Beyond the Ariel - Page 145 - diyAudio
Go Back   Home > Forums > Loudspeakers > Multi-Way

Multi-Way Conventional loudspeakers with crossovers

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 5th July 2007, 08:17 AM   #1441
gedlee is offline gedlee  United States
diyAudio Member
 
Join Date: Dec 2004
Location: Novi, Michigan
Quote:
Originally posted by john k...


While I haven't worked it through, it would seem obvious that how the distortion increases would depend on the duct (horn. pipe, etc) geometry. One thing is for sure, it isn't going down. Any way, probably why they don't show the result is because it shows that most of the distortion is generated in the initial part of the horn. For example, if an infinitely long exponential horn increases the distortion by a factor of X, then a horn with mL/2 = 1, where m is the flare constant, will increase distortion by a factor of 0.63X. If mL/2 = 2, 0.86X, mL/2 = 4, 0.98X. Anything longer that mL/2 =4 doesn't contribute much to distortion, according the Beranek's result. Beranek's result also indicates that distortion goes up linearly as frequency goes up. I would imagine more modern horn theory would alter the magnitude of these results somewhat. Don’t hold me to this as I’m just quoting from Beranek’s 50+ year old text.

Fundamentally I would suspect that all of this is a result of the fact that while acoustics is based on linear theory, the assumption of linearity is just an approximation. It's a good approximation at normal SP levels, but the nonlinearity is ever present. The thermodynamics of wave propagation says that compression waves travel faster than expansion waves. In a lossless system, compression waves build upon themselves where as expansion wave disperse. So there is a tendency for sin waves to move towards something like saw tooth waves as it travels down a duct. Certainly that would represent a lot of distortion. However, at normal acoustic pressures the distance required is probably quite great and over that distance other nonlinearities most likely need to be considered, like dissipation.
I did take a lot of nonlinear acoustics theory in school because it is of paramount importance underwater. And yes a wavefront does steepen as it propagates, but not at an excessively high rate in air (unlike underwater - water being 1000 times less linear than air). And you are right that the Beranek work is old enough so as to be misleading if not downright incorrect. The distortion does not increase linearly with frequency, but then again, the statement is somewhat ambiguous so I am not sure what is really meant.

The facts of nonlinearity in air are these. In a waveguide that falres rapidly - as the OS does - the distortion is generated in only the first inch or so - not much time for these nonlinearities to grow. But the important point is that the nonlinearity is very low order, which, acording to our results and anyone else who has studied the problem, are virtually inaudible. My bet is on there not being any audible nonlinear distortion in a well designed waveguide well up into the sound levels at which the driver will fail. There are serious problems with diffraction in waveguides and these do grow with SPL level so that waveguides are not without their problems, but this is NOT due to the nonlinearity of the air. I believe that Alex Voishvillo (JBL Pro) is now also of this same opinion.
  Reply With Quote
Old 5th July 2007, 09:35 PM   #1442
Hatti is offline Hatti  Germany
diyAudio Member
 
Join Date: May 2007
Location: Germany Hannover
Hi, a lot of theorie´s in all this post´s! And it take a lot of time to read, the most of it...

I add a little more!

I wouldn´t use speaker that are bigger than 8" All this "pro" stuff has high efficient, but the cones are to heavy and to slow for high quality listen!

Also, you can built a fullrange OB without crossover and without filter! You have much more directly information and you are nearer to the original.

My OB is standing in a small room, 2 meter from the back wall and 2 meter to hear position. By this long way; the reflexion from the back wall isn´t so much importend!!!

I show my OB here in the fullrange forum: My 4x8" fullrange open baffle

Have a nice time and good luck to your OB-Projekt, Hatti
__________________
The step from genius to crazy, is very small! Please call me crazy, so I don´t have to declare!
  Reply With Quote
Old 6th July 2007, 02:46 AM   #1443
diyAudio Member
 
Lynn Olson's Avatar
 
Join Date: Jul 2004
Location: Northern Colorado
Quote:
Originally posted by gedlee


I did take a lot of nonlinear acoustics theory in school because it is of paramount importance underwater. And yes a wavefront does steepen as it propagates, but not at an excessively high rate in air (unlike underwater - water being 1000 times less linear than air). And you are right that the Beranek work is old enough so as to be misleading if not downright incorrect. The distortion does not increase linearly with frequency, but then again, the statement is somewhat ambiguous so I am not sure what is really meant.

