A3 array speakers, and an alternative driver possibilty
I bought mfishmike's pair of speakers that use four each Audience A3 drivers per side in a Dayton curved 1 cu ft. cabinet. Mike reduced the internal cabinet volume for purpose of venting - for added bass response -- he added some plywood inside the cabinets. Mike's speakers sound great.
The A3 drivers have exceptional 'palpable' mid-range clarity, and are extremely dynamic !
I haven't made any SoundEasy Lab measurements yet, but even though these A3 cabinets are ported, the bass doesn't seem to extend below roughly 100 Hz.
I was astonished to find that adding my Velodyne DD-15 subwoofer in at around 80-100 Hz blended seamlessly with the A3s (sub is located immediately behind where I sit - & on the axis midway between the A3 array speakers). I plan to do interesting experiments with these A3 array speakers -- such as changing the detachable front panel & adding a 6.5" Excel magnesium woofer on the front, in addition to the four A3s. This woofer will roll out very low at say, 125 Hz max .. It goes without saying that much trial and error listening, testing, etc. is essential.
BTW, since many DIY hobbyists are looking for less-expensive alternative drivers to the Audience A3, the Fountek FR88-EX 3" aluminum cone driver is a most promising candidate, assuming that a bi-amp'd woofer below say, 125 Hz is added (remove cone excursion demands from the 3" drivers).
Driver colorations are not as audible below about 125 Hz -- anyway, adding a woofer should be an extremely cost-effective tradeoff !
IMO, a simple 6 dB/Oct capacitor-resistor crossover at the input to the separate amplifier for the 3" drivers should be easily tolerated by the ear.
Maybe the woofer could be rolled off at even a sharper rate above 125 Hz without detrimental results. Could even use digital low-pass filter (such as a Behringer digital crossover) for input to woofer amplifier. Of course, use a higher quality amp to drive the array of 3" drivers above 125 Hz.
The Fountek FR88-EX voice coil is underhung, & has only 1.86 gm moving mass as measured by Zaphaudio.com, This less expensive Fountek FR88-EX ($18.42 ea. for qty 20 at Madisound) looks like a 'best' where driver cost is the issue compared to the Audience A3.
Zaphaudio.com did many tests of small drivers and concluded that the Fountek was the 'best performing' 3" driver tested (smooth frequency response, very low distortion, very good efficiency, aluminum cone, neodymium motor, high quality for the money). Tests seem to imply that the Fountek's sound might be comparable to the Audience A3, assuming a designer can live with the higher Fs around 100Hz, and the smaller 1.5mm Xmax -- but an advantage might be that the Fountek doesn't exhibit a treble resonant peak below 10 kHz.
Bill
I bought mfishmike's pair of speakers that use four each Audience A3 drivers per side in a Dayton curved 1 cu ft. cabinet. Mike reduced the internal cabinet volume for purpose of venting - for added bass response -- he added some plywood inside the cabinets. Mike's speakers sound great.
The A3 drivers have exceptional 'palpable' mid-range clarity, and are extremely dynamic !
I haven't made any SoundEasy Lab measurements yet, but even though these A3 cabinets are ported, the bass doesn't seem to extend below roughly 100 Hz.
I was astonished to find that adding my Velodyne DD-15 subwoofer in at around 80-100 Hz blended seamlessly with the A3s (sub is located immediately behind where I sit - & on the axis midway between the A3 array speakers). I plan to do interesting experiments with these A3 array speakers -- such as changing the detachable front panel & adding a 6.5" Excel magnesium woofer on the front, in addition to the four A3s. This woofer will roll out very low at say, 125 Hz max .. It goes without saying that much trial and error listening, testing, etc. is essential.
BTW, since many DIY hobbyists are looking for less-expensive alternative drivers to the Audience A3, the Fountek FR88-EX 3" aluminum cone driver is a most promising candidate, assuming that a bi-amp'd woofer below say, 125 Hz is added (remove cone excursion demands from the 3" drivers).
Driver colorations are not as audible below about 125 Hz -- anyway, adding a woofer should be an extremely cost-effective tradeoff !
IMO, a simple 6 dB/Oct capacitor-resistor crossover at the input to the separate amplifier for the 3" drivers should be easily tolerated by the ear.
