Rational Acoustics



Dr. J
November 12th, 2010, 11:14 AM
Hello -- I haven't seen much activity on the forum in awhile so I thought I would spark up a conversation.

I currently have D.A.S. refrence Series 12.64 tops. When I purchased them -- I thought they were bi-amped or at least had a switch to go either full range or bi-amp. No switch. Since I didn't want to void the warranty -- I gave them a go "as is" and I was pretty happy with them.... at least compared to my old JBL's. I want to take them to the next level.

I am now considering bi-amping them anyway because I keep hearing that bi-amping is the only way to go. My question is: Would you bi-amp them? AND where would you set the XO point.

Here is a spec sheet: http://www.dasaudio.com/index.asp?pagina=productos&subpagina=1&galeria=158&producto=419&numPagina=1&lang=en

The XO point is currently set at 1.8kHz

They have 12 inch speaker with a 4 inch voice coil & a 3 inch titanium Neodynium driver on the horn. The exit is 1.5 inch on the horn 60x40 rotatable.

One more thing on the XO point. I want to stay as far away from "Beaming" as I can & a buddy sent me this article on PSW on "Beaming"

Check it out: http://www.prosoundweb.com/article/knowing_cone_drivers_how_they_work_understanding_k ey_data_specs/P4/

On page 4 -- they go into frequency BEAMING. According to this Chart -- it looks like I could benefit from setting the XO point down a bit from 1.8kHz since beaming on a 12 starts in the 1300Hz range.

I know you guys deal with theis type of thing on every system you deal with. Any ideas or comments?

Thanks in advance!

Arthur Skudra
November 12th, 2010, 05:51 PM
I guess you need to ask yourself how much your time is worth to you, and whether the time spent will give you the improvement you're looking for. Yes, you can spend a few hours trying to roll your own biamp settings, however you need to have controlled measuring conditions in order to do this right (you don't want to do this with the box 40 feet up in the air). Many reputable manufacturers have the facilities, staff, and time to come up with optimal settings for their boxes, so it's best to use their recommendations. For cheaper passive products, there typically are some compromises in the design of the box that will make correcting them with a DSP and biamp amplifiers a challenge, so if you're up to it, go for it! For me, my time is better spent using slightly more expensive boxes that have been designed to minimize compromises and have trust-worthy factory settings. After those factory settings are entered into the processor, I'm already more than halfway done optimizing the system!

In our Smaart seminars, we go over the principle of interaction between devices, and the importance of minimizing cancellation/interference between these devices. The magnitude and phase curve are your friends to get that alignment right, particularly in the acoustic crossover region where the devices overlap. Not only do you need to pay attention to the on-axis response, but also the off-axis response, as adjusting your filters to make things work well on-axis may open up a can of worms off axis! (see Rane's excellent note here: http://www.rane.com/note160.html )

As to the optimal settings for your particular loudspeaker, apart from seeing the raw response of the loudspeaker, and being there "live" to make the measurements and adjustments, it's very hard to describe in a few posts here what you need to do. It's something best to look over another's shoulder. I suggest you do some reading to get familiar with the basics:

Live Sound International August 2006 John Murray "Tuning Phase" Part 1
Live Sound International October 2006 John Murray "Tuning Phase" Part 2
EAW "Processor Setting Fundamentals" White Paper by Nathan Butler
Audio Control Industrial "Crossover Networks from A to Linkwit-Riley" by Richard Chinn
Rane Note 160 "Linkwitz-Riley Crossovers: A Primer" by Dennis Bohn
and of course the "EV PA Bible" is worth a look as well.

Dr. J
November 15th, 2010, 12:56 PM
As always -- thanks for your advice Arthur. I am certainly not on the expert or pro scale here but mostly deal with small to medium sized clubs. Most of which do not have installed systems. That means I have to take my system in and set it up every time.

One thing about factory settings: I have tuned a half a dozen systems here in my town and I always try the factory tunings initially. What I am finding is that if I don't have the controller used (same model) that was used to do the factory tuning -- I can't even get a close resemblance to what they claim using Smaart.

I have the DAS tops that I pointed out earlier in my post here and I used their recommended tunings. I didn't like the sound at all & running smaart -- confirmed to me that there were some pretty big flaws going on that needed addressed.

