Murphy Corner Line Array (MCLA)

[Jim Griffin]

Jerrald,

John has freely shared his design and we should all appreciate his expertise and efforts. But, I am not an advocate of a full range driver line array as I feel that you compromise the low and high frequency ends of the band. EQ helps but EQ alone will not eliminate these issues. Best bet in my mind is to use a three way line array system (or effectively a two way line array with active subs to carry the low end of the band). The builder of a line array needs to decide whether he can live with the compromises of a full range driver design but with no crossover.

Jim

[raiderone]

Jerrald,

John has freely shared his design and we should all appreciate his expertise and efforts. But, I am not an advocate of a full range driver line array as I feel that you compromise the low and high frequency ends of the band. EQ helps but EQ alone will not eliminate these issues. Best bet in my mind is to use a three way line array system (or effectively a two way line array with active subs to carry the low end of the band). The builder of a line array needs to decide whether he can live with the compromises of a full range driver design but with no crossover.

Jim

I really am enamored with the sound of line arrays, but I am not sure exactly why. It started with the Infinity RS1B many years ago. Then Pipedreams and recently Dali Megalines. I'm pretty sure it is the low distortion and effortless dynamics, but it might be the way line arrays propagate sound into the room. Sorry for the rambling, my question is if we only have room for one array, which one would be easier to crossover to: Low array + conventional speaker or Mid array with tweeter and subs? Thanks.

[richidoo]

Jim, thanks for sharing your paper. I enjoyed reading it, although I think it reinforces my concern about phase error being an issue at high freqs with this non-curved design.

A friend of mine has Pipedreams, which use separate tweeters, he says the high freqs are excellent, no sign of smear or incoherency. Unfortunately he lives too far away for me to audition, and it would not equal auditioning MCLA anyway. I would just like to understand the benefits and compromises of the MCLA design in the high freq region. Testing shows 1kHz to be very low distortion, with HF spectra likewise low. The aluminum drivers do not seem to have any propensity to ring in this high sensitivity application.

Jim, what specifically are the issues you have with full range drivers covering high freq and low freqs? I can imagine you mean the large drivers will beam at very high freqs, and not big enough for low freq linearity. I think the low freq response will be excellent when EQ is applied. I'm planningnto use 30 drivers per side and hope to achieve 25Hz flat response at moderate levels that i listen to music from 3 meters away. And the beaming of high freqs might actually reduce the perception of phase error at higher freqs, but will lihgt up the walls more than the direct sound, so some wall treatment may be needed.

I am especially attracted to the MCLA because of the full range no crossover feature. I have built other crossoverless speakers and I am convinced of their musical superiority, but always open to learning something new!
Rich

[John L. Murphy]

A friend of mine wants to build some PA speakers for his barn, and I was wondering if a set of MCLA-style arrays might do the job. The barn is approximately 20x30, with walls that are about 15' high before you hit the slanted sections that form the roof. For reasons of cost, the arrays would likely be limited to 18 drivers per side; as such, the benefits of running floor-to-ceiling would be lost (though with the slanted roof, I'm not so sure how well they'd work anyway.)

Assuming a subwoofer was crossed in at about 80hz, do you think this would work?

I think it would be a great system provided it is EQ'd for a flat (or other targeted) response.
I would suggest they go in the corners of the shorter (20') wall.

John

Thanks John for your reply. Do you think adding some more drivers say 5-6 to each side would help or is adding drivers a diminishing return?

I got two drivers and made a mini-MCLA to get an idea of the sound. My ears tell me the highs seem a little dull. Remember, I love ribbons. Any thought on a tweeters just for the upper octaves? I understand not having an xover is part of the love using a full range speaker. Maybe bring in an array of tweeters some where above 4khz (highest note on piano) which should keep any xover problems well above voice and most instruments. Maybe a small mylar on a bracket between and in front of each full range? I guess cost comes into question by doing this. Even a cheep tweeter would almost double the cost of the speaker. Well, just an idea. I didn't think of the cost while I was typing this.

Brad

Hi Brad,

Adding more drivers to extend the line closer to the (18') ceiling would help assure adequate low frequency output capability and extend the range of near field behavior.

I'm glad you auditioned the ND90 drivers. Unless you EQ'd them it is reasonable that the high end was less than perfect. You were likely hearing a response not terribly different from this:



This is the uncorrected ground plane measurement of the ND90 in a 0.1 cubic foot closed box. The upper octaves are shy and diffraction is causing the bass to droop. Not bad for a raw driver but the EQ'd flat response will sound much less colored.

