Rational Acoustics

Kip Conner
January 30th, 2011, 05:28 PM
I did a tow night show over the weekend which allowed me to run dual TF's. One pair was for the house and the other pair was for the center monitor. Although this isn't a room that I work in all the time, I have worked in here enough times that I have saved my EQ's and just recall them on the console. The monitor is mine and I have it eq's for loud, soft spoken vocals.

What I noticed in running simultaneous TF's is that the monitor was almost a perfect 180 degrees out of phase up to about 1kHz. (the monitor has been tuned for relative phase in the acoustic crossover region and the Xover point is 1kHz). Since the lower frequencies are more omnidirectional I choose to flip the polarity of the entire signal to align those frequencies. The room is small enough that sometimes I can pull down the acoustic di in the house and not miss it until I'm at -20dBu.

This room in question is a small 50x25 with the stage sticking out of the long room and there is seating on three side of the stage. Directly in front of the stage are a left/right system and I set-up a mono feed to the sides. There is a sub that lives against the wall behind the live mics. but only on side.

I guest engineer in the room when the most demanding artists are booked and it requires me to bring a better console and monitor system.http://www.athenssound.com/Crimson%20Moon%20TF%27s.jpg

Arthur Skudra
January 30th, 2011, 06:51 PM
Hi Kip,

I'm a bit confused, can you describe your measurement setup a bit more and how you captured these traces?


Kip Conner
January 30th, 2011, 11:32 PM
It's two separate TF measuring the PA (in blue) and the monitor (green). The mic that was resposible for the monitor was taped to the mic stand facing the monitor. The TF for the room was placed a little off center of the listening area.

The original measurement were done independently and treated as two separate systems. Since it's such a small performance space there's a lot of room for interaction between the two systems merging them into one. The thing that even got me thinking about it was a performer told me it sounded "funny" and it apparently had "always been this way here" on the stage specifically in the bottom end.

Of course there are several items at play- the main two having a sub behind you radiating at your feet and having your head so close to the flown passive mid-high cabinets. The performer even went as far as saying that he felt like he was in a comb filter.

Without getting into the science of comb filtering and it's spacing as we climb through the octaves I just ran each system together saw the potentiol for cancellations in the area that the performer was describing.

During the show, the TF's were getting clones of their respected feeds so the TF reference for the Monitor was getting that signal and not the FOH feed. The image I posted is live interaction during the show- not the set-up. I didn't capture the trace of the monitor feed because of time and it's easy to recognize two signals that are out 180 by the spacing in the phase trace.

The question is whether or not it's correct to flip the polarity to align both systems in a certain range knowing that you can't make it consistent throughout the entire freq response. My thought is yes since these frequencies are that the ones interact the most between the two systems.

January 31st, 2011, 04:33 AM
Hi Kip,

Maybe you could post your traces again separately with phase and magnitude smoothing set at 1/48th octave and coherence displayed. At 1/6th octave you arguably won’t be able to see any comb filtering and you might not see a phase wrap-around. Based on your story it’s hard to tell if your monitor and P.A. are in or out of phase since your measuring at two different positions in the venue. In order to tell if both systems are in phase, you would have to compare their respective individual transfer functions at the same position. The combined result and transfer function, depending on the speaker’s relative level and arrival time, should show addition when in phase or cancelation when out of phase. Your conclusion regarding the polarity however, is plausible due to the increasingly “near” omni-directional behavior of the speakers at these and lower frequencies. AFAIK polarity doesn’t change in any direction, assuming conventional speakers, regardless of the angle facing the speaker.



Kip Conner
January 31st, 2011, 08:03 AM
Yes, very true. and even where you measured in the venue.

I think what I am really wondering if given the choice align the phase response of the systems would you do it? I think that my situation is base on time and making the best choices for time. On an average day for me I am given less than 2 hours to correct as many things as I can and conduct a sound check for an acoustic trio that is really dependent on monitors. Having a console on the road helps.

Before I had some basic understanding of phase response and the importance of driver interactions in the acoustic xover I used to to do a physical alignment by picking a point and measuring the impulse repsponse of the PA and Comparing with the monitors. Almost like a point source situation, although I'm really pushing the terminology.

So I'm always looking for the best use of my time and deciding which battles to fight. If I don't have time to take several samples in the room, I'm trying to take the two things that I spend the most time (single point FOH and Mon) and make the best assumptions that I can with the little time that I have to work.

