One thing I still don't understand about line arrays is why the tower is not equal to the sum of it's parts. Especially the "hump" around 300Hz.
From a discussion I had with Fluid, regarding his TC-9 build (I hope he doesn't mind or it's not against forum rules since nothing personal was quoted):
It seems all attempts at building line arrays seem to arrive at this very similar looking EQ nightmare. A small speaker may sound great in a little sealed or ported box, yet response goes all over the place when an array is assembled. So the peerless TC9 have a reasonable response from 125Hz to above 10kHz yet the 25 driver arrays all have a midbass "hump" in the midbass around 300Hz. Since the height at 3.5" spacing is 87.5" for a 25 driver array, 300Hz amounts to a wavelength equal to 44", about half the height of a column. Now I can understand the 6dB rolloff below 100-150Hz as this driver exhibits the same properties single unit, and I can also understand that the output can get erratic beyond 3800Hz, the point at which a full wavelength becomes smaller than the driver spacing. This is largely be due to phase cancellations and off axis beaming. And at some distance from the line array, the dB rolloff moves from behaving as a line source at 3dB per doubling of distance, to behaving as a point source at 6dB per doubling. This is dependant upon frequency, with lower frequencies making this transition earlier than higher frequencies, as distance from the array increases. So I wonder if each EQ compensation curve is only good for a certain room position or listening distance, ie to change the "sweet spot" you have to change the EQ settings? Then there's the question whether or not to compensate with a sub.
The EQ stage definitely complicates things. Currently I have an amp plugged into a variety of analog sources with a pair of Polk Monitor 40s Series II bookshelves (D'Appolito or MTM arrangement) paired with an old 8" Sony Subwoofer with surprisingly deep bass response. I've been listening to some Organ CDs and the way I have the system tweaked sounds amazing. Do the line arrays really lose the point source effect (ie close your eyes and you can no longer "see" the speakers with your ears, as if you were wearing headphones) creating an even sound stage over a large area of the room, or does it become an EQ nightmare with each "sweet spot" requiring EQ tweaks???
Maybe I'm living in a pipe dream, but I'm intrigued by the concept of line arrays for home reproduction and would love to build such a project (perhaps on a smaller scale than full floor to ceiling 25 drivers, maybe a 9 driver or 16 driver system with listening area in the sweet spot would suffice), but budgetary restraints and the fact that it needs an active EQ stage are putting me off on the idea. Since the system would need active equalization at the preamp stage, I could no longer use my amplifier to select sound sources but instead would need to run all sources through a mixer prior to the EQ, prior to the amp. So instead of one "box" I'd need three.
My ideal HiFi speaker would be something with good bass extension, crisp but not harsh treble, that sounds great plugged straight into the amp with a perfectly flat EQ. The Polks fit this description nicely for me. Cheap full range speakers on the other hand typically need a moderate bass boost and light treble boost to sound good, and tend to have a warm, full sound when properly attenuated. A stock two or three band EQ will usually suffice for this purpose, but is out of the question for the line arrays it seems.
Or as it were, "the product does not equal the sum of it's parts."
Time to get off my soap box...
From a discussion I had with Fluid, regarding his TC-9 build (I hope he doesn't mind or it's not against forum rules since nothing personal was quoted):
fluid said:
It seems all attempts at building line arrays seem to arrive at this very similar looking EQ nightmare. A small speaker may sound great in a little sealed or ported box, yet response goes all over the place when an array is assembled. So the peerless TC9 have a reasonable response from 125Hz to above 10kHz yet the 25 driver arrays all have a midbass "hump" in the midbass around 300Hz. Since the height at 3.5" spacing is 87.5" for a 25 driver array, 300Hz amounts to a wavelength equal to 44", about half the height of a column. Now I can understand the 6dB rolloff below 100-150Hz as this driver exhibits the same properties single unit, and I can also understand that the output can get erratic beyond 3800Hz, the point at which a full wavelength becomes smaller than the driver spacing. This is largely be due to phase cancellations and off axis beaming. And at some distance from the line array, the dB rolloff moves from behaving as a line source at 3dB per doubling of distance, to behaving as a point source at 6dB per doubling. This is dependant upon frequency, with lower frequencies making this transition earlier than higher frequencies, as distance from the array increases. So I wonder if each EQ compensation curve is only good for a certain room position or listening distance, ie to change the "sweet spot" you have to change the EQ settings? Then there's the question whether or not to compensate with a sub.
