• Check out the Annual AK Secret Santa exchange Program, signup before November 23rd in the AK Charities & Christmas forum!

What is relative efficiency:

I don't mean to butt in here but what is? (only asking because I don't know if you're referring to the JBL's which I know little about or the Jubilee which is being used as a 2-way but shown on the Klipsch site as a 3-way)
Jubilee.

Even with a tweeter, the JBL AM6200 crashes hard up there, as well. I suspect that's because there's no headroom left to compensate the VHF:

http://www.jblpro.com/catalog/support/getfile.aspx?doctype=3&docid=653

Same waveguide, different driver, AM4200 does somewhat better:

http://www.jblpro.com/catalog/support/getfile.aspx?doctype=3&docid=644

BMS 4555 up top might get the job done.

We don't know how the QSC look-alike measures.... :dunno:
 
Last edited:
I don't know if it matters as far as the graphs are concerned but just so you are on the right page, all of the discussions here (mine) about the Jubilee are from the perspective of the Jubilee being a 2-way speaker, not 3-way. Of the 23 or so Jubilee owners, all of them except one are using the 2-way format and only one guy is using the 3-way system (as seen on their website)

For the 2-way users, the tweeter horn (K510) and midrange driver (K1132 I think) were dropped and the tweeter driver (K69) is moved down to the midrange horn (K402)

Probably doesn't matter but I'm retentive enough to want to clarify that so any casual readers can understand what is being talked about, with regard to the Jubilee.
 
Probably doesn't matter but I'm retentive enough to want to clarify that so any casual readers can understand what is being talked about, with regard to the Jubilee.
No problem. I trust readers also appreciate that we're not saying anything's better, merely having fun getting at some facts relating to different design approaches.... :thmbsp:
 
There are plenty of enthusiasts who believe Tractrix is the bee's knees; seems like Cosmos isn't one of them, however.
If I am in the sweet spot, don't turn my head and resist bouncing my head to the music, Edgarhorns are incredible... Easily sweeter than honey! Otherwise, well they are not as good..

If you want to talk about how stuff sounds, then we need to move up to JBL Synthesis for comparing apples.... ;)
Yeah, I'd like to at least try them apples.. :yes:
 
At 108 dB/W, these'll get you to the upper limits of where you want to be in a modest package:

http://www.jblpro.com/catalog/support/getfile.aspx?doctype=3&docid=653

[With max continuous 132 dB SPL as a bonus.... :tongue: ]

Zilch, I picked this, simply to have a single referrence point to ask my question....

Which chart/graph would I want to look at such that I'm comparing apples/apples to the chart I posted?

Beamwidth?
Directivity Index?
Horiz Off Axis Freq Resp?
Vert Up Off axiis Freq resp?

I personally find them a bit confusing but then... truth be told, I've always been a plug/play kinda guy and never really learned about some of these technical aspects... it either sounded good or it didn't!

Anyways, which one would I want to look at?
 
Gotcha...that is what I was suspecting.

Question for you, (several perhaps :D)

Is that a logrithmic scale being used in the JBL display? I'm thinking it is. What is the benefit/drawback of using a log scale verses a linear scale? I've got a voice in the back of my head that says a log scale will tend to flatten out the visual aspect of the line more than a linear scale will, true?

If not true, then stop here!

If it is true then would it be fair to say that a linear scale is showing more details than the log?

Would the "AM6200/95 High Power Mid-Hi freq" unit be considered beaming as well?

(I'm still trying to get my brain around this and it confuses me now that I'm seeing what appears to be two different ways of showing the same thing... trying to distill in my head which one is 'right' or perhaps I should say, more enlightening to me)

As I recall the K402 chart (which I'm not looking at as I type), I think I recall it had a 6 db drop starting at about 13K and was otherwise, pretty well behaved/flat up to that mark?

Looking at this illustrationi, it (horizontal) looks like it starts at maybe 150 degrees and drops down to 80 degrees, muddles around a bit at 80 degrees, slides back to 100 degrees then gets drops way off at maybe 13/14K?