The facts of nonlinearity in air are these. In a waveguide that falres rapidly - as the OS does - the distortion is generated in only the first inch or so - not much time for these nonlinearities to grow. But the important point is that the nonlinearity is very low order, which, acording to our results and anyone else who has studied the problem, are virtually inaudible. My bet is on there not being any audible nonlinear distortion in a well designed waveguide well up into the sound levels at which the driver will fail. There are serious problems with diffraction in waveguides and these do grow with SPL level so that waveguides are not without their problems, but this is NOT due to the nonlinearity of the air. I believe that Alex Voishvillo (JBL Pro) is now also of this same opinion.
Dr. Geddes, since you're the only one in this forum that's studied nonlinear acoustics in air and water, I have a question for you: just how bad is that "first inch or so" in a waveguide or horn?

At a SPL of 110 dB at one meter in the frequency range of 1 to 5 kHz, what is the magnitude of the purely acoustic distortion? Is it -40 dB relative to the undistorted component, -60 dB, -100 dB or what? Curious about the magnitudes involved, and their spectral appearance.

Is the distortion all even-order harmonics (2nd, 4th, 6th, etc.) or is the harmonic structure more complicated than that? Is is a noiselike modulation, or is it just harmonic components?
  Reply With Quote
Old 6th July 2007, 04:04 AM   #1444
gedlee is offline gedlee  United States
diyAudio Member
 
Join Date: Dec 2004
Location: Novi, Michigan
Quote:
Originally posted by Lynn Olson


Dr. Geddes, since you're the only one in this forum that's studied nonlinear acoustics in air and water, I have a question for you: just how bad is that "first inch or so" in a waveguide or horn?

At a SPL of 110 dB at one meter in the frequency range of 1 to 5 kHz, what is the magnitude of the purely acoustic distortion? Is it -40 dB relative to the undistorted component, -60 dB, -100 dB or what? Curious about the magnitudes involved, and their spectral appearance.

Is the distortion all even-order harmonics (2nd, 4th, 6th, etc.) or is the harmonic structure more complicated than that? Is is a noiselike modulation, or is it just harmonic components?
The nonlinearity in air is purely 2nd order (x^2), so it will only generate a second order harmonic (2x). It is purely harmonic distortion, no random modulation, but like all nonlinearities, it will have modulation effects between two sign waves of different frequencies. Unlike the nonlinearity of an amp stage, this nonlinearity acts continuously as the wave propagates. This means that the 2X that was generated early on will have another 2X generated to create a 4X, then an 8X, etc. So the longer the wave propagates at intense SPLs the greater the effective "order" will be. For a short propagtion of a resonably intense wave, there will only be a second harmonic of any magnitude.

The levels of these harmonics is highly dependent on specifics, but an intense sound wave generating about 100 dB at say six meters, in a OS waveguide, which has a fairly rapid flare, will have a 2nd harmonic about 20 dB down. At about 90 dB it will fall dramatically to about 40 dB down. The higher harmonics will fall faster than the lower ones.

In essence this kind of distortion is a sort of brick wall limitation. It will become very objectionable very quickly when the sound becomes very intense, but its effect will be negligable below that point. I'd say that to my ear, in a Summa, this point is reached at about 110 dB SPL in my room. Above that level and I can't take the sound anymore. Now this could be caused by many many things however, but at about 110 dB SPL, the system crashes. Unlike many systems, which gradualy start sounding worse, the Summas seem to hold up very well until they crash, and all of a sudden the sound is bad. This is not a typical effect in my experience.

If you want more than 110 dB SPL, then you need to get a pair (And in a few years some hearing aids!!)
  Reply With Quote
Old 6th July 2007, 06:45 AM   #1445
diyAudio Member
 
Lynn Olson's Avatar
 
Join Date: Jul 2004
Location: Northern Colorado
Dr. Geddes, the answer is much appreciated. This gives me a good idea what to look for when I start measuring.

