ProSoundWeb Community

Church and H.O.W. Forums for HOW Sound and AV - Your Displayed Name Must Be Your Real Full Name To Post In The Church and HOW Forums => Church Sound Archive => Topic started by: Rich Stevens on June 13, 2007, 11:57:17 PM

Title: Yup - there's the problem.
Post by: Rich Stevens on June 13, 2007, 11:57:17 PM
I have volunteered to administer, maintain, and upgrade our church's sound system.  Just so you know my modest SR skills mostly concern portable systems (frat houses and bars) and are 20 years out of date.

So, I've been picking through the tangle of cables piled on and around the mixer.  The system is a typical Carvin package: 24x4 Mixer, 15-band stereo equalizer, one big power amp, one small power amp, and a pair of 832 15-inch 2-way 90x45 black fuzzy speakers.  I'm positive it was purchased by someone who thought it would solve all the church's sound problems, oblivious to the fact that sound systems need to be installed.

The system feeds back at the drop of a hat, despite the fact that everything that can be turned down has been, most notably the equalizer, where many sliders are all the way down, and none are above about -8dB.

It's not hard to see where problem numero uno lies, but I made a drawing anyway to illustrate (see attached JPG).

index.php/fa/149/0/

The speakers are on the roof of the sacristy, and yes, that's a brick wall they're aimed at, between them and the congregation.  No one in the congregation is in their coverage area, but they do an excellent job of firing sound onto the roof.  

Don't let that lower light blue line fool you - if you look at the plan view, the vocal mikes are within the stage left speaker's nominal distribution. It's no wonder the system howls before the power amps' green "Signal" LEDs even think about lighting up.

OK, so I've read enough on this board and elsewhere to know the right way to go about this: hire a consultant or design/install firm, have them model the room and select speakers. Then install the speakers, analyze and tune the room with a 31-band EQ or DSP, then lock it up and throw away the key.

I agree with all that, especially if we have the budget (which we probably don't).  But I can't help thinking: the system is SO bad that whatever I (a semi-informed amateur) do, it would be such a huge improvement that most people wouldn't miss the additional benefits that could be gained by going the "pro" route.  I can't help it, I'm an 80:20 kinda guy - but does that line of thinking seem to make sense in this (rather extreme) case?  

Once I got the drawing done, I moved the speakers around on paper to look at some possibilities.  Actually the 832's aren't all that bad if positioned properly (as long as we're looking at colored lines on a piece of paper. If we're listening to them, who knows?).  I could post the results for discussion if you want, but I'm not really considering flying the 832's.  (But trust me - if I did, I'd hire a rigger - I may be an 80:20 kinda guy, but I'm not stupid.)

I guess what I'm looking for is some suggestions on a path forward.  Keep in mind that I'm an engineer, just not an audio engineer.  Is it possible (likely?) that I could use my cheesy little CAD drawings to lay out a system, fly some new mid-priced installable-rather-than-portable cabinets up there in a center cluster, maybe add a sub, and get decent results?  We might even spring for a 31-band, though I'm not sure about getting some one to analyze the room and set it up.

BTW, the system is presently set-it-and-forget-it with the board up with the band.  The Pastor wants to move it to the back and build a booth, but right now we have no one to run it, so that would be sort of pointless.  However, this is another discussion of another day.

Critique away!  (and thanks... I'm amazed at the level of knowledge, professionalism, and willingness to help on this board).
Title: Re: Yup - there's the problem.
Post by: Rich Stevens on June 14, 2007, 12:01:16 AM
Here's the plan view.
index.php/fa/150/0/
Can anyone tell me why the system feeds back so easily? Very Happy

P.S. Sorry about the size of the drawings - I tried shrinking them, but the lines start to disappear.
Title: Re: Yup - there's the problem.
Post by: Josh Rose on June 14, 2007, 03:51:30 AM
Well, if I am looking at that diagram correctly, the speakers are behind the stage. You put those speakers in front, and you will drop 75% of your feedback problems immediately.

No ammount of EQing will dissolve your feedback problems. The mains simply cannot sit behind the stage and shoot audio right back into the mics it's recieving it's feed from. It seems like whoever set the system up was not too informed.
Title: Re: Yup - there's the problem.
Post by: Chris Penny on June 14, 2007, 05:45:43 AM
I agree with Josh. One of the first rules of speaker placement in any setting, portable or installed, is the speakers go in front of the microphones. Move these and most of your problems should disappear.