Is this more or less erratic than the K402 and from a technical viewpoint would this be considered a better graph?

(it would be really nice if all the manufacturers had a standardized way to display the data!!!)

In the vertical, looks like this starts at almost 300 degrees (?? wow) and curves around while falling down to maybe 45 degrees, spiking back up to 100 degrees then dribbling down to just under 30?

I'm mainly trying to ascertain if I'm reading/interpreting the graph correctly and then of more interest to me, (once I understand how to read it) how it compares to the K402 graph.

I don't know which representation might be 'better'...I suppose to the technical type they are both easy to understand. For someone like me who's not accustomed to reading/comprehending these things I personally find the linear graph to be far more 'accurate' to my eyes, since it's not compressing the lines like the log graphs are.

Am I off base? (you can say so if I am, I'm trying to learn and hope maybe others will appreciate my questions of ignorance)

:D
 
The design is a 3-way. Take away the VHF horn and (1"?) driver and it's no surprise the directivity is compromised in the last octave.
The "little tweeter" on the 3-way Jubilee is the K510 which has a 2" throat. The K402 actually behaves a little better than the K510 up high, except above 7kHz they're essentially identical with the same driver. The K402 has the advantage of controlled polars to a much lower frequency.

If the objective (and claim) is constant directivity with concomitant uniform power response, then the horizontal beamwidth collapsing to 50% of spec at high frequencies is clearly "beaming;" we're back in Altec sectoral radial horn territory.
The plot I'm looking at has nearly a ruler flat horizontal beamwidth from 221Hz to 13kHz. There is so little content in music above 13kHz that to call it a "beaming horn" (or really a beaming driver) seems just totally absurd to me. A beaming design is when one intentionally narrows the polars to boost on-axis response. What's happening here is a phase plug that is clover-leafing...

Indeed, you can't get there (directivity control in the top octave) with a 2" throat horn or waveguide.
Can't? Certainly the horn can't do anything about it above 7kHz, but the phase plug surely can...


Both of those beamwidths have 1/3 octave smoothing and yet are still showing a lack of any totally flat region. They also both lack pattern control much earlier than the K402 and show a downward trend throughout pretty much the whole passband, which is also verified by the advertised DI.

If the K402 is beaming, then these JBL horns are beaming more.

There are plenty of enthusiasts who believe Tractrix is the bee's knees; seems like Cosmos isn't one of them, however.

If you want to talk about how stuff sounds, then we need to move up to JBL Synthesis for comparing apples.... ;)

I don't find straight up tractrix to be very exciting at all. The K402 isn't a tractrix horn. Klipsch refers to it as modified tractrix and it doesn't sound at all the same.

The JBL Synthesis lineup costs a whole heck of a lot more, but I wouldn't hesitate putting the Jubilee up against any of them. The problem with aesthetics driven designs is that they usually have to sacrifice some level of performance to keep things small enough. There's also no way that direct radiating bass is gonna keep up with a horn.
 
Is that a logrithmic scale being used in the JBL display? I'm thinking it is. What is the benefit/drawback of using a log scale verses a linear scale? I've got a voice in the back of my head that says a log scale will tend to flatten out the visual aspect of the line more than a linear scale will, true?
Both Klipsch and JBL are using log frequency as the horizontal scale. On a linear scale, 10 kHz would be in the center, and the high frequency would be stretched out considerably, and appear smoother.

If it is true then would it be fair to say that a linear scale is showing more details than the log?
Only if the measurement methodology had sufficient resolution to record them, and the curve were not smoothed via post-processing.

Would the "AM6200/95 High Power Mid-Hi freq" unit be considered beaming as well?
See post #62. Then look at the 6200 polars to see what is happening above 12.5 kHz; a lobe appears at 16 kHz. The pattern is considerably wider just below -6 dB.

As I recall the K402 chart (which I'm not looking at as I type), I think I recall it had a 6 db drop starting at about 13K and was otherwise, pretty well behaved/flat up to that mark?
You are still thinking frequency response. It's beamwidth AT -6 dB; to see the frequency response at any particular off-axis angle, looks at the off-axis frequency response curves, remembering that the off-axis angle is 1/2 the beamwidth.

