Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Ivan Beaver]

  Listening to those 2 wedges nothing as obvious as the same sound reaching my ear at 2 different times jumped out at me.  Of course it's not like I was playing playback music through them when I was setting them up and it's not like I was expecting to hear "The old radio playing at 2 different ends of a long field effect".   I will be listening for it now though.  And it occurs to me that I probably ought to invest in a couple of things.  Like a way to send pink noise through my system?   Sine wave generator maybe? RTA mic?  I am fixing to retire the mixwiz and invest in a digital board, which will be a learning curve in itself.

Anyway, again, thanks for taking the time to explain this stuff to me even though I had an attitude.  I probably wouldn't have if I were you.

The "combfiltering effect" can easily fool lots of people. It is not that it sounds bad, but rather that without the notches it simply sounds a lot better/clearer etc.

Since our "acoustic memory" is so short, it is easy to forget what it is supposed to sound like.

That is why people say that it is not a problem.  Yeah-not a show stopper, but when it is correct, it is simple better.

As far as getting some pink noise-I am sure there are plenty of places you can download a wavefile of it.

Any RTA should have it built in that you can use.  And before we go there, an RTA will most likely not be able to show the effects of combfiltering-unless you use 1/24th resolution or so.  1/3rd oct is not fine enough.

[Tim Weaver]

Smartphones have replaced a whole toolbox worth of gadgets for me.

If you have a ipod, iphone, or android you can do a bunch of tests.

I use:
SpeakerPop (polarity checker)
AudioCalc (audio math)
RTA (an RTA)
Signal scope (FFT and O-scope)
SPL meter (looks like the rat-shack meter)
DipSwitch (for setting DMX dip switch address)
iHandy Level (bubble level)
Mocha VNC (great VNC app)

[g'bye, Dick Rees]

Shmorns in, shmorns out...

I use concentric wedges.

[Ivan Beaver]

Shmorns in, shmorns out...

I use concentric wedges.

Which reduces the interference even more.

The polar pattern remains the same and the cancellations that happen when you have a separate horn and woofer are gone.

And when you try to use 2 coax drivers, you have 2 "correct" sources so the interference is less than when you have 2 "incorrect" sources.

With a "incorrect" source you have problems within a single cabinet-and then  you add those problems to another one that has problems, so the total is worse, than if you just had 2 correct sources.

The same thing thing goes for full range cabinets.  The better/more correct each source is, the less interference there will be when you have to add to together.

OF course not all drivers allow for a proper crossover/transition.

I have tried to design some crossovers on some coaxes and the driver physical position is simply in the wrong place to get good summation.

So not all are created equal.

[Scott Wagner]

This is all great and fine. I'm sure we have all learned something here. I have. But we all know it's virtually impossible to get a singer or musician to grasp this concept. They really don't care about the laws of physics during sound check. They just want their 2 wedges (or more) just like every other gig they do, crank up the amps to 11, beat the crap out of the drums and yell at the sound guy when they can't hear the monitors.

I've tried. All I get is a stupid look. So I throw down 2 wedges and deal with it. I don't like it either.

Give them both boxes, but only turn one on. Works for me all the time... You get the "look factor" without any of the problems.

[Steve Bradbury]

If only it moves as fast as light.  Then we could just do whatever we wanted-pile up as many speakers as we want and it would just get louder and louder and not sound worse.

Ignoring for a moment the wave/particle duality nature of light, and just considering it as a wave, light is no different from sound. Add two light sources together and they will interfere producing, in the right circumstances, interference patterns. For further information look up Young’s dual slit experiment, Fresnel biprism or Newton’s rings.

 

Take 2 small single, driver full range speakers and put one on top of the other.  Put them in parallel to each other and listen at some distance away, 5' 10' 20' whatever with pink noise applied.

Have a friend move one of them back even 1 or 2 inches.

The difference SHOULD be astounding-and quite enlightening.

When two or more sound waves combine, the resultant pressure is the vector sum of the individual pressures. The vector bit just means that you consider polarity or sign, a pressure of +3 (pick your own units) combined with a pressure of -2 sums to +1. Frequency and time delays don’t come in to this, multiple sources all add the same way.

