I’d like to introduce some new, partly-developed speaker performance metrics I’ve been chewing on. These could be developed much further with the help of others. I’d love to hear your ideas.
Box types suffer from stereotypes. “Sealed boxes are inefficient” or “Transmission lines are oversized” or “Dipoles have no bass” or “Horn speakers have terrible frequency response.” And while those reputations hint at certain truths, great designs defy the stereotypes. It always comes down to quality of implementation. The less-common box types are often made with drivers that were optimized for some other box type, and sometimes fall far short of their potential.
Each box type just applies different constraints to the same fundamental limits. The laws of physics still apply.
So for example sealed and reflex designs operate on 3D volume. TLs and horns operate on length and area. Open Baffles have the interesting quality of operating on 2D surface area.
Trading off 2D surface area against 3D volume is super interesting. It means if you have a woofer with low depth (such as the Stereo Integrity BM-11) you can create a large speaker with low F3 that still takes up very little 3D volume in liters; and has a very slim profile.
This could be reduced to a “Hoffman’s Iron Law-Like” formula, a metric like:
Acoustic watts per liter per Hz (3D box)
vs
Acoustic watts per square meter per Hz (2D Open Baffle)
[these metrics are my very intuitive guesses; once properly formalized they will look somewhat different]
and these could be further refined into practical engineering heuristics, like
dB per liter per Hz (3D)
vs
dB per liter per Hz (2D)
and further
dB per kilogram per Hz (3D)
vs
dB per kilogram per Hz (2D)
and
dB Hz per dollar (3D)
vs
dB Hz per dollar (2D)
and
dB Hz per watt (3D)
vs
dB Hz per watt (2D)
This helps us formulate some interesting new questions.
If you define the problem as WEIGHT / MASS per unit of acoustical output, OB’s can become extremely attractive under certain circumstances. The stereotype is that they are expensive but commercially OBs could be quite economical.
So for example 3D volume-based box speakers always have to have 6 sides.
But 2D surface area-based speakers (dipoles) can have as few as 1 side.
A “lambda” shaped U-frame has a front panel, triangular wings, no top, no bottom, and no back. This delivers CONSIDERABLE savings in weight and shipping cost (which is a huge contributor to the price of a commercial speaker system!!!)
It’s a lot less lumber as well. You spend more on drivers and less on wood and labor for construction. The Walnut Dipoles can hold their own in SPL bass output very well with similarly sized speakers of other box types. Small but mighty.
Economically it is often less expensive to buy a 15” or 18” woofer and mount it on a big slab of live edge wood, than to buy an 8” or 10” woofer, build a 3D box and cover it with veneer. (And put a port in it.)
Plus the result is a lot closer to a piece of furniture that people will still enjoy 40 years in the future, than a conventional box where the veneer gets nicked and scratched.
Then there is the difference between voltage sensitivity and efficiency. I ran simulations and showed actual examples of a “passive bass boost” circuit which boosts dipole bass 4-7dB. It does so by essentially converting voltage into current. The speaker with LC circuit has an impedance dip at its low frequency cutoff instead of a peak. It’s not technically more efficient, but it draws more current from the amps so it’s more sensitive in the low bass. Sometimes it’s easier and less expensive than DSP.
I took an acoustics class in college and I could probably work out how to formalize these heuristics if I put a lot of hours into it… but why not make this a group project?
We have some VERY smart people who hang out in this forum, some of whom have extensive modeling, mathematics and acoustics background.
New models and heuristics like this will shed light on what precise kinds of tradeoffs are optimal for each box type. This will lead to better drivers and better-engineered systems.
What do you think?
@Juhazi @CharlieLaub @nc535 @gedlee @Studley @Balthazarp
Box types suffer from stereotypes. “Sealed boxes are inefficient” or “Transmission lines are oversized” or “Dipoles have no bass” or “Horn speakers have terrible frequency response.” And while those reputations hint at certain truths, great designs defy the stereotypes. It always comes down to quality of implementation. The less-common box types are often made with drivers that were optimized for some other box type, and sometimes fall far short of their potential.
Each box type just applies different constraints to the same fundamental limits. The laws of physics still apply.
So for example sealed and reflex designs operate on 3D volume. TLs and horns operate on length and area. Open Baffles have the interesting quality of operating on 2D surface area.
Trading off 2D surface area against 3D volume is super interesting. It means if you have a woofer with low depth (such as the Stereo Integrity BM-11) you can create a large speaker with low F3 that still takes up very little 3D volume in liters; and has a very slim profile.
This could be reduced to a “Hoffman’s Iron Law-Like” formula, a metric like:
Acoustic watts per liter per Hz (3D box)
vs
Acoustic watts per square meter per Hz (2D Open Baffle)
[these metrics are my very intuitive guesses; once properly formalized they will look somewhat different]
and these could be further refined into practical engineering heuristics, like
dB per liter per Hz (3D)
vs
dB per liter per Hz (2D)
and further
dB per kilogram per Hz (3D)
vs
dB per kilogram per Hz (2D)
and
dB Hz per dollar (3D)
vs
dB Hz per dollar (2D)
and
dB Hz per watt (3D)
vs
dB Hz per watt (2D)
This helps us formulate some interesting new questions.
