Organically shaped enclosures may be a mere fad. A few expensive loudspeakers, several higher priced home theater systems, and lots of inexpensive computer loudspeakers and mini systems are found in organic shapes. Even a few diy people have embraced organically shaped enclosures. Despite how visually appealing and “cool” organic shapes are, this fad may not be a good thing for accurate acoustic reproduction. The increased use of organic enclosure shapes may not be the product of scientific or rational thinking, but rather something akin to the culture of modern fashion.
If you do not have experience in a modern design organization, you may not have knowledge of the division of labor and hierarchy of design considerations. If you do, then you will know how a criterion like reproduction accuracy can take a back seat to aesthetic design. In support of the proposition that this has happened in loudspeaker design, I hereby examine one organic shape: the egg.
Loudspeaker diaphragms have two sides, a front side and a backside. Both produce sound. While the backside sound is usually much less accurate than the front side, we often ignore it because we stuff it into a box, or baffle it away from the front sound and bounce it off a wall into the ambient domain. This is a design blind spot and we probably should not be doing this.
Even if we enclose the backside sound, it can still impact the front side sound in at least two ways. First, the near field backside sound pressure can load the diaphragm and impact the diaphragm motion and vibration structures. The baffle, sides, and gestalt enclosure can also direct and reflect the backside sound forward and through the diaphragm material. As I demonstrated in my partial modification of the TB W23-972, the basket shapes we use, the motors we mount behind the diaphragms, and the enclosures we mount the driver within can detrimentally impact the sound through both of these mechanism. Whether just a baffle, a baffle with wings, or a full enclosure, its design is important and will impact the front side sound.
Rounded forms replacing corners with curves and tapering instead of flat ends generally characterize organic shapes. Yet such organic forms (particularly the egg) have characteristics of both parabolic reflectors and tubes. Two forms detrimental to acoustic accuracy.
If organic shapes are acoustically superior, then an organic shape (the egg) should outperform a non-organic shape (a cube). In the tradition of a null hypothesis, that is the prediction I hereby test. I have seen the public statement by a builder of a loudspeaker using an egg shaped enclosure that the driver and enclosure combination sounds “just fine to me.” That is fine, and the enclosures are visually appealing, but what do they do to the sound? Is the sound reproduction more accurate than a cube?
The following tests are the performance comparison of a six inch dimensioned cube with an egg shaped enclosure of just under six inches at the largest diameter and just under seven inches long. The enclosure volume of the cube is larger than the enclosure volume of the egg. Both enclosures were filled with spun poly damping material. Only one driver was used for the tests. That driver was moved from enclosure to enclosure. The tests were quasi-anechoic far field impulse.
The dip in response at 2.5 kHz in an otherwise flat response region below four kHz seen in the control response (cube) is caused by a backside reflection off of the basket and magnet assembly and then passing through the cone to merge with the front side decay response.
With the driver in the egg enclosure, this dip is now completely masked by a rising response to 4 kHz marked with three small narrow peaks. As seen in the decay response comparisons, the cause of this rise is a dramatic increase in the decay acoustic output below 8 kHz. The three regular peaks are characteristic of the performance of a diaphragm loaded in a tube. The more wide spread increase in decay output is most likely caused by the parabolic reflector action of the egg bouncing and focusing the internal sound forward along the central axis of the egg.
While the egg appears to smooth small variations in the frequency response, that smoothing is at the expense of increased decay region output. The sonic result is a driver less coherently accurate with more smearing and less detail available in the reproduced sound. It is also a sound that is brighter and harder, yet with less impact and less dynamics to percussive sounds.
As an interesting phenomenon beyond the original scope of the test, we once again have a complication to the idea of baffle step phenomenon. Since there is a decrease in the size of the front baffle with the egg (it completely disappears), baffle step theory would predict a decrease in low frequency output. Yet any decrease is swamped by the increase in output with rising frequency. A baffle step correction pre filter will help with the rising response just as well as it would a decreasing low frequency far field output, so there is no problem with BSC as long as it is tuned correctly, but BSC will not reduce the magnitude of the decay response in the egg shaped enclosure.
Rational thinking (or science) is falsifiable. Science, whether ordinary or revolutionary, is about making predictions that can be tested and possibly proven false. If a statement or proposition is composed so that it cannot be falsified, it is neither scientific nor rational. That is why so little weight is given to anecdotal evidence. When someone bases a conversation upon the proposition that a loudspeaker “sounds fine to me,” the rules of the conversation are irrational rules. Such statements can never be disproved. If you knowingly participate in such a conversation, then that is fine. Indeed, the majority of conversations are based upon such propositions. For me, however, I am seeking something more than the ordinary in both loudspeaker designs and in conversation threads.
