Church Loudspeaker Tuning

[Cheyenne Meyer]

This is my first post here and I am new to using REW.

I have inherited the responsibilities of "sound coordinator" at my church and I have tasked myself with finding ways to improve the overall quality of the sound (where possible). I thought an easy place to start would be fixing the main EQ on our system. My contention being that the EQ has been set flat except for the low end which has been tapered off at the lowest frequencies. I have found that most of the engineers that mix on our board have a tendency to eq the signal sources with a boost in the mid to high range. I know we have some "trouble" frequencies that present themselves

I purchased a Dayton reference mic and setup our system to use REW and took the following sample at a seat in the sanctuary which was directly on axis with our line arrays and about half way from the front row of seating.

​

My initial take away:

The high ends definitely roll off precipitously after 1kHz (by almost -14dB by the time it flattens at 12kHz). In fact I was surprised by just how non-linear the system really is. Also, the low end on the crossover has too much gain.


A few notes about our setup:

I'm running REW on a laptop using an ASIO driver to communicate over USB to our digital mixer, an A&H SQ-6. The mains out from the mixer is fed to an analog stereo EQ (Ashly GQX-3102) which is then passed to an analog stereo crossover (a dbx, we used to have a DSP for the line arrays but it recently died on us).

The line arrays are self-powered TCS audio TL1200 cabinets, 5 on the left and 5 on the right. These are 3-way, full range, tri-amped cabinets. They are rated +/- 3dB at ~60Hz to 18kHz. There is a additional single center cabinet fill that is currently unused. Our old mixer was setup for LCR but I've had issues getting it to do anything other than be a source of feedback.

Under the stage are four TCS audio TL1200S subwoofers. These are rate +/- 3dB at ~30Hz to 120Hz. They are powered by four Carvin Audio DCM2000L amps.


The questions for those who are willing to reply:

1. What are your initial thoughts on the frequency response graph?
2. What could be causing the roll off at 1kHz? Is this likely related to the size of the room? Is the reference mic in a bad location?
3. When we replace the temporary dbx crossover with a new DSP, what is the best way to 'EQ' the system? We will likely be going back to having a separate signal feed for the top two cabinets, the middle cabinet, and the bottom two cabinets. Should I test these individually?
4. What other information can I provide that is of use?

Thanks for any replies in advance!

 

[Wayne A. Pflughaupt]

Certainly is a problematic response curve. Typically room response in an auditorium is relatively flat, with only a slight house curve.

I assume you made sure the channel on the console you plugged the computer into were set for flat? What are you using as an interface for the measurement mic?

When playing a familiar recording through the system, with the channel EQ set flat, does it sound like the graph looks? If it sounds reasonably like it should, then something is amiss in your measurement system.

Is your calibration file for direct orientation (0 degrees), or 90-degrees? I’d recommend pointing the mic at the speaker array with a 0-degree calibration file.

I’ve never been a fan of separate signal feeds to the house speakers...

Regards,
Wayne

 

[Cheyenne Meyer]

Wayne, thank you for your response.

Yes, the channels used were reset with no processing on them except for gain. The measurement mic was plugged into the console and sent back to the laptop over the ASIO USB interface.

To be honest, except for the boost in the low-end, and a handful of problematic frequencies that cause ringing, I've not noticed that the room sounds like the graph looks. I do note that every one of my mixing engineers has a tendency to correct a couple dB in the 1kHz to 2kHz range and more for the higher frequencies. That makes me question my ears.

I am not sure what orientation my calibration file is for and I could not find information on the calibration file provided by Dayton Audio. It's an EMM-6. I will send them an email for clarification unless someone here can chime in and educate me.

As for multiple signal feeds, I was mostly interested in evening out the SPL in the room. The pastor has asked for a "concert" feel so loud is the word of the day. I'd like to be able to do that without leaving the first row with bloody ears. Also, the top cabinets primarily throw towards a loft area situated behind the production booth. There are reinforcement speakers up there (on a 63ms or so delay) but the gain can only be brought up so much before the sound no longer appears to come from the stage.

I will try to get another measurement in the same location with the mic pointed at the line array.

Thanks again!

