ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: 1 2 [All]   Go Down

Author Topic: Parametric EQ Technique  (Read 6735 times)

Chris Carpenter

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 316
  • Baton Rouge
Parametric EQ Technique
« on: March 02, 2011, 01:38:26 am »

My current rig is a pair of Peavey 115Hs over a pair of PV118s with a cs800 on the tops and crown xls402 for the bottoms, crossed over with a PV23X0 crossover.

In the past, I have used 2x31 graphic EQs to tune my systems. However, when i put this system together, I grabbed a low end 2x16 digital parametric and measurement mic. My question regards technique. With the graphic, I am comfortable playing with the faders until it sounds decent. With the parametric, It is much more difficult to get everything to sound good. When I use smaart, i use relatively narrow filters to even out the response curve. However, when I drop the EQ in and out of the chain, there is little noticeable improvement. I have done some light searching, but most parametric eq essays describe using parametric to tame feedback, or are relevant only to recording. Where can I find useful tutorials or techniques for tuning a system using a parametric? Or should I go back to using graphic/no eq at all?

Logged

Minka Matikainen

  • Newbie
  • *
  • Offline Offline
  • Posts: 36
Re: Parametric EQ Technique
« Reply #1 on: March 02, 2011, 06:55:02 am »

Using parametric EQ is relatively easy, but at the same time it is quite different from using a GEQ.

When you use a GEQ, typically the width (or the "Q" as it is referred) of the frequency that is affected is dependent on the cut/boost level of the fader. The more cut/boost is used, the wider the Q-value turns to. There are exceptions in this, but they are typically labelled as "constant-Q EQs" (i.e. Rane has these).

When you use your PEQ, depending on the quality, the Q-value should be fixed and independent from the cut/boost.

How I use PEQ and GEQ on my system tuning is that I look for the nominal frequencies, resonances and clear "holes" in the individual boxes (actually you typically get some information on where and how much with which Q you should do from the manufacturer of that cabinet), which I fix with the PEQ. If I still have some bands left on my PEQ I save them to look for some modes in room etc.

Typically for me there are:
- Fixed tunings for cabinets that I do with PEQ and delays to make the system "straight" (incl. needed delays over Xover frequencies etc.), which are saved for the system itself
- 1-3 different bands of PEQ that I use for artistic purposes (depending on a band, I might boost some 67Hz on the whole system, or 100Hz or something similar, or I might cut down like 670Hz from the whole system if there is need for that in i.e. vocals, or I might cut the area of 3.15kHz if guitars are screaming like blade saws on that area). Depending on how much correction is needed PEQ could be the only tool for this (I don't have today any constant-Q GEQs).
- The basic tuning for smooth sound on any room (excluding the basic system tuning) I typically do by using the GEQ. Just remember that pulling down the low end from GEQ is not the same as using a real HPF for your system. I.e. DBX2231, which is pretty common and OK tool, will mess up your phase up to 1kHz if you pull the faders upto 40/50Hz totally down, like I've seen often done. Instead of doing this, apply a real HPF @ 40Hz (or similar, which works for your application) and leave these faders to "0".

What I would recommend you to do (unless you have the knowledge, and tools to really align the system), is to select a reasonable Q-value (2.5 - 3.2) and boost the band with some 6-10dB and start sweeping frequencies from low to high and when you hit to some frequency that sounds bad turn the gain to cut, and cut is as much as it disappears and system starts to sound reasonable. If you feel you are loosing something you don't like to loose, take the Q-value to higher level (narrower band). In this case you might need to change a bit your center frequency also.
Repeat this to another band and so on.

If you end up having a lot of bands close to each other, you might want to go and use wider band (lower Q-value) to release some of the bands and to make the phase a bit more "correct" (especially the low end tools tends to mess with phase, and so do the more expensive ones also, when you make too many things with them)...

If you find something that sounds good (adds punch, clarity etc.) don't cut that frequency, but you might even think about adding it a bit... ...just keep on mind that adding frequencies in system PEQ/GEQ will lower the level of system overdriving in a similar way as too many cuts will eat the signal-to-noise ratio.

Hope this helps,
-Minka-
Logged
95% of touring is work, and if you're lucky that 5% of the time is working with gear that are in good shape and according to rider...

