You mean like in "reflexion"?
You had an excuse. I didn't realize that it was German.
Makes sense. I guess a dipole would become omni directional closer. That is, the phase dosen't do a 180 as you move behind the sub. This is what I understand as near field, the distance where the sub stops being an omni source and the phase starts doing strange things as you go around the sub.
No, a dipole is a dipole even in the near field (even in the direct field!) It gets more complex in the near field, i.e. the bass falls off slower in the near field than it does in the far field, but there is still a "basic" dipole pattern with the added complexity superimposed on top of this.
After reading this thread I've started once again to wonder how direct field at low frequencies even works. Based on markus76's results, it does work by giving flat frequency response, but how could this happen when wavelengths are longer than wall-to-wall distance? How does pressure variation work in close distances to the source?
The direct field is basically the same as the speaker would be in a free field, an inverse square law as suggested. Close enough and the room modes are not larger in effect and the direct field peaks up above the modes - mostly.
Mathematically the direct field is contained in the modal solution, but the direct field requires many many modes to contribute to create the singular form of the solution near the source. In mathematical terms the direct field converges very slowly. But one can separate off the direct field in the solution and calculate it separately to a high order and then use fewer terms in modal field thus yielding a good model in all regions.
One problem that I had with the Welti paper on multiple subs was that he just added in the direct field which is an error. He did not modify the modal field for his extraction of the direct field. It was not serious since his subs were not really very close to the listeners, but it was still wrong.
Mathematically the direct field is contained in the modal solution, but the direct field requires many many modes to contribute to create the singular form of the solution near the source. In mathematical terms the direct field converges very slowly. But one can separate off the direct field in the solution and calculate it separately to a high order and then use fewer terms in modal field thus yielding a good model in all regions.
One problem that I had with the Welti paper on multiple subs was that he just added in the direct field which is an error. He did not modify the modal field for his extraction of the direct field. It was not serious since his subs were not really very close to the listeners, but it was still wrong.
Err guys, am I the only one who hasn't forgotten about point source to plane source transistion in the near field ?
Inverse square law (6dB per distance doubling or halving) only holds true so long as the listening distance is "large" compared to the diameter of the radiator, even at low frequencies.
This is because inverse square law is based on spreading losses from a point source. Get close enough to a woofer and its no longer a point source, it begins to approximate a plane source.
Since a true infinite plane source has no fall off with distance, transitioning from a point source to an approximation of a plane source causes the increase in SPL as you get close to the woofer flatten out and no longer follow the expected curve. If it didn't the SPL would rise to ridiculous levels close to the cone, but it doesn't. (6dB more SPL 5mm from the cone of a 12" woofer than at 10mm ? Not a chance!)
I'm also not sure that I agree with the given definition of near field a few posts back in the near field versus "direct field" (whatever THAT is at bass frequencies) debate, to me it sounds like playing with semantics. I'm more inclined to agree with Markus's original suggestion that his sub is a "near field" sub, as "direct field" is meaningless to me at bass frequencies.
To me the definition of near field and far field for a "point source" is well defined - so long as the amplitude response tracks with inverse square law (at ALL frequencies, and therefore by definition the frequency response doesn't change with distance) then you're in the far field.
As you approach closer once the amplitude starts to deviate from inverse square law (less than expected rise) or the frequency response changes (really the same thing - just different deviations from inverse square law at different frequencies) then you're transitioning into the near field - where far field measurements can not be extrapolated.
It seems I'm not the only one to agree with this definition, as the following article goes into some detail about definining the cut off between near field and far field, and places it at 6 times the radius of the radiator: (page 3)
http://www.artalabs.hr/AppNotes/AP4_FreeField-Rev03eng.pdf
By this definition Markus sub would be near field.
Simon
I completely disagree that "direct field" has no meaning at LFs. It is just as meaningful as at any other frequency because the concept is frequency independent.
At the frequencies that we are talking about the source is virtually a point source and the "near field" is that of a point source not a plane source. I just do not agree with your assessment that there is a planar wave "close enough" to the woofer. That would have to be centimeters if at all. The source is miniscule compared to the wavelengths.
As your own reference makes clear, for small ka, a value of six times the radius is extremely conservative. It's more like 2 times (or less) for values of ka << 1.0 which is the case here. So in essence, the sources that we are talking about have no "near field" and Markus's situation is certainly "direct field". (I was wrong when I said that they were the same thing in Markus case, they clearly are not.)
So while I do not agree with you at all, I do see the situation more clearly and I had it wrong before. "Near field subs" is simply wrong - not semantic.
I completely disagree that "direct field" has no meaning at LFs. It is just as meaningful as at any other frequency because the concept is frequency independent.
At the frequencies that we are talking about the source is virtually a point source and the "near field" is that of a point source not a plane source. I just do not agree with your assessment that there is a planar wave "close enough" to the woofer. That would have to be centimeters if at all. The source is miniscule compared to the wavelengths.
As your own reference makes clear, for small ka, a value of six times the radius is extremely conservative. It's more like 2 times (or less) for values of ka << 1.0 which is the case here. So in essence, the sources that we are talking about have no "near field" and Markus's situation is certainly "direct field". (I was wrong when I said that they were the same thing in Markus case, they clearly are not.)