Personally, I love the sound of a good full range driver when EQ'd flat in the top end. But there is no reason not to implement the corner line array as a 2-way system if you feel you must. Smaller drivers (say 1" domes) would broaden the (horizontal) high frequency dispersion in the top octaves compared to the slightly narrower dispersion of the 2.5" piston of the ND90. But remember that your listening angle is limited to lass than 45 degrees with the corner line array. This compensates for some degree of the concern about horizontal dispersion; vertical dispersion is not an issue since the high frequency response is expected to be near constant at any height within the length of the array.

If you were to implement the system as a 2-way I would encourage you to keep the crossover frequency as low as possible to minimize the overall audibility of the crossover. While you may gain increased horizontal dispersion at high frequencies you introduce the audible veil of a crossover along with a heap of added cost, complexity and failure modes. The simplicity of the full range crossover-less array is sweet.

Since you have two ND90's try a test where you compare listening to one driver full range to listening to two drivers crossed over at say 1k Hz. The crossed over pair SHOULD sound the same as a single driver...but does it??? Is the crossover audible??? How audible???

John

<snip>John, I am just about ready to go ahead. But I have one more question. I am not sure if this is what Jack was talking about above with doppler effect, or if it is what you mention in your article regarding comb filtering. I thought you were talking about wall reflections, but maybe you meant direct sound as well. Specifically, what I am concerned about it the phase error between the ear level driver and the highest driver is huge at high frequencies. 13kHz is 1 inch wavelength, and the different distances to ear between closest and farthest drivers are several inches. So the drivers will have interference and comb filtering, as will also happen from wall reflections. Is this something that I should be concerned about? I worry that high frequencies will sound smeared. I am familiar with the digital EQ methods and will use RTA to verify flat response.

Thanks for your input!
Rich

Hi Rich,

Any time you audition a sound source via multiple pathyways at once you introduce very complex comb filtering into any single point measured response. No loudspeaker can escape this reality unless it is auditioned in an anechoic chamber. And then, consider that listeners do not prefer the way music sounds without the usual room reflections (and attendant comb filtering). Listeners prefer the complex sound of a rooms compared to the single path (comb free) sound of an anechoic chamber. So you can relax your concerns about comb filtering and interference among the drivers at different distances. The sound from the multiple drivers all arrives well within the 20 to 30 millisecond fusion time window our ear-brain system employs to identify sound events.

For me, the subjective effect is that the line array (with its near field behavior) actually sounds more focused and detailed than any other speakers. In my testing I do not hear changes as I move my head while listening. Certainly there is no audible "flange" effect as is clearly heard when electronically combining two signals with different time delays. Such as when used as a studio sound "effect". To me, percussive clicks are defined as well as (or better than) on the best point source systems I've heard. While recorded material is compelling in this regard, it is playing my guitar through the arrays at close range that really convinces me that the line arrays can render the briefest percussive clicks and pops with absolute precision. I've worked professionally with quite a few electric guitar players over the years and have learned that they have an unusually intimate connection with the "sound" of their "rig" and are typically sensitive to little distortions that audiophiles playing recorded music would never notice in a thousand years of listening.

Regards,

John

[richidoo]

Thank you for sharing some additional subjective impressions, John.
Rich

[FredT]

Here's the link to a picture of the almost completed pair of ND-90 line arrays I'm building. The design is similar to Johns MCLA's except the room I'll be using these in doesn't have any corners, so the enclosures will be positioned a few feet out from the wall and will not take advantage of close-to-corner placement. I'll be using a subwoofer, and equalization & crossover to the sub will be accomplished by a dbx DriveRack PA+. Also, each speaker consists of two 12-driver stacked enclosures for easier transport.

http://fredt300b.smugmug.com/Hobbies...64865296_eqa7i

[blair]

Too bad you will not have those for LSAF!

Blair

[John L. Murphy]

I'd like to announce that the MCLA web site has been revamped and new project documents are available for download.

http://www.trueaudio.com/array

The original project document .pdf was created from multiple source documents and had the look of a printed web page, which is exactly what it was. The new document has been redesigned in M/S Word and the layout is much improved. The document name is unchanged in order to preserve all links to it.

I have also condensed the full 51 page document into a 12 page Project Brief. This is the document handed out at the 2010 Midwest Audio Fest. We had hoped to demonstrate a working pair of MCLA enclosures but there were assembly problems with one enclosure and the demonstration was cancelled. Perhaps next year we can demo a complete MCLA system.

Regards,

John

[oodesigns]

Hi, I am new to this group...

I am intrested in making these also... but I have the same issue as another member, i quote:

"I've had the MCLA's running for 4 days now but my first impressions have not been that great, mainly because of the distance between corners in my room. My corners are 17 ft. apart along my front wall, and my listening position can be 12 ft. max back from the front wall. As a result I'm getting too much separation between the left and right channels.

In the MCLA article, I can't see where John mentions anything about optimum spacing between speakers. I don't have any experience with using equalizers. Would optimum equalization settings minimize this problem?"

