ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: 1 2 [All]   Go Down

Author Topic: Crossover Shapes  (Read 5853 times)

Adam Pace

  • Newbie
  • *
  • Offline Offline
  • Posts: 31
    • Yarbrough's Music, Inc.
Crossover Shapes
« on: September 20, 2011, 04:23:39 am »

Hello Folks,

Well, excitement stirs as we are preparing for the arrival of our new QSC KW rig. With the new rig, comes new processing toys. We have chosen the BSS OmniDrive. Where before, the trusty ole' Rane crossover did a phenomenal job, the new OmniDrive opens fun new doors of adventure, specifically, in Crossover Shapes.

The Rane didn't offer any options in shapes, you just dialed in the frequency and level of each band and added delay if needed to the sub stacks. Now, I read about all the different shapes you can use in your crossover design. Although I am looking forward to the added accuracy in flattening and optimizing the system, I am a little overwhelmed at all the different shapes, Butterworth, Bessel, Linkwitz-Riley, etc.

In reading about them, I am leaning towards the Linkwitz-Riley 24dB/Octave slope, but there is also the Butterworth. So, I thought I would check in with the knowledge pool and see what ya'll use and why? What applications call for certain shapes?

Can't wait to get it out and tuned, and any advice is, as always, greatly appreciated.

Thanks,

Adam P.
Logged
Adam Pace
Productions Manager

Yarbrough's Music, Inc.
6122 Macon Road
Memphis, TN 38134

Phone: 901-761-0414
Fax: 901-257-7773
Email: apace@yarbroughsmusic.com
Website: http://www.yarbroughsmusic.com/

Timo Beckman

  • Full Member
  • ***
  • Offline Offline
  • Posts: 141
  • Rational Acoustics/Isemcon/Fulcrum Acoustic NL
    • Timo Beckman Geluid
Re: Crossover Shapes
« Reply #1 on: September 20, 2011, 05:40:25 am »

As far as i know there are also the NTM filters in there .
These have a simular behaviour as a eliptic filter . I played around with them a bit and where i needed delay (and pol. rev. if needed) with the standard filters i just had to type in the frequency to the x-over zone and that was it .
I'm using to BSS BLU80 audio network processors instead of the omni drive but the x-overs should be the same as the omni-drives .

http://timobeckmangeluid.wordpress.com/2011/06/01/oefenen-met-fase/ 

I did post on my blog but this artical is still in dutch and i did not (do not) have time to translate it in to english yet . (Most articles on my blog are Dutch/English and have a lot to do with alligning systems so check it out and let me know what you think) .

If you start playing with filters make sure you also have the means to measure the response of the system in phase and frequency response .

I just went trough a SIM set-up and data viewer file and i found a small project with all sorts of filters combined from 6db/oct to 24 db/oct plus the sum of the filters . It might take a few day to put it on the net but if you sent me a e-mail i'll let you know where to find them .... If you're interested  :)
« Last Edit: September 20, 2011, 07:10:22 am by Timo Beckman »
Logged

Peter Morris

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1221
Re: Crossover Shapes
« Reply #2 on: September 20, 2011, 09:06:23 am »

Hello Folks,

Well, excitement stirs as we are preparing for the arrival of our new QSC KW rig. With the new rig, comes new processing toys. We have chosen the BSS OmniDrive. Where before, the trusty ole' Rane crossover did a phenomenal job, the new OmniDrive opens fun new doors of adventure, specifically, in Crossover Shapes.

The Rane didn't offer any options in shapes, you just dialed in the frequency and level of each band and added delay if needed to the sub stacks. Now, I read about all the different shapes you can use in your crossover design. Although I am looking forward to the added accuracy in flattening and optimizing the system, I am a little overwhelmed at all the different shapes, Butterworth, Bessel, Linkwitz-Riley, etc.

In reading about them, I am leaning towards the Linkwitz-Riley 24dB/Octave slope, but there is also the Butterworth. So, I thought I would check in with the knowledge pool and see what ya'll use and why? What applications call for certain shapes?

Can't wait to get it out and tuned, and any advice is, as always, greatly appreciated.

