[Rick Craig]

Isn't the "proximity effect" a characteristic of directional microphones only?  I don't think omni condenser mics have a proximity effect.

It happens with recordings / live sound using cardioid or bi-directional microphones too close to the source.

[Rick Craig]

There's also something called proximity effect where the low frequencies can be exaggerated by certain types of microphones. I think something similar happens when measuring an array and have observed a low frequency tilt that can vary depending on the measurement distance.

Proximity effect AFAIK, exists in all mics. Condenser mics are much less prone to it than dynamic mics. (I believe calibration and testing mics are all condenser type) It tilts up bass response when placed in near proximity, and rolls off bass with distance.

I'm not sure how that would make for getting a real reading 1/4" away from a driver. It would seem that proximity effect would skew the results, no?

When we mic bass drums for example, we have often have to mic from a distance in order to get proper bass and fundemental pitch, along with tone/overtones. The theory being that the low waveforms are so long that they are missed entirely by close miking. I'm not sure on the theory, but in application, close miking never yields realistic bass response on instruments.

Cheers

Sorry I didn't explain myself better. Omnis that are typically used for measurements don't exhibit the proximity effect but there is a relationship between the distance of the mic. / size of the source / frequency range. That's why the nearfield technique has an upper frequency limit that's dependent on the cone diameter. You can check out Don Keele's AES papers to explain how it works and also Joe D'Appolito's "Testing Loudspeakers".

[FredT300B]

Hello all, this is becoming a very interesting thread, but it's getting a bit off the topic "why line array'. I started a new thread because I want to learn more about speaker measurement, especially as it applies to line arrays. I forgot to mention it in my thread-starting post, but I'm referring to the measurement of line arrays' frequency response above 200hz or so, and not the low frequency response.

[Danny Richie]

I just had to comment on a couple of things here.

Vertical imaging...

In a sense both of you guys are right about this. Denny is right in that it is not something that is an effect picked up during recording. Meaning that a mic placed higher up or something will not place an image higher up in play back. No speaker will change this either.

There is the effect though and it is a result of a time arrival of the signal from the two speakers. I easily hear effects come directly from spots high up on the ceiling verses effects way down low in the sound stage. Q sound recordings are especially good about this effect. Typically down low images are forward and up front while higher up sounds are deeper and further back in the sound stage.

If you unplug one speaker you will quickly note that the image height is limited to the speaker and driver placement. It will not extend above or below the speaker in any way. Just give it a shot and you'll see.

Now a line source will not in any way diminish this effect in fact the varying degrees of height in image is typically better with the line sources than point sources.

The next question I have is that I seems to notice that most line arrays are built with small drivers (less than 8"), so aren't these drivers limited by their inherent limitation to produce frequency below say 50Hz??

Not at all. Low end extension has little to do with driver size. I'll elaborate more on that below in another response.

There seams to be some misconceptions that I will address directly as some of what was posted was really directed to me.

The bottom line is that the bass extension is not going to be any different whether you have one woofer (like a compact 2-way) or several woofers in an array.

This is not at all true. The longer the line of woofers the lower they will play due to the collective gain effects in the lower region. For instance: I designed a group of line sources for Epiphany Audio. Most of you have seen or read about them. The Absolute Sound reviewed two of them and gave them an Editors Choice award and a Golden Ear award.

They used a 4" woofer that I had custom built for that application. A single woofer ported low and in a bit of an oversized box could hit a -3db of 57Hz. The 6/6 model used 6 of them per speaker and would hit a -3db of nearly 50Hz (at least low 50's). The 12/12 model used 12 woofers and easily hit -3db figures in the mid to upper 40's (about 45Hz). The larger 20/21 (20 woofers) would hit a -3db range in the upper 30's.

The only accurate way to measure the true low frequency extension of a system is to do a near field test where you place the microphone a 1/4" from the woofer (sealed box system) or take a sum of the woofer's output and the port's output in a ported design.

This really doesn't work at all for line source speakers. One reason is mentioned above and involves the collective low frequency gain due to the length of the array.

