Rational Acoustics



Dr. J
June 22nd, 2011, 11:24 AM
Last night I finally had the chance to do some fooling around with my system and noticed some things that really made me feel like I know nothing!

First off - from the Amp attenuators OR XO Output topic -- I started with my gain structure. There is a tiny little TRIM knob on the back of my mixer that apparently controls the input signal into my Driverack. I usually have it backed down a bit but the thing keeps getting bumped and moved every gig and I don't notice it until the DR starts clipping or the overall volume doesn't seem like it supposed to be.

The company that made the board says that knob needs to stay wide open BUT my DR won't allow it. Why do they put knobs on things if they are meant to stay wide open? Makes no sense at all.

I decided to leave it wide open and put tape over it so it doesn't get moved anymore. This really increases the signal into the DR. The DR INPUTS are way hotter than the outputs because of the signal being split.

I had to turn the INPUTS down by -9dB which actually made it read +6dB on the input meter and that put it right into balance with my Top box. The sub OUT meters were so far down that I increased the XO output gain until it read in balance with the others.

Now with my board approaching clipping -- my DR Inputs are in balance with the Outputs and they track together pretty good never reaching the clip point in the DR.

Next, I knew to not dare take the system out to a show and think it was all going to be a great experience so after finding my NEW clip points on my amps (which did change all of them) I fired up Smaart to take a look.

Sure enough my trace was all jacked up! I knew it would be so I started the whole tuning thing over again removing ALL EQ.....etc... started with a clean slate.

I used DaveG's advice and used the only 4 post PEQ's I had and went for driver correction. With a Full Range Top -- that isn't much. I placed the top up pretty dang high so I could avoid as many reflections as I could. This was also out in my backyard (huge) in the grass. I first knocked down the peaks and was left with nothing but dips.

I am always curious as to what dips get EQ'd or not so I started moving the mic around a bit OFFAX in the horizontal and vertical plane. I turned out some dips needed Eqing and others didn't so I left them.

I got the TOP box done to a point where I felt like it was the best tuning session I ever done. I even took the mic (for fun) and walked it backwards ONAX and everything above 1K NEVER budged. I even went back 30 feet. (Community tunes their speakers at 39 feet out I believe)

Next, I knew I couldn't leave the box up in the air like that at a show because it wouldn't fit in some places due to the ceiling -- I set it back down on top of two ground stacked subs. This is how I run it in clubs because we have to get in and get out very quickly.

ALL I did was lower the Top box down and set it on my sub boxes and suddenly it looked like I hadn't tuned the thing at all! Well...... everything from 650Hz on up was fine BUT the lower frequencies changed. Some of those DIPS I filled in the air NOW were showing boosts when the box got closer to the ground. They were actually boosted above 0dB on the magnitude trace. There were some cancellation dips as well.

I went right back and put the box back up high again -- it all goes away. Lower it back down like I would have to do at a gig -- trace goes to H. Keep in mind here that I did move the mic in respect to the box position -- placing it between the horn and the mid. So NO OFFAX readings.

So How should I handle this? It seems to me I am dealing with the fact that I will never get it right for more than one location. If it would change like this in the free field with grass -- then I could only imagine what it would look like inside an acoustically horrible club.

I don't know how using an anechoic chamber or even free field tuning helps when everything changes in the wild.

Should I have tuned the TOP box on the sub exactly the way I would be using it at a gig? OR do I IGNORE those changes because I know it is good when the box is up high enough from surfaces?

What I ended up doing because I was stumped -- I only took down the peaks and left the dips. It pretty much was the exact same frequencies that had dips up in the air that now had peaks closer to the ground.

I am assuming when you measure in a venue that you probably only take down Peaks and not fool with dips much. Would that be correct?

Oh -- to top this discussion off and made me quit for the night was after I finished the tuning -- I grabbed one of my other TOPS (supposed to be Identical) and did a swap. Fired up the trace........... I was hoping for the exact same result...... NOPE -- the whole dang thing was jacked!

Now I feel like I know nothing. Two nice tops that are identical that read different.