Maybe the woofer could be rolled off at even a sharper rate above 125 Hz without detrimental results. Could even use digital low-pass filter (such as a Behringer digital crossover) for input to woofer amplifier. Of course, use a higher quality amp to drive the array of 3" drivers above 125 Hz.
The Fountek FR88-EX voice coil is underhung, & has only 1.86 gm moving mass as measured by Zaphaudio.com, This less expensive Fountek FR88-EX ($18.42 ea. for qty 20 at Madisound) looks like a 'best' where driver cost is the issue compared to the Audience A3.
Zaphaudio.com did many tests of small drivers and concluded that the Fountek was the 'best performing' 3" driver tested (smooth frequency response, very low distortion, very good efficiency, aluminum cone, neodymium motor, high quality for the money). Tests seem to imply that the Fountek's sound might be comparable to the Audience A3, assuming a designer can live with the higher Fs around 100Hz, and the smaller 1.5mm Xmax -- but an advantage might be that the Fountek doesn't exhibit a treble resonant peak below 10 kHz.
Bill
Bill,
I saw those speakers for sale. They look really interesting. Glad to hear that sound great.
Do you know how many feet of hook up wire would be needed for a speaker like yours that uses 4 A3 drivers per speaker?
This would be helpful to know.
I look forward to finding out the results of your experiments with different drivers in the mix.
-Nick
I saw those speakers for sale. They look really interesting. Glad to hear that sound great.
Do you know how many feet of hook up wire would be needed for a speaker like yours that uses 4 A3 drivers per speaker?
This would be helpful to know.
I look forward to finding out the results of your experiments with different drivers in the mix.
-Nick
that fountek has a bit of hash on the waterfall plot from 1-4khz, I've seen comments of the peerless tc9 that has a similar waterfall as commenting that it sounds a little "hazy."
klemminger said (about the fountek fr-88ex) "In direct comparison with Vifa TC9-FD-18-08 and Visaton FRS8 i prefer the Fountek/Dynavox. It's well buildt and sounds clear and detailed. For the price i can recommend everybody to give it a try."
That looks to be a sharp driver.
Maybe in the far future.
Fountek FR88 3" fullrange driver - DIY Mobile Audio
or as a raal loudspeaker clone
HTGuide Forum - Fountek FR88-EX
Seems we ar getting a bunch of $20-$30 good drivers in the 3" realm.
Bill13, I'm glad you are happy with your a3's.
Norman
klemminger said (about the fountek fr-88ex) "In direct comparison with Vifa TC9-FD-18-08 and Visaton FRS8 i prefer the Fountek/Dynavox. It's well buildt and sounds clear and detailed. For the price i can recommend everybody to give it a try."
That looks to be a sharp driver.
Maybe in the far future.
Fountek FR88 3" fullrange driver - DIY Mobile Audio
or as a raal loudspeaker clone
HTGuide Forum - Fountek FR88-EX
Seems we ar getting a bunch of $20-$30 good drivers in the 3" realm.
Bill13, I'm glad you are happy with your a3's.
Norman
Wiring DIY speaker using A3 drivers & consider the new "2+2" passive radiator design
nicksgem10s,
The four A3s in the Dayton 1 cu ft cabinet are wired in parallel (resulting in a four ohm load) -- Allowing for some hookup wire slack from the rear-mounted speaker terminals to the removable front baffle-panel, perhaps you may get by with only 8 feet of wire -- but get 10 ft. 'just in case'.
However, you might ask mfishmike (Audience A3 dealer/speaker DIY expert) for his opinion(s) -- you may want to go with the impressive "2+2" with passive radiator in a smaller conventional rectangular shaped cabinet.
A '2010 Show' report includes a picture of this new Audience "Clairaudient" style experimental "2+2" prototype. Reportedly, this little speaker received some "rave" comments -- I think 'mfishmike' posts the 2010 Show pictures on his website, and supplies A3 drivers, advice (I understand) & passive radiator parts, etc.
