Against many of my instincts, I'm thinking about cardoid* or dipole midbass for my next front mains.
The rough outline of the speakers I'm contemplating be as follows:
below 120Hz - multisubs (closed box)
120ish-400ish Hz - two 7" drivers with Fs=~50Hz and Qts=0.85), which I'm considering making dipole or cardoid
400ish Hz up - 5.25" concentric driver, ~120º pattern through the midrange.
However, this setup is a multichannel setup in the living room of my loft, so there are placement issues. (My system is designed to play music, with anything involving pictures on the screen that's not a concert being a tertiary concern.) The baffle of center channel can't be more than about 15" from the wall behind it. So, for midbass only, can close-to-wall placement work for OB/cardoid?
*Now, how to get cardioid radiation with just two front-firing drivers in this range is another issue, but one I plan to study before asking questions about.
The rough outline of the speakers I'm contemplating be as follows:
below 120Hz - multisubs (closed box)
120ish-400ish Hz - two 7" drivers with Fs=~50Hz and Qts=0.85), which I'm considering making dipole or cardoid
400ish Hz up - 5.25" concentric driver, ~120º pattern through the midrange.
However, this setup is a multichannel setup in the living room of my loft, so there are placement issues. (My system is designed to play music, with anything involving pictures on the screen that's not a concert being a tertiary concern.) The baffle of center channel can't be more than about 15" from the wall behind it. So, for midbass only, can close-to-wall placement work for OB/cardoid?
*Now, how to get cardioid radiation with just two front-firing drivers in this range is another issue, but one I plan to study before asking questions about.
Bump!
Since I'm contemplating a cardioid midbass (and have similar placement constraints), too: No opinions on Pallas's questions?
I know JohnK's simulations at Boundary effects and some related statements in Holmes's and Backman's papers on cardioids, which imply a beneficial effect of placing a cardioid speaker (instead of a monopole) near the front wall because there is no destructive interference from mirror sources on the 0° axis.
But while researching I also stumbled upon a (rather rambling) discussion of the topic on a German forum where some initially plausible objections were raised concerning the directivity pattern resulting from front wall induced interference. The upshot is that the "quadrupole" resulting from the mirror sources has even wider dispersion than a monopole with "uncontrolled" directivity. This discussion (in German) can be found here: http://www.hifi-forum.de/index.php?action=browseT&forum_id=104&thread=20203
How should those two perspectives be prioritized? Can any of the more advanced members around here shed some more light on the aspects involved?
Since I'm contemplating a cardioid midbass (and have similar placement constraints), too: No opinions on Pallas's questions?
I know JohnK's simulations at Boundary effects and some related statements in Holmes's and Backman's papers on cardioids, which imply a beneficial effect of placing a cardioid speaker (instead of a monopole) near the front wall because there is no destructive interference from mirror sources on the 0° axis.
But while researching I also stumbled upon a (rather rambling) discussion of the topic on a German forum where some initially plausible objections were raised concerning the directivity pattern resulting from front wall induced interference. The upshot is that the "quadrupole" resulting from the mirror sources has even wider dispersion than a monopole with "uncontrolled" directivity. This discussion (in German) can be found here: http://www.hifi-forum.de/index.php?action=browseT&forum_id=104&thread=20203
How should those two perspectives be prioritized? Can any of the more advanced members around here shed some more light on the aspects involved?
Thanks for joining the discussion, Rudolf! As I'm quite new to the cardioid discussion camp, please bear with me while I'm trying to wrap my head around this stuff.
As far as I understand, the model behind JohnK's simulation and your statement is this: When placing a (working) cardioid before a reflecting wall, the resulting mirror sources should be unipolar, too. The reflected sound from these mirrored cardioid is equivalent to the reduced rear or side output from the original source (minus losses), so the interference between original and mirror sources in front of the speakers should be reduced (or even eliminated when only front wall reflections and the 0° axis are being considered, as in JohnK's case). The following illustration from Holmes (JAES Vol. 34 No. 12, p. 985) should illustrate this model:
On the other hand, people in the cited thread and the one following it (to which you contributed, too, IIRC) argued as follows: The reduction of a a cardioid's rear output occurs through destructive interference between the sound from its front and back drivers. This assumes unhindered radiation in both directions (front and back), i.e. distances to reflecting surfaces far greater than the wavelengths involved: When modeled and mirrored separately, both the reflected front and back radiation from the cardioid's sources will take part in the interference in a rather chaotic manner and destroy the interference pattern forming a cardioid responce.
The posts I'm specifically referring to are these:
http://www.hifi-forum.de/index.php?action=browseT&forum_id=104&thread=20203&postID=26#26
http://www.hifi-forum.de/index.php?action=browseT&forum_id=104&thread=20555&postID=20#20
Both try to illustrate this point by presenting simulations of cardioids near walls, showing a collapse of the cardioid pattern.
