ProSoundWeb Community

Sound Reinforcement - Forums for Live Sound Professionals - Your Displayed Name Must Be Your Real Full Name To Post In The Live Sound Forums => LAB Lounge => Topic started by: John Woodfield on December 26, 2012, 10:18:42 pm

Title: Crossover Help
Post by: John Woodfield on December 26, 2012, 10:18:42 pm
Preface: I have very little knowledge of crossover settings and how things are supposed to be set, please bear with me.

Presonus Studiolives->DRPA PA->AMP->Speakers

I currently have the DRPA split so I can run my mains through one side and keep my subs aux driven.

For my mains I am using TOA HX-5's. TOA has recommended a crossover setting of HPF BW (24 dB/oct) at 90Hz. From my research, primarily here, it seems that the consensus would be running it a bit higher 100-130Hz to protect the speakers.

I've figured out where to place those settings in the DRPA.

Here is my question, based on the settings TOA recommends for the mains slightly modified to say 100Hz what do I set the low end of the crossover to for the subs so I end up with transparent response? I wasn't sure if I use the same BW 24dB/oct at 100Hz to basically duplicate the HPF or if I should go lower,  higher or something completely different.

Thank you in advance for the assistance.
Title: Re: Crossover Help
Post by: Ivan Beaver on December 26, 2012, 10:33:24 pm
Preface: I have very little knowledge of crossover settings and how things are supposed to be set, please bear with me.

Presonus Studiolives->DRPA PA->AMP->Speakers

I currently have the DRPA split so I can run my mains through one side and keep my subs aux driven.

For my mains I am using TOA HX-5's. TOA has recommended a crossover setting of HPF BW (24 dB/oct) at 90Hz. From my research, primarily here, it seems that the consensus would be running it a bit higher 100-130Hz to protect the speakers.

I've figured out where to place those settings in the DRPA.

Here is my question, based on the settings TOA recommends for the mains slightly modified to say 100Hz what do I set the low end of the crossover to for the subs so I end up with transparent response? I wasn't sure if I use the same BW 24dB/oct at 100Hz to basically duplicate the HPF or if I should go lower,  higher or something completely different.

Thank you in advance for the assistance.
It is not that simple.  it is pretty simple-if you are dealing with perfect sources-which you are not.  So the PROPER settings will depends on a lot of factors-of which we cannot even start to guess at-unless we have some good measurements of both the subs and the full range cabinets.

Exactly what do you mean "transparent"?  Most people prefer the subs to be a bit louder than the full range cabinets.  But not always.  How much louder-it depends.  That difference will also affect the crossover freq.

Also the electrical and acoustical crossovers are very rarely the same-hence the need for the acoustical measurements-especially if the subs are run louder.

Sorry.  But there is no "simple" answer.  There is of course a "simple wrong answer".

Title: Re: Crossover Help
Post by: John Woodfield on December 26, 2012, 10:40:14 pm
It is not that simple.  it is pretty simple-if you are dealing with perfect sources-which you are not.  So the PROPER settings will depends on a lot of factors-of which we cannot even start to guess at-unless we have some good measurements of both the subs and the full range cabinets.

Exactly what do you mean "transparent"?  Most people prefer the subs to be a bit louder than the full range cabinets.  But not always.  How much louder-it depends.  That difference will also affect the crossover freq.

Also the electrical and acoustical crossovers are very rarely the same-hence the need for the acoustical measurements-especially if the subs are run louder.

Sorry.  But there is no "simple" answer.  There is of course a "simple wrong answer".

Ivan,

I mistakenly figured there was a simple formula of where to set the bottoms since the frequency of the tops has already been determined.

When you say "measurements" are you referring to the specs of the boxes or measurements taken in-room?

Volume is not the concern here. The frequency and slopes is what I need to determine.

If you can point me in the right direction, even a starting point that isn't perfect, I'll work with it.
Title: Re: Crossover Help
Post by: Jay Barracato on December 26, 2012, 10:51:20 pm
I may have missed it but did you ever tell us what the subs are?

If you actually want the top and sub to work together through the crossover range you need to know the phase characteristics of both. Anything else is a random guess.

Unfortunately many people use aux fed subs as an excuse to totally ignore the crossover. In that case all the crossover settings are doing is getting rid of out of band signal. If that is okay with you leave the sub filter where it is and see how it sounds. You are not likely to hurt anything by sliding the frequencies around a bit.
Title: Re: Crossover Help
Post by: John Woodfield on December 26, 2012, 11:18:47 pm
I may have missed it but did you ever tell us what the subs are?

If you actually want the top and sub to work together through the crossover range you need to know the phase characteristics of both. Anything else is a random guess.

Unfortunately many people use aux fed subs as an excuse to totally ignore the crossover. In that case all the crossover settings are doing is getting rid of out of band signal. If that is okay with you leave the sub filter where it is and see how it sounds. You are not likely to hurt anything by sliding the frequencies around a bit.