Looking at this illustrationi, it (horizontal) looks like it starts at maybe 150 degrees and drops down to 80 degrees, muddles around a bit at 80 degrees, slides back to 100 degrees then gets drops way off at maybe 13/14K?
Yes, "Drops way off" meaning gets narrower at (and below, presumably,) -6dB.

Is this more or less erratic than the K402 and from a technical viewpoint would this be considered a better graph?
Is K402 just the horn, of the full system? AM6200/95 shows the total system; only that above the crossover frequency represents behavior of the waveguide. Similarly, K402 is out of control below 200 Hz, and also "muddles around" at 80° for a bit.

(it would be really nice if all the manufacturers had a standardized way to display the data!!!)
They do. We don't have comprehensive data on Jubilee, because Klipsch doesn't publish it.

In the vertical, looks like this starts at almost 300 degrees (?? wow) and curves around while falling down to maybe 45 degrees, spiking back up to 100 degrees then dribbling down to just under 30?
Again, you're looking at system polars, two waveguides in combination. Similarly, K402 does not acquire its specified pattern control in the vertical until 1.4 kHz.

I'm mainly trying to ascertain if I'm reading/interpreting the graph correctly and then of more interest to me, (once I understand how to read it) how it compares to the K402 graph.
Then go look at the other two JBL examples I provided above for comparison.

I don't know which representation might be 'better'...I suppose to the technical type they are both easy to understand. For someone like me who's not accustomed to reading/comprehending these things I personally find the linear graph to be far more 'accurate' to my eyes, since it's not compressing the lines like the log graphs are.
It IS log; they're not showing you the grid, is all....
 
Last edited:
The "little tweeter" on the 3-way Jubilee is the K510 which has a 2" throat. The K402 actually behaves a little better than the K510 up high, except above 7kHz they're essentially identical with the same driver. The K402 has the advantage of controlled polars to a much lower frequency.
It'd be good if we had some data to discuss.


The plot I'm looking at has nearly a ruler flat horizontal beamwidth from 221Hz to 13kHz. There is so little content in music above 13kHz that to call it a "beaming horn" (or really a beaming driver) seems just totally absurd to me. A beaming design is when one intentionally narrows the polars to boost on-axis response. What's happening here is a phase plug that is clover-leafing....
See other Jubilee polars attached below, not quite so flat, with the horizontal beaming beginning considerably lower, at 3.5 kHz. Perhaps you know the difference; I surely don't?

It's been 30 years since the music stopped at 13 kHz. Just when I get to making it happen to 20 kHz 2-way, they push it to 50. Go figure.


Can't? Certainly the horn can't do anything about it above 7kHz, but the phase plug surely can...
Not this one, obviously.


Both of those beamwidths have 1/3 octave smoothing and yet are still showing a lack of any totally flat region. They also both lack pattern control much earlier than the K402 and show a downward trend throughout pretty much the whole passband, which is also verified by the advertised DI.
Size matters; JBL does it with a pair of smaller waveguides in this instance. 2352 has the horizontal down as well:

http://www.jblpro.com/pub/obsolete/23525354.pdf

3722 isn't too shabby, either:

http://www.jblpro.com/catalog/support/getfile.aspx?doctype=3&docid=620


If the K402 is beaming, then these JBL horns are beaming more.
Which ones, specifically?


I don't find straight up tractrix to be very exciting at all. The K402 isn't a tractrix horn. Klipsch refers to it as modified tractrix and it doesn't sound at all the same.
Un-modified tractrix beams at what frequency?


The JBL Synthesis lineup costs a whole heck of a lot more, but I wouldn't hesitate putting the Jubilee up against any of them.
Jubilee costs how much?


The problem with aesthetics driven designs is that they usually have to sacrifice some level of performance to keep things small enough.
JBL makes big stuff, too. Alas, we don't all live in the Academy theater.


There's also no way that direct radiating bass is gonna keep up with a horn.
I have no clue what "keep up" might mean. If you mean keep up with compression driver mid/high in terms of transient response, we can play dueling data on that one, too.

