SY said:
I bought that book a few years ago, and a friend borrowed it, and promptly moved to Australia, (Mark, I will find you one day
). I suppose I ought to get round to getting another copy really.All
I will admit that my experiences are totally subjective, but every domestic line array I have heard has had a very small sweetspot, much more so than conventional TMW setups. However, Jim's paper is once again challenging my ideas, so I can see I have yet another project for the future. Isn't this diy lark fun!!!
Tag Mac pulls out of hi-fi and Seventh Veil is out of the Kleenex business.
Apparently, someone complained to eBay about the auction and they have banned it. Sigh! According to the feedback I was getting, many people found it amusing. I was really looking forward to the next auction in the series as well.
The 'humour police' are very much alive and well and they're living somewhere in the hi-fi community.
For those interested, I believe that Asda, Sainsbury, Tesco, Safeway and Boots (among others) are alternative sources for the product. Whether it works as well as mine .....
Sadly I have to report that my company is "considering its position" with regard to the sale of the Kleenex high end audiophile accessory.Ian J said:
Apparently, someone complained to eBay about the auction and they have banned it. Sigh! According to the feedback I was getting, many people found it amusing. I was really looking forward to the next auction in the series as well.
The 'humour police' are very much alive and well and they're living somewhere in the hi-fi community.
For those interested, I believe that Asda, Sainsbury, Tesco, Safeway and Boots (among others) are alternative sources for the product. Whether it works as well as mine .....
I would believe a small sweet spot in a line source if they are electrostatic or perhaps some planar sources. But if the line is an array that is realized with inherently wide dispersion drivers plus if it is operated in its near field, a narrow sweet spot will not be an issue. In fact, the sweet spot will be wide enough that some listeners (who are used to small, head-in-a-vise point sources) will comment on the diffuse nature of the image. Literally, the image of near field line array can approach the entire area between the speakers. With such a near field array the image will be solid and within the sound stage all of the various sound elements will be located in the correct locations.
>I'd draw the line (bad pun) this way: if at the intended listening distance and range of listening heights, the power law is closer to 1/r and the spectral balance is independednt of height, what you've got is a line source.
====
Right, and by default, this means no XOs in the line and with enough discrete elements to create so much comb filtering that our internal audio processor sums its response to the equivalent of a single large point source, a salient point that in recent years seems to have dropped by the wayside when these are discussed. As good as JG's or any other's power tapered arrays may sound, they aren't line arrays, ergo their 'sweet spot'/power response shifts with frequency based on each summed driver's polar response, just like any other WMTMW.
The one true line array I auditioned (McIntosh XR290) had no 'sweet spot' in a typical sized listening room, instead sounding very close to a live event WRT dynamics/transient response, with moving around the room having little affect on its power response and tonal (spectral) balance.
The downside is cost and building complexity since it requires at least a three way vertical array of many appropriately sized drivers.
GM
====
Right, and by default, this means no XOs in the line and with enough discrete elements to create so much comb filtering that our internal audio processor sums its response to the equivalent of a single large point source, a salient point that in recent years seems to have dropped by the wayside when these are discussed. As good as JG's or any other's power tapered arrays may sound, they aren't line arrays, ergo their 'sweet spot'/power response shifts with frequency based on each summed driver's polar response, just like any other WMTMW.
The one true line array I auditioned (McIntosh XR290) had no 'sweet spot' in a typical sized listening room, instead sounding very close to a live event WRT dynamics/transient response, with moving around the room having little affect on its power response and tonal (spectral) balance.
The downside is cost and building complexity since it requires at least a three way vertical array of many appropriately sized drivers.
GM
GM: D'accord. There's a lot of stuff out there labeled "line source" that really ain't. I go about 8 feet of active area, no crossover.
I think that the width of the sweet spot will definitely depend on both vertical AND horizontal dispersion. That's one issue I'm looking at for extending the bass and max SPL of my speakers with a line of small mid-woofers.
I think that the width of the sweet spot will definitely depend on both vertical AND horizontal dispersion. That's one issue I'm looking at for extending the bass and max SPL of my speakers with a line of small mid-woofers.