The EQ stage definitely complicates things. Currently I have an amp plugged into a variety of analog sources with a pair of Polk Monitor 40s Series II bookshelves (D'Appolito or MTM arrangement) paired with an old 8" Sony Subwoofer with surprisingly deep bass response. I've been listening to some Organ CDs and the way I have the system tweaked sounds amazing. Do the line arrays really lose the point source effect (ie close your eyes and you can no longer "see" the speakers with your ears, as if you were wearing headphones) creating an even sound stage over a large area of the room, or does it become an EQ nightmare with each "sweet spot" requiring EQ tweaks???
Maybe I'm living in a pipe dream, but I'm intrigued by the concept of line arrays for home reproduction and would love to build such a project (perhaps on a smaller scale than full floor to ceiling 25 drivers, maybe a 9 driver or 16 driver system with listening area in the sweet spot would suffice), but budgetary restraints and the fact that it needs an active EQ stage are putting me off on the idea. Since the system would need active equalization at the preamp stage, I could no longer use my amplifier to select sound sources but instead would need to run all sources through a mixer prior to the EQ, prior to the amp. So instead of one "box" I'd need three.
My ideal HiFi speaker would be something with good bass extension, crisp but not harsh treble, that sounds great plugged straight into the amp with a perfectly flat EQ. The Polks fit this description nicely for me. Cheap full range speakers on the other hand typically need a moderate bass boost and light treble boost to sound good, and tend to have a warm, full sound when properly attenuated. A stock two or three band EQ will usually suffice for this purpose, but is out of the question for the line arrays it seems.
Or as it were, "the product does not equal the sum of it's parts."
Time to get off my soap box...
Look at those on and off axis curves. Imagine sitting to the right of the right speaker AND still hearing the left speaker. Also, to me its all about dynamic range, but as has been noted before, the source recording/mixing/mastering is very critical.
My biggest problem with line arrays is that the line length (measured in wavelengths) varies according to frequency.
So, you get -3dB per doubling of distance for some frequencies and not others. The result is that as you move away, the overall tonal balance goes wrong.
Floor-to-ceiling arrays are effectively infinite, so that works okay. Anything smaller, there'll be problems.
Chris
So, you get -3dB per doubling of distance for some frequencies and not others. The result is that as you move away, the overall tonal balance goes wrong.
Floor-to-ceiling arrays are effectively infinite, so that works okay. Anything smaller, there'll be problems.
Chris
Exactly. They are still essentially a point source for bass. I get the floor to ceiling concept (25 drivers x 3.5" spacing ideal for modern domiciles with 8 foot ceilings).
The MTM arrangement is popular with HiFi speakers for good reason because it yields a no BS arrangement in a manageable size as a balanced point source on the horizontal plane, sacrificing vertical imaging for horizontal, excluding it's popularity as a center channel design with poor imaging when laid horizontally...
I noticed a lot of commercial line arrays for home use employ smaller drivers (smaller driver = less low end output) and often feature dual arrays with crossover. They seem to favor an inverted "J curve" design opposite that typically employed at concert venues and auditoriums, with the column is mounted on the floor rather than the ceiling, vertical at the base then tapers backwards before terminating some 5-6 feet above the floor.