In my case I have no choice but to have the speaker further apart than I will be sitting. Will this make this design reduntant? or using could using other features of the equalizer help.

[auracle]

Hi John,

Just curious if you've had the chance to experiment with sound absorbing materials on the walls adjacent to the line arrays to see what impact it might have on imaging, tonal balance, and a sense of depth.

Having experimented a fair amount myself with line arrays, I agree wholeheartedly with some of your earlier observations about the actual percentages of direct/reflected sound in point source configurations versus line source. Given the comments I've read in various places, I think this aspect of line arrays is one of the most overlooked - concentrating the source of excitation in one place in most real environments produces greater contrast in both reflection time and intensity - often making reflected energy more noticeable.

[Chris Roemer]

Hi, I am new to this group...

I am intrested in making these also... but I have the same issue as another member, i quote:

"I've had the MCLA's running for 4 days now but my first impressions have not been that great, mainly because of the distance between corners in my room. My corners are 17 ft. apart along my front wall, and my listening position can be 12 ft. max back from the front wall. As a result I'm getting too much separation between the left and right channels.

In the MCLA article, I can't see where John mentions anything about optimum spacing between speakers. I don't have any experience with using equalizers. Would optimum equalization settings minimize this problem?"

In my case I have no choice but to have the speaker further apart than I will be sitting. Will this make this design reduntant? or using could using other features of the equalizer help.

You'll probably get more responses, but most traditionally, the best stereo listening occurs at a distance back that's from about 1x to 2x the speaker separation. In order to use some "corner line arrays", I'd think that they were designed to be used in the corners on the short wall.

chris

[oodesigns]

Thanks Chris

I was just woundering if there where any electronic ways to improve perceived stereo, when sitting close to speakers.

[Chris Roemer]

Thanks Chris

I was just woundering if there where any electronic ways to improve perceived stereo, when sitting close to speakers.

I think on some boomboxes there's a "Wide" switch that must throw (only) the tweeters out of phase to "spread" the sound. It's not right, though.

Chris

[John L. Murphy]

Hi, I am new to this group...

I am intrested in making these also... but I have the same issue as another member, i quote:

"I've had the MCLA's running for 4 days now but my first impressions have not been that great, mainly because of the distance between corners in my room. My corners are 17 ft. apart along my front wall, and my listening position can be 12 ft. max back from the front wall. As a result I'm getting too much separation between the left and right channels.

In the MCLA article, I can't see where John mentions anything about optimum spacing between speakers. I don't have any experience with using equalizers. Would optimum equalization settings minimize this problem?"

In my case I have no choice but to have the speaker further apart than I will be sitting. Will this make this design reduntant? or using could using other features of the equalizer help.

oodesigns,

I'm glad to hear you are interested in building your own MCLAs! Now, regarding your concerns... If stereo loudspeakers are spaced too wide the image will draw to one side very quickly as you move away from the center line making for an unstable image. The wider the speaker’s angle of separation the worse this effect will be. In general the MCLAs seem to provide a wide sweet spot but like any speaker, if they are spaced too far apart the imaging can become unstable.

I have my MCLAs placed across the narrow end of my 11.5 'x 27' room with my listening distance ranging from 5 feet (playing guitar) to 25 feet when listening at the very back of the room. They are consistent sounding all around the room but the sweetest area for listening to stereo recordings is the range from about 8 to 12 feet from the front wall. Because of the 45 degree angle of the enclosure the arrays axis cross at half the separation distance or about 5.75' in my case. A distance of 8' is about the forward limit for critical stereo listening and provides the widest sound stage I would want to hear for any extended time. The traditional "equilateral triangle" listening distance is 10' (.866 x separation distance) for my setup. Listening at 10' is very enjoyable and provides the traditional (60 degrees wide) stereo sound stage. At progressively greater listening distances beyond 10' the sound stage grows narrower but the imaging remains excellent. Even from the back of the room the arrays convey strong imaging with performers locations remaining firmly defined as you move about. If I had to pick a single listening position it would probably be the traditional equilateral triangle geometry at 10'.

The member you quote above, Geo had the arrays 17 feet apart with vaulted ceilings. I could not tell if he was using the Behringer EQ or not. Without the EQ the arrays are muddy sounding and certainly are not very impressive. He said he could listen at a maximum of 12 feet back which is less than the equilateral triangle distance of about 15' (for his setup). I would expect them to sound excellent even if the image was a bit wide. I wonder if a listening distance of 12' placed him right at the rear wall of the room. The vaulted ceiling would definitely not give the infinite line reflections you get when the ceiling is parallel to the floor. My first thought is that this room would likely reduce the low frequency output compared to the same room with ceiling parallel to floor.

Geo, can you tell us more? Did you EQ the system using the Behringer EQ? Of the three settings I give (flat, Sm Rm-X and X curve) which did you try?