Personally I would just for now move them down somewhere near the front of the stage centered as much as aesthetically permitted.

Title: Re: Yup - there's the problem.
Post by: Ira White on June 14, 2007, 08:27:59 AM
I'll third that short of a pro install. Since your seating is only eight rows deep, you should be able to get acceptable coverage for now from the floor, and that would get the dispersion down where it isn't all bouncing off that huge brick wall facing. Just get the stands up at least 6' high. The church would have to live with the more conspicuous placement until they can afford to do it right. For now it's a choice of either acceptable sound or looks. Take your pick.

You can probably set them to each side of the stage, or spread them farther out to each wall. Just see which works best for your coverage needs and cable runs. With proper hardware and mounting, you might even be able to safely secure them to the walls if that is more aesthetically pleasing. With a pro install, speakers with a proper dispersion characteristic for such a small area can be suspended and directed down into the seating to avoid reflections off the walls.

Oh, I forgot, you might buy an inexpensive wireless hearing-assistance system and ask everyone to buy their own receivers. Then you could get as loud as you want with no feedback or reflections! Laughing

(That was meant to be goofy. Sometimes I just need the release.)
Title: Re: Yup - there's the problem.
Post by: Gary Creely on June 14, 2007, 08:28:44 AM
in terms of the placement of speakers they should be just infront of the front end of the stage. As I say on here all the time, call a contractor.

You do not need to have a contractor model the room in order to have a dramatic improvement in your systems performance. You could find a contractor that could come in and fly an new set of speakers for like 3k or so.
Title: Re: Yup - there's the problem.
Post by: Rich Stevens on June 14, 2007, 10:14:53 AM
First, a shout out to Gary from an old Philly boy; raised in Abington, educated at Drexel, spent my young adult years in West Chester, lived for a while in Allentown.  Don't get me started on the Houston vs. Philadelphia thing.

Thanks all for your input. I'm thinking 80:20 isn't such a bad approach to take at this point, and we might be able to afford new cabs & installation.

What are your suggestions on the subject of EQ?  Assuming for the time being that I'm going to tune it by ear, is it worth going to a 31-band?  


Title: Re: Yup - there's the problem.
Post by: Ira White on June 14, 2007, 12:18:55 PM
Definitely a 31-band if not a digital processor with all the reflective surfaces you're dealing. Just make that part of any speaker upgrade and install services, and the integrator can aid in tuning the system with an analyzer.
Title: Re: Yup - there's the problem.
Post by: Rich Stevens on June 15, 2007, 10:06:02 AM
Where do we draw the line between a 31-band and a DSP though?

Now that I've asked the question, let me try and answer it myself: If we're dealing with a single mono cluster with passive crossovers and one big power amp, I'm gueesing a graphic is good enough.  For any level of complexity beyond that (in my case bi-amping would be the next logical step) jumping to DSP is preferrable to adding more discrete processors.

Am I close?
Title: Re: Yup - there's the problem.
Post by: Ivan Beaver on June 15, 2007, 04:31:18 PM
So how is a DSP better than a 31 band because of reflective surfaces?  I fail to see how the surface types have anything to do with the choice of system eq.

Title: Re: Yup - there's the problem.
Post by: Gary Creely on June 15, 2007, 11:30:18 PM
Rich,

If you are going to replace the speakers I would at least have an EQ in the package, if the budget allows go with a dsp. I am not sure what the reflective surfaces and DSP have to do with each other, but you can solve a lot of issues with good dsp.

Do not set it by ear. If you are going to set it by ear I would suggest not using one at all. If you have a company do it the should rta the system as part of the deal.
Title: Re: Yup - there's the problem.
Post by: Brad Weber on June 16, 2007, 04:52:39 AM
What a mix of good suggestions and comments apparently overlooked!

Moving the speakers forward to be at or in front of the front edge of the stage will certainly help.  Not only will it get the stage itself out of the pattern, but it will reduce issues due to reflections off the wall currently just behind the speakers and will get them much closer to the congregation, allowing you to turn down the volume.