These smaller designs would be friendlier to consumers and blend better into living spaces, however they would need a dedicated sub located in a discrete location (due to room loading effects at sub-100Hz frequencies, low end bass is non-directional and ignores the 6dB per double distance rule, instead creating a fairly even pressure throughout enclosed or partially enclosed living spaces), and as well look like a broken arch rather than a monolith.
I thing the inverted concert design has practicality for a living space, but the arch creates a torroidal radiation pattern losing the 3dB advantage of the line array. I do believe the IDS-25 style design is a bit excessive (far cheaper to brew your own as people like Fluid can attest) for a living space unless you have a dedicated room for theater and/or listening.
I had an idea floating about in my mind for building a 5.1 surround system using mini line arrays, though the center channel would be a pita to build. Basically an L/R with 9 drivers each, Surrounds with 4 drivers each placed slightly behind and to the sides of the seating area. They would be adjusted for optimum height with the ear placed in the center. Now for the center channel. Since a single full range wouldn't have the SPL or tonal matching to the 9-arrays, yet cannot be placed vertically due to blockage of the TV, A center with 9 drivers must lay horizontally. This would be terrible for anyone not seated squarely in front of the center channel if aligned in a straight line, so we create a torroidial radiation pattern with drivers spaced 4 inches apart as sides of a 24-gon, each driver angled 15 degrees apart with the spread from the 1st and 9th drivers at 120 degrees, creating a pie wheel repesenting one third of a circle. Ideally this platform would be wide enough to support the base of an HDTV, if not wall mounted behind it. The torroidial radiating area wouldn't quite have a 3dB rolloff but it would cover a broad listening area with sound coming from the TV with correct loudness.
Incorporating a suitable EQ curve into all 5 chanbels would be a PITA since most dedicated units only accept stereo inputs and most HDMI video equipment does not permit splitting discrete decoded signals prior to input into the amp. Older pre-hdmi amplifiers had discrete 5.1 analog inputs for PC sound cards but those have gone the way of the dodo. One could use the Pro Logic II downmix and run it through the EQ stage as stereo before sending it back to the reciever to extrapolate the encoded channels, but doing so would be a bit of a kludge and a disservice to the soundtrack.
I have a vivid imagination.......
Last edited:
after having heard and built a fair range of loudspeaker configurations and driver technologies in over 50yrs, my short answer to the question in the title of this thread would be
"nope, there is no best - period - all have some compromises and/or penalties of practicality "
"nope, there is no best - period - all have some compromises and/or penalties of practicality "
Yes, like with all speaker concepts there are compromises. Though with a bit of help and care line arrays can be amazing. Check out the reviews I have linked in the first post of my thread. That should give you an idea of what is possible.
A wide sweet spot certainly is possible with arrays. Personally I'd never go with short arrays, as they would bring bigger compromises and/or problems than the full range floor to ceiling array do. Much more theory about that can be found in my thread.
For processing I use a PC with a dedicated sound card (Asus Xonar Essence ST) and a separate DAC for the main arrays. That way I can serve up anything I want to my TV/speakers for home theatre and use dedicated filters for every audio branch I use.
Basically giving me full control.
I have no subs (yet) and no centre channel either, Home Theatre is 4.0 for me. That means a phantom centre and the line arrays playing the sub bass.
EQ is part of the deal with arrays, no way around it for full range arrays, but there are solutions for every problem 🙂. If you look here you'll find a theoretical analogue solution for the needed EQ which should work (not tested yet).
A wide sweet spot certainly is possible with arrays. Personally I'd never go with short arrays, as they would bring bigger compromises and/or problems than the full range floor to ceiling array do. Much more theory about that can be found in my thread.
For processing I use a PC with a dedicated sound card (Asus Xonar Essence ST) and a separate DAC for the main arrays. That way I can serve up anything I want to my TV/speakers for home theatre and use dedicated filters for every audio branch I use.
Basically giving me full control.
I have no subs (yet) and no centre channel either, Home Theatre is 4.0 for me. That means a phantom centre and the line arrays playing the sub bass.