Hmm ... looking forward to getting SoundEasy running on my PC, and examining the spectral characteristics of a compression driver on a waveguide. The information about an abrupt "brick wall" is especially interesting - it's at a very high level, but it's there. I would guess on a generic PA-grade phasing plug the onset of the "brick wall" could easily be at a lower level.

I'm beginning to wonder if there really is some kind of modulation noise with compression drivers - a measuring system with a 100~120 dB dynamic range should reveal if it's hiding in the shadows. This kind of program modulated "noise" is troublesome in digital systems where the jitter gets into the DAC convertor chip, which typically has little resistance to this kind of error, and translates the time distortion into amplitude distortion. (When sample widths at the encoding and decoding ends are different, even by very small amounts, amplitude distortion results, and can measured on high-dynamic range measuring system. The spectral artifacts frequently exhibit close-in 100 or 120 Hz sidebands from PS-induced jitter.)

At the Physical Training clinic today, I used a cane for the first time, and got about 100 feet before the onset of fatigue. Also some stair-training on low stairs, getting me ready for going up and down stairs without going on my butt and using my arms to lift myself up and down. Maybe by late summer I can finally get some gear downstairs in the basement and start making some measurements on the PC with the ACO Pacific 1/2" microphone. These are things I want to see for myself.

Gary Pimm and John Atwood have shown me how to make close-in spectral measurements with 100~120 dB on-screen dynamic range, reach into the clutter, and look for spectral components lurking in the noise. Fairly standard stuff for amplifier design, where 100/120 Hz IM sidebands can appear in all kinds of places.
  Reply With Quote
Old 6th July 2007, 07:46 AM   #1446
gedlee is offline gedlee  United States
diyAudio Member
 
Join Date: Dec 2004
Location: Novi, Michigan
Quote:
Originally posted by Lynn Olson
I would guess on a generic PA-grade phasing plug the onset of the "brick wall" could easily be at a lower level.

I'm beginning to wonder if there really is some kind of modulation noise with compression drivers -

This effect was hypothesized when I was working with B&C. We set up a subjective test to scale its importance - the results were published in JAES about 18 months ago.

We found that in about 30 blind listeners no one could detect nonlinear "distortions" in a compression driver at any level up to the thermal limit of the devices. We tested across three manufacturers. In essence nonlinear distortions were not a factor in audiblity, however there were statistically significant audible frequency response differences.

At that time, I had a patent application on a way to reduce turbulence in a compression driver thus reducing random flow modulation effects. After the results of this study I abandoned this application - what would be the point of a patent solving a non-problem?
  Reply With Quote
Old 6th July 2007, 10:10 AM   #1447
mige0 is offline mige0  Austria
diyAudio Member
 
mige0's Avatar
 
Join Date: Jan 2005
Location: Austria, at a beautiful place right in the heart of the Alps.
Hi

Quote:
The nonlinearity in air is purely 2nd order (x^2), so it will only generate a second order harmonic (2x). It is purely harmonic distortion, no random modulation, but like all nonlinearities, it will have modulation effects between two sign waves of different frequencies. Unlike the nonlinearity of an amp stage, this nonlinearity acts continuously as the wave propagates. This means that the 2X that was generated early on will have another 2X generated to create a 4X, then an 8X, etc. So the longer the wave propagates at intense SPLs the greater the effective "order" will be. For a short propagtion of a resonably intense wave, there will only be a second harmonic of any magnitude.

Earl, what you describe here is quite an unique behaviour I wasn't aware of as yet. Thanks for pointing out so clearly.

Interesting that in that blind listening tests no one was able to discover. It would mean that mankind DIDN't evolve kind of crosscorrelation detector for that to gain additional information about the surrounding where the sound was originating from.
Not that bad, just another weird theory to bury




Quote:
In essence this kind of distortion is a sort of brick wall limitation....... If you want more than 110 dB SPL, then you need to get a pair (And in a few years some hearing aids!!)
110dB ( A-weighted LEQ ) at 6m is quite common at even small pop concerts not necessarily driven by huge multiple driver line arrays but rather by a single (or double) driver stack.

Considering that OS wave guides are MUCH more shallow than the usually used horns, I would expect MUCH more 2nd HD than the 10% you measured on OS's and subsequently also considerable amounts of higher order HD, am I right in this?


