ProSoundWeb Community

Sound Reinforcement - Forums for Live Sound Professionals - Your Displayed Name Must Be Your Real Full Name To Post In The Live Sound Forums => LAB: The Classic Live Audio Board => Topic started by: Xandy Whitesel on June 04, 2018, 08:28:29 pm

Title: hockey arena sub configuration recommendations
Post by: Xandy Whitesel on June 04, 2018, 08:28:29 pm
I have a european hockey arena type tour coming up in the fall and would love to hear from people about recent sub implementations that were successful.  On this tour the arenas are 6-11k capacity and will need 180 degree coverage that you imagine a typical arena show needs.

PA is d&b.  An initial recommendation I received is 18-ish B2 sub array and 6/side flown Jsub.  This is not set in stone yet and jiggle-able; i.e. I can probably add to this sub count.

I have mixed 3 arenas so far and was never satisfied with sub coverage.  All were a standard ground stacked delayed arc.  At the most recent one we started with 6 stacks of 2 B2 (omni) delayed and spaced across the front of the stage on the floor.  With the delay in, definition of transient low end was ridiculously terrible.  Taking all the delay out resulted in an increase in definition everywhere but it was still bad, and power alley city.  I presumed that omni mode subs means we're spewing LF into the 3rd of the arena behind the stage and with the relatively low arena ceiling it just wraps around and causes havoc and reckoned that a cardiod array would be better.

We had a little time and gumption that day and switched from the 6 stacks of 2 B2 in omni to 4 stacks of 3 B2 in CSA with 4' gaps and no delay.   That made a world of difference, presumably because we were keeping LF from rolling around the 3rd of the arena that is behind the stage and were also narrowing the array so presumably widening the horizontal coverage.  Coverage on the floor was presentable and the perceived RT of LF much lower.   Coverage in the side seating tier was not terrible but definitely not good, far from what I aim for and believe to be attainable.

As good as an omni sub arc seems to work in a shed, it has so far failed me in arenas.  I would like to hear about configurations that others have recently implemented and felt generally happy with in a 180 degree arena setting.  I've read up on dave rat's designs but I've never met or spoken with anyone that has implemented any of them.   They require quite a quite a few extra boxes which I'm not opposed to if that really gets us good coverage.   But I don't understand his theories quite enough to convert the design to B2/B22 and scale down to half the boxes.  D&B/EDS seems committed to only the basic omni or CSA sub array, so I need to be confident and knowledgeable if I'm to suggest and implement an alternative.

Anybody got a good story?
Title: Re: hockey arena sub configuration recommendations
Post by: Xandy Whitesel on June 14, 2018, 01:35:59 pm
No arena stories available, apparently.
Title: Re: hockey arena sub configuration recommendations
Post by: Luke Geis on June 14, 2018, 02:50:11 pm
Arena's are not fun for that reason. You don't get to have your cake and eat it too with subs. You have enough subs that you could do a L/C/R with an endfire array on the L/R sides. I have seen L shaped arrays on the sides too in order to get coverage on the sides and provide some directional control. Depending on the stage type, a cardiod array in the center cluster could help with the build up in the rearward plane. I would see what it looks like in a prediction program to have an L/C/R systems with endfired subs on the sides with some outfill and a cardiod sub cluster in the center. With stacks of two you may only utilize 15 subs instead of 18? If you want to get more clever you could do a cardiod setup for the outfills and keep all 18 subs in use? I would play with the prediction software though. I like playing with sub arrays and I'm not afraid to deploy them. The hard part is breaking away form the norm. Get creative and see what you can come up with.
Title: Re: hockey arena sub configuration recommendations
Post by: Xandy Whitesel on June 26, 2018, 05:18:15 pm
Looking for comments from anyone that has ever implemented any of these.

I've been doing some modeling in arraycalc, specifically aiming for wide coverage and rear cancellation.  Rooms are going to be 60' stages, 6k-12k cap arenas, and 180 degree coverage, 2 to 3 levels (1 to 2 balconies).  Big.  It seems like all of the venues will have space on each side of the stage to pile stuff.  I'm hoping to have 24 B22's to play around with, plus a flown 6/side J sub, but I haven't wrapped my brain around that part yet.  What are they for and won't they just mess everything up further?

I've come up with the following designs for a 60' stage and 24 B22's.  Most of the rooms are regular oval shaped arenas, but Oslo Spektrum is super wide and shallower.  B22's have at least 3db of directionality built in to them, so I've modified the basic designs a little to compensate.