What do I have to do...... run thru two different tuning sessions for the left and right side of the PA -- mark everything as such? Geesh! :D

lundbergsound
June 23rd, 2011, 04:12 AM
I don't know how using an anechoic chamber or even free field tuning helps when everything changes in the wild.

Should I have tuned the TOP box on the sub exactly the way I would be using it at a gig? OR do I IGNORE those changes because I know it is good when the box is up high enough from surfaces?

<snip>

I am assuming when you measure in a venue that you probably only take down Peaks and not fool with dips much. Would that be correct?

The free field / anechoic sort of optimization is meant for an individual loudspeaker or loudspeaker system, i.e., aligning individual drivers and achieving a flat free-field magnitude response. Flat free-field responses are helpful in determining the cause of frequency response anomalies, that is to say, once in the field, you know that the environment or loudspeaker to loudspeaker interactions are the cause of any deviation from flat. In the case of a full range box (either self-powered or with a passive internal crossover), the manufacturer has hopefully taken care of this. In your case, free-field measurements are useful for comparing two boxes that you expect to be the same (and making them the same), and you can also treat the subwoofer and top as a single loudspeaker system if they will always be in the same position relative to each other, and do your phase alignment and EQ for that top and sub system.

Once in the venue, the measurements and adjustments you make should be for loudspeaker to loudspeaker and loudspeaker to environment interactions (some people argue against this, but audience members do hear these interactions...). These are typically LF attenuation due to various forms of loading with surfaces and other loudspeakers, whereas MF and HF are isolated in this regard. You may choose to boost the HF to counteract atmospheric effects; I say that's fine if you have the headroom and care to adjust it as these conditions change.

Regarding EQing dips, in my opinion, it depends on the type of dip. Your coherence trace and/or observing several mic positions will help you determine if the dip can be fixed with EQ or if it is due to acoustic cancellation that you cannot fix. If you have high coherence in a dip across several listening positions, and it seems to respond to EQ, then why not!

Hope that begins to answer some of your questions; I'd be interested to hear others' responses as well.

Good luck!

-Daniel

Dr. J
June 23rd, 2011, 07:37 PM
Thanks for chiming in Daniel. I got to measure again today and it finally hit me why I get different results on occasion. It is because they are different. It IS my Tops.

No matter how meticulous I am with my measuring......... I mean I figure if I do it consistently the same way everytime -- then I am either wrong ALL the time or maybe pretty close most of the time.

ALL of my fullrange boxes are different. My D.A.S. Tops have different magnitude & phase resonses. I have Four monitor wedges I use across the front of the stage...... EVERY single one has its own response even though they are same make and model.... Unbelievable.

I guess this is why I have been racking my brain for so long with all this stuff. I run out of time and quit for the night.... go back out the next day grab (apparently a different cab but same make and model) and can't believe why it would be so different.

I have FANE subs (I guess I should say DIY FANES) I built a bunch of single 18 boxes so I decided to measure those today for consistency.

Every single one had identical magnitude and phase response. They also require very little EQ if at all. They have a nice rounded bump peaking at 70Hz ..... and sound very nice.

I took a look at the crossovers today in my top boxes and one of them has some noticeable capacitor cracking going on. The boxes do get hot at times so maybe there is too much heat going on and cracked one of the caps. It doesn't look good.

The crossover area is where the boxes differ. Below and above --- they track the same. At the crossover -- they differ BIG time. I got to fix that and now is the time to Bi-amp.

Can you tell me a little more about the coherence trace?

I have heard it is my friend.
I have heard it said that it is like a polygraph test to see if it "Believes" a certain area.
I have also heard it said that the coherence trace is the "third leg" of the magnitude & phase combination...

I want to learn more about it but I get confused on some things, for example: I have a flat response going on but at 600Hz -- the coherence is pointing down a bit. The maginitude and phase trace appear to fine, there is nothing within range of the box to reflect with it, there is no wind....etc..

I have noticed it dropping down in places like the lower frequencies 200-400Hz....

I use the coherence trace for setting my level when I measure. when it gets up to +15 -- I don't push the system any louder. If it is windy out -- the coherence trace is all over the place. Makes sense..