Do you mix mono or stereo?  If mono, you might try one speaker covering each seating area, the forward corners of the stage look to be in about the right positions for that.  If stereo, you will have to turn the speakers aimed in so that each covers the entire seating as evenly as possible.  Another thought for mono mixing might be to try just one speaker located about where they currently are in Section (maybe closer to the wall behind them), but centered.  This will probably still greatly limit gain from mics on stage, but might help with the mics off stage left and eliminating interaction between the two speakers.

Think about some acoustical treatments for the rear wall and perhaps even ceiling (it looks like you could be getting some interesting reflections off the ceiling) rather than trying to band-aid the results with audio equipment.  Much better to eliminate problems that to try to overcome them.

Having all the EQ faders down 8dB is the same as turning down the system 8dB, but with some negative effects thrown in for free.  Try turning down the amp(s) and pushing the EQ faders back up.

Which brings up the fact that you mentioned there being two amplifiers.  Are they both used?  Can you tell us how they are wired?  Are the speakers bi-amped or are there subwoofers or perhaps ancillary speakers somewhere else?

A 1/3 octave equalizer would be better than the 2/3 octave you have now.  A parametric EQ or DSP would be even better.  However, unless you are familiar with using these or hire a qualified person to do the system tuning, you might not get that much more out of it.  It sounds like you are not locked into doing this all at once.  I would try getting the most out of what you have before buying a different EQ or anything like that.  Getting the speakers better located and aimed, perhaps treating the rear wall and adjusting the system you have for proper gain structure will benefit you no matter what you then do for signal processing.  You may find that you are satisfied with the system at that point or may get enough improvement that the church decides to invest more money in it.
Title: Re: Yup - there's the problem.
Post by: Ira White on June 16, 2007, 10:21:24 AM
Ivan Beaver wrote on Fri, 15 June 2007 16:31

So how is a DSP better than a 31 band because of reflective surfaces?


I actually meant either choice was necessary due to the reflective surfaces, but I see where the syntax could be taken either way. However, DSP tends to be more precise than the average 31-band unless you're purchasing an expensive high-end model, and the DSP option would offer more flexibility in the future if the system expands in any way.
Title: Re: Yup - there's the problem.
Post by: Ivan Beaver on June 16, 2007, 06:48:59 PM
I am still failing to see what differance an eq will make in a room with reflective surfaces vs one that is dead?

The only way you can eq a room is with a wrecking ball or bulldozer.
Title: Re: Yup - there's the problem.
Post by: Rich Stevens on June 17, 2007, 12:59:53 AM
Brad Weber wrote on Sat, 16 June 2007 03:52

Getting the speakers better located and aimed, perhaps treating the rear wall and adjusting the system you have for proper gain structure will benefit you no matter what you then do for signal processing.  You may find that you are satisfied with the system at that point or may get enough improvement that the church decides to invest more money in it.


I agree.  I'm going to try to improve the positions of the current speakers as a stop-gap measure and I'm sure it will give a big step in the right direction.  Unfortunately, any position is going to be a compromise unless we fly them, and I'm not going to recommend we fly the Carvins - they simply weren't designed for it. That's why the concentration on new speakers, and the EQ or DSP that would have to come with them.
Title: Re: Yup - there's the problem.
Post by: Brad Weber on June 17, 2007, 09:36:48 AM
Rich,

Big picture, what is your experience with audio system design, wiring practices, setting system gain, programming DSPs, system tuning, etc.?  What relevant tools do you have?  Might you end up buying the main equipment yourself but have pros hang the speakers, wire the system and adjust and tune the system?  And after that, then have no system warranty, documentation, etc.?  If so, then there may be advantages to simply hiring a pro to do it all, including design, and provide you documentation, warranty their work and so on.  You might find this, http://www.audiosystemsgroup.com/3Times.pdf, interesting reading.

Flying speakers in that space looks like it could be an interesting challenge.  Based on your drawing, it looks like they might have to be dropped down quite a bit to be in an effective location.  And a center cluster at the front of the stage would have to have very wide coverage, an arrangement for which it can be difficult to get even coverage over the seating.  Then there are all the practical considerations such as whether the speakers might block sightlines to some important element or cast shadows or interfere with ductwork or diffusers.

