2" Driver Line Arrays - How to Angle Parallel Drivers?

[axes4less]

I have successfully built several line arrays in the past 5 years, following the advice found on these forums. The first, and maybe still the best, uses the infamous 4" NSBs. I have (18) drivers mounted in each column. They are still being used by my classic rock cover band here in Austin, TX as our only speakers.....similar in concept to the Bose PAS solution. We use no backline guitar/bass amps or monitors, just these towers sitting slightly behind us. They are amazing, and several friends in other bands have now also employed similar solutions. (I still have ~100 NSBs boxed up in my studio....what a great driver they were for $0.59!)

2 years ago I also built two 36 driver arrays using GENTO 2" full range drivers. (SP99023A). They are currently in use at my church for our praise band.

I want to try another build using small 2" drivers, since for live band use, I can have open microphones several feet closer to the towers before any feedback problems appear. Don't get me wrong, even the NSB towers can have a mic within 6-7' directly in front of a tower and usually there is no problem. But with the smaller drivers...I can literally get within 1-2' and typically don't have any feedback.

But since the 2" drivers have a lower power rating, I want to "double up" on the number I use. AND, this time I wanted to test the idea of mounting them side-by-side, and creating an angled baffle with the hopes that I could cleanly increase the width of dispersion.....especially in the higher frequencies. While the single column of drivers has excellent sound to the sides, once you move about 30 degrees off-center, you can begin to notice some fall-off of the higher frequencies.

So...my question to this team of experienced experts is this? Should I create a concave baffle...wherein the soundwaves would immediately cross each others? - or- Would it be better to create a convex baffle, hoping to have the the soundwaves couple at their edges? I have attached an illustration to help the readers understand my question. It would seem that effects from comb filtering might be worse with the concave solution.....but the phenomenon of 'coupling' that occurs with Line Array solutions still amazes me....so maybe there is something unique physics-wise that would make that angling better???

[3way]

Re: 2" Driver Line Arrays - How to Angle Parallel Drivers?

Don Keele has published interesting work on Broadband Constant Directivity line arrays that I found very educational. A Google search of Keele's papers and patents can explain the concept better than a short post. If I had ~100 NSBs boxed up in my studio I would explore a CBT.

If you decide to construct a CBT, please report your experience.

Best Regards..............

http://www.linkwitzlab.com/Keele%20-...er%20Arrays%20...

Introduction to CBT
Loudspeaker Arrays
A Presentation Given to a Joint Meeting of the
Boston Sections of the
Audio Engineering Society,
Acoustical Society of America, and the
Boston Audio Society.
January 14, 2010
By
D K l
1/14/2010 Keele ‐ CBT Boston Presentation 1
Don Keele

[billfitzmaurice]

Re: 2" Driver Line Arrays - How to Angle Parallel Drivers?

Would it be better to create a convex baffle, hoping to have the the soundwaves couple at their edges? I have attached an illustration to help the readers understand my question. It would seem that effects from comb filtering might be worse with the concave solution.....but the phenomenon of 'coupling' that occurs with Line Array solutions still amazes me....so maybe there is something unique physics-wise that would make that angling better???

Concave. It minimizes CTC distance for minimal combing.

[bolland83]

Re: 2" Driver Line Arrays - How to Angle Parallel Drivers?

I would vote for concave, it would put the center of the drivers closer together thus minimizing the comb filter effect. I may be wrong about this but it would seem that the dispersion should be about the same in either configuration. The main difference being in driver center to center spacing.

Edit: Bill ^ and I must have posted a reply at the same time, haha!

[axes4less]

Re: 2" Driver Line Arrays - How to Angle Parallel Drivers?

Thanks for the replies.

I'm googling lots more articles...and starting to become more concerned about the effects of "beaming" by having the drivers 'side-by-side'. But everything I've read so far seems to support the advice to go with a concave vs. convex baffle mount....in that it would further shrink the c-t-c distance of the drivers and therefore further minimize comb filtering. Would appreciate further comments/advice on the "beaming" concerns. The whole reason for exploring an angled baffle mount for the side-by-side 2" driver array was in hopes of achieving wider "quality" dispersion. If in doing so....I instead cause mid-high frequency SPLs to "beam" on axis....then any performer (or audience) who is ~30 degrees off-axis will be even more disadvantaged in what they "hear". ???? So if this is a real concern, maybe I should dump the idea of parallel drivers in a column array altogether?
It is interesting to me that the BOSE column arrays never have side-by-side drivers...but the new solutions DO angle each successive vertical driver Left/Right several degrees.

