Jordan JX6 full-range line array

[Colin]

Okay, an evening and morning of experimentation with the toe-out idea. This morning was interesting because I could use a Radio 3 programme which combines music and presenter speech to a good broadcast quality.

Toe-out with the JX92 gives a slightly fuller sound, especially on speech, and a broader soundspace whilst I stay on the centre line. This is particularly so if I move away from the line between the speakers. But ... as soon as I move left or right, the image collapses to the nearer speaker. It doesn't have to be a large move, either.

Putting the speakers back to the recommended toe-in compresses the sound somewhat so that it is more between the speakers but moving left or right keeps the image in place. There's also a fair amount of depth behind the speaker line.

Interesting experiment and other set ups and - more especially - other drivers may alter give a different result. In particular, I wonder if more conventional drivers or woofer/tweeter combinations would work better with the toe-out idea. What speaker set up are you using Scott?

 

[ScottG]

Okay, an evening and morning of experimentation with the toe-out idea. This morning was interesting because I could use a Radio 3 programme which combines music and presenter speech to a good broadcast quality.

Toe-out with the JX92 gives a slightly fuller sound, especially on speech, and a broader soundspace whilst I stay on the centre line. This is particularly so if I move away from the line between the speakers. But ... as soon as I move left or right, the image collapses to the nearer speaker. It doesn't have to be a large move, either.

Putting the speakers back to the recommended toe-in compresses the sound somewhat so that it is more between the speakers but moving left or right keeps the image in place. There's also a fair amount of depth behind the speaker line.

Interesting experiment and other set ups and - more especially - other drivers may alter give a different result. In particular, I wonder if more conventional drivers or woofer/tweeter combinations would work better with the toe-out idea. What speaker set up are you using Scott?

Thats about right for the subjective response.. there is a constraint on moving about, but I usually find that you can move your head from side to side a bit and without gross deviations. Moving the speakers a bit closer together while still being "toed-out" should improve the center fill and the depth plane a bit. Note that it won't improve depth a great deal if your speakers are operating properly. My guess is that the enclosure you have for JX6 is sub-par - and is causing you a loss in depth perspective. (i.e. be wary of airflow resistance in the cabinet - and to reduce back pressure consider a "lossy" cabinet - i.e. an aperiodic enclosure.)

IMO the real problem with the JX6 is that it doesn't have enough rise in treble response, and conversly that it doesn't have enough linear off-axis behaviour. Ideally I'd like to see a flat on-axis response at the 60 degree mark out to at least 13 kHz, with a linear rise approaching +12db at 13 kHz. As an example here is a tweeter with farily flat on-axis response to 13 kHz (and beyond that actually):
http://www.tymphany.com/datasheet/printview.php?id=115
Now imagine it as if its 60 degree axis-response was the flat response and every thing was elevated up proportionatly (..ie. its on-axis behaviour was elevated +12db at 13 kHz).
Then rotate the driver "outward" so that the listener is on the tweeter's 60 degree axis.

There IS a solution howevr for the JX6.............. EQ.!

I personally use a Bit-88 tube based SRS product that applies eq. in a similar fashion to achieve this treble-rise. (..it also performs some phase tricks to stabalize imaging in the depth plane and provides BSC compensation eq..) It isn't perfect, but it does allow me to rotate my main loudspeaker, (a heavily modified panasonic loudspeaker - single low value cap speaker - essentially a 9-10" full range driver with a cone tweeter high-passed at around 5 kHz and a piezo for freq.s above 15 kHz), about 15 degrees off of the "0" axis (..or said differently "toe-out" of 15 degrees from a forward fireing position). With this combination I get stable imaging (recording dependent) left of left speaker and right of right speaker and the ability to "move about", (i.e. no "head-in-vise" phenom), at my listening position without suffering image shift. I get deep soundstage and imaging effects, BUT what I don't get is "image density" - and this is due to dampening factor (which is to high), and in particular polar "combing" caused by such a large diameter driver operated that high in freq.. (..note I also have several other project loudspeaker - including a 166esr based system with dipole midbass drivers, and a cheap array.. but I always come back to the panasonic for its particular blend of "sonic" features.)

What I find interesting about all of this is that the mode of operation I've noted is superior for typical "high-end" 2 channel/one listener listening, while Ted's method is superior for frontal speakers on a Home Theater set-up or for two channel background music listening. With the latter configuration you have a "lock" on centeral imaging - allowing you multiple listening positions. Despite this I have yet to see this fully incorporated into a Home Theater design. (..go figure.)

 

[soongsc]

The toe-in does not really require a rise in frequency response. What the toe-in does is that at center listening position, you are equally off axis to both speakers. When you move one way, you are moving closer to center axis of the opposit speaker thus the increased response will some what compensate for the increased distance.