Bill
nicksgem10s,
The four A3s in the Dayton 1 cu ft cabinet are wired in parallel (resulting in a four ohm load) -- Allowing for some hookup wire slack from the rear-mounted speaker terminals to the removable front baffle-panel, perhaps you may get by with only 8 feet of wire -- but get 10 ft. 'just in case'.
However, you might ask mfishmike (Audience A3 dealer/speaker DIY expert) for his opinion(s) -- you may want to go with the impressive "2+2" with passive radiator in a smaller conventional rectangular shaped cabinet.
A '2010 Show' report includes a picture of this new Audience "Clairaudient" style experimental "2+2" prototype. Reportedly, this little speaker received some "rave" comments -- I think 'mfishmike' posts the 2010 Show pictures on his website, and supplies A3 drivers, advice (I understand) & passive radiator parts, etc.
Bill
Fountek might make a great midrange crossed to a RAAL ribbon
Initially when I caught "speaker upgrade fever" again, the RAAL 140-15D ribbon tweeter looked promising for a high end style DIY speaker design.
However, I just wasn't happy with using 5" to 7" drivers for midrange up to 3+ kHz or so (want to try quasi-2nd order crossover). The transparency and CSD waterfall of the RAAl seemed far better than what's available from most mid-woofers.
Then a rave review of the '8+8 Clairaudient' no-crossover fullrange array speaker that uses the Audience A3 drivers.
Despite the measured response peak around 8 kHz, the A3 driver reportedly offered audiophile-quality sound from mid-bass al the way up to high trebles.
I'm still not sure that adding a RAAL ribbon tweeter to an A3 array wouldn't improve the trebles. Anyway, an A3 driver array, by itself, is expensive enough, already.
I guess eliminating the loudspeaker crossover obviously (?) outweighs any S-O-A enhancement the RAAL might make?
However, crossing over a RAAL might be more cost-effective if an array of cheaper Fountek FR88-EX drivers is used? The CSD and 'sonic-signature' of the aluminum-cone Fountek 3" might be more compatible with ribbon tweeters than typical larger cone bass-mid drivers? Use a woofer crossed below say, 125 Hz when using Fountek 3" midrange drivers.
Wish that the RAAL ribbons weren't so expensive -- but would like to try anyway. So far, can't even find used RAAL ribbon tweeters -- RAAL owners evidently must be happy.
Bill
Initially when I caught "speaker upgrade fever" again, the RAAL 140-15D ribbon tweeter looked promising for a high end style DIY speaker design.
However, I just wasn't happy with using 5" to 7" drivers for midrange up to 3+ kHz or so (want to try quasi-2nd order crossover). The transparency and CSD waterfall of the RAAl seemed far better than what's available from most mid-woofers.
Then a rave review of the '8+8 Clairaudient' no-crossover fullrange array speaker that uses the Audience A3 drivers.
Despite the measured response peak around 8 kHz, the A3 driver reportedly offered audiophile-quality sound from mid-bass al the way up to high trebles.
I'm still not sure that adding a RAAL ribbon tweeter to an A3 array wouldn't improve the trebles. Anyway, an A3 driver array, by itself, is expensive enough, already.
I guess eliminating the loudspeaker crossover obviously (?) outweighs any S-O-A enhancement the RAAL might make?
However, crossing over a RAAL might be more cost-effective if an array of cheaper Fountek FR88-EX drivers is used? The CSD and 'sonic-signature' of the aluminum-cone Fountek 3" might be more compatible with ribbon tweeters than typical larger cone bass-mid drivers? Use a woofer crossed below say, 125 Hz when using Fountek 3" midrange drivers.
Wish that the RAAL ribbons weren't so expensive -- but would like to try anyway. So far, can't even find used RAAL ribbon tweeters -- RAAL owners evidently must be happy.
Bill
Crossover discussion moved to here: http://www.diyaudio.com/forums/multi-way/161300-crossover-phase-behaviour.html
dave
dave
Plans to improve DIY A3 line array spkrs
Good idea to form separate discussion of speak crossover network coherence, etc.
Anyway, more about DIY Audience A3 line-arrays:
An update to what I'm planning with DIY A3 arrays in Dayton curved cabinets.
I'm very impressed with the sound of these A3 drivers, but as currently employed, they sound a bit too bright.