Tom Young

  • Global Moderator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 620
Re: Parametric EQ Technique
« Reply #2 on: March 02, 2011, 07:34:10 am »

My current rig is a pair of Peavey 115Hs over a pair of PV118s with a cs800 on the tops and crown xls402 for the bottoms, crossed over with a PV23X0 crossover.

In the past, I have used 2x31 graphic EQs to tune my systems. However, when i put this system together, I grabbed a low end 2x16 digital parametric and measurement mic. My question regards technique. With the graphic, I am comfortable playing with the faders until it sounds decent. With the parametric, It is much more difficult to get everything to sound good. When I use smaart, i use relatively narrow filters to even out the response curve. However, when I drop the EQ in and out of the chain, there is little noticeable improvement. I have done some light searching, but most parametric eq essays describe using parametric to tame feedback, or are relevant only to recording. Where can I find useful tutorials or techniques for tuning a system using a parametric? Or should I go back to using graphic/no eq at all?

When you are measuring with Smaart and apply a filter or several filters with the parametric EQ, do you see the results in your Smaart trace ?

Have you read the Smaart tutorials ? Do you employ smoothing ?

Probably the best (and only) book that will help you with all aspects of loudspeaker system measurement is Bob McCarthy's excellent treatise on (and titled) sound system design and optimization.
Logged
Tom Young
Electroacoustic Design Services
Oxford CT
203-888-6217

Doug Fowler

  • Member since May 1995, 2nd poster on original LAB, moderator on and off since 1997, now running TurboMOD v1.826
  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1695
  • Saint Louis, MO USA
Re: Parametric EQ Technique
« Reply #3 on: March 02, 2011, 10:10:15 am »

My current rig is a pair of Peavey 115Hs over a pair of PV118s with a cs800 on the tops and crown xls402 for the bottoms, crossed over with a PV23X0 crossover.

In the past, I have used 2x31 graphic EQs to tune my systems. However, when i put this system together, I grabbed a low end 2x16 digital parametric and measurement mic. My question regards technique. With the graphic, I am comfortable playing with the faders until it sounds decent. With the parametric, It is much more difficult to get everything to sound good. When I use smaart, i use relatively narrow filters to even out the response curve. However, when I drop the EQ in and out of the chain, there is little noticeable improvement. I have done some light searching, but most parametric eq essays describe using parametric to tame feedback, or are relevant only to recording. Where can I find useful tutorials or techniques for tuning a system using a parametric? Or should I go back to using graphic/no eq at all?


Post a screenshot of the magnitude display, if you please.

Logged
"It's got electrolytes.  It's got what plants crave."

Chris Carpenter

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 316
  • Baton Rouge
Re: Parametric EQ Technique
« Reply #4 on: March 02, 2011, 12:06:05 pm »

Minka: Thanks for the advice. I read about the sweeping technique after I posted the topic. I'll definately give it a try. My EQ's Q value is measured in 60ths of an octave; are the values you posted your Q in the reciprocal of the octive (ex. Is Q 2.5 = 1/2.5 of an octive?)?

Tom: Yes, I can see a definitive link between adjustments on the filters and the graph. My issue is that once Ive evened out the response as best I can, the system doesn't seem to sound any better. Different, sure, but it doesnt sound any closer to my linear monitors or headphones. I've read a good many tutorials and I do use smoothing. I havent glanced through design and optimization yet, but it is definately on my to do list.

Doug: Ill try to do that tonight.
 
I also set up a meeting with one of our theater tech instructors. I'm thinking maybe I can get them to let me use their high end measurement mic and rta. Did I mention this is all stored in a dorm?  8)
Logged

Doug Fowler

  • Member since May 1995, 2nd poster on original LAB, moderator on and off since 1997, now running TurboMOD v1.826
  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1695
  • Saint Louis, MO USA
Re: Parametric EQ Technique
« Reply #5 on: March 02, 2011, 12:19:47 pm »

Minka: Thanks for the advice. I read about the sweeping technique after I posted the topic. I'll definately give it a try. My EQ's Q value is measured in 60ths of an octave; are the values you posted your Q in the reciprocal of the octive (ex. Is Q 2.5 = 1/2.5 of an octive?)?

Tom: Yes, I can see a definitive link between adjustments on the filters and the graph. My issue is that once Ive evened out the response as best I can, the system doesn't seem to sound any better. Different, sure, but it doesnt sound any closer to my linear monitors or headphones. I've read a good many tutorials and I do use smoothing. I havent glanced through design and optimization yet, but it is definately on my to do list.