However, I don't have the DAS controller that was probably used to come up with those tunings. I have the DBX DR260 as my FOH system controller.

My conclusion is that system controllers seem to react waay differently than others.

I tuned another system that was all EAW and the guy had Peavey system controller and was insisting on using the EAW tunings BUT I hated the sound of the system. I convinced him that we could try a NEW patch and just let me go thru it and then he could compare & if he didn't like it -- he could just ditch it.

He absolutley loved the final product & couldn't believe his system "WOKE" up like that.

And to provide a final example -- I use Yamaha Club Series monitors (because they can take a lot of club abuse) & they are cheap. In my DR there are tunings for them in there -- I call them up and have a listen. Not what I had in mind.

I then run smaart on them and come up with a trace that looks like someone has been chewing on a rat tail file.

That has been my experience with tunings and system controllers that were most likely not used for the posted tunings.

Now whether or not to go ahead and Bi-amp my tops? I know I will not be able to get a better fit for the full range tops than what DAS came up with but if I were to bi-amp -- I would like to see if there is anything more to gain.

Thanks for the links and article recommendations. The RANE article on LR filters was the best I had ever seen on the topic.

I think I could go with that and align the two. With my budget and being on a small scale compared to you guys -- It is all I have. I HAVE to optimize my system because there is nothing else to go on.

Hopefully someday I will get to graduate to a well respectable planned out system that has been tuned as far it could go out the door BUT I got to learn how to make smaller multiple product systems and make them better as well.

Thanks again for the input Arthur.

Arthur Skudra
November 16th, 2010, 09:47 AM
Yeah, most processor settings are specific to manufacturer supplied DSP's. There has been a few threads out there on PSW, one even recently showing the differences between processors, depending on what algorithms the manufacturer chose in developing their software. In some cases, the differences are very noticeable as you have observed. At least you have some kind of starting point to work with and refine further.

I hear ya on having to use what you have on hand. Doesn't hurt to try and see if you can improve things with a biamped configuration. A good starting point is to use LR filters for your crossover point in a 2 way, use eq filters post-crossover to take care of any anomalies within each output device. I also find that correcting any "bumps" in the mag response beyond the passband of the driver can sometimes help improve the overlap between devices as well if they are significant enough. Above all, spend lots of time experimenting and comparing different DSP combinations using music to find the best one for your situation. You know you will have "arrived" if you purposely reverse the polarity of one of the drivers, and get a very deep cancellation notch in the crossover region as a test to prove both devices are aligned correctly.

Dr. J
November 16th, 2010, 11:53 AM
Thanks Arthur. Hey -- one thing about "Post Crossover PEQ's". You mentioned using Post XO Eq's to fix anomalies. I have 4 Post XO EQ's available for every output (Low, mids & highs). Is there anything particularly wrong with using Pre XO PEQ's IF I need further smoothing of the response? I have 9 of those available as well (Pre XO).

I downloaded the EV PA Bible and read it last night. Very insightful & it covers a lot of ground in just 16 pages. I never paid much attention to polar responses all that much but the booklet opened my eyes to it. Good stuff!

I am currently saving for Smaart 7 & the classes & hopefully soon I will be able to move onto a much higher level with system optimization.

I do so much measuring at home that I have at least 10 separate tuning sessions on my DAS tops alone. I then compile all of tunings into spreadsheet and study them. I get some variation BUT there a several problem areas that end up being exact almost everytime.

I have come to the conclusion that I am either consistently RIGHT or consistently WRONG.....:D Well -- at least I have the consistency part down..:D

I am pretty sure I am on the right track because the system sounds good.

Thanks again!

Dr. J
November 16th, 2010, 11:57 AM
I forgot to thank you for the alignment confirmation technique, "Flipping the polarity to check my work". Awesome.

Arthur Skudra
November 16th, 2010, 12:30 PM
Yeah, that's my bone of contention with the DR260, only 4 PEQ's on each output! So any filters that are close to the crossover point I will use the 4 post XO output PEQ's, anything further away from the crossover point I will use the pre XO input PEQ's, in addition to any shaping or "flavour" eq you wish to apply to the box.

The PA Bible is really cool, don't forget to download and read the supplements as well, lots of good information there!

