[Greggo]

Danny and fellow enthusiasts,

Have been following the "circles"  for many years now and thinking about line arrays.  Would love some feedback on this concept:

16 x Neo8 mounted sideways, edge to edge on a minimal baffle that is 9 inches wide and 60 inches tall

6 x GR 8" servo subs in separate tower cabinet that is also 9 x 60 inches.... (a stereo pair of them, placed beside and within a foot or two of the Neo8 cabinets)

My theory is that the good old original Neo8 offers some unique advantages.  Primary is the ability to play through the top range without the need for a tweeter (though I am not sure if the Neo8-S would offer the same advantage).  Rumor has it that since the coupling between drivers is stronger as you go down in frequency and weaker as you go up, that the usual notch filter is not required on the Neo8 as the balance smooths out over an array....with 16 of them sitting sideways, I would assume this happens with even greater effect.

The other advantage is my own assumption from reading around here and other places is that this really strong coupling between drivers if arranged this way would in fact allow them to be crossed over as low as 200-300 Hz, which puts them within the upper limits of the servo subs.  I would imagine a digital crossover would be preferred for a steep slope on the Neo8 to minimize any low end distortion or stress.

Such a tall line with such close spacing should also overcome the limited vertical dispersion of the Neo8 (which is now the horizontal plane dispersion for this theoretical system), with the line array pressure spreading out the high frequencies for a good balance of coverage at all frequencies, or does anyone think the power response here would be a bit wonky?  Or do we all believe in the "half cylinder radiation" effect of tall line arrays?

If the above proved to be true, then even though the Neo10 offers better clarity and dynamic range (but needs a tweeter), the sheer numbers in force of the Neo8 array would deliver rather impressively on this front as well.

I will end with the disclosure that my inspiration for this is the old Infinity Beta IRS which I heard at a hifi show in San Francisco in 1981 and it has remained as one of my top 10 most memorable moments in this hobby despite all my years of shows, shops, and home demonstrations that must be approaching 1,000 experiences by now.  The other is the cult like following of full range drivers and stats over the years, how folks rave about the beauty of voices and instruments when no crossovers are involved and/or the wave launch is from such a consistent physical device.  Even though 16 Neo8 per side is not cheap, catching a decent price on a big order like 32 of them is not as far out of range as many other exotic approaches would be.

Any thoughts?

[nickd]

Not sure you thoughts on the Neo 8's bass response are correct. I don't think you can get enough usuable bass at 250 Hz even with 16 of them. Just thoughts though. If you have 16 of them laying around, might be fun to experiment.

[Greggo]

Yeah, aside from power response, crossover point would be the "make or break" for something like this... maybe it would be mitigated by switching from Neo8 to Neo8-S, but then cost goes up and I have not heard much about the 8-S working without a tweeter versus all the good things about the regular 8 making it work OK.  B&G spec sheet mentions going as low as this for media application with EQ assuming very low SPL.... So that is why I am wondering if 16 of these things side by side would couple so strongly and have enough dynamic range to cut it at reasonably high SPL dedicated HiFi listening room or modest home theater environments.

Even a limited experiment on this concept with 8 or 9 wired up is a thousand dollar exercise, so before I would even attempt to start working towards that I was hoping Danny or anyone else who have worked with the usual vertical orientation of Neo8 in an array would comment on any findings regarding lowest possibly crossover point.  I am assuming it was problematic with 6 or 9 of them arranged vertically otherwise the LS6 and LS9 may have turned out differently as Danny has mentioned keeping their crossovers above 800 Hz since he doesn't like to go lower unless he can go below 300 Hz and avoid messing up anything in the midrange of 300-800. 

[Hank]

I would suggest you consider the second tower be a line of Danny's mid-woofs rather than the 8" subs.  You could cross over to the Neo 8's high enough to avoid a possible problem.   The line of mid-woofs would each have relatively small cone displacement and operating in this pure pistonic range  would produce absolutely minimal cone/surround distortion.  Then, do a couple of dual 12" servo subs for low bass and movie LFE.