Assuming that the originally recording had perfect frequency response, since the speakers are placed at a specific location not equal to the real instrument location, thus the response is not faithfully restored because the position relative to the ear is not the same. This is one reason why when you look at binaural playback circuits, it actually rolls of the high end a little. However, this roll-off will not be required if we are listening off-axis.

Sometimes personal preference lean toward a little more high end, and this is fine.

 

[bcherry]

Colin,

He also mentioned that the 4-unit array does the job perfectly well and the only real benefit from going to a longer array would be higher power handling for larger rooms, which seems to be what Brian is aiming at.

OK, thanks. Did he mention anything about increased combing effects with the 9driver array? I'm about ready to get some cabinets made to move this project along but don't want to take one step forward and another back.

ScottG,
The Bessel array. I could also do a 3x3 array. No problems sitting far enough back for that but would there be a minimum number of drivers for this? I searched but couldn't find any Bessel project details on this site. Maybe I'm not using the right key words.

Brian

 

[ScottG]

The toe-in does not really require a rise in frequency response. What the toe-in does is that at center listening position, you are equally off axis to both speakers. When you move one way, you are moving closer to center axis of the opposit speaker thus the increased response will some what compensate for the increased distance.

Assuming that the originally recording had perfect frequency response, since the speakers are placed at a specific location not equal to the real instrument location, thus the response is not faithfully restored because the position relative to the ear is not the same. This is one reason why when you look at binaural playback circuits, it actually rolls of the high end a little. However, this roll-off will not be required if we are listening off-axis.

Sometimes personal preference lean toward a little more high end, and this is fine.

Right - thats Ted's method (toe-in from listener axis), AND not only is the farther speaker higher in direct spl, but the near speaker is LOWER in direct spl (..all within the 2kHz and up passband).

The higher gain I spec'ed for the "toe-OUT" version is for two reasons:
1. the axis FROM the listener's axis is MUCH further than the "toe-in" version. The "toe-in" version may require only 15-to-20 degrees "inward" from the listener's axis. The "toe-OUT" version may be 60 degrees or more from the listener's axis. To maintain a flat-on-listener axis response - you will need a LOT of gain on the driver's axis to compensate for the loss in spl off-axis (..assuming the driver isn't "omni" horizontally at higher freq.s).
2. a side benefit here is the increase in average response at these higher freq.s for distances greater than 1 meter. When you do this however, it becomes important to have an extended low freq. response to subjectivly "balance" the sound.

The problem with both methods - and indeed stereo sound seperation with L+R and R+L in general - is that you hear not just the listener's axis summed with a room averaged response, but that you also hear ALL the other axis's (horizontally that is) - with the "frontal" 180 degrees relative to the listener being the most important. (..your brain acomplishes this by referencing the spl to time - and this is one major reason why we perceive imaging cues at higher freq.s.. i.e. the wavelengths are shorter in distance and thus are easier to distinguish in time - down to the low micro-second range. Conversly we perceive phase (as opposed to spl) at lower freq.s as important for imaging because direct sound is easily distorted by boundries like the floor because of long wavelengths, but the phase reversal from the boundry reflection in time is easier to detect.)
Note that this is why traditional directive loudspeakers at higher freq.s are a bad idea. (..usually justified with the notion that they are superior because they reduce room reflections - never stopping to truely understand how they perceive an increase or decrease in room reflections at these higher freq.s, nor considering the loss in averaged spl relative to the low freq. response from the listener's position.)

(..I could go into more detail *why* the "toe-OUT" version is more accurate for a typical one centered listener, but it would take a lot more time and I've spent to much already. sorry.)

 

[ScottG]

Colin,


OK, thanks. Did he mention anything about increased combing effects with the 9driver array? I'm about ready to get some cabinets made to move this project along but don't want to take one step forward and another back.

ScottG,
The Bessel array. I could also do a 3x3 array. No problems sitting far enough back for that but would there be a minimum number of drivers for this? I searched but couldn't find any Bessel project details on this site. Maybe I'm not using the right key words.

Brian

There are only two legit bessel arrays:
1. the 5 driver array
2. the 25 driver array
(..perhaps there is an even greater multiple that will work as well - but I suspect its a moot point.)

vertically, the only practical bessel array is the 5 driver array. If a 25 driver array was used strictly for vertical orientation then you would not be able to get far enough away from the line for the comb filtering to desist (..generally 10 times the line length - i.e. 7 foot line ='s 70 feet or more distance).

as an experiement I once wired a super tweeter set in the five driver array config.. but I have no idea what that wiring is now. The best I can do is say start searching (sorry):

http://www.google.com/search?hl=en&q=bessel+array&btnG=Google+Search

 

[vinnie]

I found this diagram for how to wire a bessel array, thanks dave.

