[FullRangeMan]

Constant Directivity from various drivers in a multiways loudspeaker is one more prob of use woofers and tweeters, if one wish Constant Directivity I suggest use Point Source ie a FR driver.

[jtwrace]

Constant Directivity from various drivers in a multiways loudspeaker is one more prob of use woofers and tweeters, if one wish Constant Directivity I suggest use Point Source ie a FR driver.

How did you get to such a conclusion? 

[Davey]

Constant Directivity from various drivers in a multiways loudspeaker is one more prob of use woofers and tweeters, if one wish Constant Directivity I suggest use Point Source ie a FR driver.

A typical FR driver is anything but a point source.
It will narrow as you increase in frequency in a "controlled" manner, but this is certainly not constant directivity.
Full-range systems have their benefits and trade-off's like anything else, but let's not get carried away. 

Dave.

[FullRangeMan]

A typical FR driver is anything but a point source.

First time I heard it.

It will narrow as you increase in frequency in a "controlled" manner, but this is certainly not constant directivity.

Its the nature of the music playback, Constant Directivity as a audio product seems not match with home audio as it implies the total manipulation of the audio signal, something not compatible with hifi.

[JLM]

Here's a link that has some good information. Be sure to read both parts.

http://www.tnt-audio.com/intervis/david_smith_e.html

As David Smith mentioned, Floyd Toole's book, "Sound Reproduction" is a wonderful reference to explain home in-room listening. 

In particular start at page 54 and learn about the Schroeder frequency (roughly 130 Hz - room dependent).  Below that particular frequency sound waves in a room behave as waves and above that frequency as rays.  The bass waves can be thought of like a hand moving to back and forth length wise in a shallow bathtub of water.  The waves created will hit the end and bounce back.  As they encounter the next wave they will either add to the amplitude, cancel out, or do something in-between.  In our listening rooms that can result in a +/- 20 dB sound pressure level.  Adjusting the position of your hand will only move where interference occurs.  Same with adjusting the frequency of your hand's movement. 

Trying to solve using EQ/DSP can only work at a single location in the room and care must be taken not to try to pull up frequency narrow dips (which can't be heard anyway) as these dips are often -30 dB, which represents a 1,000 fold wattage increase (can you say clipped amp and blown speakers?).  As Toole explains later in his book, the best answer is the use of multiple bass sources (subwoofers).  Ideally these would be semi-randomly placed away from corners with one located well off the floor.  Note that these multiple subs can be smaller.
 
This whole +/- 20 dB in-room response is independent of room size (to within any reasonable domestic limits) and shape and a huge factor that audiophiles been ignoring.  Toole's book also covers room shape as being critical and offers an acceptable range of ratios. 

Getting back to CD.  A point we've ignored is limiting off axis output.  Many of CD designs use large woofers with relatively high crossover frequencies.  If directivity is limited to 45 degrees off center) and the speakers are setup with a 45 degree toe-in front and side wall first reflections are eliminated.  The effect is to take away room contributions at the frequencies so directionally limited.  Those same designs also use horn mid/tweeters, to again limit off axis output in the same way.  Again, as David Smith mentioned, this should produce a good polar response.

[Davey]

First time I heard it.Its the nature of the music playback, Constant Directivity as a audio product seems not match with home audio as it implies the total manipulation of the audio signal, something not compatible with hifi.

I suggest to do some reading on the topics of point-sources and constant directivity radiation.  Both of those are well-defined characteristics.  Very few speakers can (or even try to) achieve either one.  And even then, it's only possible over a finite frequency range.

The OP's thread title/query of "Controlled Directivity" is very squishy and could mean just about anything, but "point-source" and "constant directivity" are much less squishy terms.

Cheers,

Dave.

[dburna]

I suggest to do some reading on the topics of point-sources and constant directivity radiation.  Both of those are well-defined characteristics.  Very few speakers can (or even try to) achieve either one.  And even then, it's only possible over a finite frequency range.

The OP's thread title/query of "Controlled Directivity" is very squishy and could mean just about anything, but "point-source" and "constant directivity" are much less squishy terms.

Cheers,

Dave.

I agree and would like to change the original posting to Constant Directivity, but I don't think I can once the thread has replies.  Will try, though.

NOTE: I guess I *COULD* change the OP after all.  Doesn't impact all the subsequent replies, but at least it is "fixed" going forward.

-dB

[DaveC113]

I'm really wanting someone to impress me with a CD speaker now. Thus far I haven't heard anything that stood out, but would like to hear the M2 at least. It has a few other things going on besides CD that I find pretty cool.