Thanks,

Adam P.

The QSC KW series comes complete with its own crossover(s) - I'm sure QSC has it worked out just fine - time and phase aligned etc. If you are using a complete KW rig all you will need is a GEQ. An Omni drive will not improve anything. Save your $$$
Peter

« Last Edit: September 20, 2011, 09:16:04 am by Peter Morris »
Logged

Timo Beckman

  • Full Member
  • ***
  • Offline Offline
  • Posts: 141
  • Rational Acoustics/Isemcon/Fulcrum Acoustic NL
    • Timo Beckman Geluid
Re: Crossover Shapes
« Reply #3 on: September 21, 2011, 07:53:40 am »

Logged

Timo Beckman

  • Full Member
  • ***
  • Offline Offline
  • Posts: 141
  • Rational Acoustics/Isemcon/Fulcrum Acoustic NL
    • Timo Beckman Geluid
Re: Crossover Shapes
« Reply #4 on: October 02, 2011, 10:50:37 am »

The settings of the manufacturer should work but allways verify them if you have the possibility .
I posted a small artical on my blog . 2 examples of mis-allignment with settings from the manufacturor .

http://timobeckmangeluid.wordpress.com/2011/10/02/tanstaafl-there-aintno-such-thing-as-a-free-lunch-statement-i-got-from-b-mccarthy/
Logged

John Roberts {JR}

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 16216
  • Hickory, Mississippi, USA
    • Resotune
Re: Crossover Shapes
« Reply #5 on: October 02, 2011, 01:06:45 pm »

I have been saying this for years, but it bears repeating. The real loudspeaker design engineers working at the manufacturer will have superior knowledge about the drivers being used and box design. They will also generally have superior test equipment and facilities to measure actual box with actual driver response to make corrections.

The only concern I would suggest is that you accurately transfer the manufacturer's advice into whatever DSP you are using. As often as not, problems are not user error, as much as (DSP) manufacturer's lack of discipline in delivering expected transfer functions.

There will be a session at the upcoming AES on this topic.

JR

PS: Perhaps manufacturers could also provide detailed point by point charts of final phase and amplitude response that they are asking for, to act as a second check or back-up for correct transfer function. 
Logged
On the internet people tell you everything "they" know, not the answer to "your" question.....  http://circularscience.com/

Timo Beckman

  • Full Member
  • ***
  • Offline Offline
  • Posts: 141
  • Rational Acoustics/Isemcon/Fulcrum Acoustic NL
    • Timo Beckman Geluid
Re: Crossover Shapes
« Reply #6 on: October 02, 2011, 02:51:59 pm »


The only concern I would suggest is that you accurately transfer the manufacturer's advice into whatever DSP you are using. As often as not, problems are not user error, as much as (DSP) manufacturer's lack of discipline in delivering expected transfer functions.

There will be a session at the upcoming AES on this topic.

JR

PS: Perhaps manufacturers could also provide detailed point by point charts of final phase and amplitude response that they are asking for, to act as a second check or back-up for correct transfer function.

Will the aes session be made available on the www because that would be realy interesting .
Also the charts would be nice if they were available .
Logged

Mac Kerr

  • Old enough to know better
  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 6333
  • Audio Plumber
Re: Crossover Shapes
« Reply #7 on: October 02, 2011, 02:59:18 pm »

Will the aes session be made available on the www because that would be realy interesting .
Also the charts would be nice if they were available .

Audio recordings of the sessions will be available for sale from AES after the convention. The technical sessions at the AES convention are not free, and neither ore the recordings. I don't know how soon after a session has happened the recordings will be available, but it shouldn't be too long, and this session is one of the first to take place.

It should be a good session, Rich Frembes has assembled a good panel of manufacturers and users.

Mac
« Last Edit: October 02, 2011, 03:18:44 pm by Mac Kerr »
Logged

Timo Beckman

  • Full Member
  • ***
  • Offline Offline
  • Posts: 141
  • Rational Acoustics/Isemcon/Fulcrum Acoustic NL
    • Timo Beckman Geluid
Re: Crossover Shapes
« Reply #8 on: October 02, 2011, 03:15:42 pm »

Audio recordings of the sessions will be available fro sale from AES after the convention. The technical sessions at the AES convention are not free, and neither ore the recordings. I don't know how soon after a session has happened the recordings will be available, but it shouldn't be too long, and this session is one of the first to take place.