If it's direct radiating bass vs. front loaded horn bass, the tradeoffs are well understood, and they have plagued Klipschorn design from the very beginning. JBL doesn't do Hartsfield anymore....
 

Attachments

  • Jubilee%20Polars.gif
    Jubilee%20Polars.gif
    33.8 KB · Views: 7
Last edited:
See other Jubilee polars attached below, not quite so flat, with the horizontal beaming beginning considerably lower, at 3.5 kHz. Perhaps you know the difference; I surely don't?
This is turning into a semantic argument, but whatever...

How much ripple do you think is acceptable over a desired passband? Is it different for the vertical and horizontal? I personally don't think +/- 12.5 degrees is a huge deal from a practicality perspective, but I do strongly feel that the horizontal and vertical beamwidths need to be held to the same tolerances.

For what it's worth, all the JBL horns you're posting don't have both the horizontal and vertical beamwidth being flat to within the same margin at the same time....and I have good reason to believe that the plots get a bit more wild when you start looking at the unsmoothed plots. Connecting the dots every 1/3 octave isn't very high resolution, and the log y-axis makes it look a lot flatter too.

Not this one, obviously.
The K69 used when those measurements were taken isn't exactly the worlds best compression driver. There are other drivers out there that are going to perform much better. Interestingly enough, the TAD4002 is a very popular upgrade, but its polars are actually worse above 7.5kHz! It's got a nice long 1.5" to 2" snout that I think makes it beam to around 20 degrees or so above 7.5kHz.

I would also argue that the polar response at the lower range (especially below 1kHz) is far more important than the polar response at the very top of the range. It's an issue of relative energy levels in the source material. Sure, the JBL horns you refer to are smaller, but they lose pattern control at around 1kHz (2 octaves higher than the K402).

Un-modified tractrix beams at what frequency?
Ummm, it's gonna be different for every horn.

Jubilee costs how much?
Less than 10k, more than 7k...doesn't come with a xover, so not sure if you should include the cost of the active xover and extra amplifier or not. I personally think active xovers are the only way to go for high-fidelity so that cost is already in my system...of course not everyone is the same. A passive xover design is available as well, but you gotta build it yourself and it ends up costing about the same as actives if you wanna do it right.

I have no clue what "keep up" might mean. If you mean keep up with compression driver mid/high in terms of transient response, we can play dueling data on that one, too.
I was actually thinking more along the lines of distortion and polar response...and comparing a hornloaded bass bin directly against a direct radiating bass bin.

...they have plagued Klipschorn design from the very beginning. JBL doesn't do Hartsfield anymore....
So what specifically are you referring to? I'm not at all a fan of the original Khorn from a sonic perspective....engineering-wise I think it's amazing, but all that doesn't matter if you don't come through on the music side. The Jub LF one ups the Khorn in my mind from an engineering perspective, but as far as music is concerned it's in a whole new league. I'm yet to hear anything that sounds better actually.
 
This is turning into a semantic argument, but whatever...
Well, no, I have no clue what the difference between K-402 and KPT-402-HF might be, but the data is not the same. It'd be good if Klipsch published data so we didn't have to rely on indeterminate sources to ascertain the performance of their products.

How much ripple do you think is acceptable over a desired passband? Is it different for the vertical and horizontal? I personally don't think +/- 12.5 degrees is a huge deal from a practicality perspective, but I do strongly feel that the horizontal and vertical beamwidths need to be held to the same tolerances.
I don't know, but I like to see the DI more flat than not, as it's the conjugate of the power response, and uniform power response is the bottom line of constant directivity. If the pattern is asymmetric with a narrow vertical pattern mitigating floor and ceiling reflections, then uniformity of the horizontal is more significant.

