medium Bass reflex (8inch)
one open baffle (2x 12 inch)
All crossed around 140 Hz.
I compared them quite some time (over a few days), also used some measurements and active EQs to flatten the responce and to extent the lower range if necessary.
In my opinion the biggest difference was, how they interacted with the room.
The open baffle acted different than all the other speakers since it has a narrow diffraction.
The bass was really "coming out of nowhere" which sounded somehow fascinating, but was lacking of power.
The closed sub interacted a bit more with room, but was quite sensitive regarding wall distance.
Not so boomy, but for my opinion still a little bit weak. somehow the fun was missing. But sure, it was very "clear".
I actually liked the 6inch ported bass the best. it was interacting quite a lot with room, but it gave power to the music.
The 8inch ported caused to many problems in my living room, since it is quit big and has a strong room mode at 32Hz with was stimulated to much by the low bass port tuning.
So, that was only MY experience in MY listening room.
But maybe its still somehow interesting.
Edit:
for the open baffle calculation and positioning, I just this spreadsheet:
OB guide/book
Last edited:
Hello Darakon
crossed at 140hz but was it a steep slope or less than 24db/oct ?
My experience is when crossed anywhere below 150hz, ''brickwall'' mode (was using 48-300db/oct active xover), when listened alone, isolated from the upper frequencies, the resonances are barely noticeable. Or at least not annoying. Room's or enclosure's rattles or driver's mechanical noises are more annoying at those frequencies! 🙂
That's another story entirely when crossed higher. Say 400-500hz and up.
In my specific case (pro audio 15 inches) I wanted to use it as a woofer-midrange, basically to cover 5-6 octave, which is quite optimistic I agree...
The problem of resonances is very obvious and probably unfixable.
But I think the problem applies to basically any 2-way and even some 3-way where one of the driver: woofer or mid, or both, are creating uncontrolled resonances that are leaking through the cone.
Many 3-ways are using woofers that are crossed at 250hz, 300hz or even higher. And the vast majority of 2-ways are crossed at much higher than that.
So unless you're going the OB way, or with a 2-way using subwoofer + compression driver (!) the resonances will be an issue. Especially audible in the 200hz-5khz zone.
And I suspect the 200hz-2khz to be the worst.
Brainstorming here, but my educated-guess would suggest that large Sd and low Mms are the worst in regards of those resonances.
Very lightweight cones.... Thin material... Medium to large cone surface...
Basically all high efficiency PA and fullrange drivers.
But then again, is a ''thick'' cone will act as a sound barrier much more than a thin cone?
Most probably, a thick ''W'' sandwich cone from some massive subwoofer driver will block high frequencies much better than a flimsy wideband cone, but to what extent? 4-5db less but just above 3,000hz? More? Less? And since it's a subwoofer and not used in the mid frequencies....
Again: I think there is no comparison possible with a 25mm plywood or MDF box wall.
Back 20 years ago, when I was doing SPL-drag racing, I remember the most successful vehicles were doing 160db and more, while having just a little ''uhmm'' noise on the outside, probably not more than 90db.
That's a 70db sound barrier!
Cars and trucks mostly made of metal and glass!
Can't believe the enclosure's walls are the culprit here. Even 15mm plywood or MDF should suffice to stop most of the sound energy from leaking outside.
Very lightweight cones.... Thin material... Medium to large cone surface...
Basically all high efficiency PA and fullrange drivers.
But then again, is a ''thick'' cone will act as a sound barrier much more than a thin cone?
Most probably, a thick ''W'' sandwich cone from some massive subwoofer driver will block high frequencies much better than a flimsy wideband cone, but to what extent? 4-5db less but just above 3,000hz? More? Less? And since it's a subwoofer and not used in the mid frequencies....
Again: I think there is no comparison possible with a 25mm plywood or MDF box wall.
Back 20 years ago, when I was doing SPL-drag racing, I remember the most successful vehicles were doing 160db and more, while having just a little ''uhmm'' noise on the outside, probably not more than 90db.
That's a 70db sound barrier!
Cars and trucks mostly made of metal and glass!
Can't believe the enclosure's walls are the culprit here. Even 15mm plywood or MDF should suffice to stop most of the sound energy from leaking outside.