Secondly, a line source will always couple with the room differently than most point source speakers. Lower woofers for example will couple with the floor and the floor reflections must be included in the measured output as it will always be there regardless of the rest of the room acoustics. The ceiling can cause gain too depending on the height of the ceiling and length of the array. All of this can effect the low end extension.

Summing a rear port and a woofer rarely gives a true resemblance of a room response either. It all depends on the room. I used to measure speakers in my anechoic chamber. From these measurements one can easily note that the rear firing port output gets absorbed into the foam behind the speaker and no port output is measured.

Most people are not concerned with a simulated low end response. It really means nothing. Most people want to know how low it will play in their room. Unfortunately all rooms are different but an average sized room and room response is very telling. Measurements will follow below, but first one more thing to note.

The array will move more air (increased cone area = more displacement); however, it you want true low octave extension you'll need a subwoofer to do that. Without active equalization the best you can expect from a ported 7" driver is a -3db point of 40-45hz. Some drivers I've seen (and measured) in other commercial arrays won't even go that low.

Again the collective gain from the length of the array yields more low end extension. Aside from that though if you are limited to a single woofer and off the shelf units then a 40 to 45Hz range is about the limit.

However it is no problem to get a 6.5" or 7" woofer to play down to a -3db point in the mid 20's. It's pretty easily actually. You just have to give up something somewhere else and in this case it is sensitivity.

The low sensitivity 6.5" woofers used in the LS series have a Fs of 34Hz and a single woofer is capable of hitting a -3db of 32Hz in a ported enclosure. So a group of 8 or 12 of them have no problem hitting -3db down points in the upper teens. XBL^ motors don't hurt either. This keeps the woofer linear even with long exertions and maintains a constant BL and driver control.

Most line sources utilizing ribbon tweeters can't be paired with woofers like these. The fragile ribbons are not capable of playing down very low or handling much power at all in a lower range. Crossover points must be kept in the 2kHz to 2.5khz range. So a woofer with a lighter weight diaphragm is needed to reach them well.

The custom planar magnetic tweeters used in the LS series can be crossed much lower. The LS-9's cross at 850Hz. This means the woofer need not play as high. So lighter weight cones are not needed.

The low crossover point also maintains a more even off axis response as horizontal lobbing effects are considerable less when crossing in the lower frequency ranges. This gives an even off axis response in both directions.

Another thing regarding the low end extension. The speed and accuracy of 12 or even 8 small woofers like those used in the LS series are way beyond any sub. I have been working on the fastest and cleanest subs currently available (more on this announcement soon). Still there is no comparison to a group of smaller drivers all sharing a small percentage of the load. Moving mass is much less than any single sub and the settling time is much faster. Plus the multiple small woofers can hit the same output levels as a single large sub and never be forced to move very far. This further reduces settling time and gives a cleaner tighter bass response.

Oh yea, here are a few measurements for you guys.

Here is a LS-6 in the middle of my listening room. These are an RTA responses.



There is a slight top end roll off in my well treated room when measuring at this distance. For comparison here is an A/V-1 brought up to the same level and in the same position. It is within +/-1.2db throughout its operating range. Obviously the blue line is the A/V-1. These were also moved over to the listening position on the right side of my room.



Here are a couple of the LS-9's. One is in the right listening position and the other is in the left listening position. 30hz and down actually kick up a little in my room.



Here are the same speakers measured with an un-gated MLS response.



Sorry, my Clio only measures down to 20Hz. Art measured his prototype LS-6's in his room with a DB meter to be -3db down in the upper teens.

Just guessing it sounds like the LS-9's go a few Hz lower than the 6's as they should.

There is plenty of room gain in any room with these but fortunately the bass management system allows some equalization for the room gain. So one could get a pretty flat response in just about any room of reasonable size.

I hope this helps.

[Rick Craig]

I guess the rules of physics are different in Texas 

So who do you believe - Danny or Don Keele? By the way, Don recently devised an array design that was quoted in Voice Coil (the main journal of the speaker industry) as being the most significant speaker research done in the last several years.

