That is very misleading. An OB just delays any reflections from the inside walls to the walls outside. In a box you are trying to contain, control, and remove the back wave. In an OB you just let it bounce around the room causing other issues.
OB is not a cure-all. One of the “boxiest” speakers i have every heard is an OB.
dave
KEF has used 3D printing and modern computer mathematic techniques to develop a meta-material to create a “maze” for the HFs. This work was, AFAIK, started by Hegeman with this loudspeaker (there is a thread [or more] on the subject). Then teh Hegeman 1, a Speaker BUilder article, and the Morrison loudspeaker, are marks in the evolution of the technique.
dave
Nice pictures. Never seen before.KEF has used 3D printing and modern computer mathematic techniques to develop a meta-material to create a “maze” for the HFs. This work was, AFAIK, started by Hegeman with this loudspeaker (there is a thread [or more] on the subject). Then teh Hegeman 1, a Speaker BUilder article, and the Morrison loudspeaker, are marks in the evolution of the technique.
dave
As important as FR, is what is the frequency response of the same driver 10 dB down, 20 dB down, 30 dB down, 40 dB down form a paying 0dB signal. If they do not mimic the 0dB response the driver will not have as good a DDR.
If your only tool is a hammer, everything gets treated like a nail. A FR measurement, as we see today, is like that hammer. It is only one of many tools. It gets leane don way more than it deserves because it is the tool we have.
dave
Last edited:
King has shown that Bradbury’s work on slowing speed of sound was not in fact what happens.
We are talking about a distribution of modes. Just like a room, but inside the box, usually simplier, and smaller in “time”. Just like a room we are trying to distribute them such as to minimize their impact. A simple rectangular rbox will have 3 primary modes, as you add internal complexity things grow more complicated.
dave
For all the interesting references to open baffles, the extremely small values referred to here are unlikely in most instances to result in much that would normally be called dipole behaviour -especially if there's a gasket present. I'd probably give the slightly more prosaic answer that the most likely outcome will be that you simply bleed off pressure, as you've introduced an air-leak -it just happens to be around the driver. Once upon a time, this was a fairly common means of tweaking output if it was overloading the room. If you go far enough, you may (will) get some cancellations, but 1.5mm is unlikely in many instances to see much in that sense.
Are you isolating the driver from the box mass. This, in the end, can greatly reduce the DRR, as now all (almost) of the rection energy/movement is limted to the basket, the distance it moves will increase, reducing DDR, burying the micro-detail under that movement.
That is why you see the drivers tightly integrated across teh box mass to reduce this movement dramatically and improve DDR.
dave
Thanks to Dave, Gnobuddy, Scottmoose and all of you who contributed so much useful information.
The more you learn, the more complex it becomes. It is understood that manufacturers have a hard time getting a "perfect" driver. However, Alpair has come a long way when it comes to full range.
The more you learn, the more complex it becomes. It is understood that manufacturers have a hard time getting a "perfect" driver. However, Alpair has come a long way when it comes to full range.
I have a faint memory of seeing at least one old speaker design where the woofer was mounted on a separate sub-panel, and that panel was mounted with spacers to stand it slightly forwards from the front face of the speaker enclosure. There was a narrow slot all around the edge of the sub-panel, between itself and the box behind it.
In that design, the idea was to use the resulting radial slot all around the speaker as the port in a ported system. Roughly speaking, the dimensions of the sub-panel set the length of the "port", and by adjusting the spacing between box front and sub-panel, you could vary the cross sectional area of the port, and so vary the port tuning.
It's an interesting idea, I think. Basically, a variant on the more commonly used wide and narrow slot-shaped port. Harder to calculate, though, as the cross sectional area of the "port" varies with radius as you go outwards from the speaker centre.
As a bonus of this approach, moving the woofer slightly forward in this way might help a little bit in time-aligning the woofer and tweeter around the crossover frequency. If you used nominally 1" lumber for the sub-panel (actual thickness 0.75"), and a 0.25" spacing from the front of the box, the woofer is mounted a full inch forward of the tweeter, giving you roughly 75uS of time delay compensation.
Our old rule of thumb was that a simple way to guess at a first-pass value of time delay compensation needed, was to start with the time sound waves take to travel the depth of the woofer cone (from dust-cap to front panel). Typically in the 100 uS ballpark.
-Gnobuddy
Put it this way: if you start out with drivers with ragged and uneven frequency responses, and a gap between usable woofer and tweeter frequency ranges, no amount of faffing about with meta-materials or golden ratios or anything else will help fix the resulting speaker.
You can't polish a turd, as a former co-worker accurately (but inelegantly) put it.
It takes a fair bit of good engineering to produce a driver with a very flat frequency response, and in practice, I've never measured a driver that was very good at flat FR, but very bad in some other key parameter. (Particularly if you also look at polar response, i.e. FR at a few key off-axis angles too.)
I don't mean to question your level of expertise with speaker design, simply to avoid leading Leif in such a way as to miss the forest for the trees. In other words, to start out by focusing on the most important parts, not the minutiae.
The starting point for a good speaker is drivers with good frequency responses, and good control of treble dispersion. Add in a well designed crossover network, put it in a well sized, stuffed, and tuned box, and you have now avoided many of the worst problems that tend to plague a lot of speakers. You already have something good, in other words.
