Subtle
I'm sorry but this is the truth about the BR enclosures, perhaps a lot people don't want to face it and will throw rotten tomatoes on me.
A sharp rolloff gives an unnatural and bloated LF rendering... those who are concerned should assume their tastes.
Last edited:
IMHO the horn below is considerably better than Klipsch K402, which I have compared it directly to with hours of listening. The directivity is kept lower in frequency and considerably higher compared to the Klipsch horn. However, it doesn't have a very wide dispersion. It's a 90x60° horn, but which is really sufficient to fill most room with correct placement.But, as I say, that is from 2008, over 10 years ago. Aside from the KPT-402-HF, which is somewhat ubobtanium, I can't seem to find anything better. What do I mean better? Like you say, proper loading to down to 500/600 Hz, suitable for a 1.4 or 1.5" exit CD that can go that low, and extend to near 20 kHz. A constant directivity design (no "head in vice" please) with near perfect on and off axis response. And maybe a bit wider pattern than 90 x 50. The M2 waveguide appears to be a contender at 120 x 110 pattern, but seems to struggle below 1 kHz, both in pattern control and rising distortion, even though it is crossed at 800 Hz... No sure if that is the limitation of the 3" D2430K CD or waveguide or both?
Oh yeah and something one can either DIY or purchase would be nice
Attachments
Both have their pros and cons.
Probably the most natural reproduction of sub and mid bass is accomplished by using FLHs.
But those introduce another set of challenges and (subjectively) may not always be to everybody's taste.
From Bjorn Omholt's (other) thread:
I'm still searching why i can so easily localize the subs, the video was not totally meaningless, the sub array sould be very, very cleaverly tuned on the same frequecy range IMHO.
The come back to very high efficiency large dipole woofers is a benediction in term of spatial rendering, the LF section is back to 3D
I known the owner of P.Audio, been to his factory in Thailand many times and measured many of his products. Never have I seen a more blatant misrepresentation of performance than his spec sheets. It was as if they were simply drawn by hand to be what someone hoped they would be.
B&C on the other hand has almost always been right on the money (there was that time in 2012 when they changed the DE250 and didn't tell anyone.) I use what measures best and it was never P.Audio.
Forget the room (audiophiles are particularly good at this) and listen the direct sound of narrow lobes in the nearfield, the sofa will not covered entierly, at the same time threre is no place for a sofa in a bedroom.
I don't think you can use sine waves for any kind of in-room listening analysis.
Low frequency modal interaction negates any useful conclusions.
And high frequency is even worse, as our two ears constructively/destructively alter the single frequency signal.
Just play some sine waves, walk around, tilt your head, and give up
And for subs, a sine wave is a very dense signal, usually generating harmonics and resonances either from the sub itself or from nearby room structures and items (sometimes distant ones too).
I would instead use bandwidth limited pink for a localization test.
As far as sealed vs BR...my experience was to try the bms 18n862 in both.
The BR was tuned to 31 Hz, and I gave the sealed the same tuning via Linkwitz transform.
Outdoors they sounded the same to the point I couldn't tell them apart. The only difference was how much louder the BR could go.
Indoors, I could sometimes tell them apart in terms of which one sounded deeper or punchier, etc...but I think that was really more about their size difference causing different room interactions.
At any rate, moving them around a little obviously dominated any differences they might have due to size.
As for multiple subs...must say I'm not a fan. I can't ever get the extraordinary tight sounding bass that I can from a good sub co-located with the mid-bass, that's perfectly time/phase aligned.
Low frequency modal interaction negates any useful conclusions.
And high frequency is even worse, as our two ears constructively/destructively alter the single frequency signal.
Just play some sine waves, walk around, tilt your head, and give up
And for subs, a sine wave is a very dense signal, usually generating harmonics and resonances either from the sub itself or from nearby room structures and items (sometimes distant ones too).
I would instead use bandwidth limited pink for a localization test.
As far as sealed vs BR...my experience was to try the bms 18n862 in both.
The BR was tuned to 31 Hz, and I gave the sealed the same tuning via Linkwitz transform.
Outdoors they sounded the same to the point I couldn't tell them apart. The only difference was how much louder the BR could go.
Indoors, I could sometimes tell them apart in terms of which one sounded deeper or punchier, etc...but I think that was really more about their size difference causing different room interactions.
At any rate, moving them around a little obviously dominated any differences they might have due to size.
