Some people are building curved enclosures (like the one in this picture) rather than the usual coffin shape. There must be a good technical reason for doing this and I am wondering if someone out there could explain.
I'm going to build a couple of three-way speaker enclosures and would make them curved if it's worth all the effort.
I'm going to build a couple of three-way speaker enclosures and would make them curved if it's worth all the effort.
Attachments
Visually curved shapes reduce the appears of size. While acoustics plays a role in curved cabinets, I believe the primary one is appearance. It looks more visually appealing, and has the effect of reducing the visual presence a speaker has in a room. My brother, who works as a communications designer, told me that while at Syracuse University, they spent multiple weeks just on the subject of how the eye wraps around various shapes, and how best to design things for certain visual impacts, or in this case, lack of visual impact.
I'm a little bemused as why you don't beleive the standing wave theory. Surely there has been enough scientific research into this. The sphere being the perfect shape for a loudspeaker.
Curved loudspeaker cabinets have been hard and expensive to make. Now with the improved technology the enclousures can be manufactured at a very resonable price.
Curved loudspeaker cabinets have been hard and expensive to make. Now with the improved technology the enclousures can be manufactured at a very resonable price.
Unfortunately the curved sided "reduced standing waves" isn't really correct... you do get a change vs. parallel wall boxes, but you don't get a significant reduction in total. Besides the real problem isn't "standing waves" as much as it is simple reflected energy. At LF the shape of the box has almost no effect. As you go higher in freq the shape starts to have an effect, but if you've got no absorption on the inner walls, then the level of internal reflected energy is so high that I'd not worry much about "standing waves" at all...
One can compare the internal effect of "shape" with the external effect of shape vs. freq response... except (again) most external surfaces can not be made absorptive, whereas internal surfaces can. And the problems are related to freq vs. dimensions, where higher freqs can be absorbed or attenuated, and lower freqs don't care about the smaller dimensions vs. freq...
I vote for the aesthetic design being the driving force, and that CNC routers make the more complex box shapes feasible from an economic and production point of view.
Fwiw, I don't subscribe to the "egg" or "sphere" is best school of thought when we're talking about interal reflections or resonances... for the external shape, there is clearly a difference, especially as the size of the egg/sphere goes up WRT frequency.
Just my opinion... ymmv.
_-_-bear
One can compare the internal effect of "shape" with the external effect of shape vs. freq response... except (again) most external surfaces can not be made absorptive, whereas internal surfaces can. And the problems are related to freq vs. dimensions, where higher freqs can be absorbed or attenuated, and lower freqs don't care about the smaller dimensions vs. freq...
I vote for the aesthetic design being the driving force, and that CNC routers make the more complex box shapes feasible from an economic and production point of view.
Fwiw, I don't subscribe to the "egg" or "sphere" is best school of thought when we're talking about interal reflections or resonances... for the external shape, there is clearly a difference, especially as the size of the egg/sphere goes up WRT frequency.
Just my opinion... ymmv.
_-_-bear
delphiplasma said:
Let's start here... the difference between a good speaker and a really good speaker is what it does WAY down (for the sake of a number, to quote Bill Perkins, 40 dB down (from the average signal)).
Curved (and slanted or pyrimidal) boxes 1st started in the high end, and it was primamrily for the sonic benefits.
They do reduce the standing waves by creating non-parallel walls. (a slant is probably better than a curve in this respect, but only marginally). Instead of only having the potential for a single high Q resonance, you end up with a much lower Q wider bandwidth resonance problem. This is much less obvious sonically, if excited. To be excited the driver needs o be emitting sound at frequencies that could excite the resonance.
That brings us to reason number 2... a curved wall is -- all things equal -- stiffer than a flat panel (much stiffer in most cases), so it resonates at a higher frequency so is much less likely to get excited (especially if the driver's bandpass does not extend up that high). As an example lets consider the big B&W 800 Nautilus series... the bass drivers are not going high enuff to exite side-to-side standing waves so the sides are all about stiffness (from an engineering POV*). You'll note that these boxes also have non-parallel top & bottom where the woofer might get high enuff to excite a standing wave. An interesting consequence of the curved sides is that it is generally easier to make them with multiple layers of thing plywood, which further stiffens things.
(and this takes us to 3, aesthetics... that these are generally more pleasing than a rectangular money coffin is what got then past the cost accountants.