It should be a good session, Rich Frembes has assembled a good panel of manufacturers and users.

Mac

If someone could post a link and the price of the recording that would be much apriciated (? i'm dutch so pardon my englisch)
Logged

Mac Kerr

  • Old enough to know better
  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 6333
  • Audio Plumber
Re: Crossover Shapes
« Reply #9 on: October 02, 2011, 03:25:51 pm »

If someone could post a link and the price of the recording that would be much apriciated (? i'm dutch so pardon my englisch)

HERE IS A LINK to the recordings from the last convention in NY. I assume the same vendor will be distributing the recordings this year.

Mac
Logged

Timo Beckman

  • Full Member
  • ***
  • Offline Offline
  • Posts: 141
  • Rational Acoustics/Isemcon/Fulcrum Acoustic NL
    • Timo Beckman Geluid
Re: Crossover Shapes
« Reply #10 on: October 02, 2011, 03:42:23 pm »

Tanxs a lot for the link .
Logged

Peter Morris

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1221
Re: Crossover Shapes
« Reply #11 on: October 02, 2011, 09:53:37 pm »

I have been saying this for years, but it bears repeating. The real loudspeaker design engineers working at the manufacturer will have superior knowledge about the drivers being used and box design. They will also generally have superior test equipment and facilities to measure actual box with actual driver response to make corrections.

...


+1 and as I said above ...
 http://www.qscaudio.com/products/speakers/KW_Series/
QSC have been there and done that ...

http://media.qscaudio.com/pdfs/manuals/KW_Series_user_manual_EN_revC.pdf

"DSP FeaturesThe KW Series features advanced DSP (digital signal processing) circuitry that performs many functions. Some functions are set at the design/ production level and are not user accessible. These functions include crossovers, time alignment, limiting and protection, thermal management
and a number of proprietary features. QSC has designed exclusive DSP functions that greatly enhance the capabilities and performance of the KW Series systems.
Proprietary DSP FunctionsExcursion Limiting: In addition to signal limiting to protect the amplifier and transducers from overload, the KW Series utilizes a proprietary limiter
that prevents woofer over-excursion. Over-excursion occurs when a voltage presented to the woofer causes the cone to physically travel too far. This builds up excessive heat, stresses the moving parts of the woofer, produces audible artifacts and distortion and reduces the woofer’s lifespan. The proprietary algorithm contained in Excursion Limiting prevents over-excursion. Voltages that will harm the woofer through over-excursion are reduced enough to prevent over-excursion without any audible compression, limiting or loss. Taking advantage of the Excursion Limiter, the DEEP™ (Digital Extension and Excursion Processing) algorithm functions as a highly musical and non-distorting low-frequency EQ circuit. More on the DEEP function is available in the EQ section of this manual.
Intrinsic Correction™: Introduced on QSC Concert/Touring products, Intrinsic Correction is a proprietary process and set of signal processing algorithms that address correctable characteristics of transducers and waveguides. The net result is that any KW Series system will present extraordinarily even and consistent energy throughout the physical listening area of the loudspeaker, resulting in a very musical, acoustically
transparent system."
Logged

Curtis H List (Too Tall)

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 88
Re: Crossover Shapes
« Reply #12 on: October 04, 2011, 07:38:21 pm »

Hello Folks,

Well, excitement stirs as we are preparing for the arrival of our new QSC KW rig. With the new rig, comes new processing toys. We have chosen the BSS OmniDrive. Where before, the trusty ole' Rane crossover did a phenomenal job, the new OmniDrive opens fun new doors of adventure, specifically, in Crossover Shapes.

The Rane didn't offer any options in shapes, you just dialed in the frequency and level of each band and added delay if needed to the sub stacks. Now, I read about all the different shapes you can use in your crossover design. Although I am looking forward to the added accuracy in flattening and optimizing the system, I am a little overwhelmed at all the different shapes, Butterworth, Bessel, Linkwitz-Riley, etc.