Getting back to oodesigns questions...you said that you not sit as far back as the speakers are separated. That might be OK. For equilateral triangle listening geometry you only need to sit at a distance of .866 x separation, that is, a bit less than the separation distance. So your setup might be just fine. What are the actual distances for your room?

Regards,

John

[John L. Murphy]

Hi John,

Just curious if you've had the chance to experiment with sound absorbing materials on the walls adjacent to the line arrays to see what impact it might have on imaging, tonal balance, and a sense of depth.

Having experimented a fair amount myself with line arrays, I agree wholeheartedly with some of your earlier observations about the actual percentages of direct/reflected sound in point source configurations versus line source. Given the comments I've read in various places, I think this aspect of line arrays is one of the most overlooked - concentrating the source of excitation in one place in most real environments produces greater contrast in both reflection time and intensity - often making reflected energy more noticeable.

auracle,

No, actually I have not experimented with absorbers in the room at all. The room is fairly live but with a pleasing sound that I actually allow to leak into my acoustic guitar and vocal recording tracks. So I have not felt a need to tame any reflections up to this point.

Regards,

John

[John L. Murphy]

I think on some boomboxes there's a "Wide" switch that must throw (only) the tweeters out of phase to "spread" the sound. It's not right, though.

Chris

In this case oodesigns might need a "narrow" switch to try to compensate for the wide placement. It occurs to me that a partial blend to mono might help but even that would not help the case of a centered vocal which is sucked to the side too rapidly as you move off axis of too widely spaced speakers.

It actually sounds like oodesigns has just adequate separation to achieve an equilateral triangle listening setup. That should not be too wide.

Thanks for the comments Chris.

Regards,

John

[oodesigns]

Hi, John

thanks for the information I will be anaylizing it more carefully.

My room is pretty small about 360cm * 270cm - I based in the uk where the rooms are not big!.

One thought I had off the top of my head was to angle the speakers at greater than 45 degrees towards the center of where I will be (standing or sitting) .

The way I thought was to base the cabinet on a pipe which can be rotated, or would the completly destroy the whole concept.

plus as I am sooo close to the speakers would it be helpful to phase correct each driver based on my location, this would occur a cost of a electronics per speaker (ie active time alignment - with a small amp per speaker - this might be a bit nuts but then a system with 25 drivers per unit is quite over the top any way).

[John L. Murphy]

Hi, John

thanks for the information I will be anaylizing it more carefully.

My room is pretty small about 360cm * 270cm - I based in the uk where the rooms are not big!.

One thought I had off the top of my head was to angle the speakers at greater than 45 degrees towards the center of where I will be (standing or sitting) .

The way I thought was to base the cabinet on a pipe which can be rotated, or would the completly destroy the whole concept.

plus as I am sooo close to the speakers would it be helpful to phase correct each driver based on my location, this would occur a cost of a electronics per speaker (ie active time alignment - with a small amp per speaker - this might be a bit nuts but then a system with 25 drivers per unit is quite over the top any way).

It sounds like nice intimate setup for listening to music!

You would want to place the arrays in the narrow corners of the room 270 cm apart. For my 45 degree enclosure the axis of the two enclosures would cross at 135 cm from the front wall. The equilateral triangle setup would place you back (.866 x 270 =) 234 cm from the front wall. Your room would seem to be a good candidate for the MCLA.

There is no reason not to experiment with the angle of the enclosure in the corner. Basing the enclosure on a pipe makes sense. Note that you may well end up with a different enclosure volume per driver which would affect the required corrective EQ.

For a similar pipe enclosure design see Darren Kuzma's line array project using a 4 inch pvc pipe as the enclosure:

http://www.partsexpress.com/projects.../Kuze3201.html

Considering the reflected images of the arrays in the floor and ceiling I don't think it is desirable to apply delay to any of the drivers. I actually tried the experiment of delaying the signal to drivers in my MLA8/16 arrays back in the 1980s. I never found a delay configuration I liked better than the un-delayed array. For example, if you attempted to focus the array at the listening position each reflection of the array would be miss-focused and spoil the whole plan. Likewise with amplitude tapering the array; each reflection would also be tapered messing up the whole tapering arrangement.

Let us know how your project progresses.

Regards,

John

[oodesigns]

Thanks John for the information

One Question about the "Behringer Ultra-Curve Pro DEQ2496." my better half would like to know if the rack mounts can be taken off as she thinks they do not look good, if not I will have to look for a small 1 unit high case.

[John L. Murphy]

Thanks John for the information

One Question about the "Behringer Ultra-Curve Pro DEQ2496." my better half would like to know if the rack mounts can be taken off as she thinks they do not look good, if not I will have to look for a small 1 unit high case.

I checked my unit and the rack mount ears are part of the chassis...so no, they are not removable (unless you want to look inside the chassis).

Regards,

John