In answer to your question about half cylinder radiation, I for one believe the line array wave launch is definitely a half cylinder.

Seems like a cool project - please keep us posted.

[Greggo]

Thanks Hank, but at that point I may as well turn the Neo8 back to vertical orientation (no need for overkill) and then put the woofers in the same cabinet to minimize c-t-c distance and prevent combing, and now we have the LS9 : )

It will be while before I could really take on a project like this, if ever, but the idea was to keep it very simple (2-way) and come close to the sound of a full range stat but more dynamic and somewhat affordable for a state of the art DIY project.  Again, taking the public advice from Danny and many other speaker builders I follow on the internet, if you can't cross below 300 Hz you should cross above 800 Hz, so there lies the dilemma.  That is why I like the SEOS waveguide stuff as well, with that crew mostly keeping their crossover point between 800 and 1500 Hz and then rolling in subs below 80 Hz if needed.

Following many line array builders, it seems ideal to create a 2-way system that works well without subs and leaves the 1200 Hz and above all to the same closely spaced set of drivers.  I was just thinking of how ideal it would be to take that down to 250 Hz, hence my original post.

[Danny Richie]

Well, you have overlooked a few problems that are created by stacking Neo 8's on their side.

You wind up with an 8" wide tweeter. That will have the off axis response of an 8" wide woofer. Basically as soon as you move left or right you have no highs.

In the LS-6 and LS-9 I use a custom Neo 8 that allows for the front surface area to be only 5/8" wide. That is the reason the off axis response is so goo.

There are no high quality digital crossovers out there yet either. HAL is working on one but some of the pieces of that puzzle are still under development.

And you get gain in the lower region from the coupling effect of the drivers but you don't gain a lot of extension. By the time you flatten the response back out you are not too much lower than before.

Sounds like you'd like a new design that I am working on for Mocking Bird Audio. If all goes well it can be seen this year at RMAF. I can tell you about only via phone. I am not putting anything on line about it just yet. That's not my call to make. 

[Greggo]

Danny,

Thanks for the post and the offer... I am just kicking ideas around so I will wait, with keen interest, to see what new design you have with Mocking Bird when the info is released.  You seem confident that 16 Neo's per side just won't have the muscle down to 300 Hz or lower.  Even moving up to Neo8-S (price jump is not as bad as I thought)?

Am I incorrect in assuming that horizontal dispersion improves with drivers tightly spaced in a line array?  Even 8 inch woofers    I would think that the those high frequencies getting jammed together in the vertical plane would start to spread out across the horizontal plane.  If it really is a problem, I would think some felt/foam device mounted over the outside edges of the driver (kind of like the RAAL vertical dispersion foam adjusty thingys) improve the highs and leave the mids to low mids free to forth and couple with their sister drivers in the array (and minimize the SPL efficiency hit)

Also, regarding crossover... Would I really need one?  If I used six of your 8" servo subs per tower, I would have three each on one of two servo amps per side, and those amps would handle the low pass to the subs at 300 Hz with their built in electronic filters.  I could then just put an line level active brick wall high pass set at 300 Hz (or maybe even 350 Hz) in front of the main stereo amp and with a little tweaking get good results without actually doing a full crossover design/build.  OK, I am really pushing the limits of my novice understanding of all this, but it still seems to be a unique solution with some potential benefits...again, keeping it a simple two-way system and having a seamless set of transducers operating without crossover from 300 Hz and above.

Hope I am not coming across as stubborn here, just trying to get a full dose of sunlight on the concept before moving on...

[Danny Richie]

Thanks for the post and the offer... I am just kicking ideas around so I will wait, with keen interest, to see what new design you have with Mocking Bird when the info is released.  You seem confident that 16 Neo's per side just won't have the muscle down to 300 Hz or lower.  Even moving up to Neo8-S (price jump is not as bad as I thought)?

They will never really get down that low.