Comments on Line Arrays
GM said:
Right, and by default, this means no XOs in the line and with enough discrete elements to create so much comb filtering that our internal audio processor sums its response to the equivalent of a single large point source, a salient point that in recent years seems to have dropped by the wayside when these are discussed. As good as JG's or any other's power tapered arrays may sound, they aren't line arrays, ergo their 'sweet spot'/power response shifts with frequency based on each summed driver's polar response, just like any other WMTMW.
JG Replies:
If you correctly design a line array, then it does function as a line source (no comb lines) within the parameters for driver spacing and such (see my white paper for the details). My latest efforts use less power tapering than the earlier design that you heard last year so there is less unevenness in the vertical image. The tradeoff with power tapering is how to control sound bloom yet not reduce the vertical image size so much that you just have a point source. My latest efforts are impressive in their ability to create a true line source sound with very stable but wide image plus yield even power responses to both vertical and horizontal ddirections. The crossover is executed where the drivers have very broad horizontal dispersions so the blend is near seemless.
The problem with no crossover discrete driver arrays is that they have even severe compromises on driver spacing which leads to even more restrictive power dispersion in especially the vertical and ultimately even within the horizontal plane.
Bottom line is that my latest array is a true line array and definitely not a WMTMW by any means.
GM Said:
The one true line array I auditioned (McIntosh XR290) had no 'sweet spot' in a typical sized listening room, instead sounding very close to a live event WRT dynamics/transient response, with moving around the room having little affect on its power response and tonal (spectral) balance.
The downside is cost and building complexity since it requires at least a three way vertical array of many appropriately sized drivers.
JG Replies:
You said in your first paragraph that the best way to go was to have no crossover for a true line array yet your example of a true line array is a three way design. The XR290 is an excelllent design but very expensive and complex as you point out. I'll agree that line arrays can best create a live event WRT to dynamics and transient response.
Jim
GM said:
Right, and by default, this means no XOs in the line and with enough discrete elements to create so much comb filtering that our internal audio processor sums its response to the equivalent of a single large point source, a salient point that in recent years seems to have dropped by the wayside when these are discussed. As good as JG's or any other's power tapered arrays may sound, they aren't line arrays, ergo their 'sweet spot'/power response shifts with frequency based on each summed driver's polar response, just like any other WMTMW.
JG Replies:
If you correctly design a line array, then it does function as a line source (no comb lines) within the parameters for driver spacing and such (see my white paper for the details). My latest efforts use less power tapering than the earlier design that you heard last year so there is less unevenness in the vertical image. The tradeoff with power tapering is how to control sound bloom yet not reduce the vertical image size so much that you just have a point source. My latest efforts are impressive in their ability to create a true line source sound with very stable but wide image plus yield even power responses to both vertical and horizontal ddirections. The crossover is executed where the drivers have very broad horizontal dispersions so the blend is near seemless.
The problem with no crossover discrete driver arrays is that they have even severe compromises on driver spacing which leads to even more restrictive power dispersion in especially the vertical and ultimately even within the horizontal plane.
Bottom line is that my latest array is a true line array and definitely not a WMTMW by any means.
GM Said:
The one true line array I auditioned (McIntosh XR290) had no 'sweet spot' in a typical sized listening room, instead sounding very close to a live event WRT dynamics/transient response, with moving around the room having little affect on its power response and tonal (spectral) balance.
The downside is cost and building complexity since it requires at least a three way vertical array of many appropriately sized drivers.
JG Replies:
You said in your first paragraph that the best way to go was to have no crossover for a true line array yet your example of a true line array is a three way design. The XR290 is an excelllent design but very expensive and complex as you point out. I'll agree that line arrays can best create a live event WRT to dynamics and transient response.
Jim
Hi..planning a line array as well, hence my interest.
Would it be true to say that the transient response/dynamics (is there a difference?) is due to the number of drivers, rather than the topology?
I have to admit, the closest I ever came to listening to a line array was the Mission 753 (WWWWT)..That sounded very fast! (Apart from ESL's)
Cheers,
Bas
>JG Replies:
If you correctly design a line array, then it does function as a line source (no comb lines) within the parameters for driver spacing and such (see my white paper for the details).