After much deliberation of alternative ideas, good way to go is to another four rear/side facing Audience A3s per speaker, for BSC. This should effectively lower the trebles relative to the midrange, and give more bass power.
After all, Audience Inc. doubles the number of drivers by adding rear/side facing A3s in their "4+4", "8+8", "16+16" 'Clairaudient' family of speakers.
Putting the cabinets near room boundaries (walls) for more bass is not an option in my room.
Further improvement: can try fastening some DIY shaped material like foam/cloth padding to the sides of the DIY speakers' front baffles -- to acoustically extend the speaker's width from 9" to say, 15" in order to lower the BSC frequency -- achieve a wide, curved front baffle like in the Clairaudient speakers. However, afraid that this approach won't look good.
Electrically, I like the idea of making a custom (preferably passive) R-C network to be used at the inputs to the power amplifiers. Op-amps might have to be used if need more voltage gain, isolation of filter components, etc.) This custom passive/active network can be designed using "SoundEasy" or other circuit simulation program.
This circuit can do several adjustments to the speaker's frequency response such as:
1. BSC
2. High Pass filter starting at lowest frequency to limit driver excursion, and to protect the drivers.
3. Provide gentle roll off (contour) upper midrange thru trebles if desired - including the famous desirable psycho-acoustic dip at 3 kHz (see Linkwitz OB speaker website).
4. Bass boost (fixed amount)
5. Use high-value ferrite inductors to add R-L-C notch filter in the network at amp inputs to compensate for A3 drivers' 7-8 kHz resonance(set filter Q = 10 approx.). I have a collection of such inductors up to 680 millihenry -- obtained from Mouser and at other suppliers -- they are physically small. Hysteresis and other typical nonlinearities should not be a problem for this particular application (very small signal currents throught the inductors, & very small frequency band affected, etc.).
Bill
Good idea to form separate discussion of speak crossover network coherence, etc.
Anyway, more about DIY Audience A3 line-arrays:
An update to what I'm planning with DIY A3 arrays in Dayton curved cabinets.
I'm very impressed with the sound of these A3 drivers, but as currently employed, they sound a bit too bright.
After much deliberation of alternative ideas, good way to go is to another four rear/side facing Audience A3s per speaker, for BSC. This should effectively lower the trebles relative to the midrange, and give more bass power.
After all, Audience Inc. doubles the number of drivers by adding rear/side facing A3s in their "4+4", "8+8", "16+16" 'Clairaudient' family of speakers.
Putting the cabinets near room boundaries (walls) for more bass is not an option in my room.
Further improvement: can try fastening some DIY shaped material like foam/cloth padding to the sides of the DIY speakers' front baffles -- to acoustically extend the speaker's width from 9" to say, 15" in order to lower the BSC frequency -- achieve a wide, curved front baffle like in the Clairaudient speakers. However, afraid that this approach won't look good.
Electrically, I like the idea of making a custom (preferably passive) R-C network to be used at the inputs to the power amplifiers. Op-amps might have to be used if need more voltage gain, isolation of filter components, etc.) This custom passive/active network can be designed using "SoundEasy" or other circuit simulation program.
This circuit can do several adjustments to the speaker's frequency response such as:
1. BSC
2. High Pass filter starting at lowest frequency to limit driver excursion, and to protect the drivers.
3. Provide gentle roll off (contour) upper midrange thru trebles if desired - including the famous desirable psycho-acoustic dip at 3 kHz (see Linkwitz OB speaker website).
4. Bass boost (fixed amount)
5. Use high-value ferrite inductors to add R-L-C notch filter in the network at amp inputs to compensate for A3 drivers' 7-8 kHz resonance(set filter Q = 10 approx.). I have a collection of such inductors up to 680 millihenry -- obtained from Mouser and at other suppliers -- they are physically small. Hysteresis and other typical nonlinearities should not be a problem for this particular application (very small signal currents throught the inductors, & very small frequency band affected, etc.).
Bill
I'm sorry I missed this thread - don't spend enough time on the full-range forum.