Somewhere on the internet you can download a goldmine of articles/tech notes published by Altec Lansing. Google is your friend. If you're a SynAudCon member, you can get all their past newsletters online as well. OK, now I have overloaded you with homework!!! :D

Dr. J
November 16th, 2010, 01:09 PM
Great advice Arthur. I don't mind all the articles. I really like this stuff. I read McCarthy's green bible last winter and still read it regularly. yeah -- it is way out there (at least for the novice) BUT as i go back thru it -- it makes more and more sense.

I just got to looking at my Presonus FireStudio Mobile device and I may be able to use more than one measurement mic with it. I think. It only has two input gain knobs on it BUT I guess if the signal is strong enough on the others (internally) it may work. I will have to see.

Do you know anything about the Smaller Presonus devices? I have the StudioLive mixer as well & I know it would make a great interface for Smaart 7 BUT I have no idea if it is possible to use a SL Mixer as my interface AND as the mixing console at the same time. It seems like I would be routing the main out signal back into the board somehow and It would be a circular type of thing plus I don't want to torch anything.

Thanks again!

Arthur Skudra
November 16th, 2010, 10:53 PM
I don't see a problem in using your Firestudio Mobile with two mics, and run your reference signal into one of your line level inputs on the back. The StudioLive can also act as a very flexible preamp for Smaart, just be careful in making your assignments. As long as the interface device has ASIO or CoreAudio drivers, you should be fine using it with Smaart.

Dr. J
November 17th, 2010, 04:25 PM
I don't see a problem in using your Firestudio Mobile with two mics, and run your reference signal into one of your line level inputs on the back. The StudioLive can also act as a very flexible preamp for Smaart, just be careful in making your assignments. As long as the interface device has ASIO or CoreAudio drivers, you should be fine using it with Smaart.

Ok cool. I thought the FireStudioMobile would work that way.

One more thing on my board though....... It connects to my laptop thru Firewire.

So help me here: Let's just use one measurement mic for simplicity. Measurement mic 1 goes into Channel 1 (input) on the mixer. It is picking up the pink noise source from the speakers.

My main out on the board has to be split so I can get a reference signal BUT my board is my interface so it seems that I would have to go out of the main -- split the line and then route the reference line back into my board ...say channel 2 (Input 2). so it goes thru my board and eventually goes back out of the main out and repeats itself.

Is this Okay?

IF I had two of these mixers -- I would just use one hooked to the PA and the other as my interface just to get the tuning done BUT I am trying to find a way to use the mixer as my main console AND Interface at the same time.

All in one deal with the capability of 16 inputs. Hmmm..... I don't know.

Thanks Arthur

I told Presonus to send you a board (Rational) so you guys could figure it out.:D

Arthur Skudra
November 17th, 2010, 10:55 PM
The simple solution to your reference channel issue is using an external pink noise generator that not only goes to the mains, but also appears as an input on your interface. Rational has the Noise Stick (http://www.rationalacoustics.com/store/test-and-measurement/rational-noise-stick.html), but really any pink noise generator will do, even a pink noise track on iPod. (look for the Bink Audio Test CD on the web for all kinds of useful test signals you can burn to a CD or import into an iPod)

I don't know the particulars of the firewire interface in the StudioLive, I'm sure there are others here that have better knowledge of the console. I would assume all the inputs and outputs appear directly in the firewire interface. Can you not unassign a channel from the mix bus for your loopback, yet still have it appear on the firewire interface? Something you should investigate.

Harry Brill Jr.
November 18th, 2010, 06:52 PM
The EV PA bible is several issues spanning several years and far more than 16 pages. You should get the entire thing.

Why plug the reference mic into the mixing console at all? Just go straight into your firewire box. Take the reference using a Y cable from the main out if you like as long as the Y is BEFORE the system processing.

Post frequency dividing network (notice I didn't say crossover) EQ is used for driver correction ONLY. Pre frequency dividing network EQ is used for summation correction between pass bands that are overlapping, and tonal shaping. It's also used to correct for coupling caused by using more than one speaker or by the walls in a room. It's important to distinguish these two.

If you need more than 4 post EQs per output, buy a better system. Seriously that is a lot of EQ. Even the KF850 only needed 3 per 1 or 2 of the outputs (hf and mf).

There is nothing inherently wrong with the DR260. It should get the job done.

