ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1] 2 3 4   Go Down

Author Topic: question for the pro's on Comb Filtering  (Read 18514 times)

Sharyn Ferrick

  • Newbie
  • *
  • Offline Offline
  • Posts: 31
question for the pro's on Comb Filtering
« on: September 08, 2012, 10:23:57 PM »

SO we have been taught that Comb Filtering is a bad thing, but the question remains, in a live orchestra/band setting with no amplification, why is Comb  filtering not an issue between instruments of the same type. If you look at the violins, horns, bass etc it would look to the casual observer that there should be  major comb filtering, and by implication a negative impact on the sound
Sharyn
Logged

Mac Kerr

  • Old enough to know better
  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 7552
  • Audio Plumber
Re: question for the pro's on Comb Filtering
« Reply #1 on: September 08, 2012, 10:36:14 PM »

SO we have been taught that Comb Filtering is a bad thing, but the question remains, in a live orchestra/band setting with no amplification, why is Comb  filtering not an issue between instruments of the same type. If you look at the violins, horns, bass etc it would look to the casual observer that there should be  major comb filtering, and by implication a negative impact on the sound
Sharyn

For comb filtering to take place the two signals interacting with each other must be identical and close in level as happens with 2 speakers right next to each other. Neither of those conditions is likely to be met in an orchestra. If all the violins were exactly identical, you may as well have just 1 and amplify it. Even if they are close to identical, the interaction of that many sources will be very complex, masking any negative impact.

Mac
Logged

John Roberts {JR}

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 17176
  • Hickory, Mississippi, USA
    • Resotune
Re: question for the pro's on Comb Filtering
« Reply #2 on: September 09, 2012, 10:53:04 AM »

+1 what Mac said...

For comb filtering to be problem in sound reinforcement involves 2 identical copies of the same sound source. In an orchestra, if two instruments were identical enough, and played similarly enough to create any comb filtering, it would be on purpose and OK. While as Mac noted, they would never be perfectly identical.

There are many cases of naturally occurring comb filtering, that we don't even notice (like combing in our outer ears). Comb filtering between mismanaged identical sound sources, degrades the frequency response and is undesirable. 

JR
Logged
Cancel the "cancel culture". Do not participate in mob hatred.

Patrick Tracy

  • SR Forums
  • Hero Member
  • *
  • Offline Offline
  • Posts: 513
  • Boulder, CO, USA
    • Boulder Sound Guy
Re: question for the pro's on Comb Filtering
« Reply #3 on: September 09, 2012, 11:52:52 AM »

There is comb filtering, but because the sounds don't stay the same for more than a few cycles at a time the filtering is not constant enough to be noticed. A constant state or slowly changing interaction is more objectionable than a transient one.

George Friedman-Jimenez

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 532
  • NYC
Re: question for the pro's on Comb Filtering
« Reply #4 on: September 09, 2012, 12:40:04 PM »

+1 above comments. In a little more physical detail, comb filtering also requires the proper constant (in time) phase relationship between the 2 interfering sounds. This makes it very unlikely that, for example, 2 violins playing the same note would produce noticeable ongoing comb filtering since their relative phase differences would change with every note, and even during notes if their pitches were not exactly identical and relative SPLs and complex harmonic structures were not constant in time.
« Last Edit: September 09, 2012, 12:42:26 PM by George Friedman-Jimenez »
Logged

Sharyn Ferrick

  • Newbie
  • *
  • Offline Offline
  • Posts: 31
Re: question for the pro's on Comb Filtering
« Reply #5 on: September 09, 2012, 02:27:41 PM »

For comb filtering to take place the two signals interacting with each other must be identical and close in level as happens with 2 speakers right next to each other. Neither of those conditions is likely to be met in an orchestra. If all the violins were exactly identical, you may as well have just 1 and amplify it. Even if they are close to identical, the interaction of that many sources will be very complex, masking any negative impact.

Mac

So does that mean that some sort of subtle manipulation/alteration/Processing  of the sound from each speaker might avoid comb filtering?
For instance how important is level? delay? 
Sharyn
Logged

Mac Kerr

  • Old enough to know better
  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 7552
  • Audio Plumber
Re: question for the pro's on Comb Filtering
« Reply #6 on: September 09, 2012, 02:48:05 PM »

So does that mean that some sort of subtle manipulation/alteration/Processing  of the sound from each speaker might avoid comb filtering?
For instance how important is level? delay? 
Sharyn

Not likely. The best way to eliminate comb filtering between 2 speakers is to deploy them in a way that minimizes the effect. This can be done most easily by spreading the array to minimize overlap. This is why it is recommended to not use speakers with wide horizontal coverage if you plan on using more than 1 speaker per location.