Subs:
LX-18B Model

http://www.wharfedalepro.com/Home/Products/CLASSICPRODUCTS/LXLXESERIES/LX18BLX18BE/tabid/113/Default.aspx
Title: Re: Crossover Help
Post by: Tim Perry on December 26, 2012, 11:19:25 pm
John, I looked at the manual then set up recommended filter characteristics in the Studio live.  The result looks a bit odd to me.

In particular the relatively narrow  cut at 2.5 K right in the vocal range seems somehow wrong.

My suggestion at this point is for you to contact a TOA applications engineer (not a sales guy) for additional input.

 
Title: Re: Crossover Help
Post by: John Woodfield on December 26, 2012, 11:24:14 pm
John, I looked at the manual then set up recommended filter characteristics in the Studio live.  The result looks a bit odd to me.

In particular the relatively narrow  cut at 2.5 K right in the vocal range seems somehow wrong.

My suggestion at this point is for you to contact a TOA applications engineer (not a sales guy) for additional input.

How were you able to properly set the Q in the Studiolive?
Title: Re: Crossover Help
Post by: Tim Perry on December 26, 2012, 11:36:43 pm
How were you able to properly set the Q in the Studiolive?

I just dialed it in using the PEQ to get this visual representation.
Title: Re: Crossover Help
Post by: John Woodfield on December 26, 2012, 11:44:18 pm
I just dialed it in using the PEQ to get this visual representation.

How come I don't have adjustable Q on mine or am I blind?
Title: Re: Crossover Help
Post by: Tim Perry on December 27, 2012, 12:00:10 am
How come I don't have adjustable Q on mine or am I blind?

This is a pic of the computer screen that hooked up via firewire to the Sl24.

This one is a little closer to the "factory recommended setting"
Title: Re: Crossover Help
Post by: John Woodfield on December 27, 2012, 12:03:22 am
This is a pic of the computer screen that hooked up via firewire to the Sl24.

This one is a little closer to the "factory recommended setting"

I'm not worried about that, I'm trying to figure out how you have adjustable Q on your Studiolive when I can't find it on mine.
Title: Re: Crossover Help
Post by: Tom Young on December 27, 2012, 07:29:45 am
I'm not worried about that, I'm trying to figure out how you have adjustable Q on your Studiolive when I can't find it on mine.

Perhaps you do not understand the term "Q", which is (sort of) the same as "bandwidth". DSP manufacturers employ either Q or bandwidth, or allow you to select one of these,  in parametric equalizers and (in some cases) high- or low- pass filters and to characterize their shape (width).

Your digital mixer probably displays the filters in fractional octaves and you may be able to change this to display in (or using) Q.

RTFM   ;)

Q employs a mathematically-derived number to describe the sharpness of a resonant circuit (aka: filter) whereas bandwidth is provided as a fractional octave (1/3, 3/4, 1/1, 2/1, etc. ocatve) or with a decimal (0.35, 5.00, 8.00, 1.00, 2.50, etc. octave).  With Q, the smaller the number the wider/broader the filter.

In my experience, the Q of a filter is nonintuitive. But this may be because for several decades (before DSP) the equalizers (and analyzers) we used displayed filters as fractional octaves and I have no reference for Q values burned into my brain. But a 1/3-octave filter, or notch, is inherently easier to visualize than the same thing described as having a Q of 4.32. Right ?

You can find numerous articles on Q, as well as cross-reference charts of Q versus fractional-octave, on the internet. Here are two:

http://www.rane.com/note170.html

http://www.astralsound.com/parametric_eq.htm
Title: Re: Crossover Help
Post by: Ivan Beaver on December 27, 2012, 08:03:23 am
Ivan,

I mistakenly figured there was a simple formula of where to set the bottoms since the frequency of the tops has already been determined.

When you say "measurements" are you referring to the specs of the boxes or measurements taken in-room?

Volume is not the concern here. The frequency and slopes is what I need to determine.

If you can point me in the right direction, even a starting point that isn't perfect, I'll work with it.
Measurements in the room would be best-as the physical distance between boxes can also greatly affect how the crossovers and delay times are set.

As Jay said-the phase response is one of the main concerns.  You can't get the amplitude right-unless the phase is also right.

Manufacturers curves would be a small start-but generally there is not enough information there to align the different devices-especially when they are from different manufacturers.

Also remember that there is getting it "correct" and close enough.  Most people are happy with "close enough".  Especially since you cannot get it correct for the whole audience at the same time.  You have to choose a seat-(which is most important-the operators seat-the Pastors seat-an average set and so forth) and use that as your basis-and the rest of the room will be "off", but that is part of the compromise of providing audio for more than one person.
Title: Re: Crossover Help
Post by: John Woodfield on December 27, 2012, 09:02:54 am
Perhaps you do not understand the term "Q", which is (sort of) the same as "bandwidth". DSP manufacturers employ either Q or bandwidth, or allow you to select one of these,  in parametric equalizers and (in some cases) high- or low- pass filters and to characterize their shape (width).