Cheyenne

 

[Wayne A. Pflughaupt]

The measurement mic was plugged into the console and sent back to the laptop over the ASIO USB interface.

Your connection scheme is not correct. The mic should be plugged into the audio interface (presumably you’re using one with a mic pre-amp and phantom power). The interface is connected to the computer via USB,,then the USB output connected to the console. That’s the only way the calibration file is going to work.

I am not sure what orientation my calibration file is for and I could not find information on the calibration file provided by Dayton Audio. It's an EMM-6.

I’m confident that the file is for direct-orientation. If there is only one file supplied, that is what it will be.

To be honest, except for the boost in the low-end, and a handful of problematic frequencies that cause ringing, I've not noticed that the room sounds like the graph looks. I do note that every one of my mixing engineers has a tendency to correct a couple dB in the 1kHz to 2kHz range and more for the higher frequencies. That makes me question my ears.

You can’t really tell from the live sound. As you noted, the sound guys are adding compensation, and there may be additional compensation coming from the musicians’ gear from the stage. A familiar recording (with the console channels EQ’d flat) would be the correct way to judge. If it has overpowering bass and the highs make you go, “I could have sworn the drum kit had cymbals,” then yes it sounds as bad as it measures. I expect that you’re going to find that a good recording will sound worse than the live sound.

I will try to get another measurement in the same location with the mic pointed at the line array.

That should improve things on the new graph above ~ 2 kHz.

Regards,
Wayne

 

[Cheyenne Meyer]

My connection scheme is no different then if I was plugging the mic into an interface such as a Scarlett 2i2. My interface simply happens to have 32 ins and 32 outs (as seen by the USB ASIO driver) and is attached to a beefy digital mixing console. I have phantom power on the mic (which is available on any channel should I so choose). The interface runs at 24bit/96kHz.

My comments on the sound quality were what I think about the system when I run pre-mastered, familiar content through it. My fear is that I'm too used to how the system (including the room) "colors" the sound passing through it.

I think you are right about my measurement angle being incorrect. I'll repost here with an update as soon as I can.

Cheyenne

 

[Matthew J Poes]

My connection scheme is no different then if I was plugging the mic into an interface such as a Scarlett 2i2. My interface simply happens to have 32 ins and 32 outs (as seen by the USB ASIO driver) and is attached to a beefy digital mixing console. I have phantom power on the mic (which is available on any channel should I so choose). The interface runs at 24bit/96kHz.

My comments on the sound quality were what I think about the system when I run pre-mastered, familiar content through it. My fear is that I'm too used to how the system (including the room) "colors" the sound passing through it.

I think you are right about my measurement angle being incorrect. I'll repost here with an update as soon as I can.

Cheyenne

You might try an orientation of 45 degrees to see if that helps.

I’ve never done live sound tuning of this scale before so I can’t comment. Acoustically the room should have far fewer early reflections that combine with the steady state response. That should lead to a flatter response if the speakers have a naturally flat response.

Speakers like you are using are quite directional and their coverage angle can create dead spots at high frequencies. I would imagine aiming them becomes pretty critical to ensure their high frequency response isn’t being artificially rolled off relative to the mid due to the coverage.

I would also take a lot more measurements throughout the room. I have done some commercial cinema setup work and we always measure at a handful of locations. This helps us be sure that everything is aimed right and the response is consistent over various locations.

You could try flattening the response out by turning down the bass and adding a shelf filter for the treble. If that measures better, listen to some familiar material and see. It might sound more neutral or it might sound excessively bright. That might help you get your bearings for the final tuning. My guess is that this slope you have is actually preferable. Often live venues with a flat response can sound ear peircing.

http://assets.madebydelta.com/docs/senselab/publications/Live-concert-sound-quality.pdf

Looking at this article it is clear that none of the venues have a flat response. Many have a pretty tilted response which seems to be noted as preferable.

@DanDan does more of this for a living and may have more to say on the matter. When it comes to large venues i would defer to him.

 

[Tony V.]

Live sound in a church is my specialty, Im going to come at this from a very different angle. I agree with the above regarding REW but looking at your room you have a lot of live surfaces in there (a sound mans worst nightmare) I am guessing that your stage is also very live meaning even without the mains on at all you get a lot of sound from the stage? with a db meter what do you get just from the stage as the band/worship team is playing with mains off at the center of the room?