There's one significant problem with what you say Danny. If you have a line that's not long enough to support the nearfield effect of the 20-40hz octave (which all of your designs qualify for) then the in-room response of the lower octave should actually be reduced not increased in your measurements. One of the main reasons line arrays are used in home and pro audio is that they interface better with the room. If you have to apply passive "bass management" then you obviously have a box that's tuned wrong and / or a problem with your crossover design. A good design will work well in a variety of size rooms regardless of the room gain because the low frequencies are evenly distributed in the nearfield. Frequencies below the nearfield cutoff will be reduced, similar to the effect of dipole woofers and reducing the amount of low frequency gain.

[KCI-JohnP]

I guess the rules of physics are different in Texas

You got that right! aa

[bwaslo]

So who do you believe - Danny or Don Keele?

Both!  I'd think that a line array measured at 1/4 inch will have the same problem as measuring a large woofer to higher frequencies in the nearfield (which Keele will tell you doesn't work so well). Since the long dimension of the array is a lot larger than the diameter of the single woofer, the "high frequencies" for the array where this starts to happen would be lower than they are for the woofer, getting down to the tens of Hz.  Or wouldn' it?

Also, a line array that goes essentially floor to ceiling should have a nearfield effect (at listening position) to even very low frequencies.  The image effect from ceiling and floor (their reflections from those surfaces acts like additional drivers) makes the array act as if it is much longer.

[Daygloworange]

Meaning that a mic placed higher up or something will not place an image higher up in play back. No speaker will change this either.

Yup, I agree. Those ambient drum recordings I sent you were recorded with a pair of mics in a stereo configuration during some tracking tests. The heights of the ambient mics vary from 2' off the floor to over 11' off the floor. You will not hear drums 11' in the vertical soundstage. 

I easily hear effects come directly from spots high up on the ceiling verses effects way down low in the sound stage. Q sound recordings are especially good about this effect.

Q Sound recordings are a whole different kettle of fish.

Typically down low images are forward and up front while higher up sounds are deeper and further back in the sound stage.

I'll flip flop that around and add a little, as I think it's a better description. Forward and up front sounds will appear to be lower in vertical height. Lower frequency sounds will also appear to be lower, with the exception of some very forward (and close mic'ed) high frequency sounds, as they will appear to originate right at the tweeter

Deeper and further sounds will appear to be higher in vertical height. Higher frequency sounds often appear higher in vertical height.

I don't know enough about the topic of beaming and directionality of drivers, but I suspect that it plays a role in what people describe as vertical height in 2 channel recordings. Particularly in the mid to high frequencies.

If you unplug one speaker you will quickly note that the image height is limited to the speaker and driver placement. It will not extend above or below the speaker in any way. Just give it a shot and you'll see.

That's actually a good test.

I also suggested to flip your speaker upside down (if practical) while keeping the same tweeter height. You'll see no difference in perceived vertical imaging.

Now a line source will not in any way diminish this effect in fact the varying degrees of height in image is typically better with the line sources than point sources.

I've spent a nice bit of time with some line arrays a while back, and the spatial imaging around instruments was astonishingly well defined. Spatial information was crystal clear. The dynamics were amazing, and very visceral. The leading edge of transients was also astonishing.

I'll know again in the coming weeks just how true that is.  aa

Cheers 

[Danny Richie]

Okay Rick,

I am going to over look your snide remarks and try to help you out here.

Think about what you are saying here Rick and reason it out.

If you have a line that's not long enough to support the nearfield effect of the 20-40hz octave (which all of your designs qualify for) then the in-room response of the lower octave should actually be reduced not increased in your measurements.

So the line array will have a REDUCED output below 40Hz? Is this compared to a point source? Put a group of woofers together in an array and get REDUCED output below 40Hz?

One of the main reasons line arrays are used in home and pro audio is that they interface better with the room.

Come on now Rick. Think it through. As frequency increases floor to ceiling interaction is decreased and room interaction is minimized. However as frequency decreases room interaction increases.

bwaslo is 100% correct.

The image effect from ceiling and floor (their reflections from those surfaces acts like additional drivers) makes the array act as if it is much longer.

This added line length is a coupling effect of the ceiling and floor. It is a little like taking a sub woofer and placing it in a corner. You suddenly get more output as it couples and loads against the walls.