Controlled dispersion patterns are helpful but I do agree it's no guarantee of great sound. I also don't think constant directivity waveguides/speakers are as good as traditional horns, imo and ime... The drivers are where it's at, by far the most important part of a speaker... waveguides won't make mediocre compression drivers sound good.

[rajacat]

If one wanted to try a waveguide speaker for a relatively modest investment it would be hard to beat this Jeff Bagby designed kit. If you wanted to tweak the kit just substitute some of the crossover parts with premium components and maybe stuff the interior with No Rez.
http://www.diysoundgroup.com/waveguide-speaker-kits/fusion-12.html
If you browse around the store and forum you'll find plenty of help for the very easy assembly process.

Jeff Bagby is a well known for his quality designs so there's minimal risk that you're throwing your money away.
There's a 15" version too but by a different designer.

[dburna]

If one wanted to try a waveguide speaker for a relatively modest investment it would be hard to beat this Jeff Bagby designed kit.
Jeff Bagby is a well known for his quality designs so there's minimal risk that you're throwing your money away.
There's a 15" version too but by a different designer.

Looks interesting.  DIY not going to happen these days because of the time commitment, but I always learn something from looking at these designs.

This is the kind of information I wanted to uncover, however.  Anyone can (and many do) put a "salad-bowl-with-a-hole-in-it" waveguide on a driver and call it CD.  I wanted this thread to highlight the designers that "do it right" and put sound engineering principles into play.

-dB

[Rick Craig]

In a blind test (like they do at Harman) how well do you think full-range driver systems would fare? How about the Linkwitz dipole designs? Let's hear your opinions.

[DaveC113]

Looks interesting.  DIY not going to happen these days because of the time commitment, but I always learn something from looking at these designs.

This is the kind of information I wanted to uncover, however.  Anyone can (and many do) put a "salad-bowl-with-a-hole-in-it" waveguide on a driver and call it CD.  I wanted this thread to highlight the designers that "do it right" and put sound engineering principles into play.

-dB

M2 has the most advanced CD waveguide, but a basic CD WG is just a conical with an OS throat transition. Using other kinds of horns is not "doing it wrong", I wouldn't consider one much better than the other, it depends on application and personal preference. Application being the room and whether the goal is to provide sound for a broad area or a smaller area.

Most of us are concerned with smaller areas and there's plenty of excellent non-CD horns that work great for home use like Avantgarde, Cessaro, Odeon and many more. You may even prefer them as imo there are advantages of a constant curvature like LeCleach over an OS/conical.

Honestly, I don't know of any high end speaker manufacturer with a "salad-bowl-with-a-hole-in-it" waveguide?

[DaveC113]

In a blind test (like they do at Harman) how well do you think full-range driver systems would fare? How about the Linkwitz dipole designs? Let's hear your opinions.

Poorly, but only because of the setup.

Marten Logan got trashed in the preference testing but setup was the same for them vs any other speaker. Dipoles/OB need more room from the front wall and maybe a different amp would have been better. Lots of folks like MLs, electrostats, and full rangers...  I don't think their preferences are wrong but I do not think they would be preferred in the context of Harman's preference testing.

[sonicboom]

Looks interesting.  DIY not going to happen these days because of the time commitment, but I always learn something from looking at these designs.

This is the kind of information I wanted to uncover, however.  Anyone can (and many do) put a "salad-bowl-with-a-hole-in-it" waveguide on a driver and call it CD.  I wanted this thread to highlight the designers that "do it right" and put sound engineering principles into play.

-dB

Agreed. In the spirit of looking into designs that "do it right" here are a few suggestions including some that have already been mentioned above:

Audiokinesis
Pi speakers
DIY Soundgroup Fusion and Tempest
Various designs named 'econowave' such as in the link bellow (scroll down to the middle of the page):

http://techtalk.parts-express.com/forum/tech-talk-forum/34467-flex-your-pcd-mettle/page5



If you wanted to try a CD speaker and not worry about the hassle of return shipping and the related charges, you could head down to your local Guitar Center or Musicians Friend and pick up a pair of JBL LSR308's. Their 30-day return policy would IMO probably be the most hassle-free way to sample this kind of speaker and see what you think.

