Ken,
That statement is based on my understanding of the operation of both horns and dipoles as well as the pulse response comarisons posted earlier by KWY. To make it easier, here it is again:
As you say, it would not be balanced if this were not previously provided.
As far as this discussion goes, I think it is also fair to say that I have no bias towards either horns or dipoles. My intention is in fact to eventually build both and to use each for what they do best.
That statement is based on my understanding of the operation of both horns and dipoles as well as the pulse response comarisons posted earlier by KWY. To make it easier, here it is again:
An externally hosted image should be here but it was not working when we last tested it.
As you say, it would not be balanced if this were not previously provided.
As far as this discussion goes, I think it is also fair to say that I have no bias towards either horns or dipoles. My intention is in fact to eventually build both and to use each for what they do best.
Konnichiwa,
First, you are going to hate me for that and I can already hear Magnetars evil laugh, but in the kind of room you are talking about you will most likely be better off using horns, as dipoles or unipoles will invariably require a huge size to be effective. Even my "modest proposal" with a 21" + 21" acive unipole Sub and 15" Coax Wideband would strike me as by far too small to operate as sensible music system in a room seating 200.
For genuine "high fidelity" I'd probably recommend a central stack of a few Shearer Horns, to get enough mouth surface and hight for good pattern control at low frequencies.
Shearer 2 x 15 Zoll Bass Horn
The Shearer deployed singly makes for a genuine 50Hz Horn, stack about 3 or 4 of them and you get close to the magic "20Hz" for a hornloaded system. With six to eight sealed box loaded 15" woofers you can then apply a little EQ to get still some output at 16Hz for 32' stopped organ pipes.
To go with that I'd likely deploy a pair of BIG satellite Horns, maybe something like a pair of Oris 150 with suitable drivers per side (not neccesarily Lowthers - just nice suitable cone driver to cover up to around 6KHz) plus suitable supertweeters (progagly an array is needed) and maybe 4 X 10" or 8 X 10" Dipole Bass Systems (or short front only horn cabinets like the Altec 210) as "fill in" (fairly narrow band from maybe 80 - 160Hz) to co-operate with the main horn array.
Anyway, your room problem screams "HORN" (mostly) at me, while those of more normal High Fidelity requirements (eg. rooms from 20 squaremeter to maybe 80 squaremeter) are better served by dipole/unipole systems, unless really large architectural horns are acceptable. As said, Horses, Horns and Dipoles for courses.
You might find this system of interest:
5 Channel 20,000 Watt Home Hi-Fi System of Dick Burwen (Mark Levinsons mentor)
I have to say that I have taken a dislike to any of these folded forms, compared to a modified W-Frame as shown below.
The modified (or perhaps better "the true") W Frame has less problems with resonances than any of the systems with right angles suffer.
BTW, for all those who feel that dipoles can't go loud, this page may be interesting....
"Based on Siegfried's "Phoenix" this is an open baffle pro-audio system using Eminence drivers."
Sayonara
paulspencer said:
First, you are going to hate me for that and I can already hear Magnetars evil laugh, but in the kind of room you are talking about you will most likely be better off using horns, as dipoles or unipoles will invariably require a huge size to be effective. Even my "modest proposal" with a 21" + 21" acive unipole Sub and 15" Coax Wideband would strike me as by far too small to operate as sensible music system in a room seating 200.
For genuine "high fidelity" I'd probably recommend a central stack of a few Shearer Horns, to get enough mouth surface and hight for good pattern control at low frequencies.
Shearer 2 x 15 Zoll Bass Horn
The Shearer deployed singly makes for a genuine 50Hz Horn, stack about 3 or 4 of them and you get close to the magic "20Hz" for a hornloaded system. With six to eight sealed box loaded 15" woofers you can then apply a little EQ to get still some output at 16Hz for 32' stopped organ pipes.