 

[Wayne A. Pflughaupt]

My connection scheme is no different then if I was plugging the mic into an interface such as a Scarlett 2i2. My interface simply happens to have 32 ins and 32 outs (as seen by the USB ASIO driver) and is attached to a beefy digital mixing console. I have phantom power on the mic (which is available on any channel should I so choose). The interface runs at 24bit/96kHz.

Sorry for misunderstanding about your connection scheme. I never imagined REW would recognize a digital mixing console, what with all the reports in years past of people having trouble getting fancy multi-channel interfaces to work.

My comments on the sound quality were what I think about the system when I run pre-mastered, familiar content through it.

Good to know.

The manual for your Ashly GQX-3012 (excellent equalizer, by the way) notes on p.8 under the section Large Room Equalization: ”As sound travels long distances through the air, high frequencies are attenuated more than low frequencies. In general, large rooms benefit from some low frequency roll-off [and] high frequency boost...”

That’s been my experience as well. The first time I tried to EQ the system in a medium-sized church sanctuary, I immediately dialed in the same room curve I had with my home system, and it sounded severely bass-heavy. I ended up with fairly flat response, with a slight boost below I think 50-60 Hz before it sounded right. (I’m at a loss to explain the graphs on the paper Matt linked – most of them look totally insane to me for large venues.)

From the perception you have of the way the system sounds with your reference recording, it appears that it certainly could use some adjustment. I assume this is an active multi-way system, so I’d suggest bypassing the EQ for starters, and turning up gains for the the high freq horns up a few dB, and the subs down a few dB. Keep tweaking the highs and subs levels with your reference material until you get things sounding reasonably balanced.

You can then fine-tune things with the equalizer from that point, and after that tweak based on live sound, if needed. If you find yourself constantly adjusting the highs and/or lows of the various channel inputs, that would indicate the speaker gains need further adjustment.

Regards,
Wayne

 

[Cheyenne Meyer]

Thank you all for your replies!

@Matthew J Poes

Thank you for the link to the paper. I found it's contents to be quite enlightening. I do plan on taking a day soon to take a bunch of measurements throughout the room to get a better idea of where the real problem frequencies are at. My OP was mostly made because I hear so much about having a "flat" response in a room, it never occurred to me that a large live venue might need a different point of reference.


@Tony V.

You are correct, the sanctuary is almost exclusively hard flat surfaces. They even removed the carpet (which to be honest was ready to be replaced). I do have to remind my engineers that it wont sound the same during rehearsal than it will when the room is full of people, which leads to my next question...as I go about measuring our system at different locations, how valid is this information going to be compared to during a service?

A couple years ago we moved to a silent stage setup so all the performers have an IEM that they can adjust with their own mix. We used to have a sound-proof amp cabinet behind the stage but the high quality of amp modeling units available these days made that more hassle than it was worth. Stage noise is essentially drums and vocalists.

You can get a better idea from some of the pictures on our webpage and facebook page: http://discoverychurchelgin.com/


@Wayne A. Pflughaupt

No apology necessary. I'm glad my lack of experience with REW has helped to keep me from being dissuaded in pursuing this particular setup. I do have a Scarlett 2i4 but it is strapped up in the lighting trusses being used as an audio interface for a Mevo camera. As a programmer I'm actually surprised that anyone would have an issue using any ASIO interface unless the driver was very poorly written.

Also, I can confirm that you are indeed correct. Dayton Audio replied and said the calibration file for the EMM-6 was created for 0° measurement.

The array cabinets themselves are indeed active, but curiously the built-in circuitry does not contain any external method of adjusting individual speakers or even by groups. See the picture below...

The only adjustments I can make are on a L/R EQ basis and a crossover for the subs. As I mentioned previously, we used to have a DSP (an Ashly 4.8SP) that I am still looking to get replaced which was far superior than my $150 dbx crossover that I have temporarily patched in. My first concern when I saw what the frequency response in the room was like, was that my crossover was creating an issue at higher frequencies. Given Matt's link I am glad that at least appears not to be the case.

Cheyenne

 

[Tony V.]

