[Brian Cheney]

Gove us some time to set up production, it will look so much better than handcut prototypes which is all we can do right now.

[PLMONROE]

Is this also adaptable for use on RM-40s with the older spiral tweeters?

[Brian Cheney]

RM40's of any vintage can take the upgrade.  In addition to the waveguides new lower Q PR's are required as the system sounds too bassy without them.

[dwk]

The technical issues intrigue me as I am a bit curious about the details of true CD performance being possible from such a shallow device. Then again, depending on radiation angle (dispersion), this may very well be possible. Undoubtedly it is as Brian would not stake his reputation on "snake oil." Still, I'm curious...but then I usually am.

-Bob

Well, unless I'm completely missing the point, I think the confusion arises due to the fact that this device doesn't attempt to *limit* dispersion at all. Instead, it *broadens* dispersion at HF, so that it matches the wide dispersion of the panels at LF.  This cuts to my objection to the term 'waveguide' as in any other use that I'm familiar with (EM, acoustic or optical), the purpose of the waveguide is in fact to limit dispersion.

So, I can easily see how this achieves constant directivity - it's constant directivity over ~180 degrees, exactly as you'd expect from a very narrow line source. Read up the thread, and this is exactly what Brian is claiming.

So, when compared to conventional point-source style speakers, this will probably have much more uniform dispersion. Brian is certainly not alone in suggesting that this is a very desirable property (Linkwitz suggests this in his Pluto for example)

IMHO, though, CD without pattern control is only half the game. Without a very large room or extensive room treatments, launching that much energy in all directions will bring the room into play pretty significantly - precicely what conventional waveguides are aiming to avoid.

[_scotty_]

As you do not utilize the last octave of the midrange driver in your designs
CD in the midrange driver at 20Khz is uneccessary and irrelevent.
It still looks like the the easiest way of achieving CD is to crossover to a tweeter before there is a directivity problem in the first place.
By crossing over to a tweeter earlier to achieve CD you also avoid the potential cavity resonance problems your approach may engender.
Scotty

[Duke]

My guess is that the device has three primary components, of which the thin vertical slot is the most obvious.  The slot itself would function as a diffractive device, expanding the radiation angle in the horizontal plane.

The second element of the design is also plainly visible, but easily overlooked:  The flat face of the device.  This flat face operates as a 180 degree waveguide down to the frequency where the baffle is 1/2 wavelength wide.  In fact speaker baffles can be considered to be 180 degree waveguides, but usually the drivers mounted on them beam enough that the baffles only function as waveguides over relatively small portions of the spectrum.   My guess would be that the effective radiation pattern of the device is 180 degrees, down to roughly 500 Hz.

Finally, there's whatever's going on behind the device.  I would guess that two functions are performed by the backside of the device:  First, directing as much of the sound energy as is practical into the slot; and second, absorbing unwanted sound energy that would be detrimental to fidelity.   Perhaps there is a funnel-like shape directing sound towards the slot, and then absorptive material on either side to absorb the sound that doesn't easily make it into the mouth of the funnel.   This is purely speculation.  

Duke

ps - how close did I come, Brian?

[Aether Audio]

dwk,

IMHO, though, CD without pattern control is only half the game. Without a very large room or extensive room treatments, launching that much energy in all directions will bring the room into play pretty significantly - precicely what conventional waveguides are aiming to avoid.

Thank you!

Duke,

The second element of the design is also plainly visible, but easily overlooked: The flat face of the device. This flat face operates as a 180 degree waveguide down to the frequency where the baffle is 1/2 wavelength wide. In fact speaker baffles can be considered to be 180 degree waveguides, but usually the drivers mounted on them beam enough that the baffles only function as waveguides over relatively small portions of the spectrum.

...and thank you as well.  

Looks like there's some pretty smart fellers out there.  We would all do well to consider such informed  comments when considering these rather complex issues.  Maybe they're right, maybe not... BUT - The more you know...  Uhhh, well, being totally ignorant usually isn't a good thing anyway!

[_scotty_]

It might be helpful to look at the specific dispersion  problems the RM40
has and then look at the results that Brian discloses for his invention.
The existing design has the dispersion properties mentioned by Duke before
Brians invention is applied up to the point at which the emissive width of the midrange array is equal to or greater than  the wavelength being radiated before the transition to the tweeter takes place. This  problem area of the exisiting design is less than octave wide and extends from approximately 4500Hz to the crossover frequency of the tweeter at 6.9kHz.
Brians invention, which I will henceforth call an Aperture Mask seeks to solve this discontinuity by limiting the width of the midrange to about 2/3" thus making it roughly equal in width to the tweeter.
I do not consider this radiation discontinuity to be a critical flaw in the design
but the off axis behavior would be bothersome in some rooms.
 My understanding of the RM40s design flaws and potential solutions are neccessarily limited and I may have overlooked something.
The consequences of applying Brians Aperture Mask to solve the aforementioned problem have been spelled out in Brians comment
where he said

RM40's of any vintage can take the upgrade. In addition to the waveguides new lower Q PR's are required as the system sounds too bassy without them.