 

[Norris Wilson]

Bi-polar array?

Hi Brian

Forgive my inexperienced butting in.

But, if you are mainly looking to increase the output capability of the JX-6 full range four driver array.

Have you thought of just putting the four driver array cabinets back to back to see if a bi-polar array would be worth your effort?

Possibly then, you would forgo any combing effects and detrimental EQ problems.

Just a silly thought.

Norris Wilson

 

[bcherry]

ScotG

There are only two legit bessel arrays: 1. the 5 driver array 2. the 25 driver array

Vertical 5 then. OK searching.

as an experiement I once wired a super tweeter set in the five driver array config

Any comments on the result? Did it really perform like a single driver?

Vinnie:

I found this diagram for how to wire a bessel array

Thanks. That's a start. It appears only the second version with 5drivers will give a usable impedance. Some sites have said each driver in a Bessel array must be in their own 'tuned cabinet. Any associated information on suggested cabinet?

Norris:

Have you thought of just putting the four driver array cabinets back to back to see if a bi-polar array would be worth your effort?

Appreciate your input too. I hadn't thought of that as was aiming for a system that would exploit the off-axis response of the JX6 and be mounted on the wall as suggested on Ted Jordan's site. But....thinking about it, would the drivers fire straight forward and back or two boxes configured as per Ted's 30degree edge mounting? Interesting.

Brian

 

[pikkujöpö]

WOWand WOW


This is off topic,but I happened to read what ScottG (thanks)was writing about toeing-out the Jordans.

I have had loads of trouble in getting my MLTL48" (jx92s) sound right.What has bothered me most are the vocals in some recordings,even with moderate sound levels the vocals sound too "harsh and sharp"to my ear.The problem is propably room related as my other speakers do not sound any better.

Now I desided to try the toe-out method,and it works wonders .When toed-in the best sound was at about 65degrees,but the soundstage was somewhere there in front and not very involving.
Now with plenty of toe-out the vocals are sweet and it feels like I'm inside the soundstage,the feeling is not far from the 6Ch stereo mode (which I never used as the rear speakers are too different).

It sounds like the speakers are not as "transparent" with toe-out,but I have listened only for couple of hours and the speakers may not be correctly placed yet.

I hope my ears aren't playing tricks on me ,cause the Jordans sounded really fabulous and sweet today

 

[bcherry]

Line Array simulator

Anyone seen this one? Or any other software that can predict line array implimentations? I can only get it to work with rectangular drivers.

Also reading some conflicting info on supposed 'cylindrical wavefront'. According to this site:

The common misconception regarding line arrays is that they somehow magically enable sound waves to combine, forming a single "cylindrical wave" with special propagation characteristics. Under linear acoustic theory, however, this is impossible: the claim is not science but a marketing ploy.

ProSound

They have some interesting pictures.

Am I lost?

Brian

 

[Colin]

I hope my ears aren't playing tricks on me ,cause the Jordans sounded really fabulous and sweet today

Not off topic at all - and btw, thanks to Scott for taking the time to share info with us. Facinating stuff.

What toe-out did you use? (Straight ahead or farther out?) I tried it with my 48 JX92s with the results described earlier but I suspect I may be limited by reflections from side walls as I have the enclosures by walls. More experiments needed, perhaps. I have been pretty satisfied with the stereo image to date in the conventional Jordan position, they give a good sense of the wall disappearing and a room beyond and behind the speakers. Expanding the image beyond the enclosures hasn't been too much of a concern as the speakers are against side walls and, psychologically, that might not work.

BTW, Brian - don't read too much about the arrays. Ted's been working with them for thirty years and is very stong on the basic physics. try the array as desribed and see what happens. I think the 9-array is prbably only needed to get it to the same impedance as the single unit. If you're happy halving or doubling impedance, 8 will be fine. If you get HF problems, then you could try the 12 and, as discussed earlier, keep the centre 4 fullrange and roll off the outer two arrays around 500Hz or less. This would get you the increased power handling. (Although a chepaer way would be to roll off the 4-unit array a little higher and cross to a woofer.)

 

[Colin]

And the pro-sound site seems to be talking PA applications and large venue arrays.

 

[bcherry]

(Although a chepaer way would be to roll off the 4-unit array a little higher and cross to a woofer.)