However, with identical signals, especially simple ones like sine waves, the pressure at any time and point is predictable. Start at zero or minimum then you get +max, zero, -max, zero and then go round again. With two or more static sources, the relative points along the wave (or phase) are always the same. When added the places where the signals add constructively (+ add +) always stay the same and so do the places where they add destructively (+ add -). This leads to the interference patterns that we love to talk about and call combfiltering.

Noise of any colour is supposed to be random. If you sum two random numbers you get another random number. If you delay one of the sources you just add two different random numbers which again combine to a random number. Random means there is no phase information. The chances of interference patterns or combfiltering effects will be the same with or without delay; it is all random. The experiment proposed above should not appreciable alter the sound.

The following image was created in Adobe Audition and simulates two sources with and without delay added to one of them. If I can find an easy way of uploading the audio file I will do so. Two pink noise sources were used (one for the left clip the other for the right clip). The noise was loaded into two tracks and mixed down to a final track. Half way through each clip one of the tracks was delayed 1msec (about 1ft delay) by splitting the clip and moving it along the timeline. Both tracks are mixed together at all times.



The clip on the left shows considerable cancelation at specific frequencies and the result is clearly audible. The clip on the right shows no change when a delay is added and sounds similar all the way through.

Note that the original tracks mixed were mono, the picture above shows the stereo mixdown not the two mono tracks, this is because I didn’t change the default mixdown setting. This has no effect on the result.

This is not a trick and the only difference is how the pink noise was generated. Two seemingly identical experiments produce two completely different results. Without being clear what you are measuring the results can sometimes misleading.

Horns of any size side by side (even touching) will have all kinds of cancellation problems, because the drivers are so far apart.

If you have two drive units coupled to two wave guides and arrange them in two configurations as shown below, i.e. drive units together or drive units apart? Which setup will give the better results?

For the purpose of this experiment any deficiencies in the response of the wave guide are ignored and assume that the propagation time of the wave from diaphragm to exit are identical.

[Ivan Beaver]

Ignoring for a moment the wave/particle duality nature of light, and just considering it as a wave, light is no different from sound. Add two light sources together and they will interfere producing, in the right circumstances, interference patterns. For further information look up Young’s dual slit experiment, Fresnel biprism or Newton’s rings.

 
When two or more sound waves combine, the resultant pressure is the vector sum of the individual pressures. The vector bit just means that you consider polarity or sign, a pressure of +3 (pick your own units) combined with a pressure of -2 sums to +1. Frequency and time delays don’t come in to this, multiple sources all add the same way.

However, with identical signals, especially simple ones like sine waves, the pressure at any time and point is predictable. Start at zero or minimum then you get +max, zero, -max, zero and then go round again. With two or more static sources, the relative points along the wave (or phase) are always the same. When added the places where the signals add constructively (+ add +) always stay the same and so do the places where they add destructively (+ add -). This leads to the interference patterns that we love to talk about and call combfiltering.

Noise of any colour is supposed to be random. If you sum two random numbers you get another random number. If you delay one of the sources you just add two different random numbers which again combine to a random number. Random means there is no phase information. The chances of interference patterns or combfiltering effects will be the same with or without delay; it is all random. The experiment proposed above should not appreciable alter the sound.

The following image was created in Adobe Audition and simulates two sources with and without delay added to one of them. If I can find an easy way of uploading the audio file I will do so. Two pink noise sources were used (one for the left clip the other for the right clip). The noise was loaded into two tracks and mixed down to a final track. Half way through each clip one of the tracks was delayed 1msec (about 1ft delay) by splitting the clip and moving it along the timeline. Both tracks are mixed together at all times.



The clip on the left shows considerable cancelation at specific frequencies and the result is clearly audible. The clip on the right shows no change when a delay is added and sounds similar all the way through.

Note that the original tracks mixed were mono, the picture above shows the stereo mixdown not the two mono tracks, this is because I didn’t change the default mixdown setting. This has no effect on the result.

This is not a trick and the only difference is how the pink noise was generated. Two seemingly identical experiments produce two completely different results. Without being clear what you are measuring the results can sometimes misleading.

If you have two drive units coupled to two wave guides and arrange them in two configurations as shown below, i.e. drive units together or drive units apart? Which setup will give the better results?