From Brians statement it appears that sufficient energy was lost to the absortive countermeasures employed to control the cavity resonance that the Aperture Mask caused that just turning up the pots on the tweeter and midranges could not fully compensate for the loss of output.
This is what neccessitates adjusting the bass output via changing the Q
of the PR's. It would be nice if the change in efficiency were documented
and I do not doubt that the improved radiation pattern sounds better to some listeners.
Scotty              

 edited to reflect changes in design not present on website
Thanks, warnerwh

[JoshK]

scotty, I think the point your missing about the xo point is that BC uses only quasi-second order filters so the xo point is not really the point for which the panel or tweeter is being effective.  So even though you drop the xo below 5500, the panel is still beaming above.  I think the part of the point BC is trying to overcome is CD, especially in the overlap area where either of the two drivers are say >30db down.  This should help the dreaded lobing/combing problems some for BC's filters.  

I think my concern is the same as DWK's and stated again in a post by marbles.

[ScottMayo]

RM40's of any vintage can take the upgrade.  In addition to the waveguides new lower Q PR's are required as the system sounds too bassy without them.

Does it make sense to eventually migrate in the direction of separate bass enclosures, firing bass in all directions? Despite years of people saying you can't tell where bass is coming from, the fact is, sometimes you can. A downward firing woofer is presumably as CD as it gets.

I'm also wondering how this affects imaging. Of course, I'm going to get some and find out, but a few folk must know already.

Distance from the speakers has a lot to do with imaging - sit a few feet to one side, and the magic of imaging fades. In part because right now, you lose some of the HF information that's so crucial to how ears find things, but in part because a few ms of delay in getting half the information skews things. There's a plane you can sit in where you are equidistant from the speakers - and it's right down the middle, between them. Nothing can change that. CD might make the side HF falloff stop, but it can't do anything about the speed of sound. So with this change, what does someone off center perceive?

Finally - mounting. It's obvious that being off a 1/4" when mounting the device is going to make a mess. Will it bolt directly to the drivers somehow (which would at least guarantee positioning), or is retrofitting these to existing speaks going to be tricky? I'm looking at the RM/x and hoping I won't be driling into the finish.

[warnerwh]

I'd think something like this can mount the way our grill cloth does.

Scotty: The crossover from the mid to tweeter panels has been at 6.9 for some time now.

[Marbles]

I'd think something like this can mount the way our grill cloth does.

    That's why it's over the woofers....I hate it when the answer is so obvious and I miss it  

[ScottMayo]

I'd think something like this can mount the way our grill cloth does.

That might get interesting on an RM/x, if you want to get the tweeter involved, which presumably you would. It might also look very cool, done right. Brian, any plans here?

My room is properly treated - bring on that wider dispersal, I say.

[Brian Cheney]

Thanks all for the many excellent comments.  John Casler has forwarded a 1987 disclosure with the same general idea but different implementation and lacking some features of the invention.  "Sawtooth" Chris thinks we will have no problem differentiating from the prior art.

The waveguide does more than just achieve CD.   It is a source of CONSTANT WIDTH from 280Hz to 20kHz.  And the width is very small.
And the waveguide is flat (good comment there).  

I have so far used it in live and heavily damped environments with good results.  The waveguide is easily removed and replaced so you can hear the differences easily. You can direct away from many room problems by rotating the system on its axis.  Some room treatment helps firm up the image, but that's true of any speaker.

The patent also discloses a differently configured waveguide for cone dynamic/dome tweeter systems.

Without CD EQ the trebles change and have less sparkle but also more definition in space.

The betatesters and early adopters are going to have lots of fun discovering the advantages and limits of the technology.

I may not make a $million off of this, we shall see.  Audio is a hobby, and although I make my living this way, it remains one for me too.

[Xi-Trum]

The patent also discloses a differently configured waveguide for cone dynamic/dome tweeter systems.

It's always a good idea to have many patents around the original invention... just to prevent patent poaching.  

Anyway, all this sounds very intriguing to me.  I'm still faithfully hanging on to my original RM40.  Looks like I'll be rewarded.  

[JoshK]

How do you avoid resonance in the slot?

[Brian Cheney]

Only a very small portion of the slot is reflective, the balance absorptive.

Can we keep the technical questions to a minimum?  I have a tendency to give away the store.

[John B]

So with this change, what does someone off center perceive?

The sweet spot    I know, too generalized an answer.  What I will do tonight is pick an LP or CD which has numerious musical instruments and vocals on the recording; I will start out in the sweet spot, and then change my listening locations to far left and far right, and report in detail what I preceive.  I was going to hold off on these type of reports till I got my own production model, as the "CD device" has been changed slightly from the Beta version I have.   However, the functionality should still be close enough to where this exercise with the Beta version might answer some questions for you guys.

[ctviggen]

It's a very nice looking design.  I'd like to hear it.  It also seems very simplistic, in the sense that you just put something over your drivers.  (At the same time, it bothers me for some reason that there's something over the drivers.)  I'd really be interested in seeing diagrams of off-axis response with and without slot.

[Brian Cheney]

It took a whole day for someone to ask for measurements?

I'll supply two sets: one I make in my semi-anechoic environment, and another set from Bascom King (formerly in charge of Audio Magazine's speaker reviews) when he has the time.