Manufacturer's settings are generally spot on for a SINGLE BOX. When using more than one box you need to go to that INPUT EQ and make some adjustments. The input EQ should not have been touched by the manufacturer except in the case of SYSTEMS (4-8 line array boxes,....8-16 line array boxes, etc). The main difference DSP to DSP is the way the parametric EQs are SHAPED. Frequency should be correct. Depth of cut should be close. The WIDTH is what you will want to generally play with to get them right. Measuring the PA with Smaart is no way to make sure they are correct. Too many acoustic variables. The only way to use Smaart to check the settings is to compare to the factory DSP directly. This works well. Keep in mind not all manufacturers have good sounding settings. I have yet to hear a DAS I liked. Just my opinion.

Dr. J
November 19th, 2010, 11:14 AM
Thanks Harry -- Yes I will download the rest of the EV Bible.

Well -- I was trying to find a way to use the console because I have 16 inputs available. Not that I would set up 16 measurement mics or anything but I thought it would be cool to figure out how to make it work if I ever needed to.

I have the Firestudiomobile device that has two Inputs on it with gain knobs. All other inputs on it are line level and I can't adjust them. If the signal is strong enough -- It may work fine. As a local sound guy and musician -- I probably won't get past 3 measurement mics. Using one is hard enough:D

I don't know what to say about using different system controllers & factory speaker tunings. If I go with the DAS tunings -- then there are only going to be one or two for the tops alone with .5 -1.5dB cuts at best. I find it hard to believe actually. Listening to it that way doesn't make me feel much better either.

I hope manufacturers aren't embarassed to post tunings thinking that the customer may perceive it as flawed in some way............ so they land on the conservative side?????

Yes the DAS tops are not very expensive. I feel they are a great bang for the buck though at $800 a piece. They are plenty for the bars I play in & a nice step up from my old JBL's. I am in the poor acoustic bar world.

So POST parametrics (frequency divider) are for driver correction?
Pre Parametrics are for summation? tonal responses? Summation would suggest everything that is going on over the crossover area.
What about speakers that have built in crossovers? That would be post amplifier? My tops are not bi-amped.

Thanks for your input Harry! I will just keep learning all I can.

Arthur Skudra
November 19th, 2010, 03:46 PM
If you need more than 4 post EQs per output, buy a better system. Seriously that is a lot of EQ. Even the KF850 only needed 3 per 1 or 2 of the outputs (hf and mf).Quite the generalization Harry, I know of several respectable professional loudspeakers that have 5 or more filters for driver correction, in fact, EAW recommends 5 filters for the KF850z biamp on the HF, go look on their website! ;) While less is more when it comes to equalization of loudspeakers, if "more" really does improve the response of the box, why throw the baby out with the bathwater in saying that you shouldn't need more than 4? Based on the few DAS's that I've heard, I can tell you that it probably will need "more" than "less" driver correction eq to make it sound half decent. :p


There is nothing inherently wrong with the DR260. It should get the job done.Not saying there is anything wrong with the DR260, I've used plenty of them on my projects, and for the money, it's hard to beat. Obviously with a fixed architecture they chose what DSP resources go where, and unfortunately the number of output filters was compromised IMHO. Nevertheless you can get around that with careful use of the input filters.

As with the spirit of the OP, it's using your existing resources to their maximum potential, and knowing your limitations and working around them. The fine art of managing your compromises, DAS speakers and DR260's included! :)

Dr. J
November 22nd, 2010, 05:32 PM
Thanks Pepe -- I actually just read about that very topic somewhere in the articles Arthur pointed out. I will keep everything in mind that you talked about & give it my best run. :D

Thanks!

Dr. J
November 23rd, 2010, 10:26 AM
Pepe -- I will see if I can find anything on the DAS Tops about beamwidth.

Currently my tops are passive with a fixed XO point of 1.8. It seems like a good starting point for me. I am certain that DAS thought that 1.8 was a great place for the XO point as well.

My original question was to bi-amp or not & is there anything to gain from it. It sounds like there is something to gain from it as long as I follow some pretty strict rules.

You say make sure the beamwidths match for a particular frequency between the two transducers -- sound good

I have even played with the XO point on my old JBL's to where they didn't require any delay alignment and called it a good place to start.

There is a lot to consider here I guess. Thanks again for your help.