Am I incorrect in assuming that horizontal dispersion improves with drivers tightly spaced in a line array?  Even 8 inch woofers    I would think that the those high frequencies getting jammed together in the vertical plane would start to spread out across the horizontal plane.  If it really is a problem, I would think some felt/foam device mounted over the outside edges of the driver (kind of like the RAAL vertical dispersion foam adjusty thingys) improve the highs and leave the mids to low mids free to forth and couple with their sister drivers in the array (and minimize the SPL efficiency hit)

Unfortunately that is very incorrect. Driver spacing has little to do with horizontal dispersion. It is about the width of the diaphragm. Anytime a driver is playing a wavelength that is less then the width of the diaphragm then it will begin to beam. By laying them on their sides you have a 8" wide tweeter. That won't work. Jamming more of them together does not make anything spread out across the horizontal plane.

The reason that some of the RAAL tweeters add foam on the top and bottom is to limit the height of the diaphragm and improve vertical dispersion. Otherwise if you were to stand up you'd notice all the highs dropping out.

Also, regarding crossover... Would I really need one?  If I used six of your 8" servo subs per tower, I would have three each on one of two servo amps per side, and those amps would handle the low pass to the subs at 300 Hz with their built in electronic filters.  I could then just put an line level active brick wall high pass set at 300 Hz (or maybe even 350 Hz) in front of the main stereo amp and with a little tweaking get good results without actually doing a full crossover design/build.  OK, I am really pushing the limits of my novice understanding of all this, but it still seems to be a unique solution with some potential benefits...again, keeping it a simple two-way system and having a seamless set of transducers operating without crossover from 300 Hz and above.

If I were to take a measurement of the tweeters by themselves with no crossover then you'd notice the response will be in the +/-10db range.    And you'd have a huge rise in the lower range shooting up well over 100db and much less output up top. The crossover isn't just acting as a transfer function to roll the end of the drivers range off at a certain point. It also controls the actions of the group of drivers within their pass band.

[Greggo]

Danny,

Thanks for the great feedback on this... lessons learned and I will move on to some other ideas to keep in mind for a future project as well as keep an eye on your circle here to see what you are up to now then.  It looks like your LS9 is still the closest example to my ideal DIY project, but with a decent digital crossover to get around the need for special Midbass drivers you created for the design.  Of course, the simplest approach would be to just get a full blown kit, but I want to have a little more pressure put on me to figure out a final solution.

Thanks again.

[Danny Richie]

Danny,

Thanks for the great feedback on this... lessons learned and I will move on to some other ideas to keep in mind for a future project as well as keep an eye on your circle here to see what you are up to now then.  It looks like your LS9 is still the closest example to my ideal DIY project, but with a decent digital crossover to get around the need for special Midbass drivers you created for the design.  Of course, the simplest approach would be to just get a full blown kit, but I want to have a little more pressure put on me to figure out a final solution.

Thanks again.

When a decent digital crossover is available that will not hold the performance back then I'll be the first to own one. As of right now they don't exist. A high quality DAC and separate passive network is still the top of the mountain.

[Ric Schultz]

A line of neo 8s will never have as good of highs as a line of neo 3s.  You just have lower mass and also the highs are not intermodulated by the lower midrange when you have separate mids and highs.  So a line source of neo 10s combined with a line source of neo 3s is going to be super killer.  The neo 10s have great horizontal dispersion to around 3K and there the neo 3s take over.  With a line source of at least 40 inches you can cross over as low as 2K with a 12db passive xover and still keep the distortion low in the tweeters (remember a 40 inch line source uses 12 neo 3s).  The cost on a 40 inch neo 10/neo3 line source is only $600 more than your original idea.  Now you have neo 10s down to 200hz and 12 neo 3s giving way better highs with super horizontal dispersion.  Of course, longer line sources are even better......but more money.  There is no single driver that I know of that is super for mids and highs (especially if you want the driver to go as low as 200hz to match the incredible GR Research servo woofs).  The B&G large planars don't have nearly as good of highs as the neo 3s.....in fact, Genesis uses those large B&Gs in their super expensive speakers but use a line source of ribbon tweeters along with them.  The LS9 is an incredible speaker, but has limitations.  An open baffle speaker (where the neos can be used dipole) combined with the servo woofs (also open baffle) will simply be much better.  This is why Danny's monster Serenity line source speakers will be made that way.  However, DIYing allows it to be mucho cheaper. 