You are correct, the sanctuary is almost exclusively hard flat surfaces. They even removed the carpet (which to be honest was ready to be replaced). I do have to remind my engineers that it wont sound the same during rehearsal than it will when the room is full of people, which leads to my next question...as I go about measuring our system at different locations, how valid is this information going to be compared to during a service?

A couple years ago we moved to a silent stage setup so all the performers have an IEM that they can adjust with their own mix. We used to have a sound-proof amp cabinet behind the stage but the high quality of amp modeling units available these days made that more hassle than it was worth. Stage noise is essentially drums and vocalists.

You can get a better idea from some of the pictures on our webpage and facebook page: http://discoverychurchelgin.com/

Hi Cheyenne, Beautiful building by the way.
In ear certainly helps and thats a good step in the right direction. As you noted all those hard surfaces does mean alot of reflection and even your chairs are not absorbent it seems. A full room of people is certianly going to change your readings as well
I dont think eq is the biggest issue here. I think acoustics needs to be addressed and I know thats not a simple answer but really is the key part of the problem. I would start with building some large acoustic panels simply by building a 2x4 frame, fill with insulation and cover with burlap. Hang those above the seating at a 45degree angle. Carpet on the floor would be a big help also.
I would also build some decorative acoustic panels and place them on the front wall behind the band/worship team.

I hope that gets you started for now.

 

[Cheyenne Meyer]

Tony,

We do love our building, it is definitely part of the character of our church. As you can probably guess it's a fairly old building, it was originally opened as a silent movie house in Dec. 1920. I've actually been contemplating the topic of acoustic treatments. In some respects it's a hard topic, in part because everybody loves the beautiful brick walls and stone columns. I think the head of production is on board with the idea though so I'll keep my fingers crossed.

I'm assuming you mean DIY insulated panels such as these:

Cheyenne

 

[Tony V.]

Yes, exactly like those.

You don't need to cover every surface just strategically placed. It would give you the most bang for buck in solving your sound issues. Panels suspended from the ceiling would also really help. Just don't hang them flat they need to be at angles.

 

[Oldmics]

Cheyenne

I am an owner of the TCS- TL1200 passive versions and am also familiar with the self powered units that you have.

I acquired my self powered units probably from the installer/vender of your system.
I personally am not a fan of any self powered units but thats a conversation for another place and time.
My goal here is to help you along.I can only address the concerns of the top TL1200 boxs

The way to correct the imbalance in the output devices is to go the amp manufactor and download the software to gain access to the control parameters necessary to accomplish your desired results.

Your amps are from Digmoda and your amps are the DDC552 units 3 way in the top boxs
Link for Digmoda is below

http://www.digmoda.com/

You should consider using the newest test noise source from Meyer Sound Labs. It is called M-NOISE.
The short explanation of this newest test proceedure is that it takes into account the crest factor of live music that has not properly been addressed by pink noise tests.

Link to M-NOISE is below

https://m-noise.org/

Simply adjusting your speaker processer (whatever it may be) is only going to allow for equalisition.This wont allow to properly utilise full capacity of the individual componet schemes availiable in the cabinets.
Once the system has been optimised for all three frequency bands then your measuring can begin.

This endeavor is somewhat advanced and if you have not done it before you may want to call in the pros in your area who installed the system. I can put you in touch with them if you like.

My system is setup as a passive system using a Dolby/Lake LP4D12 and good old heavy Crest amplifiers.I did go thru a bunch of speaker processers before getting the results that I wanted with the Dolby/Lake.
I"m way critical and I will say this is a good sounding rig.
TCS never got there respect and were short lived in the Pro marketplace.
Most of the guys out there who bashed the TCS1200 system are not skilled enough to setup a crossover on a system from scratch in this plug and play world.

Fire back if you have any questions. Good luck!
Oldmics

 

[Cheyenne Meyer]

Oldmics,

Thank you for the extremely helpful info. I was somewhat flummoxed by the lack of information on the DSP that was on the cabinets themselves so this is very good info to have. I have downloaded the software as well as the m-Noise test source. Having read the procedure, i agree, it is somewhat advanced. I'm happy to give it a shot though I will likely have questions. I would be happy to take info on any pros in my area who have experience with these, feel free to leave me a message and thanks again!