To go with that I'd likely deploy a pair of BIG satellite Horns, maybe something like a pair of Oris 150 with suitable drivers per side (not neccesarily Lowthers - just nice suitable cone driver to cover up to around 6KHz) plus suitable supertweeters (progagly an array is needed) and maybe 4 X 10" or 8 X 10" Dipole Bass Systems (or short front only horn cabinets like the Altec 210) as "fill in" (fairly narrow band from maybe 80 - 160Hz) to co-operate with the main horn array.
Anyway, your room problem screams "HORN" (mostly) at me, while those of more normal High Fidelity requirements (eg. rooms from 20 squaremeter to maybe 80 squaremeter) are better served by dipole/unipole systems, unless really large architectural horns are acceptable. As said, Horses, Horns and Dipoles for courses.
paulspencer said:
You might find this system of interest:
5 Channel 20,000 Watt Home Hi-Fi System of Dick Burwen (Mark Levinsons mentor)
paulspencer said:
I have to say that I have taken a dislike to any of these folded forms, compared to a modified W-Frame as shown below.
An externally hosted image should be here but it was not working when we last tested it.
The modified (or perhaps better "the true") W Frame has less problems with resonances than any of the systems with right angles suffer.
BTW, for all those who feel that dipoles can't go loud, this page may be interesting....
"Based on Siegfried's "Phoenix" this is an open baffle pro-audio system using Eminence drivers."
Sayonara
RHosch said:Well, at least you did get rid of that hideous Karlson loaded bass rig, so I suppose there is some hope yet for you!
I still have them -- they sound excellent when mated with an open back midrange -- Nice small footprint, controlled directivity into the bass, no over excursion problems like you'll find in an open baffle "bass" system. Goofey looking? For sure!
Have you built or owned a Karlson and listened to it compared to what you have now or are you just condemning the way it looks? Or maybe are you just using a text book to determine it's home entertainment value?
What do you use to reproduce music in your home?
An externally hosted image should be here but it was not working when we last tested it.
To ever posted this -- What is a shock test???????
Why would anyone use an amplifier with a source impedence of 15 ohms???
I'd like to see this at 20 volts input where the bandwith of all the examples is limited to being exactly the same along with harmonic distortion measurements below 150 cycles
Also it would be interesting to see the in room response 20 to 200 cycles....
paulspencer said:
I didn't criticise I others I just stated what appears to be fact. If other's have not built the systems and used them than that is just the way it is. They are basing their beliefs on theory not practrice.
Where are my statements that are not correct and where is the proof that compliments this?
Also - What exactly would you like me to prove? What will it take for you to be 100 % convinced that my beliefs are the same as yours?
Kuei Yang Wang said:
You pretty much described my system -- except I use a compression driver 1k to 8k instead of a wide range mid -- 140 to 1k is handled by the JBL 2123 in a short tractrix horn -- I have found wide bandwith horns are about as useful (well maybe a little more) as open baffle subwoofers. IOW I prefer the drivers to be fully horn loaded throut the intended bandwith -- kinda hard to do at the extreme bottom but if the room is solid enough it's possible.
Actually this works quite well in a 32 square meter room.
Konnichiwa,
I posted it. A "shock" test is nowadays known as "step-response".
The high source impedance was needed as the voicecoil was used as sensor. The scan is from the old Briggs book.
And I would like to see a RT60 vs. Frequency plot for your system/room, but we cannot always have what we want and have to make do with what we got.
To me too.
Large snipping of system description (note, my recommendation was for a large, 5m high "Theatre" type room seating 200, not a 32 Squaremeeter living/listening room).
Hmmm. 32 squaremeter seems to dictate to me a limited size room where horns with only the mouth surface of around 4 18" Drivers (according to a post further up by qxlxp) would do much to maximally energise room modes, especially when placed in corners.
Did you ever measure the RT60 at low frequencies? With a (sufficiently) directional LF system the common rise at low frequencies is minimal.
BTW, this in no comment on your preferences in sound, but strictly on the technical "absolute" issues involved.
Sayonara
Magnetar said:
I posted it. A "shock" test is nowadays known as "step-response".