January 31st, 2011, 12:47 PM
Hi Kip,

IMHO I’d might consider, taking time to phase align monitors to P.A. if they would receive identical feeds reproduced at equal loudness. Only then the interaction is most severe. The logical point of alignment for me, would be ear height of the artist I guess, since he/she is the only one who has benefit of this alignment. However the down side is considerable. You can’t bring the P.A. closer to your artist for evident reasons, so you’ll have to delay the monitors. If you take into account that the low’s often lag behind the mid and high’s in both P.A. and monitors you might end up with several decades of milliseconds of delay. This could become unpractical. Any differences between the two feeds due to i.e. dynamics and/or equalizing alters the relationship in such way that it might become useless anyway. AFAIK in most situations the bulk of LF comes from the mains, meant to cross several dozen feet and more across the room and only a fraction of that distance to the artist. If you add the power alley effect, assuming you’re standing in the middle and not using a cardioids, you might end up with a P.A. that easily trumps your monitors in this part of the spectrum. This also raises the question whether these frequencies are needed in a monitor mix, besides drum, bass guitar fills etc. where LF timing becomes equally important as intelligibility and intonation. I tend to listen and setup my monitors with the mains turned on. More so because one talented monitor engineer, who’s name I regrettably have forgotten, once said that good monitor mixing is completing-the-picture at the rear side of the P.A.. A metaphor that makes a lot of sense to me. Also it cleans up my mix, less spill and plosives near the front of the stage and my monitor level is overall lower. With two hours to spend I might put the majority of my effort in adjusting the P.A., assuring myself and the audience of a “fair” frame of reference by means of measuring. This way I like to think what I'm to expect of my vocal microphone. With the mains turned on, using my own voice I’d adjust the monitors by ear to taste. Making sure that mains and monitors are in phase/polarity is something I’d could also check back in shop.



Kip Conner
January 31st, 2011, 07:47 PM
it's a great opinion and fairly true to my thoughts, it just happened to be in this case I could flip the polarity of the monitor and get two independent systems working together in the area where they tend to intercede the most. As for monitor level, I almost always leave the house one and bring them up to a blending situation. it gives you the feeling of it being big without isolating someone to the point that they need to ask for more than is necessary to make up for the fact they are miles ahead of where they should be in the big picture.

thanks for the input.

February 1st, 2011, 03:50 AM
Hi Kip,

Youíre welcome. Itís nice to exchange thoughts. Being fully aware that everyone has and is entitled to their own opinion, I find it always difficult to formulate them without appearing schoolmarmish.



Kip Conner
February 2nd, 2011, 10:14 AM
It's why I don't go to the Live Audio Board to post anything, I find that people that hang around here know what they are talking about or are here to learn something in a non-confrontational manner.

I look forward to more idea exchanging.

Arthur Skudra
February 2nd, 2011, 11:14 AM
Cool stuff! I would add that if you can do a multi mic setup, I would put one mic on the stage at the talent's position, one further upstage listening to another monitor, one at the front of the audience, and measure the effects of flipping the polarity of the stage monitors in all positions. Sometimes when you solve the problem for one position, you create another elsewhere, and depending on the off axis response of the monitors, that might be a possibility here, even a remote one.

Of course you could suggest your talent uses in-ear monitors!! :D

Calvert Dayton
February 3rd, 2011, 04:44 PM
I'd definitely like to see this with less smoothing on it. Maybe you could just post TRF files of the two traces instead of a screen shot. (I think that pretty much everyone here has Smaart.)

Also, what exactly are the mains? Are they 3-way boxes?


Brad Harris
February 4th, 2011, 11:59 AM
Of course there are several items at play- the main two having a sub behind you radiating at your feet and having your head so close to the flown passive mid-high cabinets. The performer even went as far as saying that he felt like he was in a comb filter.

As primarily a monitor guy, I deal with this every day. Some venues are better (large outdoor venues with no direct reflections back to the stage) to 'cool' looking venues with everything coming back on stage before (it seems) to the audience.

I've usually always treated this with delay (on the monitor), and a little complimentary EQ.

In fact the last gig I did, I've been thinking about seeing if I can measure what it is that I'm hearing when I'm adjusting the delay, and what it is that I'm looking for on the screen to make it a quicker process.

But thats going to have to wait a few weeks as I'm out on tour and heading home for a late christmas.

February 10th, 2011, 09:33 PM

If I understand the conversation to date, I have questions about the validity of the posted measurement.
The system engineer has a responsibility to understand the frequency and phase characteristics of the main system for the benefit of the audience. The measurements taken in support of that understanding are made from the audience area. Point the speakers at the audience, make it sound good. Easy enough.