If you define the problem as WEIGHT / MASS per unit of acoustical output, OB’s can become extremely attractive under certain circumstances. The stereotype is that they are expensive but commercially OBs could be quite economical.
So for example 3D volume-based box speakers always have to have 6 sides.
But 2D surface area-based speakers (dipoles) can have as few as 1 side.
A “lambda” shaped U-frame has a front panel, triangular wings, no top, no bottom, and no back. This delivers CONSIDERABLE savings in weight and shipping cost (which is a huge contributor to the price of a commercial speaker system!!!)
It’s a lot less lumber as well. You spend more on drivers and less on wood and labor for construction. The Walnut Dipoles can hold their own in SPL bass output very well with similarly sized speakers of other box types. Small but mighty.
Economically it is often less expensive to buy a 15” or 18” woofer and mount it on a big slab of live edge wood, than to buy an 8” or 10” woofer, build a 3D box and cover it with veneer. (And put a port in it.)
Plus the result is a lot closer to a piece of furniture that people will still enjoy 40 years in the future, than a conventional box where the veneer gets nicked and scratched.
Then there is the difference between voltage sensitivity and efficiency. I ran simulations and showed actual examples of a “passive bass boost” circuit which boosts dipole bass 4-7dB. It does so by essentially converting voltage into current. The speaker with LC circuit has an impedance dip at its low frequency cutoff instead of a peak. It’s not technically more efficient, but it draws more current from the amps so it’s more sensitive in the low bass. Sometimes it’s easier and less expensive than DSP.
I took an acoustics class in college and I could probably work out how to formalize these heuristics if I put a lot of hours into it… but why not make this a group project?
We have some VERY smart people who hang out in this forum, some of whom have extensive modeling, mathematics and acoustics background.
New models and heuristics like this will shed light on what precise kinds of tradeoffs are optimal for each box type. This will lead to better drivers and better-engineered systems.
What do you think?
@Juhazi @CharlieLaub @nc535 @gedlee @Studley @Balthazarp
Last edited:
...and @CharlieLaub, your baffleless OBs on wire mesh screens occupy VERY few liters of space and are very low in weight. Which is a significant advantage in some situations.
Last edited:
I very much wish I heard some of the open baffles at the last pe competition, maybe with my own music choices, didn't find any speaker engaging as I didn't care for the music choices.
My own ob experiments left me feeling like it was a neat thing to play with but just didn't really match my boxed speakers. I've got some pa380s and dx25+ wg300 even.
My own ob experiments left me feeling like it was a neat thing to play with but just didn't really match my boxed speakers. I've got some pa380s and dx25+ wg300 even.
While I’d take wide OB 18” woofers all day long, logistics comprised of available space and domestic discord make them impossible in my home. As the cost of square footage continues to increase disproportionate from just about everything, i can’t see the allure of OB being adopted by many…..and that’s a shame……all of it.
Given the lower cost of watts these days and bass systems not needing much beyond brute force power, the sealed system IMO is the predominant long term survivor of bass alignments……small boxes with built in high power class D amps and user adjustable DSP. I use a SWARM approach in my own home family room with 4 diy examples of these…..smooth in room response down to 18hz at 110db for the cost of around $2k consuming around 6 square feet of total floor space.
Given the lower cost of watts these days and bass systems not needing much beyond brute force power, the sealed system IMO is the predominant long term survivor of bass alignments……small boxes with built in high power class D amps and user adjustable DSP. I use a SWARM approach in my own home family room with 4 diy examples of these…..smooth in room response down to 18hz at 110db for the cost of around $2k consuming around 6 square feet of total floor space.
Last edited:
Two things attracted me to OB. One was the weight savings meant I could implement a high SQ speaker that I could move without straining my back. The other one was that OB requires less treatment in the room, provided one has a big enough room to space it adequately from the wall, which I now do.
One could make a smaller system using cardioid to control directivity, but it would be heavier. The cardioid could be placed closer to the front wall, but it would be more subject to boundary interference, particularly sidewall and ceiling. One nice thing about cardioid is that there are closed boxes within which to hide the amps and DSP. In OB, they have to be separate.
Any system would benefit from multiple subs. The problem is usually waf. If we had good DIY wireless sub connectivity, that might issue might go away.
Unfortunately, OB's wide baffle and requirement to be spaced out from the walls relegates to a niche despite its compelling advantages where it fits. The problem is it fits only a subset of spaces and may fail waf. Developing a metric that highlights its advantages isn't going to change that.