As far as what speakers and whether subs might be appropriate, more on what you do would help.  It looks like you have a band but it is also a fairly small room, so do you mic everything or only certain instruments?  Is it a high level 'energetic' service?  Do you have wireless mics and people that tend to move around a lot when using the mic?  If you mic the kick, DI the bass, etc. then one or more subs might be appropriate, but that also introduces a crossover and more system tuning into the picture and thus makes the installation and tuning more complex.

Since you identified a center cluster, I'm assuming that a mono mix is possible.  Would it be feasible to temporarily sit just one of the Carvin's on the brick wall at the center of the room with it aimed down the center of the room at the last row of seating and then sending signal to just that one speaker?  You might have to have someone hold the speaker to test this, but it could get some improvements and emulate a step towards what a center cluster would do (not saying it would be the same, just an interim step between what you have and a center cluster).  It is also a location that could probably be made more permanent if it does provide significant improvements.
Title: Re: Yup - there's the problem.
Post by: Rich Stevens on June 17, 2007, 03:48:48 PM
Big picture, I have substantial experience, but mostly with portable (i.e bar band) systems and 20 years out of date.  No DSP experience - I was away having a family while they were being invented.  I'm an electrical engineer by profession, was an electrician before that, and have put together several systems for my own band and others over the years, and was known to be a pretty good mixer when I wasn't playing guitar. So, most of the equipment and wiring doesn't bother me.  My weak areas are new technology (DSP) and the finer points of speaker design/placement in permanent installations.

If we fly speakers, admittedly they will be waaaaay up.  The peak is about 32 ft high and there is a line-of-sight issue.  There is   a circular stained-glass window in the front wall, above the altar - it's top is at 26 ft off the floor, so any flown cluster would have to be above that.  No ductwork problems, but good point about casting shadows - I'll have to check possible locations against the lighting.

Our music is not exactly high-energy.  It's usually 2 or 3 guitars: one electric, one acoustic, and one who switches back and forth.  Acoustics go direct, the other guitarist uses a modeler direct.  Mine is the only amp on stage - a Fender Champ putting out a more-than-sufficient 6 watts. We run the bass through the PA, mostly because there is no stage amp.  We do not presently mike drums.  The musicians do not move around, we just stand in front of out mikes and play and sing.

We do mix in mono.  We can't move a single speaker to the center of the top of the wall - not shown on the drawing there is a large cross on the wall that sticks up about 6 ft above the top. So, if speakers are going on the top of the wall, they'll most likely need to be at the outboard ends.  I think they'd have to be pointed inward to keep away from the band - won't that cause cancellation problems in seats where people can hear both speakers?  The brick wall is still a bit farther back than I'd like - better than where they are now, but still behind most of the mikes.

Title: Re: Yup - there's the problem.
Post by: Rich Stevens on June 17, 2007, 04:55:38 PM
Brad got me thinking "what if" we tried some of the fast, cheap, and easy solutions.  Like, what if we just move the cabs forward, and atop the wall?  It's bound to be better than pointing them at the wall.

Something like this:
index.php/fa/154/0/

Title: Re: Yup - there's the problem.
Post by: Rich Stevens on June 17, 2007, 05:00:43 PM
...and here's the elevation.
index.php/fa/155/0/

I assume this is a lot better, just due to the fact that it improves directional control.  The cabs at least have a straight shot at the congregation (without first bouncing off the ceiling and walls) and mostly avoid the band and altar.

Is crossing the cabs toward the center a bad thing?
Title: Re: Yup - there's the problem.
Post by: Mike Sessler on June 17, 2007, 05:49:15 PM
You may get better coverage, but your gain before feedback will almost surely go down. Like others have said, you need to get the speakers in front of the stage. Start with that and work up from there...
Title: Re: Yup - there's the problem.
Post by: Rich Stevens on June 17, 2007, 06:02:44 PM
That puts us back to the "flying" idea...
index.php/fa/156/0/
Title: Re: Yup - there's the problem.
Post by: Rich Stevens on June 17, 2007, 06:04:49 PM

index.php/fa/157/0/
Title: Re: Yup - there's the problem.
Post by: Brad Weber on June 17, 2007, 06:46:26 PM
Mike Sessler wrote on Sun, 17 June 2007 17:49

You may get better coverage, but your gain before feedback will almost surely go down. Like others have said, you need to get the speakers in front of the stage. Start with that and work up from there...