Once the high end exposed the concept to the general public, we find that the price point keeps coming doen, and in a lot of cases it is purely the looks doing it... a curve created with a kerf for instance throws away alot of the stiffness advantages. But not always... look at the PE boxes... they have rectangular ones that jut aren't very good (they need lots of added bracing). To get the new cury ones, the sides are made of 6 lyers of 1/8" MDF (ie MDF plywood) and the walls are dramatically stiffer. These are quite reasonable boxes -- i'd replace the baffles with ones made of ply, but the target audience is unluckly to have the facilities to do that. The MDF is on the soft ide and it is fairly easy to damage them.
dave
If done well, curved designs are stiffer than flat-walled ones. The diffraction properties of their surfaces are different as well.
To quote a famous audio guru: "No one builds pressure vessels and submarines with flat walls - so why should a speaker be built with flat walls ?"
To be honest: I also build my speakers with flat walls for two reasons:
- simplicity
- lazyness
Regards
Charles
To quote a famous audio guru: "No one builds pressure vessels and submarines with flat walls - so why should a speaker be built with flat walls ?"
To be honest: I also build my speakers with flat walls for two reasons:
- simplicity
- lazyness
Regards
Charles
Totally. A spherical internal shape just creates one single honking great resonance. Worst possible shape unless the driver used doesn't operate at the resonance's frequency (or, presumably, have any harmonics there)
B&W seem to do something clever by cutting a hole in the sphere opposite the driver and attaching their exponentially tapered tube to it (as in the nautilus speakers). Not entirely sure how it works, but it looks interesting.
B&W seem to do something clever by cutting a hole in the sphere opposite the driver and attaching their exponentially tapered tube to it (as in the nautilus speakers). Not entirely sure how it works, but it looks interesting.
Baffle step
I'm surprised that no-one has mentioned the baffle step yet, so I'll chirp in...
Spherical enclosures have a very smooth baffle step response - they can, in theory, be equalized to remove the effect of the baffle step completely.
Enclosures with a flat baffle show very uneven transition between 2PI and 4PI radiation and may be considered inferior in this respect.
I'm surprised that no-one has mentioned the baffle step yet, so I'll chirp in...
Spherical enclosures have a very smooth baffle step response - they can, in theory, be equalized to remove the effect of the baffle step completely.
Enclosures with a flat baffle show very uneven transition between 2PI and 4PI radiation and may be considered inferior in this respect.
Somehow the idea that all of this depends on wavelength, and therefore dimension seems to be lost in the discussion. It's important, and critical if you want to make an intelligent decision or understand it at all.
As far as -40db down being important, I'd say that it is, but show me an enclosure that has the rear reflected energy down -40db? That's not so simple to do. Impossible the way most enclosures are built internally - what's in them. In fact the lower in freq u go, the less effect anything you do inside the box has...
So what are we talking about?
And, let's understand that a standing wave means nothing unless it is radiated through the diaphragm and out of the box... what we're really talking about is reflected energy, and if a standing wave at some frequency will cause a peak or dip in the response.
Let's take a look at the dimensions vs. frequency vs. wavelength for a standing wave in the audio band, and see what frequency the box dimension starts to be an issue... and what absorption can be effective (and how much) at that freq before worrying about fancy internal shapes.
Put this way, volumes of air don't notice non-parallel walls until the frequencies get fairly high, at which point if you have not yet absorbed the rear radiation, imho, you've built a poor enclosure...
Or, which would you rather have - a volume taken away by a curved side or filled with a highly absorptive material that works lower in frequency as its thickness dimension increases??
You decide.
I know what I would pick.
And, stiffness is only valuable if the material does not re-radiate sound via transmission, no matter what the frequency. It could be argued that if you had a choice between an enclosure that radiated no mid and higher frequencies, but radiated LF, compared to a cabinet that radiated mid and highs (stiff, eh?) and few lows that the former would sound better overall.
Obviously the ideal enclosure has no baffle step, radiates nothing, and has zero reflected internal energy... anyone got one?
_-_-bear
PS that's monkey coffin not money coffin...
As far as -40db down being important, I'd say that it is, but show me an enclosure that has the rear reflected energy down -40db? That's not so simple to do. Impossible the way most enclosures are built internally - what's in them. In fact the lower in freq u go, the less effect anything you do inside the box has...
So what are we talking about?
And, let's understand that a standing wave means nothing unless it is radiated through the diaphragm and out of the box... what we're really talking about is reflected energy, and if a standing wave at some frequency will cause a peak or dip in the response.