R.e. the CBT...fascinating solution that I am enjoying learning more about. However, for the solution I am working on....which is portable sound reinforcement for live bands in small to medium size venues.....I don't think it is practical. Here's why > If I understand the solution correctly, the array must be curved as in the illustrations and photos I've seen so far. The club venues we play in Austin typically range from 1000-2000 sq ft...and the stage you have to set up on is sometimes as small as 16 x 24. So the ability to place tall/thin arrays directly against the back wall is a real advantage. A "C" shaped CBT array would simply take up too much space, let alone the complexity of moving it from gig-to-gig.

[billfitzmaurice]

Re: 2" Driver Line Arrays - How to Angle Parallel Drivers?

the whole reason for exploring an angled baffle mount for the side-by-side 2" driver array was in hopes of achieving wider "quality" dispersion. If in doing so....I instead cause mid-high frequency SPLs to "beam" on axis....then any performer (or audience) who is ~30 degrees off-axis will be even more disadvantaged in what they "hear". ?

And there's the rub. Above the frequency where the drivers mutually couple dispersion of two drivers horizontally placed is half that of a single driver, so more often than not the result is the opposite of the intent.

[axes4less]

Re: 2" Driver Line Arrays - How to Angle Parallel Drivers?

Thanks Bill.
So I'm back to the drawing board...sticking with a "single wide" tall array....and instead still trying to figure out how in my home woodworking shop I can design/cut a baffle that alternatively angles the drivers Left/Right about 15 degrees of each other...similar to what the current BOSE PAS solution does.
Probably will require some pretty fancy router work to shave a 15 degree incline every-other driver space :D ...wondering how much improvement in the off axis I could expect to achieve .

Maybe I just instead cut a single "veneer" board from 1/2" MDF at 15 degree angle....drill the driver holes...and then cut it apart into "shims" to glue in place (alternating angles) underneath the drivers ...which would thereby cock every other driver Left/Right.

Any better or add'l ideas out there?

[DougP]

Re: 2" Driver Line Arrays - How to Angle Parallel Drivers?

Double up on the number or arrays you are using. Go from 2 to 4 arrays. It also has the advantage of keeping the boxes lighter and easier to move.
Doug

[axes4less]

Re: 2" Driver Line Arrays - How to Angle Parallel Drivers?

Having multiple arrays that are sending out the same source material would cause comb filtering problems at the points where their sound waves intersect. And....the performers and audience will lose the psycho-acoustical benefit of the "cocktail party effect". So things would actually be worse ....not better.

Unfortunately, I still constantly have to remind musicians I play with that it is actually BAD to have their sound split between multiple towers. It's much better to hard-pan to the tower that's closest to them...and have nothing coming out of the opposite tower. (We run LEFT and RIGHT towers for our band)

It's a common misunderstanding,.....a hangover from years of traditional PA systems and running stereo PA mains/monitors/backline amps etc. You have to reshape your mind :-)

If you don't understand these concepts...it's important that you learn. Go read Jim Griffin's white paper > http://www.audioroundtable.com/misc/nflawp.pdf

[axes4less]

Re: 2" Driver Line Arrays - How to Angle Parallel Drivers?

Sorry...Jim Griffin's paper only addresses some of this. The more extensive explanation of why hard-panning is important can be found on white papers written by the BOSE engineers that developed the PAS system.

[billfitzmaurice]

Re: 2" Driver Line Arrays - How to Angle Parallel Drivers?

Thanks Bill.
So I'm back to the drawing board...sticking with a "single wide" tall array....and instead still trying to figure out how in my home woodworking shop I can design/cut a baffle that alternatively angles the drivers Left/Right about 15 degrees of each other...similar to what the current BOSE PAS solution does.

The reason they do so is that a 2.5" driver can't do well off axis above 8kHz, and rather than use a proper 2-way with tweeters they employ that offset. Why? Because it's cheaper that way.