What if you have a peak going on that needs to be taken down but the coherence trace is flat as can be over that area? That is where I get confused.

Arthur Skudra
June 24th, 2011, 10:59 PM
Last night I finally had the chance to do some fooling around with my system and noticed some things that really made me feel like I know nothing!
Ah you discovered the paradox that keeps biting me in the rear...The more you learn, the more you realize how little you know…sorry, but you're in too deep now to turn back! :D

I'm curious what mixing console you are using that has a trim output control? Look at the specs, usually manufacturers will publish at what dB level the input or output will clip. Often the constraint is the input clipping level of the DSP, often requiring an external resistor attenuator in order to maintain the best noise floor in your system.


I took a look at the crossovers today in my top boxes and one of them has some noticeable capacitor cracking going on. The boxes do get hot at times so maybe there is too much heat going on and cracked one of the caps. It doesn't look good.

The crossover area is where the boxes differ. Below and above --- they track the same. At the crossover -- they differ BIG time. I got to fix that and now is the time to Bi-amp.
Yep, sounds like you have some passive crossover issues to contend with, sometimes a weak point in many passive designs, especially with age and/or abuse. Even one tired capacitor is enough to throw things off, remember we're talking about lots of interaction in this frequency range. Fix your crossovers, or go with an active crossover.


Can you tell me a little more about the coherence trace?

I have heard it is my friend.
I have heard it said that it is like a polygraph test to see if it "Believes" a certain area.
I have also heard it said that the coherence trace is the "third leg" of the magnitude & phase combination...
Yes, when you have averaging turned on, it compares each "slice of data" to each other in the buffer, and at frequencies where there is inconsistency from one slice to another, you will see lower coherency being displayed. This is often due to random background noise and reflections/cancellations in the measurement. If you see a broad range of poor coherency, it's indicating that your Smaart delay time is off, or that a driver is turned off or doesn't even cover a portion of the audible spectrum.


I want to learn more about it but I get confused on some things, for example: I have a flat response going on but at 600Hz -- the coherence is pointing down a bit. The maginitude and phase trace appear to fine, there is nothing within range of the box to reflect with it, there is no wind....etc..

I have noticed it dropping down in places like the lower frequencies 200-400Hz....Yep, as you go lower in frequency, the window of time is increasing, and room interactions are gaining greater influence in your measurements. Hence why going multichannel and measuring with three to four locations helps even out these position dependent anomalies.


I use the coherence trace for setting my level when I measure. when it gets up to +15 -- I don't push the system any louder. If it is windy out -- the coherence trace is all over the place. Makes sense..
When I measure a system, depending on the noise floor of the space, I will increase the level of the system until I see no improvement in the coherence trace, and usually stop there unless it's a very loud show where measuring at higher levels is justified. If it's windy out, turn on the delay tracker to compensate.


What if you have a peak going on that needs to be taken down but the coherence trace is flat as can be over that area? That is where I get confused.If coherency is good at the frequency where you see the anomaly in the magnitude response, then by all means make corrective action, as you have good confidence in the data at this frequency. It's when the coherency is low (below 20%) at a particular frequency where you should exercise caution and avoid making corrective measures at those particular frequencies.

Here's another clue: when you see the phase curve going downward from left to right at a particular frequency of interest and you see a peak that needs to be flattened at this frequency, more than likely that would be a good place for corrective action.

It's all about context, when making choices on what to do, you want to look at the live impulse response, phase, magnitude and coherence simultaneously so you can make informed decisions.

Dr. J
June 25th, 2011, 04:51 AM
Thanks for the response Arthur -- very helpful. I have the Presonus StudioLive mixer. They put the little trim knobs on them for some reason.

Tonight's show went very well. I initially used a 16 track recording I did a while back to load my mixer down to try & set the gain structure. I have been striving for more of a "Unity" type of mixing strategy where I don't worry about getting everything as hot as I can. The StudioLive is a digital mixer so achieving the hottest signal possible isn't recommended.