I spent quite a lot of time with the Audience A3 line array, the 8 driver model. It's a very good speaker and the driver is remarkable. I like line arrays up to a point. To me they sound very natural and clean, like a really good horn system. But too many line arrays sound "too big." Everything is over sized. Some people like that, I don't.
But the Audience line array sounded quite natural to me. At least the 8 driver version. I never heard the bigger one.
The only obviuos flaw I heard was a disturbance in the ~300Hz range. Like you might hear on a 10" woofer. No surprise, as the A3 driver is just about the same size as the dustcap on a 10".
I have 1 of the A3 drivers, wish I had 2. A great driver, but I really do feel it's overpriced. Too bad, as at a more reasonable price, they might sell a lot more and make a lot of people happy.
My 2 cents worth.
I spent quite a lot of time with the Audience A3 line array, the 8 driver model. It's a very good speaker and the driver is remarkable. I like line arrays up to a point. To me they sound very natural and clean, like a really good horn system. But too many line arrays sound "too big." Everything is over sized. Some people like that, I don't.
But the Audience line array sounded quite natural to me. At least the 8 driver version. I never heard the bigger one.
The only obviuos flaw I heard was a disturbance in the ~300Hz range. Like you might hear on a 10" woofer. No surprise, as the A3 driver is just about the same size as the dustcap on a 10".
I have 1 of the A3 drivers, wish I had 2. A great driver, but I really do feel it's overpriced. Too bad, as at a more reasonable price, they might sell a lot more and make a lot of people happy.
My 2 cents worth.
Audience A3 array - estimate compensation using equalizer, or Behringer 2496 digital
Audience A3 array - can estimate compensation using a Behringer 2496 digital to determine requirements for a passive network at amp inputs.
Perhaps the 300 Hz problem could also be addressed.
My old 1971 vintage Soundcraftsmen 20-12 Equalizer is the least colored that I've heard. Still works - it's a passive design using toroidal and ferrite-core inductors. Reviewers said that this old equalizer was sonically transparent when all the adjustment sliders were set to zero.
I would only use multiband equalizers to get a rapid, vague idea of a how boost/cut in each pass band might sound with an Audience A3 array. I believe that the available filter Q (determines shape of each boost/cut curve) is usually a fixed value depending on the equalizer's number of adjustable bands.
I'm interested in placing a passive R-L-C circuit in front of the power amps to address the A3 driver's roughly 6 dB resonant peak at the input to the power amp (with a high-pass RC to limit cone excursion).
Placing passive correction components at the amp input still allows to amp to exert maximum control of the drivers (keep driving impedance low) - better than using big passive components connected to the driver array.
Some may disagree, but making 'good' small frequency spectrum adjustments using a circuit at the power amp inputs, should not degrade the wonderful A3 full-range audiophile sound quality .. on the contrary would expect overall perceived improvement.
For BSC and bass response, I still would pursue adding rear/side drivers, and a wide cabinet baffle (say, 15" front baffle width like in the Clairaudient speaker family). Don't rely on an electrical correction network for fixing everything !
IMO, a better approach to developing speaker correction networks would be to use the inexpensive Behringer DCX-2496 digital processor which you can easily connect to your computer (RS232 serial) to graphically see, and hear, the effect of adding a notch filter, & spectral contouring, high-pass filter, BSC filter, etc.
The Behringer is actually easy to use due to it's intuitive graphical computer interface. For example, notch filter Q can be dialed-in the result seen/heard simultaneously (The A3 resonant peak has a Q roughly = 10 -- I used my Behringer 2496 to very quickly visually estimate the required notch filter's Q, so far).
Yes, the digital Behringer processor is a bit colored, but still very good to point you in the right direction - it's not just for crossovers.
After you determine your desired overall network response characteristics, you could use a crossover network simulation program to design an equivalent (passive) network to place at the inputs to your power amps (use SoundEasy BoxCad, or other simulation program).
Bill
Audience A3 array - can estimate compensation using a Behringer 2496 digital to determine requirements for a passive network at amp inputs.
Perhaps the 300 Hz problem could also be addressed.
My old 1971 vintage Soundcraftsmen 20-12 Equalizer is the least colored that I've heard. Still works - it's a passive design using toroidal and ferrite-core inductors. Reviewers said that this old equalizer was sonically transparent when all the adjustment sliders were set to zero.