Greetings
Michael
  Reply With Quote
Old 6th July 2007, 10:36 AM   #1448
gedlee is offline gedlee  United States
diyAudio Member
 
Join Date: Dec 2004
Location: Novi, Michigan
Quote:
Originally posted by mige0

Interesting that in that blind listening tests no one was able to discover. It would mean that mankind DIDN't evolve kind of crosscorrelation detector for that to gain additional information about the surrounding where the sound was originating from.

110dB ( A-weighted LEQ ) at 6m is quite common at even small pop concerts not necessarily driven by huge multiple driver line arrays but rather by a single (or double) driver stack.

Considering that OS wave guides are MUCH more shallow than the usually used horns, I would expect MUCH more 2nd HD than the 10% you measured on OS's and subsequently also considerable higher order HD, am I right in this?

The reason that we can't hear low orders of distortion is that the ear masks these things. In nature there is no transmission path that would generate these nonlinear effects and so our hearing would have no need to accomodate them.

110 dB is not too common in a small room such as we are talking about here.

Actually the OS situation is the opposite of wht you are saying. OS waveguides flare rather rapidly minimizing the sound level and the nonlinearity very quickly. Virtually all of the nonlinearity is in the compression driver as these flare very slowly. But we showed that this distortion was not significant, so the distortion added by the OS waveguide would not be significant either.

Nonlinear distortion in these devices (drivers and waveguides) is simply not a factor at any but the very highest SPLs. Far more important are the diffractions in the waveguides (and drivers), but certainly in typical horns with intentional diffraction for directivity control. The diffraction effects are also level dependent and more than likely dominate the sound perception in 90% of the devices out there today - maybe even 100%, even at very high SPLs.
  Reply With Quote
Old 6th July 2007, 11:48 AM   #1449
mige0 is offline mige0  Austria
diyAudio Member
 
mige0's Avatar
 
Join Date: Jan 2005
Location: Austria, at a beautiful place right in the heart of the Alps.
Hi

Quote:
Actually the OS situation is the opposite of wht you are saying.

Earl, ok, I reread your posting and saw that I got confused. You already did that listening tests with HORNS.

This listening tests didn't focus on even order HD but rather on IM ( so I don't have to bury my theory ) ?




Quote:
110 dB is not too common in a small room such as we are talking about here.
I agree 100%
But Lynn is going to redefine the range of "common listening levels in small rooms"!

But seriously – I was actually asking about your findings here in the context of increased SPL's found at concerts.



Quote:
Far more important are the diffractions in the waveguides (and drivers), but certainly in typical horns with intentional diffraction for directivity control. The diffraction effects are also level dependent and more than likely dominate the sound perception in 90% of the devices out there today - maybe even 100%, even at very high SPLs.
My "visualisation" of diffraction in horns ( and wave guides ) is that there is continuous diffraction as long as the wave front follows the horn curve ( = low "reflection" due to the creation of low level second sources along the boundary ) and kind of abrupt diffraction at the mouth.
I would expect the mouth to behave like lets say the "water surface" when seen from underneath at diving = lots of reflection.

Could you give a understandable pin point description about the HOM issue please ?

Greetings
Michael
  Reply With Quote
Old 6th July 2007, 03:27 PM   #1450
MBK is offline MBK  Singapore
diyAudio Member
 
Join Date: Apr 2002
Location: Singapore
Earl,

as an attempt to summarize, to my understanding:

There are two mechanisms leading to distortion perception:

a- nonlinearity of compressed air at very high SPL's. Usually not dominant in audibility;

b- internal WG or horn reflections, including HOMs, which are not minimum phase and are perceived as though they were nonlinear distortions. Usually significant audibility in conventional WG / horn designs.

Ad a, this also applies to the internal compression chamber of a compression driver, but the distances are usually too short to generate significant nonlinearities.

Ad b, these effects are hard to measure, though they might be viewable on a CSD to some extent. The generation process might include the compression chamber, but is usually confined to the proximate throat area of a WG. Mouth reflections may occur in addition to that, but are of a so far not well determined significance. Minimization of all these effects through OS profile, rapid flare rate, and foam.

Just trying to get a handle on the issues.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



New To Site? Need Help?

All times are GMT. The time now is 11:50 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2