Hi,
depends on design of the actual driver of course, but since you talk about "likelihood" (and most real world models on the market) i would tend to answer "yes" frankly and simply ...
One key in making better coax drivers is IMO to free them from the burden of lowest frequency reproduction and use a dedicated Woofer/Subwoofer for this. There is little sense in talking about "point sources" as beeing desirable e.g. in the mid to upper bass region, where a multiway can fullfill that requirement - if it was one - quite good.
Then the coax driver's (LF) membrane structure and shape, basket shape, suspension shape etc. can be made to better meet those requirements mentioned like e.g.
- aim for smooth structures having lower diffraction (e.g. less excursion makes flatter surround possible ...)
- having the dispersion pattern of woofer (of the coax) match that of the tweeter (better) at crossover
- avoid cone breakup by keeping diaphragm smaller compared to bending wavelengths (keep Krm lower)
- ... more to mention
Last edited:
Another is to make the horn bigger to allow the woofer to be one:
9864
http://alteclansingunofficial.nlenet.net/proloudspeakers/duplex/9864A-Data.pdf
9894
GM
Hi GM,
the manufacturer's dispersion patterns e.g. of the Altec 9864 look very uniform in a wideband manner.
Of course you are right, that general approach will also work in matching the radiation patterns between woofer and tweeter at XO ...
But i cannot see why using a coax driver in that particular kind of design is even necessary ?
You could even use separate drivers for LF/HF easily IMO ?
Then also approaches like Tom Danley's Synergy Horn can be regarded as being "very coax" , which is a more recent approach to similar goals IMO.
the manufacturer's dispersion patterns e.g. of the Altec 9864 look very uniform in a wideband manner.
Of course you are right, that general approach will also work in matching the radiation patterns between woofer and tweeter at XO ...
But i cannot see why using a coax driver in that particular kind of design is even necessary ?
You could even use separate drivers for LF/HF easily IMO ?
Then also approaches like Tom Danley's Synergy Horn can be regarded as being "very coax" , which is a more recent approach to similar goals IMO.
Last edited:
I am currently undertaking a project utilizing the JBL 328CT coaxes (normally in-ceilings, but the 8"s are nice, and the 2412s.... well... not exactly well regarded). I'm hoping the 2412 have the reputation they do because of a 1" diaphragm (undersized) and utilization crossed to 15"s at 1600hz in the cheap JBL pro stuff TRX/JRX not because they're truly bad drivers. The 8" seems like a solid pro JBL quality driver and the XOs seem pretty well-designed.
Part of the challenge is that DSP has made time-alignment more viable now where before that was always a shortfall. I don't see it as much of an issue in surrounds, which is where these are planned for.
Part of the challenge is that DSP has made time-alignment more viable now where before that was always a shortfall. I don't see it as much of an issue in surrounds, which is where these are planned for.
The newer B&C's seem to be pretty good performers. I've been using Ciare coax's with great success for the past couple years.
I think the Martin CDD series have a lot of potential for domestic use. They put a lot of work into the throat area. Even appears to be some foam in the throat a la Geddes. They seem to get positive accolades from anyone who hears them.
I think the Martin CDD series have a lot of potential for domestic use. They put a lot of work into the throat area. Even appears to be some foam in the throat a la Geddes. They seem to get positive accolades from anyone who hears them.
Greets!
These were designed before digital delay and are superior to the traditional separate HF horn + woofer systems and of course for prosound or very large HT apps, though a few folks have used these in HIFI apps.
Agreed, where depth is available, the unity concept renders all other coincident designs obsolete, just included these as traditional alternative design options/historical relevance for those who might want to fabricate different horns/WGs for existing drivers as some have done.
Forgot to point out that the original Altec duplex had a relatively larger true multi-cell horn/optimal 1 kHz XO for a smooth transition between the two Vs the rest with its little fake 'eggcrate' WG/1.5-or 1.6 kHz XO, so still one of the best, if not the best, co-ax to date in this respect. Why it was changed is a mystery other than possibly to cut cost as this all transpired during WWII and once accepted as 'good enough', became the 'bible' till the advent of its Urei variant.
GPA's 3 series has a larger WG as a compromise size between the original two, though still has the too high 1.5 kHz XO.
GM
And then there's coax drivers that measure quite good. The newer Uni-Qs for example.
Conventional multi way designs have lobes which lead to spectrally distorted reflections. Does it matter? Probably. How much? Unknown.
There's a lot of (educated) guesses what those priorities are but what performance characteristics really lead to objectively superior results is still largely unknown. Raising, falling, constant directivity? What's the role of adaption? What's the role of the room? What's the role of recording/miking techniques?
pnix said:
I linked that KEF "Uni-Q" example already as an example for "one of the better ones".