I have to disagree.  Getting the speaker coverage in front of the mics is best but just getting the speakers to where they are aimed at the seating rather than hitting everything but the ceiling indirectly will likely improve the gain before feedback go down, especially for the vocal mics that are shown stage left.

If you mix mono you don't want such extreme overlap from the speakers.  The center aisle helps as that's where the worst interactions would be, but it looks like you could almost get away with one 90 degree horizontal speaker if it is located on or at the short parapet wall.  It appear that it would greatly shoot over the mics.

My concern with a center cluster is that with your constraints it starts becoming the voice from above.  You may be able to get the coverage and even intelligibility, but it may be very disconcerting as far as localization.

Rich, keep in mind that while you can grossly show the nominal speaker coverage with angles such as you have done, the coverage a) is three dimensional, the coverage is based on the actual distance from the listener, which will be a combination of both the horizontal and vertical distances (e.g. the coverage in plan for a 90 degree speaker that is 5' in front of the seating and 25' overhead is that from a speaker 25.5' away from the front row, not one 5' away, and for your 23.5' depth, a speaker 34.3' away for the back row),  b) is not nice clean angles like that, it is usually more ovoid, c) is defined by the -6dB points, it's not as though the sound stops there and different speakers respond differently outside that angle (some may actually go back up in level at certain angles), and d) the coverage is very frequency dependent, it is usually much larger than the nominal angle at lower frequencies and can vary at other frequencies, some speakers even flip the horizontal and vertical pattern at some frequencies.  So the angles such as you show do give a general idea and are very helpful, but don't show an accurate picture.
Title: Re: Yup - there's the problem.
Post by: Rich Stevens on June 17, 2007, 07:56:10 PM
Brad Weber wrote on Sun, 17 June 2007 17:46

...it looks like you could almost get away with one 90 degree horizontal speaker if it is located on or at the short parapet wall.  It appear that it would greatly shoot over the mics.


Do you mean just the stage left speaker, where it's shown on the wall?

Brad Weber wrote on Sun, 17 June 2007 17:46

My concern with a center cluster is that with your constraints it starts becoming the voice from above.  You may be able to get the coverage and even intelligibility, but it may be very disconcerting as far as localization.
Ding! NOAH....

Brad Weber wrote on Sun, 17 June 2007 17:46

Rich, keep in mind that while you can grossly show the nominal speaker coverage with angles such as you have done....

Well, no matter what I do, I try to make sure it's gross.  Seriously, I know all that, but working in 2-D, I had to draw a line somewhere, so why not the "nominal" angles?  I actually did a bit of trig and corrected the angles (at least partially) on the overhead cluster drawing.  I didn't bother on the top-of-wall version since the angles were much shallower and I was in a hurry.
Title: Re: Yup - there's the problem.
Post by: Ira White on June 18, 2007, 09:13:38 AM
Ivan Beaver wrote on Sat, 16 June 2007 18:48

I am still failing to see what differance an eq will make in a room with reflective surfaces vs one that is dead? The only way you can eq a room is with a wrecking ball or bulldozer.


Sorry I was late getting back on that question. I thought that was familiar terminology. EQing the room simply implies that you are having to tune the speakers for all the generated room characteristics, not just the speaker response. In a well designed studio control room for example (a controlled environment which could be compared to a "dead" live sound environment), you don't have to EQ for the room or top-notch studio monitors properly designed and configured for the space.

Conversely, a reflective room is going to be demanding when it comes to tuning out the predominant frequencies that are reflected and enhanced around the room, no matter how accurate the speakers might be. Is this not obvious or am I missing something in your question? Confused
Title: Re: Yup - there's the problem.
Post by: Brad Weber on June 19, 2007, 07:57:44 AM
Rich Stevens wrote on Sun, 17 June 2007 19:56

Seriously, I know all that, but working in 2-D, I had to draw a line somewhere, so why not the "nominal" angles?  I actually did a bit of trig and corrected the angles (at least partially) on the overhead cluster drawing.  I didn't bother on the top-of-wall version since the angles were much shallower and I was in a hurry.