[bdp24]

Ric---There is no single driver that you know of that is super for mids and highs? Not even any ESL? Did you like what Danny did with the P-Audio coax in the Super-V? Sorry to put you on the spot!

[Ric Schultz]

The only Electrostats for DIY use that I know of are the ERAudio ones from Australia.  You could stack them for a line array.  The problem with many electrostats is dispersion.  The flat ones have no dispersion (one person speaker...no dancing around).  They also don't have the dynamics of the Neos.  So, if you do not mind the beaming and have an amp that can drive them then a stack of the ERAudio mini panels or the new Essence Electrostats might be a super duper incredibly transparent speaker.  You can buy the Essence 1200 for about $4600 delivered and then make a line source of 6 8 inch servos next to it and cross over the 8s around 200hz.  You would use a cap before the main amp to roll off the electrostats at 200hz.  Such a speaker would probably be really good if you don't mind the beaming and they will never play 110+db.  Also, you need a good amp like a modified Job 225 to drive them properly (2 ohm load at 20K).  The Essence speakers have not been reviewed and are new but are the latest version from the guy that made the Final Electrostats.   

A line source of Neos will play as loud as you want, are super reliable, incredibly transparent and give you dispersion so more than one person can enjoy the sound and would be much more suitable for home theater use and for those in need of an occasional rock and roll fix.....not to mention Mahler symphonies at 105db.  A pair of LS9s can also play crazy loud.

Danny has said the Neos are way better than the P-Audio coax.   I believe him.

[Hank]

Just Real Music might be a 'stat DIY resource:  http://www.justrealmusic.com/content/home1.htm  I've been paying attention to the ER Audio 'stat panels, but using enough of them for a tall line array would be VERY expensive.

[mojave]

When a decent digital crossover is available that will not hold the performance back then I'll be the first to own one. As of right now they don't exist. A high quality DAC and separate passive network is still the top of the mountain.

If using a Mac or PC, you can have a software based digital crossover that will work with any DAC. In my opinion, an issue isn't the digital crossover, but the limited number of DAC's with more than two channels (even though I own 3 different multi-channel DAC's in addition to 3 multi-channel soundcards ). Another issue is the extra cost of good amplification.

[SoCalWJS]

The only Electrostats for DIY use that I know of are the ERAudio ones from Australia.  You could stack them for a line array.  The problem with many electrostats is dispersion.  The flat ones have no dispersion (one person speaker...no dancing around).  They also don't have the dynamics of the Neos.  So, if you do not mind the beaming and have an amp that can drive them then a stack of the ERAudio mini panels or the new Essence Electrostats might be a super duper incredibly transparent speaker.  You can buy the Essence 1200 for about $4600 delivered and then make a line source of 6 8 inch servos next to it and cross over the 8s around 200hz.  You would use a cap before the main amp to roll off the electrostats at 200hz.  Such a speaker would probably be really good if you don't mind the beaming and they will never play 110+db.  Also, you need a good amp like a modified Job 225 to drive them properly (2 ohm load at 20K).  The Essence speakers have not been reviewed and are new but are the latest version from the guy that made the Final Electrostats.   

A line source of Neos will play as loud as you want, are super reliable, incredibly transparent and give you dispersion so more than one person can enjoy the sound and would be much more suitable for home theater use and for those in need of an occasional rock and roll fix.....not to mention Mahler symphonies at 105db.  A pair of LS9s can also play crazy loud.

Danny has said the Neos are way better than the P-Audio coax.   I believe him.

No argument here - completely agree. The speed, transparency, detail, and tone are better and more natural sounding to my ears. Only place the P-Audio is better (to me in my listening room) is the soundstage depth that is created, and this is a biggie for me.

[bdp24]

Guess I'd better get it in gear and get some Neo 3's! I love ESL's, but am all too aware of their SPL and dispersion limitations. Rock n' Roll speakers they are not! Anybody need a vintage snare drum or kit? 25 sets and 50 snares to choose from.