Using multiple speakers in a array doesn't add much in overall level except at low frequencies, but it adds plenty of sound quality issues at mid and high frequencies. One speaker with the proper coverage will always sound better than multiples on that speaker in an array. Add speakers for coverage, not for loudness.

Mac
Logged

Ivan Beaver

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 9534
  • Atlanta GA
Re: question for the pro's on Comb Filtering
« Reply #7 on: September 09, 2012, 03:54:53 PM »

So does that mean that some sort of subtle manipulation/alteration/Processing  of the sound from each speaker might avoid comb filtering?
For instance how important is level? delay? 
Sharyn
It is IMPORTANT to truly understand what causes comb filtering.  maximum cancellation (and summation to some extent) occurs when 2 signals that are equal in level arrive at the listening position at different times. 

The difference in arrival times will determine both where the combfiltering starts and how wide it is between the "teeth" of the comb.

So it is the actual delay that CAUSES the notches.  Adding more delay will simply start it at a lower freq.

If the levels are not equal-then the interference (notches) will be less.  So to have the least amount of cancellation-turn one speaker completely OFF.  Hence the whole idea of using a SINGLE loudspeaker to cover a particular area.

When more than one physical loudspeaker (or arrivals of the same freq from different devices within a single loudspeaker) arrives at the same position-there will be combfiltering-since there is no way that 2 loudspeakers can occupy the same physical position in space.

The thing is most people don't even know what it sounds like.  They think it is some "horrible notchy" kind of thing.  They hear it all the time-and think nothing of it-UNTIL they hear the sound WITHOUT it.  NOW it sounds clearer-more distinct-open and all kinds of other terms.

While in reality-it makes the sound sound duller than when it isn't there.  And adding a HF boost is NOT going to fix the dull sound due to combfiltering-but that is what most people use to try and "fix" it.

You CANNOT fix time domain issues with eq (the energy part of the equation or "TEF cube").  Take a block and write the words TIME-ENERGY-FREQ on 3 sides.  Now when you turn the block around you can see how the different "measurements" can affect each other and what they can show.  Such as an impluse give time and energy-but no freq. Freq response give energy and frequency- but no time.  And so forth.
Logged
A complex question is easily answered by a simple-easy to understand WRONG answer!

Ivan Beaver
Danley Sound Labs

PHYSICS- NOT FADS!

Sharyn Ferrick

  • Newbie
  • *
  • Offline Offline
  • Posts: 31
Re: question for the pro's on Comb Filtering
« Reply #8 on: September 09, 2012, 05:44:13 PM »

Great information as usual.
The question came because I was looking at different Speaker placements
I currently have 6 EV zxa5 60x60. I have been using two of them l &r side-by-side angled for minimal overlap, and then the remaining two as delays further back in the venue on the sides,  All are mounted up high angling down for coverage.   I have avoided  placing 3 per side  in an "array" due to my understanding that this would introduce more comb filtering and  no advantage.  So these comments seem to confirm that.   I also have 4 LS800P's that currently I set up center stage and can adjust placement for some steering ability
Thanks
Sharyn
Logged

Mac Kerr

  • Old enough to know better
  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 7552
  • Audio Plumber
Re: question for the pro's on Comb Filtering
« Reply #9 on: September 09, 2012, 07:19:26 PM »

Great information as usual.
The question came because I was looking at different Speaker placements
I currently have 6 EV zxa5 60x60. I have been using two of them l &r side-by-side angled for minimal overlap, and then the remaining two as delays further back in the venue on the sides,  All are mounted up high angling down for coverage.   I have avoided  placing 3 per side  in an "array" due to my understanding that this would introduce more comb filtering and  no advantage.  So these comments seem to confirm that.   I also have 4 LS800P's that currently I set up center stage and can adjust placement for some steering ability
Thanks
Sharyn

You are currently doing the right thing.

Mac
Logged

ProSoundWeb Community

Re: question for the pro's on Comb Filtering
« Reply #9 on: September 09, 2012, 07:19:26 PM »


Pages: [1] 2 3 4   Go Up
 



Site Hosted By Ashdown Technologies, Inc.

Page created in 0.033 seconds with 24 queries.