Dr. J
November 23rd, 2010, 12:50 PM
:D:D:D: Ha -- I mean what you said about finding the beamwidths sounds good to me. Sounds like a good plan. Sorry :D

Thanks Pepe

Arthur Skudra
November 24th, 2010, 12:18 PM
Pepe -- I will see if I can find anything on the DAS Tops about beamwidth.You'll find the spectrograph to be your friend here. Set up a mic on axis to the loudspeaker, then rotate the box to see the beamwidth of each individual driver. Perhaps open up the high pass crossover filter on the HF down to about 1000 Hz so you can see the full effect of the waveguide's control of the HF. You could also take snapshots of the magnitude response, displace them by either 3 or 6 dB down, then rotate your speaker to match the level of the live trace to the displaced stored trace, and that will be the angle of coverage of the loudspeaker.

Dr. J
November 24th, 2010, 12:45 PM
Arthur -- Let me get some clarification here. When I hear Beamwidth -- I keep thinking about how the frequencies get narrower as frequency rises. Kind of like the phrase "BEAMY". Is that correct?

It is like you are saying Beamwidth and Coverage pattern (-6dB off axis) mean the same thing. Help set me straight here.

Just to elaborate on your post a bit. I first tried to master the ground plane technique using one mic outside and one stack. After I was able to repeat that multiple times over month or so and get the same results -- I felt good about it. I also like the way the system sounded of course.

Next -- I ventured out into doing averages. This seemed really hard to do with moving the mic around so I eventually decided it was MUCH easier to leave the TOP in place on a pole up in the air pretty high and the mic straight out on axis.

I took an on axis shot and then an off axis shot (-6dB) and averaged the two together. That gave me an averaged STATIC trace. I then fired up the live trace and simply turned the speaker back and forth to give me the closest match to the static BUT only I am able to keep my Live trace going and not have to worry about moving the mic. Instead I get to move the box --- waay easier.

From this position -- I equalized the system.

Now -- It sounds like what you are saying is to do this same type of measuring ONLY on each driver separately to find the coverage area Or Beamwidth AND then repeat on the other driver as well to locate a match between the two.

Am I half way close to what you are talking about?

Thanks Arthur!

Arthur Skudra
November 24th, 2010, 09:14 PM
Arthur -- Let me get some clarification here. When I hear Beamwidth -- I keep thinking about how the frequencies get narrower as frequency rises. Kind of like the phrase "BEAMY". Is that correct?

It is like you are saying Beamwidth and Coverage pattern (-6dB off axis) mean the same thing. Help set me straight here. Let me be careful on how I state my terminology. Generally when I talk about beamwidth, I'm stating the angle of coverage in the horizontal and vertical planes according to frequency, based on x number of dB down from "nominal". A lot of speaker manufacturers state a coverage angle for their loudspeakers, horizontal x vertical, but if you dig deeper, they are essentially looking at a certain number of dB down point (say -6dB) at a certain frequency, and thus stating their coverage angle based on that point.

When I think of "beamy", it's more a result of the physics involved with the speaker components themselves, as you go up in frequency, the coverage narrows. However keep in mind that a coverage may narrow, then widen at certain points, then become narrower due to certain peculiarities of waveguides and how they are designed.


I took an on axis shot and then an off axis shot (-6dB) and averaged the two together. That gave me an averaged STATIC trace. I then fired up the live trace and simply turned the speaker back and forth to give me the closest match to the static BUT only I am able to keep my Live trace going and not have to worry about moving the mic. Instead I get to move the box --- waay easier.
Hmmm, you should *listen* to the loudspeaker at the -6dB point. You'll find that you're outside the perceived coverage of the loudspeaker, not a good measurement point that would be representative of what the majority of those in the coverage area of the loudspeaker are hearing. Let me give you a hint, keep your spatial averaging within the -3dB contours of your loudspeaker. Like I said before, listen with your ears. You'll find that you're starting to get out of the coverage zone at -3 dB, not -6 dB. The -6 dB contours are helpful for helping determine how you are going to splay speakers when they are arrayed together. Listen to a speaker on it's own, and you'll find that your coverage is a bit narrower than what is stipulated with the -6dB down points.