What are the lessons of this experiment? I suppose one lesson is that there are an almost infinite number of variables in loudspeaker designing. And while all of those variables are important, some are more important than others, and which are more important varies upon the specifics of any given design. The second is the low value of anecdotal evidence. The third is embodied in an English language proverb about the pitfalls of assumptions. I will not include the proverb here since it involves scatological play on words.
In sum, do not fear the box unnecessarily and do not embrace organic shapes on faith alone.
Good designing,
Mark
If you do not have experience in a modern design organization, you may not have knowledge of the division of labor and hierarchy of design considerations. If you do, then you will know how a criterion like reproduction accuracy can take a back seat to aesthetic design. In support of the proposition that this has happened in loudspeaker design, I hereby examine one organic shape: the egg.
Loudspeaker diaphragms have two sides, a front side and a backside. Both produce sound. While the backside sound is usually much less accurate than the front side, we often ignore it because we stuff it into a box, or baffle it away from the front sound and bounce it off a wall into the ambient domain. This is a design blind spot and we probably should not be doing this.
Even if we enclose the backside sound, it can still impact the front side sound in at least two ways. First, the near field backside sound pressure can load the diaphragm and impact the diaphragm motion and vibration structures. The baffle, sides, and gestalt enclosure can also direct and reflect the backside sound forward and through the diaphragm material. As I demonstrated in my partial modification of the TB W23-972, the basket shapes we use, the motors we mount behind the diaphragms, and the enclosures we mount the driver within can detrimentally impact the sound through both of these mechanism. Whether just a baffle, a baffle with wings, or a full enclosure, its design is important and will impact the front side sound.
Rounded forms replacing corners with curves and tapering instead of flat ends generally characterize organic shapes. Yet such organic forms (particularly the egg) have characteristics of both parabolic reflectors and tubes. Two forms detrimental to acoustic accuracy.
If organic shapes are acoustically superior, then an organic shape (the egg) should outperform a non-organic shape (a cube). In the tradition of a null hypothesis, that is the prediction I hereby test. I have seen the public statement by a builder of a loudspeaker using an egg shaped enclosure that the driver and enclosure combination sounds “just fine to me.” That is fine, and the enclosures are visually appealing, but what do they do to the sound? Is the sound reproduction more accurate than a cube?
The following tests are the performance comparison of a six inch dimensioned cube with an egg shaped enclosure of just under six inches at the largest diameter and just under seven inches long. The enclosure volume of the cube is larger than the enclosure volume of the egg. Both enclosures were filled with spun poly damping material. Only one driver was used for the tests. That driver was moved from enclosure to enclosure. The tests were quasi-anechoic far field impulse.
The dip in response at 2.5 kHz in an otherwise flat response region below four kHz seen in the control response (cube) is caused by a backside reflection off of the basket and magnet assembly and then passing through the cone to merge with the front side decay response.
With the driver in the egg enclosure, this dip is now completely masked by a rising response to 4 kHz marked with three small narrow peaks. As seen in the decay response comparisons, the cause of this rise is a dramatic increase in the decay acoustic output below 8 kHz. The three regular peaks are characteristic of the performance of a diaphragm loaded in a tube. The more wide spread increase in decay output is most likely caused by the parabolic reflector action of the egg bouncing and focusing the internal sound forward along the central axis of the egg.
While the egg appears to smooth small variations in the frequency response, that smoothing is at the expense of increased decay region output. The sonic result is a driver less coherently accurate with more smearing and less detail available in the reproduced sound. It is also a sound that is brighter and harder, yet with less impact and less dynamics to percussive sounds.
As an interesting phenomenon beyond the original scope of the test, we once again have a complication to the idea of baffle step phenomenon. Since there is a decrease in the size of the front baffle with the egg (it completely disappears), baffle step theory would predict a decrease in low frequency output. Yet any decrease is swamped by the increase in output with rising frequency. A baffle step correction pre filter will help with the rising response just as well as it would a decreasing low frequency far field output, so there is no problem with BSC as long as it is tuned correctly, but BSC will not reduce the magnitude of the decay response in the egg shaped enclosure.
Rational thinking (or science) is falsifiable. Science, whether ordinary or revolutionary, is about making predictions that can be tested and possibly proven false. If a statement or proposition is composed so that it cannot be falsified, it is neither scientific nor rational. That is why so little weight is given to anecdotal evidence. When someone bases a conversation upon the proposition that a loudspeaker “sounds fine to me,” the rules of the conversation are irrational rules. Such statements can never be disproved. If you knowingly participate in such a conversation, then that is fine. Indeed, the majority of conversations are based upon such propositions. For me, however, I am seeking something more than the ordinary in both loudspeaker designs and in conversation threads.
What are the lessons of this experiment? I suppose one lesson is that there are an almost infinite number of variables in loudspeaker designing. And while all of those variables are important, some are more important than others, and which are more important varies upon the specifics of any given design. The second is the low value of anecdotal evidence. The third is embodied in an English language proverb about the pitfalls of assumptions. I will not include the proverb here since it involves scatological play on words.