Still member of a minority ...
pnix said:
But "distorted reflections" not only because of "lobing" - expecially in the vertical plane at crossover, which can be mitigated by proper design - but also because of dispersion of woofer and tweeter not matching:
Which is an issue not just associated to "multiway" vs. "multiway coax", there are many coax drivers suffering from that issue ...
Btw. a definition of "spectrally distorted reflections" is not that easy IMO:
Is constant(!) directivity needed in highs (even ...) in the vertical plane to get usefull LS-/Room Interaction ?
How to get the "room curve" right then using "perfect CD" (which virtually no speaker has ...) in a real listening-room, without sacrificing a flat anechoic response ?
(see e.g. Bruel&Kjaer room curve but also others ...)
Seems "some kind of spectral distorsion" is even needed to make a good LS ... i.e. (real) constant(!) directivity index might not be that useful in real rooms ...
Regarding the influence of "spectrally distorted reflections", say "energy response" on listener's preference:
pnix said:
Unknown just in case the work e.g. of Floyd Toole might be unknown to you ... which i do not believe.
"Best performes" in unsighted listening tests are mostly "conventional multiways" according to your interpretation of "conventional" ...
So "More data, less wank" as cited from Zilch in your footer, might be a real good thing.
Last edited:
Hi pnix,
i really think matters have arrived "beyond guessing" although there might still be room for different opinions ...
I do not think there is any "mode of recording", that could compensate for a speaker's DI curve having the wrong tendency ... or disadvantageous "vertical vs. horizontal" distribution.
This is still an interesting point to me: When specifying a flat anechoic response, one would in fact need to modify the DI curve according to the room's frequency dependent absorption in order to get a desired "room curve", right ?
So adaptive DI curves could be missing in loudspeakers since decades ...
But before getting there, a first goal - as a kind of engineering exercise - could be a loudspeaker having a DI curve rather "smooth" to meet some kind of "prefered room curve" at least in some kind of "usual" room having about average decay time (curve) over frequency.
Of course these few lines adress only few aspects of the larger picture to be looked at ...
Last edited:
1) Examples for "room curves found to be preferred":
Bruel&Kjaer:
http://i47.tinypic.com/rrt8yd.jpg
Harman International:
http://i1217.photobucket.com/albums/dd381/mitchatola/Harman.jpg
2) Some measures used for describing/estimating "listeners preference" according to Harman/Dr. Floyd Toole:
relates to anechoic response
-on-axis response
-listening window (average)
relates to reflected sound (though calculated by model using anechoic polar data)
-early reflected curve
-sound power
directivity indices related to
- the sound power and
- early reflections.
See picture below:
http://m.eet.com/media/1164613/134790-tmw04_11f1fig1a.gif
3) Harman Patent on
"Method for predicting loudspeaker preference"
http://www.google.com/patents/US8311232
Bruel&Kjaer:
http://i47.tinypic.com/rrt8yd.jpg
Harman International:
http://i1217.photobucket.com/albums/dd381/mitchatola/Harman.jpg
2) Some measures used for describing/estimating "listeners preference" according to Harman/Dr. Floyd Toole:
relates to anechoic response
-on-axis response
-listening window (average)
relates to reflected sound (though calculated by model using anechoic polar data)
-early reflected curve
-sound power
directivity indices related to
- the sound power and
- early reflections.
See picture below:
http://m.eet.com/media/1164613/134790-tmw04_11f1fig1a.gif
3) Harman Patent on
"Method for predicting loudspeaker preference"
http://www.google.com/patents/US8311232
Last edited:
If it's still a matter of opinion then objective data is missing. If you read through the available literature you'll notice the gaps. Toole even points them out in most of his papers. They looked only at single reflections in anechoic rooms. They tested only monophonic reproduction within a specific controlled acoustical environment. There's no conclusive research looking at imaging, spaciousness and envelopment in different acoustical environments with different recording techniques.
Of course not and also not what I was implying. Some recording techniques might benefit from room reflections others might not hence dispersion patterns might or might not support or negatively affect perception. This is largely unknown and no conclusive research is available.
In regards to recording techniques there's a great paper by Lipshitz, "Stereo microphone techniques: Are the purists wrong?".
The problem is that ideal reproduction properties are largely unkown and/or not well defined. Is room interaction desirable? If yes, how much? How does the ideal reflection pattern look like? Is it even possible to define optimal reproduction conditions or are there as many ideal conditions as there are control rooms and monitoring loudspeakers? If room interaction is not desirable then we would need to reduce reflections by 30dB or even more (now this is known from psychoacoustic studies) and not just the commonly suggested 10 to 15dB. In that case good reproduction at home becomes an impractical goal.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.