For the purpose of this experiment any deficiencies in the response of the wave guide are ignored and assume that the propagation time of the wave from diaphragm to exit are identical.

While it is fun to play in models in the computer-REAL LIFE TRUMPS any model, every time.

If you simply do my little test-it will be VERY EASILY to hear the differences.  If it is not-then video or lighting is the career for you.

I have done it LOTS OF TIMES-and the results are VERY noticeable (every single time)-even with a distance difference of an inch or so you can easily hear the notches/holes in the response.

In your horn situation, the one that the drivers are covering different areas is the one that will work best.  That way they are not interfering with each other (assuming the horns are large enough to control the lowest freq of interest).

The one that is combining both drivers together has a whole set of problems.

There have been various methods of combining Hf drivers on a single horn in the past (looking like your combiner).  While they get a little bit louder-it is not as much as people would want it to be-due to all of the cancellations going on in the combiner.  And the sound quality is lower (again due to the cancellations).

It can be done (combining HF drivers)-but NOT by the method you show.

[John L Nobile]

What would you call the whooshing sound when you walk towards a line array? It comes and goes. Seems to be most pronounced when you're directly in line where 2 cabinets join and very obvious with cymbals. Or is that just a bad design?

[Luke Geis]

A line array is far from a perfect tool. You are actually hearing the interaction between the two boxes. You will also notice as you walk closer or further away from the array that there is a shift in the frequency that sound like a wah wah pedal slowly being opened or closed. This is the Q shift as the distance from each of the boxes changes in relation to you. As you get further away the boxes are more equidistant in relation to you. But as you get closer the tp boxes become further in relation and you hear that interaction.

I had a recent show where the artist requested two boxes for his monitor mix, I said ok and orientated them to what I felt would give the best results. He then completely rearranged them futher apart and pointing inwards. I said that I don't think you will like that very much and that he would actually increase feedback potential. He asked if we could try it anyway and that he would like to hear why that is. After about five minutes of him complaining about why the sound seems to continually move and not getting as much level as he would like before feedback, I finally turned one off and moved it directly in front of him. He instantly said it was better. After sound check he asked the why question.

The mic has a pickup pattern. In this case a heart shaped pattern with a null directly behind the mic. This is the ideal spot to place the monitor, in the mics null. If you had a hyper mic with the nulls off to the sides then a dual monitor setup may be ideal? Then I explained that there is now two sources of identical audio coming from two places at the same time. If you so much as move a little bit you will hear the difference in time between the two and get the constant swishing of sound as it combs in or out. So in this case having two monitors pointed at a part of the mic that it can actually pick up the monitors output, combined with the comb filtering when you move is why I was reluctant to set it up that way. He got to hear first hand and relate the experience to the real world. He got what he wanted and so did I.

An artist that actually believes he can get more level when using two wedges is probably only half true. I would say that it appears louder because you are pointing sound directly at each ear. Now whether or not that artist realizes, recognizes or even cares about the comb filtering is a whole different issue. If they don't care, they don't care. Many probably hear it and don't even realize or know that they hear it? It is what it is and they just deal with it. I have found that if you can't get the GBF you need with a single box setup, having two won't make it any better. You may get the level you want, but the system as a whole will be less stable doing it.

[Ivan Beaver]

What would you call the whooshing sound when you walk towards a line array? It comes and goes. Seems to be most pronounced when you're directly in line where 2 cabinets join and very obvious with cymbals. Or is that just a bad design?

It is the "argument" that line array lover like to make.

They say that "point and shoot" boxes have the interference in the horizontal so you hear it when you walk side to side.

But since the interference in a line array is in the vertical you only hear it when climbing up and down a ladder Ha HA. Funny Funny.

HOWEVER-more people walk front to back than side to side anyway.  And when you walk front to back-you DO hear the vertical interactions.

If you were to stop at any one point, you will have the particular notches at particular freq-based on the different signal arrivals from the different cabinets.

That is until the wind starts to blow.  THEN you can hear the interactions without having to walk around.  The wind (temp gradients) will simply bring it to you and take it away at will.  You don't have to do a thing to hear all of the different arrivals.

[ProSoundWeb Community]