I would only use multiband equalizers to get a rapid, vague idea of a how boost/cut in each pass band might sound with an Audience A3 array. I believe that the available filter Q (determines shape of each boost/cut curve) is usually a fixed value depending on the equalizer's number of adjustable bands.
I'm interested in placing a passive R-L-C circuit in front of the power amps to address the A3 driver's roughly 6 dB resonant peak at the input to the power amp (with a high-pass RC to limit cone excursion).
Placing passive correction components at the amp input still allows to amp to exert maximum control of the drivers (keep driving impedance low) - better than using big passive components connected to the driver array.
Some may disagree, but making 'good' small frequency spectrum adjustments using a circuit at the power amp inputs, should not degrade the wonderful A3 full-range audiophile sound quality .. on the contrary would expect overall perceived improvement.
For BSC and bass response, I still would pursue adding rear/side drivers, and a wide cabinet baffle (say, 15" front baffle width like in the Clairaudient speaker family). Don't rely on an electrical correction network for fixing everything !
IMO, a better approach to developing speaker correction networks would be to use the inexpensive Behringer DCX-2496 digital processor which you can easily connect to your computer (RS232 serial) to graphically see, and hear, the effect of adding a notch filter, & spectral contouring, high-pass filter, BSC filter, etc.
The Behringer is actually easy to use due to it's intuitive graphical computer interface. For example, notch filter Q can be dialed-in the result seen/heard simultaneously (The A3 resonant peak has a Q roughly = 10 -- I used my Behringer 2496 to very quickly visually estimate the required notch filter's Q, so far).
Yes, the digital Behringer processor is a bit colored, but still very good to point you in the right direction - it's not just for crossovers.
After you determine your desired overall network response characteristics, you could use a crossover network simulation program to design an equivalent (passive) network to place at the inputs to your power amps (use SoundEasy BoxCad, or other simulation program).
Bill
Good ideas, all.
I'd use the DEQ2496 instead (I own both). It seems better suited to this job. And you can use a high quality DAC after the DEQ if you want to.
I'd use the DEQ2496 instead (I own both). It seems better suited to this job. And you can use a high quality DAC after the DEQ if you want to.
Panomaniac,
Subject: Audience "4+4" or 8-inline A3s? & use the DEQ2496 instead of DCX2496.
Thanks for your posts.
My Behringer DCX2496 is many years old -- Thanks, I will look into the DEQ2496 as an improvement/upgrade for this application.
Maybe upgrade/mod the DEQ's analog input components, etc., and using an outboard DAC could allow going all-digital for signal processing ... I think that some people have done this already.
Did you listen to the "4+4" Clairaudient 8-driver array -- the one with half the drivers facing to the side/rear?
This Audience "4+4" model might not over-size the musical instruments, as much as the taller speaker with all eight A3 line-array drivers mounted on the front?
Bill
Subject: Audience "4+4" or 8-inline A3s? & use the DEQ2496 instead of DCX2496.
Thanks for your posts.
My Behringer DCX2496 is many years old -- Thanks, I will look into the DEQ2496 as an improvement/upgrade for this application.
Maybe upgrade/mod the DEQ's analog input components, etc., and using an outboard DAC could allow going all-digital for signal processing ... I think that some people have done this already.
Did you listen to the "4+4" Clairaudient 8-driver array -- the one with half the drivers facing to the side/rear?
This Audience "4+4" model might not over-size the musical instruments, as much as the taller speaker with all eight A3 line-array drivers mounted on the front?
Bill
Dave,
Thanks for your post referencing "Passive Line Level Baffle Diffraction Step Compensation".
The circuit shown incorporates good ideas.
BTW, I believe that an optional high-frequncy passive-notch
-filter (e.g. around 8 KHz) could easily be added by simply putting a series R-L-C leg in parallel with the schematic's R2-C2 leg.
I think that the impedance/load of such a notch filter leg would be essentially act independently of the other compents in the example BSC passive network.
Bill
Thanks for your post referencing "Passive Line Level Baffle Diffraction Step Compensation".
The circuit shown incorporates good ideas.