In reading about them, I am leaning towards the Linkwitz-Riley 24dB/Octave slope, but there is also the Butterworth. So, I thought I would check in with the knowledge pool and see what ya'll use and why? What applications call for certain shapes?

Can't wait to get it out and tuned, and any advice is, as always, greatly appreciated.

Thanks,

Adam P.

Hi Adam,
Just a few points to hit on.

First I am not encouraging that you rip into the settings QSC is using.
When working on a totally different generic processer the pitfall of not agreeing on what “Q” is not withstanding.

Do not get confused with the fancy names.
It would help if you think of Hertz as CPS (Cycles per Second).
It’s nice to have something named after you, but it is also more confusing until it becomes 2nd nature.

The difference between a Linkwitz-Riley and a Butterworth is mostly the shape of the knee.
The way the knee is drawn so that when you add a Butterworth high-pass and a Butterworth low-pass you get a mild bump (about 3dB) at the xover point.

If you use Linkwitz-Riley you get a flat line at the xover point.

So why would you ever want Butterworth if it makes a bump?
Well this is where “phase” comes in.

If the phase at the xover point for both high-pass and low-pass is about 90 degrees (or 270 degrees) you may have a 3dB hole. The Butterworth “hump” takes care of this nicely.

PS- It may be a 6dB hole. I just can’t remember right now.
 Sorry.

So when you measure drivers this is something you want to keep track of. Write it down as you go.
Hope this helps.


Logged

John Roberts {JR}

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 16216
  • Hickory, Mississippi, USA
    • Resotune
Re: Crossover Shapes
« Reply #13 on: October 04, 2011, 09:42:14 pm »

Hi Adam,
Just a few points to hit on.

First I am not encouraging that you rip into the settings QSC is using.
When working on a totally different generic processer the pitfall of not agreeing on what “Q” is not withstanding.

Do not get confused with the fancy names.
It would help if you think of Hertz as CPS (Cycles per Second).
It’s nice to have something named after you, but it is also more confusing until it becomes 2nd nature.

The difference between a Linkwitz-Riley and a Butterworth is mostly the shape of the knee.
The way the knee is drawn so that when you add a Butterworth high-pass and a Butterworth low-pass you get a mild bump (about 3dB) at the xover point.

If you use Linkwitz-Riley you get a flat line at the xover point.

So why would you ever want Butterworth if it makes a bump?
Well this is where “phase” comes in.

If the phase at the xover point for both high-pass and low-pass is about 90 degrees (or 270 degrees) you may have a 3dB hole. The Butterworth “hump” takes care of this nicely.

PS- It may be a 6dB hole. I just can’t remember right now.
 Sorry.

So when you measure drivers this is something you want to keep track of. Write it down as you go.
Hope this helps.

I'm not changing my position that amateurs should not second guess professionals with skills and tools, but to clarify a little on your short tutorial about filter alignments, and this may be confusing too, but 4 pole L-R are simply two 2pole Butterworth alignments cascaded in series. The 3dB or 6dB characteristic at tuning frequency is related to even or odd order (number of filter poles). Since each pole is + and - 45' at tuning (90' relative to each other) we get 90" relative for 1 pole, 180' relative for 2 pole, 270' for 3 pole, 360' for 4 pole, etc.  Odd order alignments sum to unity with -3dB at tuning because of the 90' relative offset. Even order crossovers like 4 pole L-R sum to unity with both sections down -6dB at tuning. (note: old school 2 pole crossovers were 180' out so required a polarity flip to not create a notch (suck out) at the exact crossover point. Since Butterworths are normally tuned for -3dB at the crossover frequency, stacking two 2 pole Butterworths in series, auto-magically gives us the -6dB desired for L-R.

This is the cliff-notes version, and I repeat the design engineers at the speaker companies generally know what they are doing, certainly more than me.     

JR
Logged
On the internet people tell you everything "they" know, not the answer to "your" question.....  http://circularscience.com/
Pages: 1 2 [All]   Go Up
 


Page created in 0.067 seconds with 24 queries.