Setting the trim knob wide open causes my Driverack to clip at the input fairly easy. I ended up lowering the DR input a bit and raised my sub outputs and this seem to place the inputs and outputs in balance. The high outs were already in a good spot so I matched everything to it.

IN the middle of the show tonight -- I was pushing the system pretty good and checking my meters and I liked the way it looked. The DR inputs versus outputs were all equal and I never saw any clipping whatsoever. The only thing that was different were the knobs on my poweramps. I had to attenuate them a bit. I am ok with that.

Although the system was probably the best I have ever heard it tonight -- I could tell that the tops sounded slightly different from each other.

My tops are still under warranty BUT I may just bi-amp them and move on. If you saw the phase trace -- you would probably say -- move on too. :D

There is some real funky stuff there. If it is nice out next week -- I will take a screenshot for you to look at so I can demonstrate the two Boxes.

Arthur Skudra
June 25th, 2011, 11:04 AM
Setting the trim knob wide open causes my Driverack to clip at the input fairly easy. I ended up lowering the DR input a bit and raised my sub outputs and this seem to place the inputs and outputs in balance. The high outs were already in a good spot so I matched everything to it.Looking at the specs for the Presonus and the Driverack 260, the Presonus is capable of +24 dBu output level, and the DR260 is set at the factory for a maximum of +22 dBu input level, with an option to set some internal jumpers for a +14 or +30 dBu input. I'm wondering if someone had set the jumpers to +14???

From page 73 in the manual:
"The DriveRack 260 gives you the option of changing the input gain level settings. There are 3 hardware configurable gain settings. They are: +14 dBu, +22 dBu and +30 dBu. For these cases, use the following procedure to change the gain level settings. Please be advised however, that once the gain level has been changed from the factory settings, the output meters will no longer be calibrated correctly."

Dr. J
June 25th, 2011, 02:08 PM
Looking at the specs for the Presonus and the Driverack 260, the Presonus is capable of +24 dBu output level, and the DR260 is set at the factory for a maximum of +22 dBu input level, with an option to set some internal jumpers for a +14 or +30 dBu input. I'm wondering if someone had set the jumpers to +14???

From page 73 in the manual:
"The DriveRack 260 gives you the option of changing the input gain level settings. There are 3 hardware configurable gain settings. They are: +14 dBu, +22 dBu and +30 dBu. For these cases, use the following procedure to change the gain level settings. Please be advised however, that once the gain level has been changed from the factory settings, the output meters will no longer be calibrated correctly."

Well Arthur -- you may be right about that. Almost ever piece of gear I have has had some factory goof up going on.... (my tops the latest). I think the internal jumpers on the 260 are under the LID of the unit. I wrote to DBX about this a while back and the response I received states that the default position is +22dB.

They also said this: The default position is +22dB. If your mixer outputs +24 dBu max, I recommend setting the input jumpers to +30dB (this allows for enough headroom in the analog input stage) then applying 6 dB of gain in the input mixers (this is applied in DSP or the digital stage). This should make the DriveRack’s inputs (in the digital stage) and mixer clip at the same point and still allow for enough headroom in the analog input circuit.

But like you pointed out -- the meters may not be calibrated any longer according to the manual.

Alright -- I will pull the driverack out and open it up. I will be disappointed if this is messed up BUT at the sane time if the jumper is set to +14 -- then that may explain why I have trouble getting this set correctly.

In McCarthy's book -- he has a whole chapter dedicated to "Verification" & "Examination"...... never knew it ran this deep. :D

Thanks again Arthur -- I will get back to you on the results...

sweepableQ
June 26th, 2011, 10:45 AM
Hello all.

Here is a link to Bob McCarthy's blog. He has several great question and answer things going on surrounding coherence and impulse response. I can't tell you how much these have helped me personally.

http://bobmccarthy.wordpress.com/2009/12/ (the top blog entry on this page is about coherence)

http://bobmccarthy.wordpress.com/2010/02/ (phase alignment of sub blog deals with impulse etc...)

What I have found in my experience is that the nicer the speaker systems the better the tolerances are. Some pretty big differences in my JBL, EAW boxes. My Meyer boxes are much, much closer to each other!!!