In sum, do not fear the box unnecessarily and do not embrace organic shapes on faith alone.
Good designing,
Mark
An externally hosted image should be here but it was not working when we last tested it.
Illuminating work as always Mark.
You may have stated this in another thread, but, how do you arrive at the "decay spectrum?" Is it a slice taken starting some arbitrary point after T0? If so, how do you determine when to start it?
It would be interesting to apply this kind of analysis where the variable is enclosure damping. I would imagine there are some subtlties to what is going on with damping even when the frequency response becomes flatter.
You may have stated this in another thread, but, how do you arrive at the "decay spectrum?" Is it a slice taken starting some arbitrary point after T0? If so, how do you determine when to start it?
It would be interesting to apply this kind of analysis where the variable is enclosure damping. I would imagine there are some subtlties to what is going on with damping even when the frequency response becomes flatter.
Circumstantial evidence is more often far more accurate then scientific evidence, one thing that nature has or should have taught man by now is that she is a far better designer then we are, nature did not design the box, we did. That being said, I do agree that there are too many variables, and all the scientific scrutiny will likely raise more questions then answers. Most of us no- professional designers prefer to experiment with the recourses at hand, usually our ears, and others experiences. I made several different shaped enclosures and did a similar far less scientific experiment with my ears. I found that deeper non-parallel sided enclosures sounded better then shallow parallel boxes and cubes. Spheres sounded very smooth but the deeper taper polygon with lots of sound absorption material sounded best to my ears, so I went with that. I think an elongated egg shape would have done better, but at the time I had no recourses to make such an enclosure, however now I do and I’m already working on a center channel with a somewhat egg shape, much deeper although. Preliminary tests are good and I’m working on a acoustic measurement system so I can dig a little deeper, but in the end I will still rely on my ears.
My Diy Speakers
http://www.geocities.com/kingdaddykeith/Kingdaddys_DIY_Projects.html?1078331805978
My Diy Speakers
http://www.geocities.com/kingdaddykeith/Kingdaddys_DIY_Projects.html?1078331805978
were you refering to this egg?:
http://www.vaessen.com/products.html
oh by the way, my first post here! Hi all!!
http://www.vaessen.com/products.html
oh by the way, my first post here! Hi all!!
If I understand correctly, this is a single measurement taken in the far-field, presumably on-axis? Due to the complex nature of baffle edge diffraction, I would expect to see additional significant variations between these enclosure types when off-axis measurements are included.
The presented measurements tell a good story, but I suspect not the whole story. Good work... and I'd love to see a continuation of more comprehensive testing. Maybe I should get off my butt one day and do it!
The presented measurements tell a good story, but I suspect not the whole story. Good work... and I'd love to see a continuation of more comprehensive testing. Maybe I should get off my butt one day and do it!
Could you please provide more information the enclosures you chose? How they were constructed, how close they were to each other. You say that the difference in the responses is due solely to the effect of the shape, yet admit that there is a volume difference and you don't mention the materials of each.
I enjoy my creations and I have never said that the design was superior...I have stated that I do in-fact enjoy the sound, but never that they were better than something else.
Here was the disclaimer I gave in one of my design threads, because I could feel the angst in the air...
I think we'll stick together and take our chances
I enjoy my creations and I have never said that the design was superior...I have stated that I do in-fact enjoy the sound, but never that they were better than something else.
Here was the disclaimer I gave in one of my design threads, because I could feel the angst in the air...
I think we'll stick together and take our chances
what wereu using to make such excellent measurements?
I actually spent several hours last week looking for interestingly shaped large plastic containers, amphora, and similar with no success. If anyone finds some of these kinds of things, with thick plastic walls, I would be highly interested! In my school, we even had some 5 foot tall polygon plastic shapes in the playground with two holes in them for crawling around them. These would be great to get hold of! Any ideas?
Perhaps one of the best ideas is to combine the inherent stability of ovoids with their reflective properties. You should calculate whether the reflections of the speaker cone will land back on the speakers, or if they will bump into each other at a focal point and cancel each other out with a perfect effect. Was this taken care of them planning of placements?
Someone please tell me where I can get some some heavy duty large vase shaped plastic things that I can stack on top of each other to create conjoined ovoids.
I actually spent several hours last week looking for interestingly shaped large plastic containers, amphora, and similar with no success. If anyone finds some of these kinds of things, with thick plastic walls, I would be highly interested! In my school, we even had some 5 foot tall polygon plastic shapes in the playground with two holes in them for crawling around them. These would be great to get hold of! Any ideas?
Perhaps one of the best ideas is to combine the inherent stability of ovoids with their reflective properties. You should calculate whether the reflections of the speaker cone will land back on the speakers, or if they will bump into each other at a focal point and cancel each other out with a perfect effect. Was this taken care of them planning of placements?