Last edited:
You can measure the sound coming back through the cone from reflections inside the cabinet. I did this using a function generator, a microphone and a scope. Use the function generator to make clicks (square wave and a small series capacitor), then measure the them with the microphone. On mine, I got three peaks on the scope; the first was direct sound, the second was reflected off the magnet and the third was reflected off the back of the cabinet. The principle is here: https://www.diyaudio.com/forums/multi-way/321128-innerchoic-cabinets-6.html#post5400210
Brian
Brian
Oh it can, but you have to give up efficiency. Often as much as 10 dB. The bass is still there, you just have to bring the rest down to match it.
Well that is a problem. 🙂 But at 19" wide baffle with some wings can do the trick. That's too big for most people.
I've heard far too many with no bass. If we're clever it doesn't have to be that way.
According to my test, I was left with no EQ traction whatsoever in the 25-35hz region.
The acoustic short-circuit was simply too strong.
Probably around 40hz is more manageable, I agree.
That can depend on the driver Q, the baffle size and Fs. It's all in what you are willing to trade away. With modest size baffles OB can often be limited to woofer Fs, no matter how hard you try. But I have had OB woofers that were flat to near 20Hz with no EQ. BIG baffles, tho. I've been very happy with 15s on OB. 18s and 21s, even better.
But I also like bass reflex, so there you go. 😉
But I also like bass reflex, so there you go. 😉
A pair now sadly discontinued MCM 21" woofers. Just about the perfect OB woofer. Big baffle, tho.
OB bass capabilities
As I read this thread I am very disappointed concerning (understanding) what a well engineered OB of reasonable size can do. If you build a sealed or ported box with, lets say for argument, the ability to truly reach 32 Hz before beginning roll off and it does that with a half space efficiency of say, 1- 1.5 percent (92 - 94 db/watt give or take) in a typical residential environment, the box will range from around 5 to 12 cu. ft. in size. This assumes the ability to deliver a minimum of 104-108 db of out put with out exceeding X-max at 32 Hz including room typical room gain.
This can easily be done with a 20" wide shallow winged OB less than 42" high with a total front to back depth of under 16". Way smaller floor foot print than the majority of box designs. And in the typical room the OB will produce much smoother bass response with far less room interaction. It's not even remotely close. The simple low pass passive crossover required usually can be done with 3-5 off the shelf components. Crossover frequency needs to be in the 175-350 Hz range depending on the bass drivers utilized.
As for the concern the OB requires 40-45" of distance from the rear wall, a 5-12 cu. ft. box against the wall will be around 24-30" deep. And truth be told, the box will sound the most balance when place at least 12-18" away from the rear wall. So no space advantage is realized.
Plus, a large, well built box with anything approaching a decent finish is a fortune to build. Not to mention very heavy. Two Sub Zero fridges side by side.
As I read this thread I am very disappointed concerning (understanding) what a well engineered OB of reasonable size can do. If you build a sealed or ported box with, lets say for argument, the ability to truly reach 32 Hz before beginning roll off and it does that with a half space efficiency of say, 1- 1.5 percent (92 - 94 db/watt give or take) in a typical residential environment, the box will range from around 5 to 12 cu. ft. in size. This assumes the ability to deliver a minimum of 104-108 db of out put with out exceeding X-max at 32 Hz including room typical room gain.
This can easily be done with a 20" wide shallow winged OB less than 42" high with a total front to back depth of under 16". Way smaller floor foot print than the majority of box designs. And in the typical room the OB will produce much smoother bass response with far less room interaction. It's not even remotely close. The simple low pass passive crossover required usually can be done with 3-5 off the shelf components. Crossover frequency needs to be in the 175-350 Hz range depending on the bass drivers utilized.
As for the concern the OB requires 40-45" of distance from the rear wall, a 5-12 cu. ft. box against the wall will be around 24-30" deep. And truth be told, the box will sound the most balance when place at least 12-18" away from the rear wall. So no space advantage is realized.
Plus, a large, well built box with anything approaching a decent finish is a fortune to build. Not to mention very heavy. Two Sub Zero fridges side by side.
I have a pair of JBL 2269 I want to clone a pair of SUB18s with. So I'm gathering from this thread that in order to tune the boxes to 20hz flat, It will bring down sensitivity to 82db and require 5000 watts to yield 115db at a 12ft. listening position?