====
Even most wide BW drivers have combing within themselves and why they often need phase plugs.....
Really, in a round about, but technically incorrect IMO, way you're agreeing with me in that if you have enough drivers of the right properties/spacing there's so much comb filtering that the brain overloads and sums them to discrete frequencies.
This is the mechanism that surround sound is based on, i.e. a significant number of insignificant sources.
Over a fairly narrow BW comb filtering sums in an increased nearfield, but this is limited to ~1WL of the line, so a long line must have a bunch of acoustically small drivers to increase its usable BW, but the mechanism is the same, massive comb filtering. The only 'lines' that sum with no comb filtering over a wide BW are compression horn arrays. These are truly fearsome performers, as is their size/cost, but the tradeoff is increased distortion.
Not many 'free lunches' in audio.
====
>My latest efforts use less power tapering than the earlier design that you heard last year so there is less unevenness in the vertical image.
====
OK.
====
>Bottom line is that my latest array is a true line array and definitely not a WMTMW by any means.
====
Not knowing their design details, I can't comment beyond what I've already said.
====
>JG Replies:
>You said in your first paragraph that the best way to go was to have no crossover for a true line array yet your example of a true line array is a three way design.
====
Right, and there's no contradiction. Each discrete line has no power tapering, such as your Linus array does, and they are each well within the 1/3rd rule of thumb XO 'driver' spacing. To make a single line 'line' array cover a very wide BW means a row of 3/4" dia. tweeters EQ shaped to handle a much wider BW, with a substantial tradeoff in efficiency/dynamics.
====
>I'll agree that line arrays can best create a live event WRT to dynamics and transient response.
====
Hmm, I didn't say 'best' since we've been discussing arrays composed of point source drivers, but even if limited to multiple compression driver horn arrays I'm still not sure I agree. With current technology, I believe the best would be a huge fullrange Unity concept horn, but how much better it would audibly perform Vs a well done compression driven array is anybody's guess.
====
>BH:
Would it be true to say that the transient response/dynamics (is there a difference?) is due to the number of drivers, rather than the topology?
====
These are two different things, though often incorrectly used interchangably. Transient response is how well the speaker tracks the signal's leading attack/decay, while dynamics is how close to 'live' the speaker can reproduce the signal, i.e. its dynamic range without compression. IOW a speaker can be transient perfect, yet be lacking in dynamics due to low efficiency or thermal power compression.
WRT various design topologies, right, the smaller/greater quantity of drivers, the wider the usable BW and the less we're able to discern the negative effects of comb filtering on transient response, and of course the more drivers in the line, the greater its dynamic range.
GM
If you correctly design a line array, then it does function as a line source (no comb lines) within the parameters for driver spacing and such (see my white paper for the details).
====
Even most wide BW drivers have combing within themselves and why they often need phase plugs.....
Really, in a round about, but technically incorrect IMO, way you're agreeing with me in that if you have enough drivers of the right properties/spacing there's so much comb filtering that the brain overloads and sums them to discrete frequencies.
This is the mechanism that surround sound is based on, i.e. a significant number of insignificant sources.
Over a fairly narrow BW comb filtering sums in an increased nearfield, but this is limited to ~1WL of the line, so a long line must have a bunch of acoustically small drivers to increase its usable BW, but the mechanism is the same, massive comb filtering. The only 'lines' that sum with no comb filtering over a wide BW are compression horn arrays. These are truly fearsome performers, as is their size/cost, but the tradeoff is increased distortion.
Not many 'free lunches' in audio.
====
>My latest efforts use less power tapering than the earlier design that you heard last year so there is less unevenness in the vertical image.
====
OK.
====
>Bottom line is that my latest array is a true line array and definitely not a WMTMW by any means.
====
Not knowing their design details, I can't comment beyond what I've already said.
====
>JG Replies:
>You said in your first paragraph that the best way to go was to have no crossover for a true line array yet your example of a true line array is a three way design.
====
Right, and there's no contradiction. Each discrete line has no power tapering, such as your Linus array does, and they are each well within the 1/3rd rule of thumb XO 'driver' spacing. To make a single line 'line' array cover a very wide BW means a row of 3/4" dia. tweeters EQ shaped to handle a much wider BW, with a substantial tradeoff in efficiency/dynamics.