[JoshK]

Agreed. In the spirit of looking into designs that "do it right" here are a few suggestions including some that have already been mentioned above:

DIY Soundgroup Fusion and Tempest

If you wanted to try a CD speaker and not worry about the hassle of return shipping and the related charges, you could head down to your local Guitar Center or Musicians Friend and pick up a pair of JBL LSR308's. Their 30-day return policy would IMO probably be the most hassle-free way to sample this kind of speaker and see what you think.

I'm an M2 owner.  I've also visited Dr. Geddes' house and heard both his living room system and his dedicated theater (the whole push to go down this direction).   I've heard the Fusions a few times, specifically the 15" version.  They are damn good and pretty cheap.   I'd take them over a lot of more expensive speakers. 

[Randy]

I own a pair of amphion Argon 3LS floor standing version of the speaker in walnut the presenter picks up toward the end of this video. One of the design features is "constant directivity" as is explained.


https://www.youtube.com/watch?v=tUYCFdH7DP4

[Tyson]

CD's are nice and the Geddes are some of my favorite commercial speakers.  But they don't eliminate the room, or even control for it very well.  Even CD speakers need room treatments to sound their best. 

[AJinFLA]

In a blind test (like they do at Harman) how well do you think full-range driver systems would fare?

Best sounding alarm clock of all time 

How about the Linkwitz dipole designs? Let's hear your opinions.

That's an interesting question. I discussed the performance of the MLs tested by Harman, with Dr Toole and clarified a few issues. Turns out that particular ML tested had a nasty resonance in the panel. He showed me the impedance measurement. I would hesitate to translate that result to "all dipoles". It's also clear Dave Smith has misread Bech, a very common thing.
He also stated that during his time at the NRC (prior to Harman), the highest rated speaker in their blind testing program, was the Mirage M1 bipolars.
In fact, they became his personal reference speakers at home, prior to him moving to his now home.
The key of course, is understanding reflections and perception and not misinterpreting the 50+ years of research.

cheers,

AJ

[Duke]

In the spirit of looking into designs that "do it right" here are a few suggestions including some that have already been mentioned above:

Audiokinesis
Pi speakers
DIY Soundgroup Fusion and Tempest
Various designs named 'econowave' ...

Thank you for the inclusion.  Earl Geddes and Wayne Parham (of Pi Speakes) and Roger West (of SoundLab) have been my teachers, so credit to them for anything I get right. 

In my opinion, it all goes back to psychoacoustics.  Reflections can be highly beneficial if they are spectrally correct and if their arrival time (and even arrival direction) are done right.   Constant directivity over a roughly 90 degree angle (in the horizontal plane) from about 700 Hz on up, plus intelligent aiming of the radiation patterns, can work quite well.  Constant directivity down to 700 Hz or so typically calls for large sources, roughly 20 inches across, so in practice we usually make compromises. 

Imo in addition to the reflections being spectrally correct, we want a fairly long time delay - 10 milliseconds was suggested by Earl - in between the first-arrival sound and the onset of significant sidewall reflections.   It's usually not possible to get shorter time delays for the floor and ceiling bounce, but it may be possible to disrupt them somewhat with a judiciously placed coffee table and some large vertical slats on the ceiling.  Or maybe use a speaker geometry (MHM) that reduces the vertical pattern height. 

It is also desirable, from a psychoacoustic standpoint, for the first sidewall reflections from the left speaker to arrive at the right ear, and vice-versa.  When the reflections arrive at the opposite ear from the first-arrival sound, the ear/brain system tends to interpret this as spaciousness and envelopment.  This is accomplished by aggressively toeing-in our constant directivity speakers, such that the first significant sidewall bounce of the left speaker's output is off the right side wall, and vice-versa. 

So we want to get the first-arrival sound as right as possible, then have a time delay before the significant onset of reflections, then have a bunch of spectrally-correct reverberant energy coming in, ideally decorrelated with respect to the first-arrival sound as far as arrival direction goes. 

I like a bit richer reverberant field than we normally get from these relatively narrow-patterned speakers ("narrow" in comparison with most high-end audio speakers).  So I'm inclined to add a bit of additional spectrally-correct reverberant energy using additional drivers in some configuration that respects the 10-millisecond path-length-delay target. 



   

[dburna]

Thank you for the inclusion.  Earl Geddes and Wayne Parham (of Pi Speakers) and Roger West (of SoundLab) have been my teachers, so credit to them for anything I get right.     

As always, you are too modest, Duke.  Someone asked me (offline) what kinds of speakers I *DO* like, so I gave him a short list....and AudioKinesis was featured prominently.  You may have learned from the best, but you have contributed a great deal to the legacy in your own right.

-dB