Now -- It sounds like what you are saying is to do this same type of measuring ONLY on each driver separately to find the coverage area Or Beamwidth AND then repeat on the other driver as well to locate a match between the two.

Am I half way close to what you are talking about?
All I'm saying is that when you're searching for that ideal crossover frequency, pick a frequency where the coverage angles match between drivers, for the smoothest transition between drivers. Paying careful attention to the spectrograph will help determine this rather quickly.

Dr. J
November 29th, 2010, 04:50 PM
Arthur -- Thanks for the clarification. You have been incredibly helpful. I re-did my system the other day and kept the coverage areas to within 3dB. I did my averaging and got a nice smooth response. I have not bi-amped yet. That will be down the road a bit.

I noticed on-axis I have dips around 200-400Hz range. I didn't react to those but moved the microphone around to see if it remained. 8 foot out didn't change it much but when I moved it in closer -- the dip started to disappear. I am assuming ground bounce was in action here so for the heck of it -- I laid the top down on it's side and did a ground plane measurement to see if the dips stayed BUT they too disappeared.

I am still learning the basics here but what can I learn from a Ground Plane measurement vs. an on axis / off axis average?

Maybe I should ask this way: Since ground plane is a PZM style of measurement -- is there still anything to look out for? Any ground bounce or comb filtering going on? Just how accurate is it?

Also -- this was all done outdoors away from as many reflections as possible.

Thanks!

Arthur Skudra
December 1st, 2010, 10:33 PM
I am still learning the basics here but what can I learn from a Ground Plane measurement vs. an on axis / off axis average?

Maybe I should ask this way: Since ground plane is a PZM style of measurement -- is there still anything to look out for? Any ground bounce or comb filtering going on? Just how accurate is it?I recommend you review the many papers out there on the PZM microphone, a wealth of information out there. Using a ground plane measurement simply shifts the frequency of the first of many subsequent comb filter dips in the response to a high enough frequency that makes it easy to ignore in your measurements. They are still there, but high enough on the frequency scale. A valuable tool in measuring systems, and as easy to create as leaning the mic against a wall, floor, case lid, sheet of plywood, etc. If you are going to use the floor as a boundary, then make sure you do it under controlled circumstances so that nobody steps on or runs over your mic!

There's some good threads on the old EAW Smaart Support forum that would be good for you to check out:
http://forums.eaw.com/cgi-bin/ubb/ultimatebb.cgi?ubb=get_topic;f=15;t=000237
http://forums.eaw.com/cgi-bin/ubb/ultimatebb.cgi?ubb=get_topic;f=15;t=000258

Dr. J
December 2nd, 2010, 02:12 PM
Will do -- thanks for the links. I forgot about the old forum..... That will keep me busy for a while :D

Kip Conner
December 9th, 2010, 06:57 AM
J- I think that once you sit down with one of the instructors you will find more benefits to bi-amp'ing your system. I think that it will come down to whether it's going to matter to the listener.

I have a set of monitors that I use in the rental game that are ready to go in either direction. I pulled the passive crossovers out of the boxes and mounted them in aluminum boxes that I purchased at Fry's. It gives me the freedom to go either way very quickly depending on the clients funds. Most of the time on a small level we are doing it for ourselves because we know the benefits. However the cost of adding a multiple strong DSP's and more amplifiers doesn't warrant the liability from the little money the gig may bring in... or the extra weight.

As your looking at DSP's one thing you can start looking at when you are optmizing systems is the latency that occurs within them. If you are using one DSP such as a DR260 to control the entire system there's not much to worry about if you individually measure your outputs and adjust the output delay to compensate for your phase response at the acoustic crossover point. If you're having to use 2 identical DSP's (one for left and one for right)- you may want to measure the DSP itself and compare all inputs and outputs to make sure that the latency is identical.

I was once hired to not optimize a system, but to analyze and report on the system. The first thing I did was to measure the the DSP's (both EV DX38) and discovered that there was a slightly longer latency in one device but not the other. The system was an EV system and they were using the manufacturers provded settings. Unfortunately they weren't aware that one of their devices was flawed and they gave settings that caused a phase anomaly in one side, but not the other. The shift wasn't enough to cause major cancellations but the house engineer knew there was "something up with it" and couldn't explain it since all of the settings were identical.