Thanks Colin,
I have the 4 driver/woofer now but I'm used to the midrange/treble dynamics of compression horns and >10" drivers. I'm hoping the 9drivers will give more 'pop' along with the spun-silk quality of the JXr6, but it still needs cone surface area to to get that visceral impact in the mids and highs. Also crossing over the subs high lands right on a sensitive area for male voices - tenors sing down to 80hz and baritone/bass singers even lower.

Agreed about Ted's experience. So unless he's indicated something otherwise about 9 divers vs 4 then I guess best is just to move ahead with the 9 drivers in a box similar to the 30degree edge box on his site.

Brian

 

[pikkujöpö]

What toe-out did you use? (Straight ahead or farther out?)

Farther out...
After some more listening today,the best toe-out seems to be aproximately 40-45 degrees(from listeners axis).The speakers are placed 2meters away and 2meters apart,only 20cm from the side walls.

With less toe-out they just don't sound right at all(they sound rather weird),but when there is enough the stereo image becomes "solid" again.
There are some strong reflections coming from the side walls now,and I'm considering putting some absorbers to the first reflection points.

 

[ScottG]

Not off topic at all - and btw, thanks to Scott for taking the time to share info with us. Facinating stuff.

What toe-out did you use? (Straight ahead or farther out?) I tried it with my 48 JX92s with the results described earlier but I suspect I may be limited by reflections from side walls as I have the enclosures by walls. More experiments needed, perhaps. I have been pretty satisfied with the stereo image to date in the conventional Jordan position, they give a good sense of the wall disappearing and a room beyond and behind the speakers. Expanding the image beyond the enclosures hasn't been too much of a concern as the speakers are against side walls and, psychologically, that might not work.

Your Welcome!

Side wall reflection typically isn't a serious problem unless there is correlation (i.e. where the freq.'s wavelength is within the freq.). Usually correlation is a problem at these more directional freq.s (i.e. the 2kHz an up passband). I personally have one speaker that is about half a meter away from a side-wall, and don't experience signifianct alteration to the sound.

Note that the comb-filtering at these higher freq.s simply diffuses imaging (making the image "out-line" a bit more "fuzzy"). It does the same basic thing to the soundstage - and in particular effects this character more than comb filtering from an array. To a degree it also effects the expansiveness of the soundstage (..though this is far more a low freq. effect). It can however be benefical (substantially so) to absorp higher freq.s from side-wall reflections AND from front-wall reflections. Typically a tastefull decor solution can be obtained with curtains that can be "drawn" into place when listening. A better solution though is painting the room with the same material I listed for painting the inside of the box and driver frame. Note that the manufacturer has a ready-made product that will likely be better for this application - producing a better "finish":

http://www.hytechsales.com/prod150.html

 

[ScottG]

ScotG

Vertical 5 then. OK searching.

Any comments on the result? Did it really perform like a single driver?

yes.. but I created it for a horizontal array (as opposed to a vertical one). The interesting thing there is that with the horizontal orientation it sounded more like a tweeter that was aproximatly 7 inches in diameter. With a vertical array though I don't think it will have quite this effect (..but I'm not sure). Still, if you have the drivers it effectivly costs nothing to try it out.

I still don't think the bessel array will provide any significant audible contribution to this driver over a standard array. Furthermore with a standard array you can play with height/line length and listener distance to get the treble response you want at the listener's position (..see the graph on page 9 of Dr. Griffin's line array paper for more information).

 

[bcherry]

here we go - 9 driver linear array with the JR6

The test boxes arrived, honeycomb sealed cabinet, about 16L, wide baffle and cutout for 9 drivers. Will tomorrow post a pic of the cabinets. In the meantime I've tried to gain an understanding of the nearfield vs farfield combing issues with the 4 driver array. So far as I hear combing is not an issue in either case but expecting the worst and hoping for the best.

Ted Jordan was very gracious in a telephone conversation to give some guidance about the cabinet and placement, and not to worry (in our large room) about mounting the drivers on the angled edge of the box. This was a relief as I definitely wanted them in the wide part of the baffle. It's worked out very well for the 4 driver and I've kept the same cabinet depth and width, just making them taller. Ended up pointing the 4 drivers directly at the listener to get a deep soundstage to the ceiling and well outside the speakers. The arrays are not the champions in the pinpoint imaging department such as the LS35A but not noticeably lacking either.

It seems linearrays need more tweeking of positioning to get the best out of them but I'm slowly learning and getting a very smooth dynamic sound now. So looking forward to 9 drivers which will be roughly equivalent to an 8" inch driver but with the combined motor power of a strong 15" AND free-to-air like a ribbon!

Plans are to connect them as per Colin's posting from Ted's book.

Cheers
Brian

 

[bcherry]

the cabinet

.

 

[soongsc]

Looks nice. Are you going to put them in corners to get the most bass out of them?