Basic D&B style cardiod spaced arc.  Good cancellation, coverage is smooth on axis but frequency response really gets jagged once you get off to the sides.  Really inefficient compared to the others - there's just a lot of canceled energy I guess.  I normally love this setup but it seems like it falls apart trying to be so big and wide.   It seems like the coverage is smoother with minimum spacing between stacks, but I didn't go less than 2' gaps, as recommended by D&B.  D&B doesn't want you to tight pack if you are in cardiod mode.  But I really wonder about the practical difference of keeping the line to 40' or going wider to say 60'.  60' definitely looks worse in arraycal.  And then there's the whole transient argument from the cardiod aspect.  24 boxes somehow seems like not enough.

D&B single high tall-side up cardiod tight packed line.  There is no option in arraycalc for putting them on their ends with no gaps, just a solid line of subs (the middle one of every 3rd reversed for CSA) across the front of the stage, but it is pictured as OK in the D&B sub document.  So I wonder about that.  But I have no model.

Sloped tall single wide 3 deep end-fire L/R.  (sloped up (in number of boxes) since B22's are somewhat directional already - front stack has 5, middle stack has 4, rear stack has 3)  I'm wondering if the slope actually steers the beam up, which would be good for balconies maybe.  Easy to implement, just 4' spacing and some delay.  5 tall stack seems like a PIA?  I'll bet this setup is punchy.  Reasonably compact.  Off axis frequency response seems jagged.  Seems like a heavy power alley, maybe coverage is too wide in the center.  Doesn't seem like you'd want more center coverage on this setup.

Forward and Sideways End fire L/R.  a row of 3x2 boxes facing sideways/offstage extending along the side of the stage, and then another similar row 3x2 4' offstage of that. Not too bad?  Seems really wide.  Probably nice and tall coverage too for high balconies since it's just 2 box high stacks.  Really wonder about this one.  Kind of a lot of space required but seems doable since it is along the side of the stage.

Dave rat vortex L/R.  Very wide...power alley but dispersed frequency wise and power wise.  hmm!  no crazy tricks to do it besides space and measuring and delay.  It looks crazy but seems like it is basically a 2 dimensional take on endfire.  If you turn it diagonally coverage can narrow a little.  Takes up some room.  Cancellation on stage looks very good.  I've never talked to anyone that's done it, though it is talked about on the internet quite a bit and there's even an AES paper that dives in a little.   Arraycalc even pretended to understand what would happen (and it matched my expectations.)  Has anyone besides Dave Rat ever implemented this setup?  It honestly this setup looks the most promising on paper of all of these.

None of these require  polarity reversal (which D&B amps are NOT capable of).

And then I have no idea what flying 6/side J-sub does to any of this.  I haven't been able to get arraycalc to tell me anything about how the flown subs will interact.

I'd love to hear any thoughts good, bad and ugly.
Title: Re: hockey arena sub configuration recommendations
Post by: Xandy Whitesel on June 26, 2018, 05:20:53 pm
more pix.
Title: Re: hockey arena sub configuration recommendations
Post by: Xandy Whitesel on June 26, 2018, 05:22:39 pm
and 1 last design that I kind of like the least.

Wide low steered end-fire L/R.  Seems actually really good but needs a ton of space, probably more than is available.  And I wonder how much transients take a dump with the steering delay action.
Title: Re: hockey arena sub configuration recommendations
Post by: Steven Eudaly on June 26, 2018, 06:30:59 pm
I cannot speak to the B22s specifically as we're all V-Subs and J-Infras in-house, but I will point out to those following along that they need to be looking at the polar plots for comparison, not necessarily the SPL maps as the frequencies being modeled in the SPL maps are not consistent between screen shots (some at 50 Hz, some at 63 Hz, some at 80 Hz, etc).

At this point I would be asking at what frequencies are the most important LF portions of your mix? Is your kick drum response at ~65 Hz the most important aspect or is it bass guitar/synthy stuff that falls below 50 Hz? In contrast, what frequencies are the folks on stage most bothered by? At what frequencies the majority of your low end mix lives should certainly take weight in your decision.

As you mentioned, consistency in the room and rejection on stage are your primary goals. That being said, at least according to these models, I personally feel the typical virtual arc of cardioid stacks across the front of the stage provides the most consistent seat-to-seat experience and the best rejection downstage center, presumably where your lead singer lives. You could play with the delay of the outer stacks to shift those L/R rear lobes a bit if you've got musicians down there that have issues with LF spill on stage.

This setup will likely narrow your vertical LF coverage, but ideally your flown subs are deployed in such a way they help mitigate this issue. I don't have a ton of experience with flown subs, but I've usually used them to make my mains sound "bigger" allowing me to rely less on the ground subs for "body" in the 60-100Hz region and more for "impact" below 60Hz.