BTW, I believe that an optional high-frequncy passive-notch
-filter (e.g. around 8 KHz) could easily be added by simply putting a series R-L-C leg in parallel with the schematic's R2-C2 leg.
I think that the impedance/load of such a notch filter leg would be essentially act independently of the other compents in the example BSC passive network.
Bill
oversize the instruments?
I too object to that type of sound.
do you think that panel speakers sound that way ?
If you do, then a line of 8 full rangers in a line will be similar.
Norman
I too object to that type of sound.
do you think that panel speakers sound that way ?
If you do, then a line of 8 full rangers in a line will be similar.
Norman
The A3s I spent time with where the models with 8 drivers on the front.
What surprised me is that did not sound that "over-sized." Maybe a little, but not like so many line arrays I've heard. I have to say I liked them, tho a little EQ in the midrange would have gone a long way. Very articulate speakers, nice bass, good image.
Well built, too.
What surprised me is that did not sound that "over-sized." Maybe a little, but not like so many line arrays I've heard. I have to say I liked them, tho a little EQ in the midrange would have gone a long way. Very articulate speakers, nice bass, good image.
Well built, too.
Better to place half of the A3 line-array drivers to side/rear
Yes, maybe the Audience 'Clairaudient' speakers, that have ALL eight line-array A3 drivers mounted on the front, may sound a bit "over-sized".
Also, if you don't place the 8-line array speaker near room wall boundaries, BSC becomes even more important (also, with all the drivers on front baffle, much more midrange equalization may be desirable).
The upper-bass output is about the same whether you mount half of the A3 drivers on the cabinet front baffle, or all line array drivers on the front.
That's why I currently prefer the Audience speaker configuration where half of the A3 drivers are mounted facing to the side/rear for better BSC (and improving midrange response).
Currently I have a pair of DIY speakers, only four A3 drivers per cabinet; all drivers mounted on the front baffle. Array of only four drivers seem to generate properly sized musical instruments, IMO.
So, I prefer to go with a "4+4" driver line-array configuration. Next, will attempt to obtain a pair of "4+4" "Clairaudient" style cabinets -- but fear that such curved enclosures would be very expensive to have a woodworker fabricate ... very difficult to DIY.
BTW, many, many years ago, I saw somewhat similar commercial (0.5"?) thick cardboard curved speaker enclosure tubes available from a specialty company -- looked like big, shaped, rug cardboard cylinders.
Oh well, unfortunately, good things usually aren't cheap. The more you want them, the more they cost .. or so it seems.
Bill
Yes, maybe the Audience 'Clairaudient' speakers, that have ALL eight line-array A3 drivers mounted on the front, may sound a bit "over-sized".
Also, if you don't place the 8-line array speaker near room wall boundaries, BSC becomes even more important (also, with all the drivers on front baffle, much more midrange equalization may be desirable).
The upper-bass output is about the same whether you mount half of the A3 drivers on the cabinet front baffle, or all line array drivers on the front.
That's why I currently prefer the Audience speaker configuration where half of the A3 drivers are mounted facing to the side/rear for better BSC (and improving midrange response).
Currently I have a pair of DIY speakers, only four A3 drivers per cabinet; all drivers mounted on the front baffle. Array of only four drivers seem to generate properly sized musical instruments, IMO.
So, I prefer to go with a "4+4" driver line-array configuration. Next, will attempt to obtain a pair of "4+4" "Clairaudient" style cabinets -- but fear that such curved enclosures would be very expensive to have a woodworker fabricate ... very difficult to DIY.
BTW, many, many years ago, I saw somewhat similar commercial (0.5"?) thick cardboard curved speaker enclosure tubes available from a specialty company -- looked like big, shaped, rug cardboard cylinders.
Oh well, unfortunately, good things usually aren't cheap. The more you want them, the more they cost .. or so it seems.
Bill
That would be 8+8..
Hmm, someone around here postulated that an 8 driver array (frontal) would be the most appropriate for this driver.. 😛
If I were building some cabs for these, I would go down the translaminate route. So long as you have access to a plywood supplier with a CNC routing facility, then the process is very simple and looks stunning once polished up. Virtually any shape is possible, and the end result is exceptionally rigid. You could even have two seperate chambers inside the cabs (one for the forward firing drivers and one for the side/rears). You would probably want the CNC shop to designs the curves for you.