EQ is part of the deal with arrays, no way around it for full range arrays, but there are solutions for every problem 🙂. If you look here you'll find a theoretical analogue solution for the needed EQ which should work (not tested yet).
Sorry to muddy the waters, been away from audio for years.
So the OP mentioned a MTM which got me thinking, since I am doing something similar MMTMM currently and MMMMTMMMM soon. The M (midrange) is actually fullrange with possible high pass filter. The T (tweeter) is a super tweeter crossed over @10kHz+ with a capacitor...
So my question is: At what point is a design considered a line array?
So the OP mentioned a MTM which got me thinking, since I am doing something similar MMTMM currently and MMMMTMMMM soon. The M (midrange) is actually fullrange with possible high pass filter. The T (tweeter) is a super tweeter crossed over @10kHz+ with a capacitor...
So my question is: At what point is a design considered a line array?
I have not idea but it seems the crossover should be a bit lower so that it approximates the wavelenth between the speakers where sideband radiation gets fuzzy. Polk released an MMMTMMM center channel that IMO would produce very poor sideband spread patterns if listened from the sides. I believe quality built soundbars counter this by using a DSP to help shape the sound by beaming it in different directions for different channels, but there is no way to do this in a passive speaker. My guess is the tweeter and midbass would radiate fairly wide with a hole in the midtones for off axis radiation. Placed vertically, it may have good dispersion but it won't have the low end excursion of the MTM designs like the old monitor series.
Good question about arrays though.
Ronald, my post above was not intended as a critique of any well engineered or executed build - yours among them. Of the several line array types of systems I've had the opportunity to hear during the past 15yrs or so, I've yet to be blessed with the transcendental experience that would be expected from a "holy grail", so to the OP's question, my response stands.
Hi Chris, I didn't take it as such. I agree with you, every system we choose will have its own set of compromises. It pays off to choose the one that fits your needs and spend some time to optimize its performance.
For the OP's goals I'd say abandon the idea of using the short array concept and go with Synergy/Unity style speakers.
Those are a really smart MTM concept that works over a large coverage area.
If you'd still insist on arrays go floor to ceiling. Though it needs EQ, as part of the concept. They truely are awesome for HT. But no doubt, so are the Synergy designs.
For the OP's goals I'd say abandon the idea of using the short array concept and go with Synergy/Unity style speakers.
Those are a really smart MTM concept that works over a large coverage area.
If you'd still insist on arrays go floor to ceiling. Though it needs EQ, as part of the concept. They truely are awesome for HT. But no doubt, so are the Synergy designs.
Thanks Wesayso. I read the first 90-some-odd pages of your build thread, so I was right there feeling the horror when it "cracked" then following up on how you fixed the issue. Your design was truly over the top in terms of execution and effort, but gorgeous internally and externally. Honestly I believe a stuffed box like Fluid is doing would sound just as good. That appears to be basically what the IDS-25 is.
So shorter arrays at ear level are a bad idea then? How many speakers get stacked before the EQ balance goes haywire? MTM designs prove that two speakers are safe. I could go up to 4 or 6 or 8 or 9 drivers but how many can I add before the frequency response of the drivers goes south? I have a stack of four GRS-3FR-4 speakers now since I've upgraded the birdhouses to TC9s. I'll admit the TC9s do sound much better than the GRS with clearer highs and aren't midtone heavy like the GRS speakers. The TC-9 also add a bit more "punch" but really shine with a 6dB bass boost.
Back to the "dirt cheap" lazy build ideas, a vinyl fence post and 18 GRS drivers would build a respectable 8-ohm line array covering ~70 the space between the floor and ceiling (for 4-ohm drivers, you really want to build an 8-ohm impedance by doing two squares in series). Times two for stereo, but even 36 GRS drivers at $3.80 each is more than I'm willing to invest atm. Bill of materials is already at $200 for the pair not counting wiring and stuffing. Poly stuffing could be bought at hobby Lobby or somewhere and much safer to work with than fiberglass or the pink stuff. I have all the necessary tools though, assuming a 3" hole saw doesn't split the vinyl. The router would be safer for plastic work but would take way longer.