The sloped endfire and vortex methods are intriguing, as they definitely improve response to the sides, but at the cost of some fairly drastic lobes on the floor through a good portion of the "money seats." I'd be looking to give the most consistent experience within sections of seats while understanding certain areas may suffer, comparatively, but at least I know that going in to the gig and can try to alter delay times to steer that toward poor selling sections, cheap seats, etc. These are hard compromises to make as I whole heartedly want everyone to have the best show experience possible, but often necessary in effort to minimize negative effects.
Title: Re: hockey arena sub configuration recommendations
Post by: Dave Garoutte on June 26, 2018, 11:31:58 pm
I don't know if it's relevant, but a while ago, Merlyjn (sp) did an article on spacing subs a few inches apart.
Title: Re: hockey arena sub configuration recommendations
Post by: Helge A Bentsen on June 27, 2018, 05:17:13 am

I've never heard a flown sub array in Oslo Spektrum that works, but I've not been to many concerts there.
Someone might have done it.

The best sounding subwoofer coverage in that room seems to be groundstacked solutions if you can tolerate variable timing between your mains and your subs. I've only done one gig in there as a systech (JBL VTX/S28), ended up with all the subs on the floor in four end-fire arrays (60hz XO) and that was reasonable good for what we did in there.
Title: Re: hockey arena sub configuration recommendations
Post by: Uwe Riemer2 on June 28, 2018, 03:18:07 am
Aggreed on everything Steven said, but would like to add some random remarks:

Distributing sub energy 180 degrees ( Forward Aspect Ratio 1:1 and reduced energy to the rear ) is probably the most difficult task I can think of.

The line of subs in front of the stage:
The wider the line, the narrower the dispersion.
And then there is an upper frequency limit the line can reproduce: Length of the line divided by 8 gives the wavelength,
applying delay to create an virtual arc will lower this frequency limit.

Flown subs and groundstacked subs covering the same range:
I would place the groundstacked subs as close as possible to the flown ones and I would make sure, both arrays have similar impulse behavior and dispersion on their own,
otherwise I would try to highpass the flown ones and use them as low extension to the main array. d&b processing does not allow this option AFAIK.
Introducing an additional spectral crossover is not an easy task and comes with the penalty of increased group delay.

Directional subs:
These have a direction built in, so I point them where they should go, whenever space allows this.

Endfired arrays:
the first pulse of the sub closest to stage will have no cancelation to the rear until the next sub comes into play, so I think endfired should be at least 3 deep, better 4 or 5 deep.

Keep up the good work Xandy, have you thought about contacting d&b directly for assistance ?


Title: Re: hockey arena sub configuration recommendations
Post by: Jean-Pierre Coetzee on June 29, 2018, 03:33:03 am
I would look at something like Dave Rat's vortex setup on the sides with a usual cardiod setup through the middle.

He mentions some stuff about it here ( just search for for vortex it should come up.
Title: Re: hockey arena sub configuration recommendations
Post by: Xandy Whitesel on June 30, 2018, 01:49:13 pm

Thanks you'all for these comments...all tidbits of information are helpful.


As Steven mentioned, looking at the polar plot is indeed a good thing to point out and keep in constant mind.   In fact the basic spaced virtual arc looks pretty crap at 80Hz and above (presumably because of the wide line length and also the decrease in output of a B22 at the freq) but comes together below that and looks more powerful, thought at 65Hz it still doesn't look great to me considering it's not pushing hard in the +20 degree and -20 degree directions.  But the coverage below 65 is very very even.

I answer the frequency response questions probably the same as everyone else.  Yes, I want a big tight 65hz-ish push for rocking kick drum.  Yes, I still need the rest of the frequency range, the lower stuff for the songs that are bass synthy, the higher stuff for the 2nd kit's kick drum, the whole range for the bass gtr on the rock songs.  Yes, the musicians are most bothered by low end that is lobe-y on stage or that clouds up their mix (80-120) beyond discernability.  The weight of my kick moves around some, room to room.  If 70-100 is tight in that particular room I like it higher.  If it's not I'll go lower.  Super low kicks (40-50hz) kind of irritate me but I'll do it if I have to.  What I'm saying is that I think we all probably tailor decisions and compromises we make based on what is happening in the room that day.  If 40hz is a floppy mess that goes on for 10 seconds then if we're smart we avoid it.

Steven, why do you say that the "virtual arc of cardiod stacks...will likely narrow vertical coverage"?  I would like to understand the physics of that statement.  I've done a few measurements in theaters of 2 high stacks vs. 3 high that indicate that the top sub of a 3 high stack increases low end on the floor 6db but doesn't do anything for the balcony.  Is your reasoning related to this?
RE money seats... from the response I've received from my PM/BMs, seat cost goes down with elevation, which doesn't help us much with decisions: we still need wide and even coverage. 