Hope this helps.

MQ

Dr. J
June 26th, 2011, 11:27 AM
Hello all.

Here is a link to Bob McCarthy's blog. He has several great question and answer things going on surrounding coherence and impulse response. I can't tell you how much these have helped me personally.

http://bobmccarthy.wordpress.com/2009/12/ (the top blog entry on this page is about coherence)

http://bobmccarthy.wordpress.com/2010/02/ (phase alignment of sub blog deals with impulse etc...)

What I have found in my experience is that the nicer the speaker systems the better the tolerances are. Some pretty big differences in my JBL, EAW boxes. My Meyer boxes are much, much closer to each other!!!

Hope this helps.

MQ

Thanks for the info Q -- I know this takes time to grasp & interpret. I have never really altered any of my configuration settings in Smaart while measuring BUT do you find that you use certain settings for measuring your tops versus your subs?

Do you ever change your coherence settings depending on what you are measuring?

sweepableQ
June 26th, 2011, 08:04 PM
Thanks for the info Q -- I know this takes time to grasp & interpret. I have never really altered any of my configuration settings in Smaart while measuring BUT do you find that you use certain settings for measuring your tops versus your subs?

Do you ever change your coherence settings depending on what you are measuring?

I don't usually change the coherence threshold etc...I do change my expectations though, and I work hard to explain the dips in coherence. This is to make sure you are applying the correct cure for the ailment. Spending time measuring a two way enclosure and just simply raising and lowering the box and mic relative to the ground, and then measuring the offset, you can then more easily start to weed out the floor bounce on the transfer function visually.

Hope this helps.

MQ

Dr. J
June 30th, 2011, 06:37 PM
Looking at the specs for the Presonus and the Driverack 260, the Presonus is capable of +24 dBu output level, and the DR260 is set at the factory for a maximum of +22 dBu input level, with an option to set some internal jumpers for a +14 or +30 dBu input. I'm wondering if someone had set the jumpers to +14???

From page 73 in the manual:
"The DriveRack 260 gives you the option of changing the input gain level settings. There are 3 hardware configurable gain settings. They are: +14 dBu, +22 dBu and +30 dBu. For these cases, use the following procedure to change the gain level settings. Please be advised however, that once the gain level has been changed from the factory settings, the output meters will no longer be calibrated correctly."


Hi Arthur -- I opened up the 260 and the jumpers are set to +22dB.... I was kinda hoping for the +14dB setting BUT in any case the DR is set correctly.

I also took some pics of the internal crossovers in my Tops and sent them to the manufacturer. They immediately replied and asked I pull it out and send it to them so they can fix it right away. They agreed that one of the caps on the mid woofer had a blow out. Man -- that really messes with the magnitude and phase response over the crossover...

I think I am going to mount the crossovers (when repaired) into small boxes like Kip Connor did a while back so I will always have that FullRange option if needed in an emergency (like an amp going down). Throw in the box -- switch presets in the DR and off I go... good precautionary measures..

I have everything I need to go ahead with the Bi-amp project.......... The system is already wired up for it..... now if I can get some great weather to get some good readings done. :D

Dr. J
July 11th, 2011, 11:05 AM
Just an update on the bi-amp project. It turned out to be pretty nice move.

For the first time in a while -- we had spectacular weather here in central Illinois & I got to spend a lot of time measuring. I first measured each driver all by itself OnAxis and applying correction past the crossover points and then checking each driver for consistency between them. To my surprise (finally) my horn responses were pretty identical from cab to cab. I moved onto mid speakers and then subs and sure enough -- everything was a really close match.

In the meantime -- I took many OFF Axis shots to make sure I didn't EQ any dips that needed to be left alone from the OnAx response & worked my way thru combining components and re-testing.

In the end I felt that I had done more to my system than I ever had & luckily -- had a sound gig this weekend. It was time to take it out and test it.

All I have to say it WOW! Just when I thought I was maxed out on what I could do -- I found a way to take it to the next level. Thanks for everyone's help!