[axes4less]

Re: 2" Driver Line Arrays - How to Angle Parallel Drivers?

Perplexing debate. For Live Music, how much I should be concerned about the off-axis equal dispersion of frequencies abov ~8Khz? If I reference this Musical Instrument frequency chart....for the most part, the range above 8Khz is beyond the fundamental tones in most performances, and the range where we capture harmonics and definition (described as breath, air, sibilance).
http://www.independentrecording.net/...in_display.htm

I've seen several DIYer arrays, like your TLAH solution, with a long line of 1" tweeters, or planars, aligned beside their 3-4" full range drivers, but I have not yet had the opportunity to personally A/B them to the single driver arrays like I have been building.

I run a BBE X3 Crossover in my system to split between my subs and the towers, and I cross at 175Hz. With it's Sonic Maximizer function engaged at only about 30%, and I can bear witness to the improvement in definition/clarity that it brings to our live performances. So if incorporating a line of tweeters into the tower solution really does extend the off axis ability to hear those frequencies above 8khz better, I'd probably be equally sold that the extra cost would be worth it.

Would like to hear from other musicians who are using their arrays in live performances as their sole "PA system", especially if they've been motivated to a 2-way solution incorporating tweeters. I've been following John Murphy's project, know he uses his system for live music production, and chose not to incorporate tweeters, citing the concerns he had about crossover coloration...and also the fact that he was corner loading his solution, and therefore he didn't need to worry about anyone ever being more than 45 degree's off-axis. He's using the ND90-8 3.5 drivers. So maybe my 2" GENTOs are already achieving similar off-axis dispersion (tho' I know they are crap when you'd compare distortion measurements...what you get for $1.25 each).

So...I know lots of folks like to kick the BOSE engineers around. But did they take the "cheap route" to maximize profits...or did they simply decide there were very marginal returns in going the 2-way route, and instead put a few more dollars into tooling for the fabrication of the their enclosure that alternatively angles their 2.25" drivers? It's essentially the same quandary I'm asking as a hobbyist.....if I spend the money and the time, what are the chances that I'll decide later it was worth it? :D

[billfitzmaurice]

Re: 2" Driver Line Arrays - How to Angle Parallel Drivers?

Perplexing debate. For Live Music, how much I should be concerned about the off-axis equal dispersion of frequencies abov ~8Khz?

Good off-axis response to 12kHz is recommended. Above that the source material rolls off, mainly due to the range of dynamic microphones.

With it's Sonic Maximizer function engaged at only about 30%, and I can bear witness to the improvement in definition/clarity that it brings to our live performances.

IME marginal speakers sound better with BBE, good speakers worse. Take that for what it's worth.

I know lots of folks like to kick the BOSE engineers around. But did they take the "cheap route" to maximize profits...or did they simply decide there were very marginal returns in going the 2-way route, and instead put a few more dollars into tooling for the fabrication of the their enclosure that alternatively angles their 2.25" drivers?

The cheapest method is how Bose has always done it. Too much $ invested in building their products leaves less $ in the advertising budget, and they learned long ago that advertising is what sells, not 'better sound'.

[Organcat]

Re: 2" Driver Line Arrays - How to Angle Parallel Drivers?

"Would like to hear from other musicians who are using their arrays in live performances as their sole "PA system", especially if they've been motivated to a 2-way solution incorporating tweeters. I've been following John Murphy's project, know he uses his system for live music production, and chose not to incorporate tweeters, citing the concerns he had about crossover coloration...and also the fact that he was corner loading his solution, and therefore he didn't need to worry about anyone ever being more than 45 degree's off-axis."

I've just built a 14-speaker 9 lb line-array column for pro PA use with a DIY sub. I used buyout BMI hi-wave 2" full-range speakers and my horizontal dispersion is almost 180. No tweeters needed. I can't believe the results myself. Check out the projects on my page.

[mikec]

Re: 2" Driver Line Arrays - How to Angle Parallel Drivers?

Question from a novice -- since there's a limit on how close together you can put drivers, and if you get them too far apart there's other problems, is there a "not too close or too far" distance where you minimize the problems??