Here's an article that I started on that system, but never really fine tuned... there are lots of screen shots that show the latency.

http://www.athenssound.com/cct.html

Dr. J
December 10th, 2010, 01:48 PM
Kip -- Thanks for the info. I like your article too. I have not measured my 260 to see if it differed from left to right plus high, mid, low. Geesh -- I would hope that it is fine BUT then again you never know. I can easily check it I guess.

There is a certain version of the DRPA that purposely had the high outputs reversed and I found it and called DBX. They told me a certain version had it and the newer version didn't but that it was intentional nonetheless. Confusing......

Hey Kip -- I have my low end response ramped up & I noticed that more and more guys are going with a FLAT but yet Spectral Tilt. There was a definite tilt to the system you were working with.

My system is about +15 dB strong in the low end and has a tilt towards 1K & from 1K out -- is flat as possible.

Do you have any suggestions that would help me with this? You said you have done plenty of rock gigs and I do mostly Rock/Metal.

Right now I have spectral tilt with some Flatness. Bob McCarthy calls this Spectral Tilt WITH Variance. He states that if the tilt has smoothness -- meaning constant gradual slop -- basically a straight line but tilted -- you will have minimum variance which is what really matters. Also -- it is just as good as a completely FLAT response because it is the smoothness or minimum variance that matters.

What does your final trace end up looking like if you don't mind sharing.

Thanks!

Kip Conner
December 12th, 2010, 07:12 PM
I would have to dig up the quick and dirty transfer function of that room. The client had the funds to pay for the analysis and not the fix. I know, very odd. After the Production Manager left for the day I went through the processor and did a few things to show them what the potentiols were in the system. I compensated for the latency by adding some delay and lowered the outputs of the subs for a slightly more flat system along the bottom end.

My biggest complaint with overworking the subs in rooms is that the room itself tends to get muddy fast if you don't have variable HPF's on the desk and if you like to do simple two track recordings to document your show, they are going to come out a little then.

As for the Spectral Tilt in that room that shows up on the transfer function, a lot of it has to do with the PA set-up. The subs were ground stacked and the arrays were flown causing a ramp up in low end response. The rigging points weren't rated to fly the system together in terms of weight so they had to put the subs on dollies and roll them into place.

I am personally in a situation where I carry my own desk and monitors, but have to use the house racks and stacks. I find my self going for flat on the low end and 50% of the time I end bringing some of it back. The low end response of the room is tricky thing for me since I tend to work with a lot of singer-songwriters that have have baritone voices. In these shows, lyrics are what people have come for so the vocal has to be clear.

So my trace is fairly flat from 100Hz and up with special attention given from 1k to 4k to make sure the PA sounds warm. I'm an analog guy living in a digital world so I do what I can with what I am given.

PhillipIvanPietruschka
December 13th, 2010, 03:47 AM
Here's an article that I started on that system, but never really fine tuned... there are lots of screen shots that show the latency.

http://www.athenssound.com/cct.html

Kip, thank you for posting that article. With regards to the measured discrpency in phase response, did you use the same delay line setting for the left and right channels of the PA? Is it possible that the relative phase shift you measured is because the microphone is not quite equidistant to each array?

Kip Conner
December 13th, 2010, 01:12 PM
I used a laser to determine the distance from each array, however the phase shift that was occuring was due to large impart to the latency in the EV Dx38 processors. They were using 2 of them (One for Left and One for Right). The theatre production manager wanted me to document it all so I started with all of the electrical components. I measured the transfer function of each output with and without the filters in place to be sure they were seeing the a non filtered response and ran the delay locater to determine the latency in the processor. If it was a single processor with than the latency wouldn't matter as long as it was compensated for in the acoustic crossover.

The very last thing that I did with the house sound engineer was have him add delay to the to the non-latent processor (or one with less latency) to line the phase response up in the acoustic field. I only had to add the .02 to correct it. The latency and phase shift wasn't really the reason I was called in, just something I discovered along the way. Honestly, it was somehing that I never really ever considered in setting up a system since I typically don't use separate left and right processors- most stuff I do doesn't require me to set-up that type of rig. When I work on that type of rig I never the system engineer, just a guest engineer that has time to only measure one side and make some assumptions about the other. Not by choice, but necessity.

Dr. J
December 13th, 2010, 05:22 PM
Thanks Kip!