As for multiple subs...must say I'm not a fan. I can't ever get the extraordinary tight sounding bass that I can from a good sub co-located with the mid-bass, that's perfectly time/phase aligned.
It's true about the specs.
I forgot to add "ignore the plot in the specsheet" to my post.
Still, especially at the sale price, it doesn't hurt to experiment with a few of those.
For a manufacturer it's not an option.
When I say "xover" I also include the components needed to flatten the response of the waveguide. In my passive designs this is some 11-13 different components in just the HP. There are a lot of possible combinations and I found along the way that some were cheaper, or worked a little better, or various other reasons. The layout virtually never changed, only the actual component values.
In a crossover, there is a natural delay of the woofer due to the LP filter. If done correctly this puts the woofer in time alignment with a waveguide whose driver is set back from the baffle, so there is no need for delay. Even in my active crossover there is no delay in either the HP or LP. With piston sources this is not the case. Since they all are mounted in the same plane there will be different delays to each driver from the crossovers and so linear phase cannot be achieved without delays (sloped baffle being the exception here.)
Along the way I found that small changes to the component values had a big change on the phases at the crossover and tweaking these could sometimes improve the and amplitude characteristics through the crossover.
My passive designs have a very tight control over phase, it not varying more than about 90 degrees across the passband. This isn't "perfectly linear", but it is very close. With DSP one could improve that to almost 0 degrees, but I would question the need for that.
I'm still searching why i can so easily localize the subs, the video was not totally meaningless, the sub array sould be very, very cleaverly tuned on the same frequecy range IMHO.
The come back to very high efficiency large dipole woofers is a benediction in term of spatial rendering, the LF section is back to 3D
Subjectively these are hard experiments to perform because they are never quite as "clean" as you would like. There may be harmonics, or cabinet rattles, even room rattles. For music sources, I can never localize an individual sub, although there is a phantom image that is quite stable.
To your first point, theoretically the two systems would have exactly the same passband efficiency. So how can this statement possible be true?
Poor setup of multiple subs is always an issue, but done right it achieves the best bass (both measured and perceived) that I have ever heard. It is simply a matter of fact that no one sub can achieve the smooth response that multiple subs can. Done wrong and they can also make things much worse.
Your experiences are very similar to mine.
Multiple subs, as implemented by Dr. Geddes in a dedicated room, is undoubtedly a superior solution.
However, this is not always possible due to practical, or other constraints.
Sealed boxes are generally less efficient, due to the (TS) parameters needed in order to have deep bass in a sealed enclosure.
The 'bad reputation' of BR is largely fuelled by numerous examples of lousy implementations.
Last edited:
I don't understand how the two systems, BR vs sealed, can possibly have the same passband efficiency if the passbands are the same.
I mean take the bms 18n862 example...the Linkwitz transorm required 'I forget how much boost' to reach 31Hz , but it was like at least 9dB. That's a heck of a lot of power needed and efficiency loss imo!
btw, my method for determining efficiency...
Put a measured voltage on a sub (2.83v usually of bandwidth limited pink, or just have crossovers in place for full pink) .
Since pink bounces around, the voltage measurement needs to be an average for up to a minute.
Take an average SPL measurement over the same exact measurement period.
Do the math and get a real world efficiency spec.
You can also take an average current reading over the same time period and get a real world impedance spec.
I honestly don't know why this isn't used for specs...simple, and works ime/imo.
Oh, just remembered something....I've confused efficiency with sensitivity before...and just did it again....
Pls pardon, and redirect to efficiency...
Last edited:
Those "constraints" and compromises will get you every time.
The same driver in a BR and closed will have the exact same efficiency. The BR has a little lower cutoff because of a resonance placed at this point. above that resonance the two system act exactly the same. The "apparent" increase in efficiency usually comes from a high Q BR tuning yielding a slight increase in efficiency over the bandwidth of the port resonance. But this resonance also rings, which is not a good thing.
Actually efficiency and sensitivity are paramount to the same thing. One is for power and the other for voltage to pressure. Basically the two have to track one another, so your confusion lies somewhere else.
Just look at any simulation software and you will see that the passband sensitivity depends only on the BL and mass of the cone (and area). Neither of these change with a port or not so how could the sensitivity be different?
As I said, a high Q port tuning will measure in your test a bit higher because of the port resonance, but that is a very narrow bandwidth effect.