====
>I'll agree that line arrays can best create a live event WRT to dynamics and transient response.
====
Hmm, I didn't say 'best' since we've been discussing arrays composed of point source drivers, but even if limited to multiple compression driver horn arrays I'm still not sure I agree. With current technology, I believe the best would be a huge fullrange Unity concept horn, but how much better it would audibly perform Vs a well done compression driven array is anybody's guess.
====
>BH:
Would it be true to say that the transient response/dynamics (is there a difference?) is due to the number of drivers, rather than the topology?
====
These are two different things, though often incorrectly used interchangably. Transient response is how well the speaker tracks the signal's leading attack/decay, while dynamics is how close to 'live' the speaker can reproduce the signal, i.e. its dynamic range without compression. IOW a speaker can be transient perfect, yet be lacking in dynamics due to low efficiency or thermal power compression.
WRT various design topologies, right, the smaller/greater quantity of drivers, the wider the usable BW and the less we're able to discern the negative effects of comb filtering on transient response, and of course the more drivers in the line, the greater its dynamic range.
GM
GM,
Thanks for your reply.
Now for something completely different.
Dunno why but don't like the idea of placing drivers all neatly in one line!
Is there any advantage in placing the drivers in a line array....ever so slightly off center from the vertical line??? Say maximum 1 inch? Or is it then not a line array anymore.
Or the one driver 4mm higher than the previous driver and the next just 1mm away from the bottom one and the next 3mm from the driver. I.e. small enough that it is still a line array but not equi distant?
Cheers,
Bas
Thanks for your reply.
Now for something completely different.
Dunno why but don't like the idea of placing drivers all neatly in one line!
Is there any advantage in placing the drivers in a line array....ever so slightly off center from the vertical line??? Say maximum 1 inch? Or is it then not a line array anymore.
Or the one driver 4mm higher than the previous driver and the next just 1mm away from the bottom one and the next 3mm from the driver. I.e. small enough that it is still a line array but not equi distant?
Cheers,
Bas
Here's my latest foray into LAs.
Reading JGs original paper got me interested a while back and I've been playing with variations of these for months.
Pete
An externally hosted image should be here but it was not working when we last tested it.
Reading JGs original paper got me interested a while back and I've been playing with variations of these for months.
Pete
>GM,
Thanks for your reply.
====
You're welcome!
====
>Now for something completely different.
Dunno why but don't like the idea of placing drivers all neatly in one line!
====
Can't imagine why, but whatever 'floats yer boat' since you're the one that's going to be listening to them.
====
>Is there any advantage in placing the drivers in a line array....ever so slightly off center from the vertical line??? Say maximum 1 inch? Or is it then not a line array anymore.
====
It's still a line, just one with a ~4316Hz offset from top to bottom that increases with increasing frequency. Don't have a clue if it would audibly impact the stereo imaging/transient response negatively.
====
>Or the one driver 4mm higher than the previous driver and the next just 1mm away from the bottom one and the next 3mm from the driver. I.e. small enough that it is still a line array but not equi distant?
====
This would unacceptably increase lobing, making for a too 'phasy' sound IMO, but there's one way to find out for sure, so how about cobbling up some and let us know.
GM
Thanks for your reply.
====
You're welcome!
====
>Now for something completely different.
Dunno why but don't like the idea of placing drivers all neatly in one line!
====
Can't imagine why, but whatever 'floats yer boat' since you're the one that's going to be listening to them.
====
>Is there any advantage in placing the drivers in a line array....ever so slightly off center from the vertical line??? Say maximum 1 inch? Or is it then not a line array anymore.
====
It's still a line, just one with a ~4316Hz offset from top to bottom that increases with increasing frequency. Don't have a clue if it would audibly impact the stereo imaging/transient response negatively.
====
>Or the one driver 4mm higher than the previous driver and the next just 1mm away from the bottom one and the next 3mm from the driver. I.e. small enough that it is still a line array but not equi distant?
====
This would unacceptably increase lobing, making for a too 'phasy' sound IMO, but there's one way to find out for sure, so how about cobbling up some and let us know.
GM
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.