I recall a youtube video featuring someone in their garage building an array out of dirt cheap mobile audio drivers (I know of a local Pawn shop that has 3, 4, and 5" drivers galore for $1 each but I'd bet they sound pretty bad) and a fence post. Just stuff the heck out of it to dampen any resonance. I could get a bail of poly stuffing from Hobby Lobby or somewhere.
I'm still sold on the beauty and elegance of the MTM design, and still love my Polks Monitor 40s which utilize them, and although I've heard reports they aren't quite audiophile quality, they are IMO the greatest sounding speakers I've ever owned. I do love the sound of silk dome tweets as the metal ones like Klipsh use sound harsh. My ears haven't been ruined by loud music like much of the ipod generation, and ever since I was a child, I was extremely sensitive to loud noise. So people who've already trashed their ears may like a shrill metallic tweeter so they can hear the highs, but to me, it just causes fatigue and ringing. The Tritrix use MTM design with silk dome tweets and very high quality components. Once I get some spare cash, perhaps building a pair of TL Tritrix cabs would be better spent than trying to half-donkey a cheap line array without even a proper EQ stage to compensate it. I could feel good about my craftmanship and a quality sound system without cutting corners all over the place by using cheap materials and bulk drivers.
Last edited:
I wouldn't take just any speaker to make a line array. The TC9 isn't chosen because of it's favourable price alone. Basically you get the speaker you use in an array back on steroids.
Meaning every flaw will be exaggerated.
Meaning every flaw will be exaggerated.
Yeah the TC9 sounded wonderful in my birdhouse enclosure despite being undersized. I did an A/B comparison with the GRS speaker through the left and right channels of my Lapai amp and it was night and day. I wished I had bought four of them so I could build a pair of vented bookshelf enclosures. I could get them down to a respectable 63Hz f3 using a .2 cubic foot enclosure tuned to around 75-80Hz at least according to the numbers I plugged into WinISD.
Don't know what I'll do with the four GRS drivers now. Technically I've got six of them if I count the two that came with my Porta-Pi arcade cab. The arcade cab has pretty bad acoustics (but in a "retro" kind of way) so I'm not sure how much upgrading to TC9 Peerless would benefit it, plus the holes do not perfectly align. The GRS definitely do not have a flat response judging by ear, a bit harsh in the midtone region.🙄
Other thoughts, I noticed the construction of the two drivers is different. The TC9 doesn't have visible epoxy where the wires join the underside of the cone, and the spider on the TC 9 is open. The spider on the GRS is sealed to the support frame of the magnet, meaning more air is forced through the relief pinhole in the back of the driver, the entire spider area rather than just the cap.
Without knowing the reasons why it's hard to tell what such a difference really means.
If we take the 10F as an example, there's the 10F/8414G10 that has an open spider:
While the more expensive 10F/8424G00 is closed off below the spider:
All I know I really like my 10F/8424G00's 😀.
If we take the 10F as an example, there's the 10F/8414G10 that has an open spider:
While the more expensive 10F/8424G00 is closed off below the spider:
All I know I really like my 10F/8424G00's 😀.
Attachments
The reason with the 10Fs is this: the 'fullrange' is expected to be driven to much greater excusion than the 'midranges', despite similar Xmax, thus much more air being moved further(low end being used rather than rolled off-with such a wide flat response, crossovers are easy and can be gentle).
Also with the midrange versus fullrange, you aren't concerned as much with Fs being as low as possible given driver size and mass, increasing Xmax, or ensuring that treble is clear to near the end of audible range rather than an early rolloff. Then there is the "beaming" effect of FR drivers at high frequencies, which actually works to our advantage in a line array system given the proper toe in, since there is less phase cancellation and reinforcement in the vertical sideband region, if a single driver radiates less off-axis in the HF range.