Magnetar said:
The high source impedance was needed as the voicecoil was used as sensor. The scan is from the old Briggs book.
Magnetar said:I'd like to see this at 20 volts input where the bandwith of all the examples is limited to being exactly the same along with harmonic distortion measurements below 150 cycles
And I would like to see a RT60 vs. Frequency plot for your system/room, but we cannot always have what we want and have to make do with what we got.
Magnetar said:
To me too.
Large snipping of system description (note, my recommendation was for a large, 5m high "Theatre" type room seating 200, not a 32 Squaremeeter living/listening room).
Magnetar said:
Hmmm. 32 squaremeter seems to dictate to me a limited size room where horns with only the mouth surface of around 4 18" Drivers (according to a post further up by qxlxp) would do much to maximally energise room modes, especially when placed in corners.
Did you ever measure the RT60 at low frequencies? With a (sufficiently) directional LF system the common rise at low frequencies is minimal.
BTW, this in no comment on your preferences in sound, but strictly on the technical "absolute" issues involved.
Sayonara
Kuei Yang Wang said:
A step response "picture" is much dependent on the bandwith of speaker right? I assume the horns in the snapshots have limited bandwith compared to the open baffle. More info can be found here
The mouth of my subhorn (three horns used in unison) is 28 square feet into 1/8th space with concrete and the earth as room wall extensions. I sit inside the horn.
It is my understanding that RT60 is basically invalid in small rooms (depending on the expert) and pretty much a waste of time (according to the experts) trying to measure it in small rooms below 200 cycles. one example can be found here
qi said:
Thanks
Here is some info on Sensurround
I use three of the horns in mono below 50 cycles across the front wall with the mouths venting to the floor. They are loaded with 18" drivers and have slightly modified throats and backchambers (using different drivers in them) along with being throughly braced to lower the enclosure noise. Mine are the MGA model M's
An externally hosted image should be here but it was not working when we last tested it.
Konnichiwa,
To a point. Even when lowpassed there would merely be a slower "attack" and an increase of the area above the slow negative going return to zero. Still the main form of one spike with a smooth return to zero should be retained. Any failure to show such a profile indicates the presence of timedomain problems.
Note that what is measured is NOT the output from the speaker, but the cone movement of the driver, so the bandwidth limitation of the output is irelevent, as it is the stimulus to the speaker assembly by the voicecoil which is measured.
You assume rightly they have limited bandwidth, however note my above comments, the limited OUTPUT bandwidth is irelevant to the distortion of the stimulus.
First, with the crossover at 50Hz you are at about the same point where I suggest my handover to monopole Subs. This put's you pretty much close to the lowest mode in the room. Assuming 32 squaremeter is distributed as 5 X 6.5m your lowest modes are at 25 & 35Hz.
Note that these modes du to a reliable failure to be reinforced by harmonics of other modes will be comparably weak, more so as the room dimension is 1/2 of the wavelength, compared to 1 or several times the wavelength at higher modes. Hence all the REAL room mode problems happen ABOVE the range covered by your horns, moreover, I doubt that much directionality remains at such low frequencies (25..35Hz) even with your horns as the dimensions of the horn array approach those of the room, making any directivity moot.
And hence the prevailing adopted standards for European Radio/TV Sound Studios specify it with a quite narrow tolerance range (below 200Hz) for the room design.
I would say that such comments come from people who simply do not have sellable items that can manage the RT60 below 200Hz (arguably a material problem unless you use low frequency dipoles) and hence are a little bit self serving.... To quote:
"RT-60 measurements in small room acoustics below 200Hz have little significance as they are flawed by the energy build-up of room modes. "
Hmm. So, room modes do not actually influence the RT60 (the time by which the sound of a given stimulus has decayed by 60db)? Funny, I would suggest that the very presence of excited room modes is RESPONSIBLE for the increase in RT-60 and that this added sound when non should be present represents a distortion similar to that of reverb at higher frequencies, except that room modes are "one note", thus altering the percieved pitch of instrument decay.