If you have to apply passive "bass management" then you obviously have a box that's tuned wrong

Yea, a flat response into the upper teens is clearly a result of being tuned wrong. 

and / or a problem with your crossover design.

Oh yea, now I have a crossover design problem. That crossover point up there at 850Hz to 1kHz is really causing a lot of room gain. 

A good design will work well in a variety of size rooms regardless of the room gain because the low frequencies are evenly distributed in the nearfield.

Regardless of how "good" the design is it is completely at the mercy of the room. There is no one size fits all. Room loading in the lower frequency ranges takes place regardless of whether the woofers are in a line, a squar,e or all in separate enclosures. It is an effect of the size of the room and pressure created in it.

This is how a car, lacking sufficient distance to propagate a low frequency wave length, can reproduce low frequencies down into the upper teens. It is about pressurization. Now if you simply roll the windows down you loose a ton of low frequency SPL.

This is the same with any room. A large room with a lot of air space will simply not give you as much low frequency gain as a small room and small air space will.

A passive and adjustable circuit allows the end user to equalize the speakers output for their given room.

[FredT300B]

Process Check:

I'm seeing some valuable and informative debate here, but unfortunately there are also some personally insulting remarks buried in a few of the posts. Everybody please take a deep breath, then re-read these remarks from my first post on this circle:

"It's customary in the first post of a new circle's facilitator to remind everybody to observe the rules of the forum. OK, you are reminded, and there's one that's especially important. I expect everybody to get at least a B- on their report card each semester under "Works and plays well with others". Participants who are commercially involved in line array speaker design are held to an even higher standard, and they are expected to consistently get A's".

Thanks in advance for your cooperation.

[JohninCR]

Fred,

Line array discussions are always entertaining, because it separates the men from the boys in terms of who actually knows what they are talking about vs those who think they do, and the "who's is bigger" stuff below the surface is a riot.  I do, however, think the sales pitches are a bit over the top, since there are circles for that.  Talking about the benefits of LAs is fine, but to do so without addressing the issue of time smear and its effect on imaging smells of promotion by at least 3 parties in this thread so far.  FWIW, measuring array behavior in the very near field is plain silly.

[TomS]

LOVE the new circle 

...I realize these guys are certainly regrettably tip-toeing over the lines from a personal standpoint, but also in a selfish sense I've just learned a ton about how the application of science and facts to line arrays can lead to gray areas and clear differences of opinion on how to do so.  They are things I've always wondered about myself but don't have the background, tools, and/or analysis skills to research it.  These guys have tons of practical real world experience at this that few others possess, so I hope it doesn't get watered down too much because they have to play nice. 

Please guys, get real, and keep it coming.  At the very least, thank you for your participation.  This is a FUN hobby, not war, right?

Tom

[FredT300B]

...Talking about the benefits of LAs is fine, but to do so without addressing the issue of time smear and its effect on imaging smells of promotion...

Your point is well taken and it adds value to this discussion. I'm not sure it smells of promotion, but I agree your comment about the importance of time smear in line arrays is one that can't be overlooked. If you are hearing more than one driver and they aren't equidistant there has to be some time smear. This is true of any speaker whose drivers aren't time-aligned, and I suppose it's more critical in the case of a line array where several drivers that are handling the same frequency range aren't equidistant from the listener's ear. The important issue for me is how much of a disadvantage this time smear is and, since every speaker design involves compromises, does it outweigh the advantages of line arrays. 

After reading your post I measured the woofer array. The enclosure is 6' 2" tall, but that's deceiving. With the eight closely spaced 7" woofers, it's only 4' 5" from the center of the bottom woofer to the center of the top woofer. I listen from about 12' out from the baffles, so the distances to the respecitve drivers doesn't vary too much. The tweeter array doesn't matter because with the ribbons' vertical directionality the only one I'm hearing is the one that's at ear height.

For me the answer is entirely subjective: I'm not aware of the time smear caused by the different distances the sound travels from the edges of the array versus from the center, but I am aware of the enhanced soundstage, sensitivity and dynamics of my line arrays versus my point source speakers. In my case the sensitivity is a must-have because I'm driving them with a 40 watt SET amplifier, but even if they were lower sensitivity and I were driving them with a large SS amp I still would prefer the sound of the line arrays. If I could substitute a reference quality three way, like the custom designed high efficiency pair Jim Salk recently completed for my friend John Z (see them on the Salk circle) I know I would love them, but I doubt I would prefer them to line arrays.  Like I said, it's just a personal preference.