Doug: Ill try to do that tonight.
 
I also set up a meeting with one of our theater tech instructors. I'm thinking maybe I can get them to let me use their high end measurement mic and rta. Did I mention this is all stored in a dorm?  8)

Chris -

If this is your system configuration, let me offer a guess:

The interference between the top boxes is creating nulls that cannot be equalized.  If this is indeed your config, and you trying to equalize it, it does not surprise me you do not hear much improvement.  You have built yourself a sonic trainwreck that has problems in the time domain, and you are trying to fix it in the frequency domain. 

And that's a dead end, every time.

Logged
"It's got electrolytes.  It's got what plants crave."

g'bye, Dick Rees

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 7424
  • Duluth
Re: Parametric EQ Technique
« Reply #6 on: March 02, 2011, 12:20:20 pm »

Minka: Thanks for the advice. I read about the sweeping technique after I posted the topic. I'll definately give it a try. My EQ's Q value is measured in 60ths of an octave; are the values you posted your Q in the reciprocal of the octive (ex. Is Q 2.5 = 1/2.5 of an octive?)?



http://en.wikipedia.org/wiki/Equalization_filter

http://www.sengpielaudio.com/calculator-bandwidth.htm
« Last Edit: March 02, 2011, 12:35:54 pm by dick rees »
Logged
Pay no attention to the man behind the curtain...

Chris Carpenter

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 316
  • Baton Rouge
Re: Parametric EQ Technique
« Reply #7 on: March 02, 2011, 02:54:06 pm »

Chris -

If this is your system configuration, let me offer a guess:

The interference between the top boxes is creating nulls that cannot be equalized.  If this is indeed your config, and you trying to equalize it, it does not surprise me you do not hear much improvement.  You have built yourself a sonic trainwreck that has problems in the time domain, and you are trying to fix it in the frequency domain. 

And that's a dead end, every time.

Oh no, that's certainly not my setup configuration. That's just a funny pic showing how the system is floor to ceiling in my dorm.
Logged

Chris Carpenter

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 316
  • Baton Rouge
Re: Parametric EQ Technique
« Reply #8 on: March 02, 2011, 02:56:48 pm »

Logged

Minka Matikainen

  • Newbie
  • *
  • Offline Offline
  • Posts: 36
Re: Parametric EQ Technique
« Reply #9 on: March 03, 2011, 01:22:09 pm »

http://en.wikipedia.org/wiki/Equalization_filter

http://www.sengpielaudio.com/calculator-bandwidth.htm

Thanks Dick, good links for future use for me also (now I don't have to explain it myself anymore)! You beat me  :)

What comes to Doug's comment on problems in time domain, it is most likely one of your issues even, if that is not your setup.

I don't say that the top box would have a lot of issues by itself, but when you combine it to the subwoofer, you will start to get all the funny things in the area of crossover frequency.

There are systems that have been developed so that if you match the cabinets' front ends to each other, the xover point will be in good shape within time domain (and frequency, gain etc. point of view also).

If you are hooking up a system yourself and it is not a system that people at manufacturer have thought together (and system is this case includes all the things: cabinet, amps, xover/DSP settings...), you need to align the system so that it starts to work as a system, instead of 2 cabinets thrown together with high hopes.

fwiw,
-Minka-
Logged
95% of touring is work, and if you're lucky that 5% of the time is working with gear that are in good shape and according to rider...

Chris Carpenter

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 316
  • Baton Rouge
Re: Parametric EQ Technique
« Reply #10 on: March 04, 2011, 06:44:59 pm »

I think I got it figured out now. First, I had some gain issues with my interface. Now that that is fixed, the rta works much better. I attached my frequency response after tuning. I used the rta to level it out as best I could, then used the sweep technique to notch a couple frequencies that sounded bad. You can see a bit of a notch at 130Hz, where my crossover is set. Ill attribute that to the alignment discussed earlier, but i don't really have the gear to fix it now.

Oh, and I think the roll off at the top is due to the radioshack meter I'm using. Every one I've used does that.
« Last Edit: March 04, 2011, 06:46:55 pm by Chris Carpenter »
Logged
Pages: 1 2 [All]   Go Up
 


Page created in 0.059 seconds with 24 queries.