Your digital mixer probably displays the filters in fractional octaves and you may be able to change this to display in (or using) Q.

RTFM   ;)

Q employs a mathematically-derived number to describe the sharpness of a resonant circuit (aka: filter) whereas bandwidth is provided as a fractional octave (1/3, 3/4, 1/1, 2/1, etc. ocatve) or with a decimal (0.35, 5.00, 8.00, 1.00, 2.50, etc. octave).  With Q, the smaller the number the wider/broader the filter.

In my experience, the Q of a filter is nonintuitive. But this may be because for several decades (before DSP) the equalizers (and analyzers) we used displayed filters as fractional octaves and I have no reference for Q values burned into my brain. But a 1/3-octave filter, or notch, is inherently easier to visualize than the same thing described as having a Q of 4.32. Right ?

You can find numerous articles on Q, as well as cross-reference charts of Q versus fractional-octave, on the internet. Here are two:

http://www.rane.com/note170.html

http://www.astralsound.com/parametric_eq.htm

Tom,

I'm sorry, apparently I'm not being clear enough. We are both using Studiolive boards. I can't seem to find the ability to make my Q settings adjustable on the fat channel or in any options. I can either use a fixed Q or I believe the other option is a HiQ (I'm not in front of my board, they may call it something else) I have never seen the option to have a variable Q like you show on your screenshot.

I posted this same question over on the Presonus forums and I was pretty quickly asked if you are using a SL24. That is probably where the difference is. You're using a SL24 which has some better features compared to my SL16's. If that is indeed the case its another one of those annoying things where I assumed the SL16 and SL24 would both have the same feature set.
Title: Re: Crossover Help
Post by: John Woodfield on December 27, 2012, 09:08:06 am
Measurements in the room would be best-as the physical distance between boxes can also greatly affect how the crossovers and delay times are set.

As Jay said-the phase response is one of the main concerns.  You can't get the amplitude right-unless the phase is also right.

Manufacturers curves would be a small start-but generally there is not enough information there to align the different devices-especially when they are from different manufacturers.

Also remember that there is getting it "correct" and close enough.  Most people are happy with "close enough".  Especially since you cannot get it correct for the whole audience at the same time.  You have to choose a seat-(which is most important-the operators seat-the Pastors seat-an average set and so forth) and use that as your basis-and the rest of the room will be "off", but that is part of the compromise of providing audio for more than one person.

I think close enough will suffice in this application. The subs are in line with the tops but the tops are center flown and the subs are at stage level on either side. Time delay aside, I need measurements to set the crossover point in the DSP?

Let me fire up REW today and see what I can report back.
Title: Re: Crossover Help
Post by: Tim Perry on December 27, 2012, 09:46:56 am
I guess the fat channel on the 16 is less fat :)
Title: Re: Crossover Help
Post by: John Woodfield on December 27, 2012, 09:52:01 am
I guess the fat channel on the 16 is less fat :)


 >:(
Title: Re: Crossover Help
Post by: Garry Wilson on December 27, 2012, 10:55:20 am

 >:(

John,

       I bought the 16 first, later bought the 24. The adjustable Q was one of the first feature set differences I noticed. I guess more real estate helped, they also added more control to gating and compression.


Garry W.
Title: Re: Crossover Help
Post by: John Woodfield on December 27, 2012, 11:35:38 am
John,

       I bought the 16 first, later bought the 24. The adjustable Q was one of the first feature set differences I noticed. I guess more real estate helped, they also added more control to gating and compression.


Garry W.

I need to buy other board for portable use. I guess it will be the SL24. Glad I figured this out now.

Have fun on your side of the bridge. At least you're closer to Chuck Levin's over there.
Title: Re: Crossover Help
Post by: Tom Young on December 27, 2012, 12:22:37 pm
Tom,

I'm sorry, apparently I'm not being clear enough. We are both using Studiolive boards. I can't seem to find the ability to make my Q settings adjustable on the fat channel or in any options. I can either use a fixed Q or I believe the other option is a HiQ (I'm not in front of my board, they may call it something else) I have never seen the option to have a variable Q like you show on your screenshot.

I posted this same question over on the Presonus forums and I was pretty quickly asked if you are using a SL24. That is probably where the difference is. You're using a SL24 which has some better features compared to my SL16's. If that is indeed the case its another one of those annoying things where I assumed the SL16 and SL24 would both have the same feature set.

OK. I did misunderstand your situation and query. My Bad.

But I did not post anything about "my SL24". That was someone else.

BTW - once you have 4 bands of parametric EQ at hand on each input, it is very frustrating to then work with anything less, even "semi parametric" filters.

But you already know this  ;-)