The main concept I was trying to get across is that a simple 2-D view of coverage can be very misleading.  For a simple example, think of a 90 degree horizontal coverage speaker located 5' in front of and 25' above a listener.  If viewed simply in a 2-D plan view the speaker appears to be just 5' away from the listener and thus the coverage from the speaker at that point would be 10' wide.  But that is not what is really happening.  Due to the height of the speaker it is actually 25.5' away from the listener, not 5', and the nominal coverage is 51' wide, not 10'.

So in a situation like would apparently exist for your central cluster where the speaker is high overhead but only a little in front, simply looking at it in terms of the nominal coverage is useful if the speaker is at ear height, but is not real accurate for the actual situation.  Add in the real coverage, the fact that it could differ greatly with frequency, etc. and what you actually get may not at all resemble what is shown by the nominal coverage in 2-D.
Title: Re: Yup - there's the problem.
Post by: Ivan Beaver on June 19, 2007, 08:05:00 AM
Basically I hope you are not talking about applying an eq to the resultant reflection and the damage that it causes (either a bump or a dip) in the freq response at a single mic location.  

For what you "fix" at one location will more than likely cause a greater damage to all the other seats.  You have to do a detailed analysis including looking at the proper phase trace to determine if what you looking at on your software is either a minimum or non minimum phase event.  If it is non minimum you HAVE to simply ignore it-NO MATTER how bad it may look, as it is only existant at that mic position, other positions/seats will have a different response.

It is for this reason that you can't jsut simply go into a room and start whacking away at the eq hoping to fix the problem.  It might "look" better on your screen, but actually sound worse throughout the room.
Title: Re: Yup - there's the problem.
Post by: Ira White on June 19, 2007, 09:26:16 PM
Ivan Beaver wrote on Tue, 19 June 2007 08:05

Basically I hope you are not talking about applying an eq to the resultant reflection and the damage that it causes (either a bump or a dip) in the freq response at a single mic location.


Not at all. Essentially, it was a general response about the potential need for better equalization in a problem room.
On the subject of graphic EQ, I was thinking in terms of any "resonances" created throughout the room primarily from multiple reflections in the low and low-mid range. This can be tackled at least to some degree with precision EQ. Most of the other adverse elements must be addressed in terms of speaker design and placement, room design, boundary treatments, and/or other aspects of tuning.

Thanks for bringing it up so the subject will be clearer.
Title: Re: Yup - there's the problem.
Post by: Scott Helmke (Scodiddly) on June 21, 2007, 01:23:47 PM
Some good info above.  But I'd like to mention that you have the right idea - it really couldn't be any worse than the current setup, so why not try some cheap experiments?  If you can set aside  some off-hours time with the pastor and other folks and test the speaker(s) on top of the wall, on the ground at the sides, etc., you'll probably be able to get a lot of improvement for next to nothing in cost.
Title: Re: Yup - there's the problem.
Post by: Don Lanier on July 06, 2007, 01:26:31 AM
I agree with Josh, Move that speaker out in front of the microphones and your sound will immediatly improve, take it down and put them on speaker stands down on the ground and then work at resetting all your mixer and eq settings. Since its been pulled back and back you now need to reset things. By yanking all the EQs down and turning everything down you system is running with no headroom, no gain and will sound like a cardboard box. Zero everything out and start fresh, one channel at a time and get the band and musicians/singers to help. This means a lengthy soundcheck !!!!

If you dont want these speakers on the ground(stands) buy wall mounts and put the speakers on wallmounts OUT IN FRONT OF EVERY MIC,
Or hire a professional to rig them, these speakers probably arent designed for flying but do have pole mounts for use with stands or wall mounts.

The picture I saw is your classic we dont want to see speakers so lets put them in up there and who cares(Because we dont understand) what the sound guy has to deal with. Youll be amazed how good the system sounds and how little feedback youll have when the Mics are behind the speaker. Trying to visualize 3D drawings and use Polar plots to fix this problem is a waste of time for a layman who is struggling to understand the basic principles of operation let alone propagation or dispersion angles etc....Simple, MOVE THE SPEAKERS. Shocked