Is my summary of the arguments involved roughly correct, or did I miss something? If it is, how are the two lines of reasoning reconciled?
Thanks in advance for some elucidation!
As far as I understand, the model behind JohnK's simulation and your statement is this: When placing a (working) cardioid before a reflecting wall, the resulting mirror sources should be unipolar, too. The reflected sound from these mirrored cardioid is equivalent to the reduced rear or side output from the original source (minus losses), so the interference between original and mirror sources in front of the speakers should be reduced (or even eliminated when only front wall reflections and the 0° axis are being considered, as in JohnK's case). The following illustration from Holmes (JAES Vol. 34 No. 12, p. 985) should illustrate this model:
On the other hand, people in the cited thread and the one following it (to which you contributed, too, IIRC) argued as follows: The reduction of a a cardioid's rear output occurs through destructive interference between the sound from its front and back drivers. This assumes unhindered radiation in both directions (front and back), i.e. distances to reflecting surfaces far greater than the wavelengths involved: When modeled and mirrored separately, both the reflected front and back radiation from the cardioid's sources will take part in the interference in a rather chaotic manner and destroy the interference pattern forming a cardioid responce.
The posts I'm specifically referring to are these:
http://www.hifi-forum.de/index.php?action=browseT&forum_id=104&thread=20203&postID=26#26
http://www.hifi-forum.de/index.php?action=browseT&forum_id=104&thread=20555&postID=20#20
Both try to illustrate this point by presenting simulations of cardioids near walls, showing a collapse of the cardioid pattern.
Is my summary of the arguments involved roughly correct, or did I miss something? If it is, how are the two lines of reasoning reconciled?
Thanks in advance for some elucidation!
Correct, though the radiation pattern comes gradually through as frequency rises. Hence for frequencies close and above the Schröder frequency such a configuration can be desirable.
As I have recently stated in various posts the correct way to look at dipoles and cardiods at low frequency in a room is to look at them as multiple monopoles with different phase (and delays) and position. Each source contributes linearly to the room response. This is different than considering boundary reenforcement alone.
I have a cardioid midrange with > 20 dB front/back ratio and I cannot measure the front wall reflection, even though the speakers are nearly directly in front of the wall (baffle is about 30 cm from the wall). Guess it does work rather good.
Dipoles are different though. They definitely need a lot of space and/or toe-in to avoid a strong front wall reflection.
Dipoles are different though. They definitely need a lot of space and/or toe-in to avoid a strong front wall reflection.
Depends on the measurement. The low frequency reflection is definitely there though not visible.
Probably. I use it from 120 Hz up which does involve somewhat long wavelengths, but it's not really that low I guess. I've noticed an absence of a sensation of pressure or a pressure gradient between the ears when I put my head directly behind the speaker. I'm curious what a free-field measurement at different distances would show.
I want to do a better measurement some time, and perhaps a comparison with a closed box speaker. I cannot hear or measure the bass increase a normal box speaker has when you move it towards the wall.
I want to do a better measurement some time, and perhaps a comparison with a closed box speaker. I cannot hear or measure the bass increase a normal box speaker has when you move it towards the wall.
John and Markus, can you elaborate a bit on the frequency ranges where the effects you mention come into play?
Up to now, i thought that
But in a_tewinkel's case, the wavelengths are still large compared to the wall distance, and the frequency region the OP is focusing (midbass) is – at leat partly – well above the Schröder frequency of modestly sized rooms.
I guess I'm missing something important here…
Up to now, i thought that
- dipoles and cardioids should be considered as multiple monopoles only in the modal region of the room,
- a cardioid pattern should emerge as wavelengths become short compared to the wall distance(s).
But in a_tewinkel's case, the wavelengths are still large compared to the wall distance, and the frequency region the OP is focusing (midbass) is – at leat partly – well above the Schröder frequency of modestly sized rooms.
I guess I'm missing something important here…
The thing to understand is that dipoles and cardioids only achieve the theoretical radiation pattern in the free field. Free field means 3 to 10 wave lengths away from the source. If you move much closer then the difference in distance between front and rear sources means that there is a significant difference is source strength at the measurement point due to path length differences. See Dipole on axis. In a room, there is not dipole or cardioid radiation at low frequency. It's all modal.
...unless you're in the near field
Yes, that is a point I have made before as well. But you know that.
Is the wavelength the most important parameter or the effective path distance between the two sources? In my case, the former would be < 2.85 meters, the latter would be about 0.16 m.
Can you illustrate or show photos of the cab's here? I'm interested in the mechanics behind what polar(s) are being produced compared to what a cardioid pattern would normally show. I'm having some difficulty imagining a true cardioid rotation not being affected (ie. no reflection) 11" off a wall.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.