Surely, if a speaker system does not excite room modes much and if as a result the RT-60 remains with only a small rise at low frequencies in a NORMAL, ACOUSTICALLY UNTREATED ROOM that must be considered as removing distortion. Or am I missing something here?
Maybe all the long standing one note reverb from room modes is actually "a good thing" to some?
Lastly, some comments from the Sensurround introduction about the sensurround horns, though I maintain that in the case of Magnetars room size they are lacking relevance, as the horns (IMNSHO) operate ultimatly purely as monopole pressure source in the room under discussion (which is of course of little consequence at ant extent):
"Reduction of the horn mouth area was accomplished by consideration of the fact that if the horn is located in a corner, the walls and floor of the theatre form boundaries, and restrict angle into which the horn radiates. If the area into which the horn operates is small, the mouth of the horn can be reduced. With a small mouth, some attenuation occurs just above cutoff and there exists some fluctuation in frequency response. Since continuous tones were not being reproduced in this system, the un-eveness in response was not considered important."
Of course, uneven response in the frequency domain usally also implies timedomain problems, though I would assert that at frequencies as low as those discussed and in a room as small as discussed neither distortion nor timedomain behaviour is of much relevance. In fact I would argue that the room dimesions are too small to allow any significant directivity to be developed by a horn system, arguably a statement open to disagreement and based solely on experiences using Hornloaded PA systems in very small rooms (eg. practice-cellar rooms similar in size to what is being discussed here).
It will remain to be seen what the equivalent Woofer conficuration in corner mounted, equalised sealed enclosures of equal volume to the horns would provide in the 15 - 50Hz range, when compared to the existing horns in the same room and how large, if any advantage would remain for the horns and what the respective acoustic outputs would look like.
Sayonara
Magnetar said:
To a point. Even when lowpassed there would merely be a slower "attack" and an increase of the area above the slow negative going return to zero. Still the main form of one spike with a smooth return to zero should be retained. Any failure to show such a profile indicates the presence of timedomain problems.
Note that what is measured is NOT the output from the speaker, but the cone movement of the driver, so the bandwidth limitation of the output is irelevent, as it is the stimulus to the speaker assembly by the voicecoil which is measured.
Magnetar said:
You assume rightly they have limited bandwidth, however note my above comments, the limited OUTPUT bandwidth is irelevant to the distortion of the stimulus.
Magnetar said:
Magnetar said:
First, with the crossover at 50Hz you are at about the same point where I suggest my handover to monopole Subs. This put's you pretty much close to the lowest mode in the room. Assuming 32 squaremeter is distributed as 5 X 6.5m your lowest modes are at 25 & 35Hz.
Note that these modes du to a reliable failure to be reinforced by harmonics of other modes will be comparably weak, more so as the room dimension is 1/2 of the wavelength, compared to 1 or several times the wavelength at higher modes. Hence all the REAL room mode problems happen ABOVE the range covered by your horns, moreover, I doubt that much directionality remains at such low frequencies (25..35Hz) even with your horns as the dimensions of the horn array approach those of the room, making any directivity moot.
Magnetar said:
And hence the prevailing adopted standards for European Radio/TV Sound Studios specify it with a quite narrow tolerance range (below 200Hz) for the room design.
I would say that such comments come from people who simply do not have sellable items that can manage the RT60 below 200Hz (arguably a material problem unless you use low frequency dipoles) and hence are a little bit self serving.... To quote:
"RT-60 measurements in small room acoustics below 200Hz have little significance as they are flawed by the energy build-up of room modes. "
Hmm. So, room modes do not actually influence the RT60 (the time by which the sound of a given stimulus has decayed by 60db)? Funny, I would suggest that the very presence of excited room modes is RESPONSIBLE for the increase in RT-60 and that this added sound when non should be present represents a distortion similar to that of reverb at higher frequencies, except that room modes are "one note", thus altering the percieved pitch of instrument decay.