Everything else - chamfered edges, golden ratios, etc, are more about small amounts of fine-tuning to something that is already good. They make a good thing better (but won't make a bad thing good).
IME, golden ratio speakers are not really a thing outside of DIY builds. They waste far too much wood. For production speakers that have to be made in large numbers and sold at competitive prices, what I saw was panel dimensions chosen to provide the right interior volume, but also to fit well into the usual 4'x8' sheet of MDF with minimal wastage, and to have a footprint that let them sit in the sort of space they were likely to end up. For nearfield monitors, that typically means ending up on rather small speaker stands, or worse, being plopped down on a desktop.
-Gnobuddy
Golden Ratio looks cool. They are fun. Audio Note are close to GR. Some of the Devores. But the very price concious boxes will have a material optimization step in the design.
This cut for a 13 litre CGR is pretty good.
And certainly smooth & flat is nice. And if there is an XO the drivers need to overlap sufficiently so as to make for a chance with the XO desogn (the hardest part) to work.
But you see people who love loudspeakers with ragged resonse (woofers in Thor), or less overlap than you would like if you have an XO (Zu), or not that flat FR (Ecplise).
The FR measurement is helpful, but the siituation is much more complex than that one parameter. And one of those parameters (a multi-input one) is the set of compromise that need to go into the speciific loudspeaker for the particular person.
Leif has been playing with drivers that are very good.
dave
This cut for a 13 litre CGR is pretty good.
And certainly smooth & flat is nice. And if there is an XO the drivers need to overlap sufficiently so as to make for a chance with the XO desogn (the hardest part) to work.
But you see people who love loudspeakers with ragged resonse (woofers in Thor), or less overlap than you would like if you have an XO (Zu), or not that flat FR (Ecplise).
The FR measurement is helpful, but the siituation is much more complex than that one parameter. And one of those parameters (a multi-input one) is the set of compromise that need to go into the speciific loudspeaker for the particular person.
Leif has been playing with drivers that are very good.
dave
Last edited:
I wouldnt say that. To build a good baffle can be quite a challenge, because there's inherently less structure available for strength and rigidity.
Yes and most of the popular builds easily circumvent that with a "helper" woofer.
They certainly do that. I consider part of a "boxy" sound is all that noise inside the box coming back at you through the rather acoustically transparent cone area. I see that one hasnt been mentioned yet. With enough SPL inside it comes through the cabinet sidewalls as well; moreso in a thin walled, unbraced enclosure of course.
Yeah. Be it an illusion, along with the other junk gone / fixed It's something that makes me and my ears happy. Wash is a pretty ugly word; from what I hear back there it's not at all chaotic sound spraying every which way.
There's available this program called the edge, which you can use to locate your driver in a better place on the baffle than that which produces the "Open Baffle FR" response, shown. Let it be known all OB designs dont necessarily make that worst-case response...
There are many drivers on the market. Some are amazingly good. Some less good. Sometimes it's fun to get pretty good sound from a little cheaper driver.
Of course, you can not conjure as much as you like.
I have tried both expensive and cheap drivers. Carefully selected from those that exist. Can name a few.
Scanspeak, Seas, Alpair, Visaton, Monacor and Fostex. All brands have given me joy and many good listening hours. I have a passion for building speakers and trying to get the last drops of good sound by building and optimizing. Just for fun.
Of course, you can not conjure as much as you like.
I have tried both expensive and cheap drivers. Carefully selected from those that exist. Can name a few.
Scanspeak, Seas, Alpair, Visaton, Monacor and Fostex. All brands have given me joy and many good listening hours. I have a passion for building speakers and trying to get the last drops of good sound by building and optimizing. Just for fun.
I have received a lot of help from Scottmoose and Dave in my search for good sound.
It has been a blessing for me. They are unique in that they share.
If you think I take too much of your time, you can say no. I'm unemployed and have way too much time to deal with this. 🙂
It has been a blessing for me. They are unique in that they share.
If you think I take too much of your time, you can say no. I'm unemployed and have way too much time to deal with this. 🙂
Last edited:
You may be thinking of either the old R-J enclosure, the similar Eliptoflex or the slightly different though IMO related implementation by Supravox which they also call R-J. They are all basically 6th order bandpass.
I like golden ratio enclosures. They are not that wasteful of material IMO, nowhere near as much as slim and tall floorstanders. In the end, the acoustical design should trump material utilization though, outside of grossly wasteful methods like trans-lam, as pretty as it may look. Unless one plans a use for the cutouts. 😉
Have you built a box with the Golden Ratio?
If you have done that, can you describe the sound they perform?
I have built large 142 liters enlcosures for a 15" and waveguide in the golden ratio and am currently working on near-enough to GR boxes of much smaller dimension for small drivers.
I would not say they have they "own sound". They may avoid coincident modes, assuming these are withing the driver's working bandwidth, but they don't add anything of their own on top of baseline performance, which they in fact should not. IOW, I don't think they sound different from any other form-factor of competent design. This is but one element of consideration in the whole system design.
I would not say they have they "own sound". They may avoid coincident modes, assuming these are withing the driver's working bandwidth, but they don't add anything of their own on top of baseline performance, which they in fact should not. IOW, I don't think they sound different from any other form-factor of competent design. This is but one element of consideration in the whole system design.