It is hard to know what the ideal ratio of forward to side coverage is; I guess it will really depend on the shape of the venue.  1-1 sounds actually too wide to me, at least in a normal oval shaped arena where the PA will be setup to throw farther than wider, though in Oslo spektrum 1-1 looks desirable since the venue is kind of even depth to width and very wide.  I think that (in the sub frequency range) rear cancellation is just as important as coverage in an arena setting.  If sub freq is bouncing around the back of the arena then the entire venue sounds like mushy wet dogdoo.

end responses, begin notes.

My thinking has distilled down to the desire for two options.   The status quo option is: spaced and cardiod stacks across the front of the stage delayed to arc the coverage, augmented by a flying cardiod hang to widen side coverage.  Is there an alternative setup that looks as good or better if you add up and weigh all the compromises involved?  I think I may have found one, described later, but the proof will be in  the pudding.  If I have a total of two setups to audition, I can find the time, gumption, and manpower to set up and dial in them both, and decide which one seems better.

Attempting to improve on the status quo, I made a new file for the spaced arc keeping the line to 40' wide but turning the outside cardiod stacks 45 degrees.  This seems to tighten up the polar plot in the 90 degree off axis region, bringing in that lobe at 80Hz and maybe improving rear cancellation.  It would still be weak on the sides; enter a flown hang of J-subs into the equation, toed out and hung to cover 60-90 degrees off center.  But how is the lobed coverage you'd get from a flown side hang interacting with a center covering arc any different from the lobed coverage from a Left pile interacting with a Right pile?

The vortex plots seem to look promising.  But what is going on in there?  What is the vortex exactly?  How does it work?  I've distilled it down to this: it's a two dimensional end fire.  If you setup a square of subs all pointing in the same direction, and think of it as an end fire in the forward direction and another end fire in the sideways direction, you can end up with a plot resembling the vortex.  But the vortex actually has something else going for it.  With the subs oriented in different directions it has somewhat varying distances between each pair of acoustic centers.  This morphs it a bit away from end fire and a smidge towards delay steering, but also dissipates comb filtering (in the forward direction) and cancellation (in the backwards direction).  And also makes you look at the thing and wonder what the hell is going on happen when you turn it on.  What all is getting ignored, including that sub energy from one box actually has to go around the box in front of it, and can you really treat a 3'x4' box as a single point of acoustical center?  Not only that, but B22 subs are different dimensions and directionality than SB28s, so the math between my model and DR's model is not the same.  When I nudged the B22's locations to be a closer match to the acoustic center of an SB28, the plot looks worse.   If I put in delay numbers that match closer to what DR describes, the plot looks worse.  Currently the delay numbers are what arraycalc came up with to get the most forward steering, and arraycalc actually decided to use a different sub stack as it's timing start point than DR's calculations.  Which is all fine, I guess.  We're dealing with some kind of squishy seeming math anyway.

In my attempt to understand how the vortex worked, I made various models of subs in various orientations, as I said, and ended up deciding that the theory is based on an end fired end fire.  Then I stumbled on a setup that looks similar to a vortex, but simpler, and in my opinion better, at least with the B22 subs that I'll be using.  Finally, this plot looks like a potential competitor to the status quo.  And it might be slightly easier to understand and set up.  It's still a square-ish shape of 4 stacks of subs; the inner 2 stacks face out, the outer stacks face forward.  This offsets the acoustical centers somewhat, naturally steering the array slightly towards the outside.  It's about 9' deep and 8' wide.  Mucking with the delays, I arrived at a plot that looks very wide cardiod, and the rear cancellation looks excellent.  (You can widen or narrow it a bit with delay times, but rear cancellation changes along with it.)  Add everything together and the compromises still look reasonable.   (The squeeze in 40Hz off axis coverage on the LR polar plot is a freq/distance specific hole.  Or is it a crack?)  I feel good about it, but what do I know?  I know I want to know what you'all think.  Can you'all please pick this apart and point out the goods, bads and uglies of this solution?

plots in accompanying posts.

Title: Re: hockey arena sub configuration recommendations
Post by: Xandy Whitesel on June 30, 2018, 01:50:27 pm
new awesome setup??
Title: Re: hockey arena sub configuration recommendations
Post by: Xandy Whitesel on July 09, 2018, 12:26:57 pm
new awesome setup??

Had some time and space and gear after a rehearsal a few days ago, so I set this up with 4 Qsubs and walked around.  No measurements, just pumping a track and getting a seat of the pants impression, and also be convinced that math in a fancy prediction program actually translates to the real world with real subs inside a real room.  It was quite perceivable and appeared to obey the plot, super punchy in front and a lot of cancellation obeying the wide cardiod pattern.  The 80-100hz lobe in back was also noticeable.  It seemed like the pattern might be a little narrow actually, if I set this up again the times can be massaged/lowered to widen the pattern a bit if necessary.