That said, a good FR driver with flat response should also make an excellent midrange in a 3-way, provided it's sensitivity is comparable to the woofer and tweets.
That said, a good FR driver with flat response should also make an excellent midrange in a 3-way, provided it's sensitivity is comparable to the woofer and tweets.
Are Line Arrays really the holy grail of sound reproduction?
A floor to ceiling line made from a continuous source with fullrange output and decent xmax would be what is needed to get close to a holy grail type speaker. They don't exist so we are back to picking compromises.
Building a proven design sounds like a much better use of limited funds than wasting your money on a "half donkey line array with no EQ" 🙂
A floor to ceiling line made from a continuous source with fullrange output and decent xmax would be what is needed to get close to a holy grail type speaker. They don't exist so we are back to picking compromises.
I don't mind 🙂 But if you are going to quote other people from private messages or email it is often considered polite to check first as they didn't submit it to a public forum 😉
If you measure any speaker in a room at different distances and angles you will get a different response at each one. The directivity and how constant that is with frequency will set how well the speaker can be EQ'd to a target response. If there are power response issues EQ won't help. A full range line array has a good power response over a wide frequency range and takes well to being EQ'd. I don't think it is an EQ nightmare at all as you can get a pretty flat response with only a few filters. In a test I used 8 filters to get a close resemblance to the EQ posted by others. I could have only used 4 and got a pretty similar response. Any EQ scheme on any speaker will be subject to location variance.
If the constraints that a line array brings don't work with your current setup don't try and build one. There are plenty of good speaker designs that will work nicely within those constraints.
Then you don't want a line array 😀
I hope so, maybe one day I will get to hear Ronald's and I will know the answer.
To do a line array justice it can't be dirt cheap or easy.
Poly stuffing is pretty hopeless for low frequencies, the fibreglass works much better. In a sealed cabinet fibreglass is contained. There are newer types like I have that are nowhere near as bad as the pink fluffy stuff. Another choice is the Bonded Logic recycled denim. I would not waste your money on poly stuffing if you lived closer I'd happily give you a big bag of it that I had for my array, as I thought like you until I tested it.
The MTM design has it's own problems just like most others. It is hard to get a tweeter to go low enough to cross it over within a half wavelength of the mid drivers due to the separation distances. This is the compromise that most MTM designs wrestle with. All metal dome tweeters are not created equal either, if you don't like them that is up to you but how many have you really heard in well integrated systems to come to that conclusion?
Building a proven design sounds like a much better use of limited funds than wasting your money on a "half donkey line array with no EQ" 🙂
I would never quote anything said in a PM with remotely personal details, but I will ask next time anyway.
Yeah I've given up on the idea for now, but fun to watch others' builds.
A person could buy fewer better drivers and house them in a better enclosure instead of "X cheap drivers in a compromised enclosure." The primary reason why I believe arrays are used in commercial sound applications is because a single point source driver or multi-way would never have the power handling capability to fill a large stadium or auditorium, and line arrays satisfy the power handling requirements per driver and help improve building acoustics by beaming the sound defectively towards the audience. Even then, ceiling mounted J-curves are the preferred route of delivery.
My next project will probably be a kit speaker to remove the guesswork. The Tritrix TL floorstanding speakers are highly praised, and I can always use an active sub if the bass output isn't substantial enough.
Last edited:
You might want to explain that one to me. For all I know we didn't buy "X cheap drivers"... We bought drivers that are up to the task, suitable, in other words: very capable drivers.
Look around for successful array projects using other "more expensive" drivers. I bet you won't find many that turned out to be as successful as the TC9 driver is in an array. That isn't because of it's price.
And what is the compromise in the enclosure department you speak about? 😕
I'd call any kind of short array a big compromise, regardless of the driver used.
Many other more capable solutions come to mind to beat that.
- Status
- Not open for further replies.