Someone please tell me where I can get some some heavy duty large vase shaped plastic things that I can stack on top of each other to create conjoined ovoids.
Incidentally, there is virtually nothing in nature that compares to a loudspeaker. Cricket's legs and vocal chords are the closest parallel. Namely our vocal chords and parrot vocal chords. It is another dimensional sound altogether.
The ear on the other hand far exceeds any microphone ever produced, the crossover in your brain actually does tremendous noise cancellations etc to counter the reverberations of your cranium
What are the most important factors in the form of the Cabinet?
If it is just internal reverberations and padding, perhaps the optimum for sound pressure waves, which are also rather unbiological, would be a cone shaped with a rubber damper membrane at the point of the cone.
The ear on the other hand far exceeds any microphone ever produced, the crossover in your brain actually does tremendous noise cancellations etc to counter the reverberations of your cranium
What are the most important factors in the form of the Cabinet?
If it is just internal reverberations and padding, perhaps the optimum for sound pressure waves, which are also rather unbiological, would be a cone shaped with a rubber damper membrane at the point of the cone.
Mark,
this is good news for most of us trying to build "a better speaker".
while you have spoekn at length about the egg i would like to know if you have done any similar expriements with the type of enclosures of B&W Nautilus and those made my Micheal Barnes at Norh (www.norh.com).
this is good news for most of us trying to build "a better speaker".
while you have spoekn at length about the egg i would like to know if you have done any similar expriements with the type of enclosures of B&W Nautilus and those made my Micheal Barnes at Norh (www.norh.com).
I would just like to point you to the 1951 JAES article by Harry Olsen, which is sited in 'the LoudSpeaker Design Cookbook'. In that article the author tested different shapes and found the cube to have what is discribed as 'constantly undulating [FR] with 5dB variations' as opposed to the sphere which was only +/-.5dB. I don't know the exact detail, such as the construction of the box, or the method of measurement I just thought that you might find that interesting.
Joe
Joe
There are four things we have to care about, regarding enclosures (I am not talking about any sort of waveguides, in order to not make it even more complicated):
1.) Diffraction
2.) Internal standing waves
3.) Vibrations (i.e. wall-flexing)
4.) Ease of manufacture (i.e. price)
A square box only wins in category 4 !!!!!!!
Unfortunately that is the reason why my boxes so far were of this type as well !
Regards
Charles
1.) Diffraction
2.) Internal standing waves
3.) Vibrations (i.e. wall-flexing)
4.) Ease of manufacture (i.e. price)
A square box only wins in category 4 !!!!!!!
Unfortunately that is the reason why my boxes so far were of this type as well !
Regards
Charles
hello!phase_accurate said:
I think we're talking about low fr. part of sound waves.
this frequency have to be commensurable quantities about wave-length.
it means ..... metres. so,
1.) Diffraction
2.) Internal standing waves - problem for middle range of waves.
about 3.) Vibrations (i.e. wall-flexing) - we have to make box using strong materials.
Josephjcole said:
We can't, looking only at the FR measurement of a system, deduce what part of the FR variance is diffraction versus standing waves. But my response was directed at your comment about square baffle shape causing up to 5dB response variations. Here, the variable under test was only the baffle shape, and 100% of the variation was due to diffraction. That wasn't what Mark was looking at--his argument was more like "given an 'ideal' enclosure shape, do the negative effects of the internal shape swamp the positives of the external?"
Now, if you do a test like Mark did, you can take a look at the comparison between enclosures. Rather than smoothing out any humps the output is increased across a broad portion of the spectrum in the 'egg' case. This indicates to me that 'standing wave' effects are swamping diffraction effects in terms of the difference between these enclosures. Having access to more data would improve our ability to analyze it.
phase_accurate said:
This sounds fair enough, it's of particular interest to me since I'm currently agonising over the cabnet design of my 3 way main speakers.
1.) Diffraction;
So what does everyone consider to be a "good" external shape? , I'm personally leaning towards either a cylinder or truncated cone (with drivers suitably flush mounted)
2.) Internal standing waves;
This area I'm mainly only worried about the midrange, I make bass enclosures for a living so I have that area taken care of to my satisfaction, but the critical midrange (5'' peerless mid) area is still giving me concerns.
3.) Vibrations (i.e. wall-flexing);
I plan to use a spaced out fibreglass skin (external shape) over a heavily braced timber internal enclosure, with sand filling the cavity, once a suitable midrange enclosure shape is found this is pretty straight forward.
4.) Ease of manufacture (i.e. price);
hea, its diy afterall
opinions, ideas, more than welcome.
from page 23 of Olson's Acoustical Engineering:
dave
An externally hosted image should be here but it was not working when we last tested it.
dave
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.