Cheyenne

 

[Matthew J Poes]

Cheyenne

I am an owner of the TCS- TL1200 passive versions and am also familiar with the self powered units that you have.

I acquired my self powered units probably from the installer/vender of your system.
I personally am not a fan of any self powered units but thats a conversation for another place and time.
My goal here is to help you along.I can only address the concerns of the top TL1200 boxs

The way to correct the imbalance in the output devices is to go the amp manufactor and download the software to gain access to the control parameters necessary to accomplish your desired results.

Your amps are from Digmoda and your amps are the DDC552 units 3 way in the top boxs
Link for Digmoda is below

http://www.digmoda.com/

You should consider using the newest test noise source from Meyer Sound Labs. It is called M-NOISE.
The short explanation of this newest test proceedure is that it takes into account the crest factor of live music that has not properly been addressed by pink noise tests.

Link to M-NOISE is below

https://m-noise.org/

Simply adjusting your speaker processer (whatever it may be) is only going to allow for equalisition.This wont allow to properly utilise full capacity of the individual componet schemes availiable in the cabinets.
Once the system has been optimised for all three frequency bands then your measuring can begin.

This endeavor is somewhat advanced and if you have not done it before you may want to call in the pros in your area who installed the system. I can put you in touch with them if you like.

My system is setup as a passive system using a Dolby/Lake LP4D12 and good old heavy Crest amplifiers.I did go thru a bunch of speaker processers before getting the results that I wanted with the Dolby/Lake.
I"m way critical and I will say this is a good sounding rig.
TCS never got there respect and were short lived in the Pro marketplace.
Most of the guys out there who bashed the TCS1200 system are not skilled enough to setup a crossover on a system from scratch in this plug and play world.

Fire back if you have any questions. Good luck!
Oldmics

M noise would only be used for measuring maximum SPL, like setting limiters. Not for EQ or any other form of optimization (e.g. phase/time alignment) that is necessary. Its a great test signal, but it serves a very limited purpose in the overall setup process.

 

[Oldmics]

M noise would only be used for measuring maximum SPL, like setting limiters. Not for EQ or any other form of optimization (e.g. phase/time alignment) that is necessary. Its a great test signal, but it serves a very limited purpose in the overall setup process.

From the Meyer Sound - M-Noise F.A.Q.s page

Can I use the M‑Noise wav file in place of pink noise in my current workflow to tune systems in the field?

Yes, M‑Noise can be used with any source independent two channel analyzer. M‑Noise is advantageous when measuring the Maximum Linear Peak SPL but yields the same frequency response at lower SPLs as pink noise and music.

Oldmics

 

[John Mulcahy]

Yes, M‑Noise can be used with any source independent two channel analyzer.

That means it requires a dual channel FFT system, such as Smaart. You can use almost any signal that has decent bandwidth with dual channel systems, but those signals can't be used on their own as they don't have a uniform spectrum. M-Noise isn't suitable as a replacement for random or periodic pink noise in single channel systems or with RTAs. See image below, from a paper on the m-noise site, showing how the M-Noise signal rolls off from 500 Hz.

 

[Oldmics]

John

I see your point and agree.

Myself not being familiar with your REW creation I just assumed (we all know what happens when) REW was a dual device.

After downloading your program I understand the intended application and operation. I commend you for creating such a user friendly tool.

I especially like the dual freq output feature.

As you have mentioned SMAART has the dual channel approach which I have found to be invaluable in my testing.
SMAART and Iasys (for its power compression testing) have been my go to tools for system settings.

As an owner of large scale systems , my most feared concern is overdriving of a system.

It seems in todays world that large amounts of SPL somehow equals talent.
Even though we hang enough boxs to do the job , there has been more than one example of the technicians driving the F.O.H. to uncomfortable levels.

E.G. one outside show was at 115 db with 118 peaks at a 175 foot mix position.

The use of the M-Noise allows me to safely maximise the availiable db out of a rig as opposed to pink noise settings.

Not all shows are this way,but in an attempt to prevent expensive repairs I will use M-Noise with the dual FFT technology rather than a single channel analyzer.

Oldmics