Let's take a look at the dimensions vs. frequency vs. wavelength for a standing wave in the audio band, and see what frequency the box dimension starts to be an issue... and what absorption can be effective (and how much) at that freq before worrying about fancy internal shapes.
Put this way, volumes of air don't notice non-parallel walls until the frequencies get fairly high, at which point if you have not yet absorbed the rear radiation, imho, you've built a poor enclosure...
Or, which would you rather have - a volume taken away by a curved side or filled with a highly absorptive material that works lower in frequency as its thickness dimension increases??
You decide.
I know what I would pick.
And, stiffness is only valuable if the material does not re-radiate sound via transmission, no matter what the frequency. It could be argued that if you had a choice between an enclosure that radiated no mid and higher frequencies, but radiated LF, compared to a cabinet that radiated mid and highs (stiff, eh?) and few lows that the former would sound better overall.
Obviously the ideal enclosure has no baffle step, radiates nothing, and has zero reflected internal energy... anyone got one?
_-_-bear
PS that's monkey coffin not money coffin...
planet10 said:
Well, if this is a problem and everybody is going down the road of thicker and heavier enclosure walls how about thinking of something different and possibly better
I live on a very busy main road and without double glazed windows I would never get to sleep at night, it's so noisy. I am quite amazed how quiet my bedroom is, just relieving the pressure on the window catch: ever so slightly, allowing the airtight seal to just break and it is quite amazing how much noise comes galloping in through this miniscule gap.
To my mind the gap between the double glass facings is attenuating the sounds from the road dramatically. Couldn't this apply to speaker cabinet making?
By building enclosure walls with say two 5 mm sheets of ply wood with a 10/15 mm air gap in between. Of course the gap must be 100% airtight.
Or alternatively filling the gap with some inert substance such as sand.
Over to you.
Those Nautilus tweeters
The idea behind them is to make 0 loading by air pressure on the tweeter, to allow movement more freely. Like those speakers that have channels inside the speaker cabinet which makes the pressure dissipate as it goes through the chanel.
However, the curved sides imporve stiffness because they are tense. Think about it - those panels used to be flat. They they were bent. This adds tension. Meaning there will be less audiable resonance stuff. Because they are more tense, it will be harder to shift them (to create the soundy-resonancy peak).
Basically there's less resonancing to worry about
The idea behind them is to make 0 loading by air pressure on the tweeter, to allow movement more freely. Like those speakers that have channels inside the speaker cabinet which makes the pressure dissipate as it goes through the chanel.
However, the curved sides imporve stiffness because they are tense. Think about it - those panels used to be flat. They they were bent. This adds tension. Meaning there will be less audiable resonance stuff. Because they are more tense, it will be harder to shift them (to create the soundy-resonancy peak).
Basically there's less resonancing to worry about
cirrus18 said:
If you could somehow float the internal enclosure within the external, then you might have something. As it is, they would have to be physically connected at mulitple points, and therefore still transmit the resonance you're trying to eliminate.
Not to mention, how would you deal with baffle of the external enclosure?.... ....some airtight membrane..?
There have been many good points made on this thread thus far. Bear talks of the lost space (compared to a rectangular cab) which could be used for absorbtive material for better internal reflection/standing wave performance.
But, the rounded enclosures still retain the stiffness and diffraction benefits. Also, early internal reflections will tend to "bounce around" more before re-exiting, since there will be widely distributed incidence angles for the backwave components. This will mean more energy is likely retained within the cab and dissipated or at least delayed prior to re-exit.
But, the rounded enclosures still retain the stiffness and diffraction benefits. Also, early internal reflections will tend to "bounce around" more before re-exiting, since there will be widely distributed incidence angles for the backwave components. This will mean more energy is likely retained within the cab and dissipated or at least delayed prior to re-exit.
smokinghot said:
Yes, but it wouldn't be connected everywhere, perhaps only 15/20% which would have these same sound transmission characteristics as a solid walled cabinet so you haven't lost or gained anything there. The remaining odd 80% or so would have very low transmission, so therefore over all, one has gained. If my theory is correct that is.
Anybody want to test it out?
Similarly with the speaker baffle, this could be of similar sandwich construction, sealed around the edges with say half inch wide 12 mm thick strips of wood. For the speaker cut out, a piece of 12 mm thick MDF which would overlap inside the sandwich once the speaker hole has been cut out by perhaps 1 inch, leaving something solid for the speaker to be mounted to.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.