One could make a smaller system using cardioid to control directivity, but it would be heavier. The cardioid could be placed closer to the front wall, but it would be more subject to boundary interference, particularly sidewall and ceiling. One nice thing about cardioid is that there are closed boxes within which to hide the amps and DSP. In OB, they have to be separate.
Any system would benefit from multiple subs. The problem is usually waf. If we had good DIY wireless sub connectivity, that might issue might go away.
Unfortunately, OB's wide baffle and requirement to be spaced out from the walls relegates to a niche despite its compelling advantages where it fits. The problem is it fits only a subset of spaces and may fail waf. Developing a metric that highlights its advantages isn't going to change that.
I am very much with nc535. Even though my installation is "permanent" https://www.diyaudio.com/community/...slim-8-and-visaton-ws25e.364351/#post-7543539, I was able to position them well in our room. I was also able to put them together without having to resort to difficult wood working. At home, we all like them very much. They are not (yet) overengineered as some of the work presented here, but when I have the chance I can sit and listen to them for hours...
Bass reproduction is indeed still a "problem" though. Indeed, sealed subwoofers can be easily made,power and DSP are cheap. Sealed would defeat an immense advantage of these dipoles, and that is that they do not rattle our enormous (door) glasses, and they do not disturb the neighbors (we live in an appartment).
Bass reproduction is indeed still a "problem" though. Indeed, sealed subwoofers can be easily made,power and DSP are cheap. Sealed would defeat an immense advantage of these dipoles, and that is that they do not rattle our enormous (door) glasses, and they do not disturb the neighbors (we live in an appartment).
To complement my previous post about the bass leaking of sealed subs, which is unfortunately a problem for some. @cowanaudio reported positively about the advantages of (nearfield) dipole wrt bass leakage, keeping performance levels that kept him happy. IIRC he is on worldtrip now, but hopefully will be able to share more details once settled with the new system.
This is the only parameter where a Ripole Subwoofer would not be at his advantage : dB per Watt ratio. Otherwise, for all the others ratio mentioned in your post, @perrymarshall, the Ripole principle could compare very favorably to the other type of Subs.
---> Compact (mine below is 400x400x440mm, 2x12")
---> Lightweight
---> very low extension (15Hz for mine below)
---> unobtrusive but wide dispersion (effect of deep floating bass, when placed at the center of the main speakers)
---> no boxy tone or port resonance (there's non)
---> excellent damping and step / transient response (like a QTC=0.5 to 0.71 sealed enclosure)
T
If one need to compare different type of boxes in bass region, best is to compare SPL level per driver cone amplitude, in other way it can be called acoustical efficiency.
Limit is then driver Xmax.
Limit is then driver Xmax.
We can also make a metric of "SPL as function of Xmax." I emphasize that in this article I wrote in AudioXpress, discussing how to get absolutely optimal output from a bass reflex https://audioxpress.com/files/attachment/2721
I always likened an open baffle speaker to the guy trying to save his sinking boat by trying to " bail out" the gushing water with a water-glass.
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------Rick...
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------Rick...
Yes im one of those very smart people who enjoy this forum, and i have two of the most advanced auditory analyisis mechanisims known to man, my ears.they love hearing all sorts of music. I also love diy audio and the way creativity evolves on this and other forums so thanks everyone for your contributions this year, smart people or not.merry xmas and a prosperous new year to you all.k
@tubelectron : you can buy glulam in sheet form now? Last time I checked in the UK it was just being used for beams and roof trusses.
@tubelectron Do you have any more details on that sub? I seem to remember seeing it before, but not sure where.
"...the most advanced auditory analysis mechanisms....my ears..." Oh really? " We have high opinions of our human biology, when in fact we should not".
Neil Tyson Answers "Do You Believe In UFOs?" & Sets The Record Straight!! on Youtube.
---------------------------------------------------------------------------------------------------------------------------------------------------Rick...
Performance metrics are interesting, fun to talk about and occasionally useful, but they are not sound. Good sound is good sound and I have not found any one metric to define it. Qms might be the most interesting to me, but there are so many more variables.
I mostly look at BL/Mms for bandwidth and low Cms for good behavior with no enclosure. These are practical values for basic functionality and in no way equal good sound.
Interesting topic, however...
I mostly look at BL/Mms for bandwidth and low Cms for good behavior with no enclosure. These are practical values for basic functionality and in no way equal good sound.
Interesting topic, however...
I do think that one of the keys to good sound is minimizing resonances. OB systems tend to begin rolling off well above the bass resonance, which is an attribute that I like. So maybe -db of bass resonance as a parameter?
Noticed that tidbit at the bottom on your past work with the automotive industry, pretty cool stuff there!
I saw an earlier post about OB needing less treatment, be curious to hear more about that. When I tried OB I kinda felt like it revealed my room deficiencies to greater extent than my boxes speakers. I almost felt like I needed more treatment, but my driver integration might have been poor.