Surely, if a speaker system does not excite room modes much and if as a result the RT-60 remains with only a small rise at low frequencies in a NORMAL, ACOUSTICALLY UNTREATED ROOM that must be considered as removing distortion. Or am I missing something here?
Maybe all the long standing one note reverb from room modes is actually "a good thing" to some?
Lastly, some comments from the Sensurround introduction about the sensurround horns, though I maintain that in the case of Magnetars room size they are lacking relevance, as the horns (IMNSHO) operate ultimatly purely as monopole pressure source in the room under discussion (which is of course of little consequence at ant extent):
"Reduction of the horn mouth area was accomplished by consideration of the fact that if the horn is located in a corner, the walls and floor of the theatre form boundaries, and restrict angle into which the horn radiates. If the area into which the horn operates is small, the mouth of the horn can be reduced. With a small mouth, some attenuation occurs just above cutoff and there exists some fluctuation in frequency response. Since continuous tones were not being reproduced in this system, the un-eveness in response was not considered important."
Of course, uneven response in the frequency domain usally also implies timedomain problems, though I would assert that at frequencies as low as those discussed and in a room as small as discussed neither distortion nor timedomain behaviour is of much relevance. In fact I would argue that the room dimesions are too small to allow any significant directivity to be developed by a horn system, arguably a statement open to disagreement and based solely on experiences using Hornloaded PA systems in very small rooms (eg. practice-cellar rooms similar in size to what is being discussed here).
It will remain to be seen what the equivalent Woofer conficuration in corner mounted, equalised sealed enclosures of equal volume to the horns would provide in the 15 - 50Hz range, when compared to the existing horns in the same room and how large, if any advantage would remain for the horns and what the respective acoustic outputs would look like.
Sayonara
jeff mai wrote yesterday about dipoles:
“Baffles resonate too”
Yes, but in a horn the sound is potentially or likely colored by both the resonating of the cabinet, and the backwave of delayed energy from the release of sound within the cabinet.
The resonating problem can be dealt with by adequate stiffening; the delayed energy release cannot.
And today “all horns are not the same”
Which is why it would be better to pick a type or diy design that was well regarded (if not necessarily “best” or great in all areas), and base the discussion around that, rather then generalize and leave the door open for inconclusivensess.
On another note, I firmly agree with KYV’s idea of using dipoles down to about 40 Hz, then sealed boxes below.
“Baffles resonate too”
Yes, but in a horn the sound is potentially or likely colored by both the resonating of the cabinet, and the backwave of delayed energy from the release of sound within the cabinet.
The resonating problem can be dealt with by adequate stiffening; the delayed energy release cannot.
And today “all horns are not the same”
Which is why it would be better to pick a type or diy design that was well regarded (if not necessarily “best” or great in all areas), and base the discussion around that, rather then generalize and leave the door open for inconclusivensess.
On another note, I firmly agree with KYV’s idea of using dipoles down to about 40 Hz, then sealed boxes below.
This is what horn loading looks like:
This an example of where RT60 quits being relevant:
An externally hosted image should be here but it was not working when we last tested it.
This an example of where RT60 quits being relevant:
An externally hosted image should be here but it was not working when we last tested it.
Kuei Yang Wang said:
Funny you should mention the Shearer horns. I once had two of these units operating in my home. Even with the suboptimal "thrown together" set up I used - they are unique in my experience.
There is *nothing* that competes with 4m^2 of driver surface area and, make no mistake about this, I'm not talking about loudness or extension (the way I used them they fell off at around 50Hz - I used a Velodyne sub for the bottom octave.) There was an ease that I've not heard before or since. You have to hear it to know - you can't imagine it. They had faults, but this quality more than made up for them.
I'm not sure about all of the Shearer based units, but mine, made by RCA, had no rear chamber - they were designed to run open backed, mounted in a huge baffle or wall. They weighed around 200kg each and were extremely well braced.
Even these huge horns, however, I do not consider ideal. I sold them when I lost my storage space for them with the thought that someday when I have the space again, I'll build straight horns(!) of a similar size.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.