[Rick Craig]

So who do you believe - Danny or Don Keele?

Both!  I'd think that a line array measured at 1/4 inch will have the same problem as measuring a large woofer to higher frequencies in the nearfield (which Keele will tell you doesn't work so well). Since the long dimension of the array is a lot larger than the diameter of the single woofer, the "high frequencies" for the array where this starts to happen would be lower than they are for the woofer, getting down to the tens of Hz.  Or wouldn' it?

Also, a line array that goes essentially floor to ceiling should have a nearfield effect (at listening position) to even very low frequencies.  The image effect from ceiling and floor (their reflections from those surfaces acts like additional drivers) makes the array act as if it is much longer.

Bill,

Thanks for the contributions you've made to the audio world with your business. You misunderstood - I don't measure an array at 1/4". Individual woofers (7" or less) in a sealed array can easily be measured nearfield up to 200hz. Ported measurements are a little more tricky, especially if the port is on the rear of the cabinet but if you sum the responses properly you can get an accurate curve.

The nearfield coupling of course depends on the length of the line but even if you have a 7 ft. line it only extends the nearfield down to about 160hz.

[Rick Craig]

Fred,

Line array discussions are always entertaining, because it separates the men from the boys in terms of who actually knows what they are talking about vs those who think they do, and the "who's is bigger" stuff below the surface is a riot.  I do, however, think the sales pitches are a bit over the top, since there are circles for that.  Talking about the benefits of LAs is fine, but to do so without addressing the issue of time smear and its effect on imaging smells of promotion by at least 3 parties in this thread so far.  FWIW, measuring array behavior in the very near field is plain silly.

The delay issues in my opinion are not an audible problem. In some ways an array would be better than a point source because the controlled vertical directivity will resulted in more direct versus reverberant sound at the listening position. Delays from room reflections would be more audible due to the increased time of delay versus the line's delay. That's one of the main reasons why arrays are often used in live sound venues.

[JohninCR]

After reading your post I measured the woofer array. The enclosure is 6' 2" tall, but that's deceiving. With the eight closely spaced 7" woofers, it's only 4' 5" from the center of the bottom woofer to the center of the top woofer. I listen from about 12' out from the baffles, so the distances to the respecitve drivers doesn't vary too much. The tweeter array doesn't matter because with the ribbons' vertical directionality the only one I'm hearing is the one that's at ear height.

Fred,
The higher frequencies is where the problem occurs, and maybe the limited vertical dispersion of the ribbons adequately addresses it.  Unfortunately my geographic handicap has prevented me from experiencing an array with ribbon tweets.  With all that I have heard, I could clearly pick out sound coming from the topmost and bottommost drivers, with the result only being problematic with some recordings.  The image distortion it caused was most prevalent with deeper male vocals.  eg Barry White sounded like his head was at least a meter in diameter.  Even so I still highly recommend arrays.

[Danny Richie]

Talking about the benefits of LAs is fine, but to do so without addressing the issue of time smear and its effect on imaging smells of promotion by at least 3 parties in this thread so far.

JohninCR , There are no audible time smears.

I design far more point source type designs than line sources and for many different companies. I am not pro one and con the other. I listen to them side by side quite often.

I can assure you that the time differential of drivers near the top of the array verses drivers in the middle of the array (where your ear is) is minuscule. From a typical listening distance this 1" to 2" differential is in the order of about a 1/10th of a millisecond or so.

Does this smear the vocal region? Not at all. For all we know the recording engineer might have dialed in three or four times that much delay on the vocal to try and fatten them up a little.

The only place this will have an effect is in the upper harmonic ranges where wavelengths are very short. Even then as you get a little comb filtering effect causing a slight peak and dip it can easily reversed with a slight variation in listening height. You actually hear an average level due to room related effects. The limited dispersion of some taller tweeters limits interaction in the top octave any way.