There is a lot of wood wastage in translam construction, and the CNC machine time is not cheap, but neither are the A3 drivers, so I would say that the effort is worth it.
Trans-Laminated DIY Speakers [audiojunkies]
This is one of my translam efforts below. This time the laminations were verticle to allow for a transmission line.
Heroic DIY closed cabinet, or maybe alternative Open baffle
I just received Feb 2010 issue of "Audio Express" in which nicely shaped, stacked, woofer and midrange laminated-layer boxes were featured on page 6.
Heroic DIY dedication!
The external shapes of these woofer & midrange boxes were a sort of truncated double pyramid.
The shaped woofer enclosure was laboriously constructed by cutting & shaping & drilling 62 sections (62 laminated flanges) made from 22mm MDF. Clamps (25cm and 80cm) were use to affix the layers (flanges) for assembly/gluing.
Similarly, the midrange enclosure was made from 18 layers (shaped MDF flanges) glued together. The builder of these cabinets said that he used a table saw and a disk saw for cutting & a router was used to recess the drivers ... apparently, no CNC tooling was used. Four layers of matte, and two layers of gloss lacquer were applied for finishing.
For me, it was 'scary' to consider amount of DIY work evolved!
If I can't find an easier, or cheaper, way to obtain sealed curved enclosures for my A3 arrays (curved cabinets roughly 15" wide like the expensive Audience 'Clairaudient' speakers), then might consider some kind of U-shaped open baffle {OB} approach where the A3 driver arrays will only be asked to cover say, at most, 100+ Hz on up.
Use enough A3 drivers to make up for some acoustic cancellation at the low end of the range. I believe that the front-to-rear acoustic path length can be made long enough in an acceptably sized OB speaker to avoid too much cancellation from 100 Hz on up (OB's have 6 dB/octave roll off rate - see Linkwitz).
Closed enclosure is probably best for an Audience A3 array, but OB may be good alternative. See Linkwitz OB speaker construction website for basic applicable OB acoustic theory.
IMO, open-baffle allows use of lightweight semi-flexible baffle wall-materials, (also, could do flexible constrained-layers) , promoting quick & easy prototype speaker construction.
Bill
I just received Feb 2010 issue of "Audio Express" in which nicely shaped, stacked, woofer and midrange laminated-layer boxes were featured on page 6.
Heroic DIY dedication!
The external shapes of these woofer & midrange boxes were a sort of truncated double pyramid.
The shaped woofer enclosure was laboriously constructed by cutting & shaping & drilling 62 sections (62 laminated flanges) made from 22mm MDF. Clamps (25cm and 80cm) were use to affix the layers (flanges) for assembly/gluing.
Similarly, the midrange enclosure was made from 18 layers (shaped MDF flanges) glued together. The builder of these cabinets said that he used a table saw and a disk saw for cutting & a router was used to recess the drivers ... apparently, no CNC tooling was used. Four layers of matte, and two layers of gloss lacquer were applied for finishing.
For me, it was 'scary' to consider amount of DIY work evolved!
If I can't find an easier, or cheaper, way to obtain sealed curved enclosures for my A3 arrays (curved cabinets roughly 15" wide like the expensive Audience 'Clairaudient' speakers), then might consider some kind of U-shaped open baffle {OB} approach where the A3 driver arrays will only be asked to cover say, at most, 100+ Hz on up.
Use enough A3 drivers to make up for some acoustic cancellation at the low end of the range. I believe that the front-to-rear acoustic path length can be made long enough in an acceptably sized OB speaker to avoid too much cancellation from 100 Hz on up (OB's have 6 dB/octave roll off rate - see Linkwitz).
Closed enclosure is probably best for an Audience A3 array, but OB may be good alternative. See Linkwitz OB speaker construction website for basic applicable OB acoustic theory.
IMO, open-baffle allows use of lightweight semi-flexible baffle wall-materials, (also, could do flexible constrained-layers) , promoting quick & easy prototype speaker construction.
Bill
- Status
- Not open for further replies.