For what it's worth, all the JBL horns you're posting don't have both the horizontal and vertical beamwidth being flat to within the same margin at the same time....and I have good reason to believe that the plots get a bit more wild when you start looking at the unsmoothed plots. Connecting the dots every 1/3 octave isn't very high resolution, and the log y-axis makes it look a lot flatter too.
Like this one, the $6, 6" square PT waveguide in AC16?

attachment.php

The JBL plots I've cited and shown are system plots, except for the 235x horns, and the verticals include the nulls in the crossover region. They can't be the same, unless the pattern is very narrow.

The K69 used when those measurements were taken isn't exactly the worlds best compression driver. There are other drivers out there that are going to perform much better. Interestingly enough, the TAD4002 is a very popular upgrade, but its polars are actually worse above 7.5kHz! It's got a nice long 1.5" to 2" snout that I think makes it beam to around 20 degrees or so above 7.5kHz.
That's why a 1" throat is preferred when VHF pattern control is a performance objective.

I would also argue that the polar response at the lower range (especially below 1kHz) is far more important than the polar response at the very top of the range. It's an issue of relative energy levels in the source material. Sure, the JBL horns you refer to are smaller, but they lose pattern control at around 1kHz (2 octaves higher than the K402).
The objective of constant directivity above 1 kHz is to acquire control of the spectral balance and spaciousness cues in the reflective environment of an enclosed space such as a typical listening room. Studies clearly indicate that's where the action is in these respects, and it's easily seen that substantial compromises are necessary to accomplish directivity control at lower frequencies, where the wavelengths are longer. The easiest answer is to stick the sources in room corners. Seems like we've been there already, tho.

Ummm, it's gonna be different for every horn.
Yes, which is why we see stacked arrays of them of diminishing dimensions, with concomitant additional issues.

Less than 10k, more than 7k...doesn't come with a xover, so not sure if you should include the cost of the active xover and extra amplifier or not. I personally think active xovers are the only way to go for high-fidelity so that cost is already in my system...of course not everyone is the same. A passive xover design is available as well, but you gotta build it yourself and it ends up costing about the same as actives if you wanna do it right.
Indeed, I was wondering how the requisite 3.5 ms delay is accomplished in the passive crossover.

$7K - $10K gets into Synthesis territory. There are other JBL models in that range as well, but you've got to go to Tokyo to audition and be resourceful to acquire them.

I was actually thinking more along the lines of distortion and polar response...and comparing a hornloaded bass bin directly against a direct radiating bass bin.
Yes, fine; there are tradeoffs inherent in both approaches.

So what specifically are you referring to? I'm not at all a fan of the original Khorn from a sonic perspective....engineering-wise I think it's amazing, but all that doesn't matter if you don't come through on the music side. The Jub LF one ups the Khorn in my mind from an engineering perspective, but as far as music is concerned it's in a whole new league. I'm yet to hear anything that sounds better actually.
Specifically, horn-loaded bass has bandwidth limitations. Khorn never filled the gap above 200 Hz. Jubilee relies on a huge midrange horn to reach down where it needs to go; we don't know where that is yet, but this compromise clearly dictates the design. There have been plenty of players in this over the years, most all gone, though JBL still offers horn bass for SR. More generally, Klipsch's product line reflects its own assessment of the technology.

Danley just introduced a dual-12" tapped horn sub that actually gets down to 10 Hz, for just under $1K DIY. Above 80 Hz, there are plenty of options to explore with $5K - $8K left to do it.... :yes:
 

Attachments

  • AC16 Beamwidth.JPG
    AC16 Beamwidth.JPG
    34.2 KB · Views: 36
Last edited:
I'll break it down for you.

Zilch is a known JBL slappy.
DrWho is a known klipsch fanboy.

Zilch appears to have data and a firmer grasp of what's going on in that data.
 
More generally, Klipsch's product line reflects its own assessment of the technology.


That's an interesting and important observation. One must wonder if the development of the Jubilee was simply PWK working out a problem that bothered him personally but that wasn't really important anymore, kind'a like Einstein in his later years.

In any event speaking subjectively ole Coytee's system sounds about as good as any I've heard, it's right up there.
 
While this is some amazing reading, I must ask, can we just enjoy some music?
Yes, of course. It's what we do, actually, and I suspect it's playing in the background of all of the participants here.