A simple side wall reflection can be many times greater with delay of 4 to 6 milliseconds. How much does this smear the sound? Some say it needs to be at least 8 milliseconds of a delay to even hear it as a delay. Some say a little less. Whatever it is it really doesn't matter for this example. The delay of the line source drivers is not even close to that.

3.  The nature of how sound disperses from a line array leads to better in-room behavior.  Not only is floor and ceiling bounce virtually eliminated, but that also eliminates the floor/ceiling room mode from having an influence.

Floor and ceiling bounce is not eliminated. It decreases as frequency increases but it increases as frequency decreases.

I have also worked with power tapering, but this really doesn't change anything. The delay is still there and turning down the upper or lower drivers in output does not effect it. It is an effort to try and fix a problem that really isn't a problem to begin with.

Take a typical 2 way design with the tweeter mounted on the same plane as the woofer. Let's say it is a typical 6.5" woofer. Guess what?    The tweeter is further forward in time compared to the woofer just as much as tweeters in the top of a line source compared to the middle of a line source, maybe even more so in some cases.

Do you hear a time smear in a typical two way speaker too? Side by side comparisons made with a two way design that do time align the tweeter to the woofer easily shows that one can NOT hear any time related smear with the drivers not physically aligned. It is too small.

The tweeter array doesn't matter because with the ribbons' vertical directionality the only one I'm hearing is the one that's at ear height.

Hey Fred, You still hear all of them. The only area where they don't overlap is about the top half of the top octave. As frequency decreases they overlap more.

Try disconnecting all but one of your tweeters, or just measure a single tweeter. I am of coarse referring to a long ribbon. Measure it's on axis and vary the height and measure it's off axis response moving up about 6" at a time. If it is a 6" tall ribbon then measure the center of it then move up 6" or so to where you would be on axis with the ribbon that is above or below it. This is especially true if at or near listening distance. I bet you'll see very little drop off in output. Just the very top end will drop out a little.

The nearfield coupling of course depends on the length of the line but even if you have a 7 ft. line it only extends the nearfield down to about 160hz.

Hey Rick,

Down here in Texas   we can take just two woofers side by side playing the same information and they will completely couple for as low as they can play, but you know our physics might be different down here.  

I hope that doesn't get me a "B" in plays well with others but I couldn't help myself on that one.

[Daygloworange]

A simple side wall reflection can be many times greater with delay of 4 to 6 mili-seconds. How much does this smear the sound? Some say it needs to be at least 8 mili-seconds of a delay to even hear it as a delay. Some say a little less. Whatever it is it really doesn't matter for this example. 

There is a recording technique of splitting a mono signal into 2 sources separated by a delay so that you can pan them hard right and hard left with virtually nothing in the middle of the stereo image. It's a trick to get more headroom in a 2 track recording.

When the 2 sources are split, you can sum them in the center at the same amplitude (kind of like overlaying one over the other) and alter the delay between them. The harshest phasing problems are between about 3ms to 11ms (IIRC). Longer than that, and it becomes much smoother. In fact, that's what flanging, chorusing are. You'll sum two of the same signal,( usually with a 20 ms delay and up) but the difference is, is that they add LFO to the delayed signal so that there is a slight variance (sweeping up and down in time) to the delayed signal compared to the straight (undelayed) signal.

If we're talking 1/10 of a millisecond, it becomes rather moot.

There are much bigger fish to fry.

BTW, how big do the fish get in Texas? 

Cheers

[Danny Richie]

About this big.



http://urbanlegends.about.com/library/bl_2004_catfish.htm

Caught just about an hour and a half from my house.

This help?  

[Jim Griffin]

I can not top Danny's fish tale but there is a good white paper on the High Emotion Audio site that details some of the important characteristics of line arrays.  It will help you understand 'Why line array?'  They talk to the importance of frequency response, dynamics range, temporal resolution, and spatial discrimination.   

High Emotion Audio is the new company that was formed by some of the Nearfield Acoustics principals after Nearfield went belly up.  The original Nearfield Acoustics Pipedreams design lives on with High Emotion Audio.

The white paper is at:

  http://www.highemotionaudio.com/whiteppr.pdf

http://www.highemotionaudio.com/whiteppr.pdf