The monitor engineer has a similar situation in that his or her speakers point at his or her audience and the audience is the artist. Measurements in support of this study take place on the stage.

Understanding how the mains behave on the deck dictate that measurement be done onstage. A highly directional and well behaved speaker system sounds great in the audience area where it's aimed to cover. How it behaves on the stage is different and does not correlate with its behavior in the intended coverage zone. Conversely, making judgements about the house sound from the confines of the stage is idiotic. Off axis is off axis despite the number of Grammys the artist may or may not have won.

To understand the behavior of the main speaker system as it exists on stage, measure there. To understand the behavior of the monitor system on the deck, measure there. Most importantly, to measure the possible interaction between the two measure at the position of interest. In this case, the artist's home microphone position.
What we understand about comb filtering is that where equal levels of the same signal exists but with different arrival times is the prime spot for interference. The star vocal exists in both the mains and monitor mix. The level difference between the two will fluctuate some because of active mixing, different compression settings, and system non linearities, but let's assume a general parallel relationship. First criteria met.
Next, at the star's home vocal position that signal will be arriving at two different times. I think that time relationship is more stable than the level relationship. Criteria two met.

If you can make the monitor system interact constructively with the mains system FROM THE PERSPECTIVE of the artist, you will have achieved the goal. In order to do so, I'm suggesting that the systems engineer needs to grab control of the mains from his perspective, the monitor engineer needs to grab control of his rig from his perspective, AND THEN make time, amplitude, and frequency changes to make it behave more like the mains. (Left unsaid is the whole strategy of noise cancellation, or wanting to negate the behavior of the mains on the stage)

Hopefully eloquent, not schoolmarmish


Kip Conner
February 11th, 2011, 03:53 PM
Agreed. But we're not talking about a million dollar PA. We're not even talking about a 100k dollar PA- we would be talking about a less 10k dollar PA if I didn't have to supplement the room with personal gear for the slightly more demanding artists.

We're talking about a very small listening room where there is no "system engineer" and all is done from a console that I have t bring to meet the demands of the artist. It's a small enough room where the flown cabinets can almost be touched by the artist without any trouble and the audience can touch the monitor without stress. I mean touch in the physical fashion, not the sense of feeling :) The stage is 8x8 with passive JBL cabinets on all sides facing out in a 240 degree listening area. If you're in the front row of stage right you get more stage monitor than the actual fill hung to cover that zone.

So because of the close proximity to all drivers involved the PA can seem "dark" because of the less directional frequencies from the monitor and vice versa for the performer. In this room the interaction is huge up to 160 cycles- it can easily drive either system to sound "muddy" for either audience (the audience or the performer)

So, knowing that the interaction is what it is (in this case)- is it worth fixing?

Although my main artist has never one any Grammy's (we don't put stock in an organization that gives awards for auto-tuned records) he does like arena volumes, does not like the separation that ears create, does not pay me enough to mess with extra microphones to feed IEM's and like to play rooms with 300 seats or less. There are a ton of paradoxes invloves and so I have to just pick my battles.

I unfortunately deleted the files from this day but I'm scheduled to go back in there and do a better job with keeping up with the TF files and will even take a few pictures to give you an idea of the intimacy.

February 11th, 2011, 07:33 PM
You may have missed the intended sarcasm of the Grammy reference. I was speaking from experience there.
Your situation is interesting and full of paradoxes. Picking your battles wisely is key.
My shorter answer this time is "Yes!" Fixing the problem is worthwhile. First, let's define it. Is the problem that the monitor system fucks up the house or that the house fucks up the monitors? You can't fix both even though I'd guess both are problematic. Keeping you star comfortable is important and in a compact environment like what you describe, the interactions between house and monitors can be immense. If he's happy with the choice you're off the hook. Tilt things in his favor where you can. ie timing issues, tonal issues


February 12th, 2011, 06:42 PM

Next time you're in that venue would you make a measurement or two for us?
Specifically, show us a TF of the main system from a typical seating locaation in the house when you're happy with the PA.
Show us a TF of the monitor for your singer from the star's home position when you're happy with its tonality.
Thirdly, and MOST importantly, with both the house and monitor on, and from the same location as measurement #2, show us a TF with pink noise feeding both mains and monitors. I'm trying to get a feel for what the interaction is like. We would have the same signal arriving from two different sources at different times. I'm guessing you will visually verify your singer's observation of "holes in the sound" and the low frequency buildup so many sources will inevitably contribute to.
It doesn't matter if the PA cost a million or a hundred bucks; the test is the same, and for the sake of the test, you can serve as both monitor and system engineers despite your pay grade.

Happy hunting,