The subject is directivity, and for comparison, here's a 1986 tech note from Altec just posted by Robh3606 over on LHF explaining a bit of the science involved, with a conventional 3-way bookshelf as example:

attachment.php

Extrapolate that out to 20 kHz, and the beamwidth is ~20°. If you want the full sizzle, they have to be toed-in to the listening position, and there are effectively no sidewall reflections to generate spaciousness cues above 2.5 kHz.

http://www.audioheritage.org/vbulletin/attachment.php?attachmentid=42902&d=1258751993

In any event speaking subjectively ole Coytee's system sounds about as good as any I've heard, it's right up there.
Lest anyone misinterpret the discussion, I'd certainly love to have a Jubilee system here.... :thmbsp:
 

Attachments

  • Model 9 Directivity.JPG
    Model 9 Directivity.JPG
    60.2 KB · Views: 35
Last edited:
I don't know, but I like to see the DI more flat than not, as it's the conjugate of the power response, and uniform power response is the bottom line of constant directivity. If the pattern is asymmetric with a narrow vertical pattern mitigating floor and ceiling reflections, then uniformity of the horizontal is more significant.
How do you get a flat DI with a collapsing vertical coverage pattern if the horizontal is held constant?

Have you ever compared a 'collapsing vertical constant horizontal horn' against a horn that is constant in both directions (keeping all else as equal as possible)? In order to get them to sound similar, you gotta compromise the on-axis response for the beaming horn...even though it's only in the vertical. Can it sound good? Ya, definitely. Does it sound better? Nope.

Like this one, the $6, 6" square PT waveguide in AC16?
I'm talking ripple off a nominal target....the target doesn't need to be the same for both the horizontal and vertical.

The point I'm trying to make is that where you consider the horn as "starting to beam", it's what I would consider within a reasonable tolerance of nominal. And the only reason I'm trying to make that point is because the K402 polars are awesome for not having a diffraction slot.

Specifically, horn-loaded bass has bandwidth limitations. Khorn never filled the gap above 200 Hz. Jubilee relies on a huge midrange horn to reach down where it needs to go; we don't know where that is yet, but this compromise clearly dictates the design. There have been plenty of players in this over the years, most all gone, though JBL still offers horn bass for SR. More generally, Klipsch's product line reflects its own assessment of the technology.
The Jub LF is solid up to 1kHz...

I might argue that Klipsch's product line ultimately reflects what people look for when they choose a speaker.
 
How do you get a flat DI with a collapsing vertical coverage pattern if the horizontal is held constant?
Didn't say collapsing, rather, "narrow." 40° vertical is narrow. 90° x 40° is an asymmetric horn/waveguide. Typical vertical listening window is half that or less.

Have you ever compared a 'collapsing vertical constant horizontal horn' against a horn that is constant in both directions (keeping all else as equal as possible)? In order to get them to sound similar, you gotta compromise the on-axis response for the beaming horn...even though it's only in the vertical. Can it sound good? Ya, definitely. Does it sound better? Nope.
K402 is collapsing by half in the horizontal, from 100° to 50°.

I'm talking ripple off a nominal target....the target doesn't need to be the same for both the horizontal and vertical.
I'm saying I care less about the vertical than the horizontal. I want the vertical narrow enough to mitigate floor and ceiling bounce and stable within the listening window.

The point I'm trying to make is that where you consider the horn as "starting to beam", it's what I would consider within a reasonable tolerance of nominal. And the only reason I'm trying to make that point is because the K402 polars are awesome for not having a diffraction slot.
Neither do PT waveguides. Do we see a diffraction slot in the 6" square $6 PT waveguide I have cited as extreme contrast/comparison here?

attachment.php

The Jub LF is solid up to 1kHz...
Suggesting that the humongous K402 with a 2" throat compression driver that'll play down to 500 Hz but not do so well above 7 kHz is gratuitous? May we see the data on the Jubilee LF, please?

I might argue that Klipsch's product line ultimately reflects what people look for when they choose a speaker.
And I would agree.... :yes:
 
Last edited:
Back
Top Bottom