So while I do not agree with you at all, I do see the situation more clearly and I had it wrong before. "Near field subs" is simply wrong - not semantic.
Last edited:
Maybe so, but the near field is only in the first several centimeters. So by what you are saying there is neither a near field nor far field. Maybe we should call it the "Markus field"? You situation is clearly the direct field.
There being no "near field" is consistent with my room model. Basically adding in the proper "near field" terms does not change things very much in most cases. there are some situation that change, but for a damped room the near field is not very significant.
There being no "near field" is consistent with my room model. Basically adding in the proper "near field" terms does not change things very much in most cases. there are some situation that change, but for a damped room the near field is not very significant.
You may have a point - if you consider a woofer on a finite sized baffle (instead of the infinite baffle assumption in the article I linked) I believe that the effective radiating size for calculating the far field to near field transition distance becomes 3x the width/height of the baffle, rather than 3x the diameter of the driver. (At least in frequency ranges where diffraction is occurring)
So even though your driver might only be 20cm in diameter, if you put it on a 50cm wide baffle the baffle diffraction becomes part of the source radiation and increases the distance at which far field to near field transition occurs - this is evident in that baffle step loss from a baffle (particularly a large one) is not "fully formed" until you're quite some distance from the baffle - more than a metre for most baffles.
Point being that the physical size of the driver is not the whole story if there is re radiation of energy nearby.
If the myriad of reflections contained within the room boundaries increase the effective source size are you really in the far field ? Maybe not.
In a room with fairly reflective walls and woofers only at one end does the average bass level fall off with inverse square law with distance or does it fall off much slower ? (Of course determining the "average" level in a room full or standing waves is not trivial) If it falls off significantly slower than inverse square law you can't be in the far field.
Does a distributed sub setup mean that the effective source size is nearly the size of the room, thus you are always in the near field (of the virtual source) at any location in the room ? If you really were in the near field anywhere in the room then the bass response wouldn't vary much in amplitude at different locations in the room with no obvious trend of reduction in bass in a certain direction away from the "source" ?
Just stirring the pot a bit...
Stirring the pot, but incorrect. The near field for the infinite baffle would likely be further than for a box because the radiation can spread out all the more rapidly. In either case its still only a few centimeters and Markus situation is still not "near field".
Only sources can have near fields, not rooms. And as I discovered adding in the near field terms into my room modes calculations it is only an effect at extremely close distances. There simply is not a near field - by any definition that I know of - for a subwoofer. There can be a direct field if the room is damped well enough that there is not a substantial reverberant field. But even that goes away for an undamped room, except that one can see some filling of the nulls from the direct field close to the source.
Only sources can have near fields, not rooms. And as I discovered adding in the near field terms into my room modes calculations it is only an effect at extremely close distances. There simply is not a near field - by any definition that I know of - for a subwoofer. There can be a direct field if the room is damped well enough that there is not a substantial reverberant field. But even that goes away for an undamped room, except that one can see some filling of the nulls from the direct field close to the source.
Before I get jumped on, let me elaborate a bit further.
Before we can answer "are we in the near field or far field" we need to define of what - an individual driver, or an array of drivers ?
For example imagine an array of four 12" woofers spaced two metres apart centre to centre located at the four corners of a square, and operated at a low enough frequency that they're less than half a wavelength apart.
From a very large distance out in free space they will all look like one point source - by all definitions you are in the far field. Inverse square law will prevail.
Get much closer along the central axis of the four and you will start to enter the near field - based on listening distance and the spacing of the drivers which form a virtual source at the centre of the four with a diameter of approximately the centre to centre spacing of the drivers...
But this is the near field of the array not necessarily of the individual drivers. The array size is much larger than an individual driver.
As you travel along the centre axis of the array towards the plane of the array the increase will flatten out because you are in the near field of the array yet you might still be in the far field (or nearly in the far field) of an individual driver.
Now move sideways towards an individual driver and you will enter the near field of an individual driver and experience a "local maximum".
My point being that you could simultaneously be in the far field of an individual driver, but in the near field of the array of multiple drivers in a multi sub configuration.
(In fact in the case of a woofer in each corner of a room and the listener at the centre you are at a maximum distance from all the individual drivers so have no near field effect from an individual driver but you are at the centre of the virtual source that the array forms)
Markus's point is is there any difference between an array of drivers in a room and an array of virtual drivers formed by the reflections in the room boundaries. Do the virtual sources mirrored in the walls form a physically large array which has its own near field and far field characteristic at a macro level over an above the individual drivers ?
Am I making sense ?
Before we can answer "are we in the near field or far field" we need to define of what - an individual driver, or an array of drivers ?
For example imagine an array of four 12" woofers spaced two metres apart centre to centre located at the four corners of a square, and operated at a low enough frequency that they're less than half a wavelength apart.
From a very large distance out in free space they will all look like one point source - by all definitions you are in the far field. Inverse square law will prevail.
Get much closer along the central axis of the four and you will start to enter the near field - based on listening distance and the spacing of the drivers which form a virtual source at the centre of the four with a diameter of approximately the centre to centre spacing of the drivers...
But this is the near field of the array not necessarily of the individual drivers. The array size is much larger than an individual driver.
As you travel along the centre axis of the array towards the plane of the array the increase will flatten out because you are in the near field of the array yet you might still be in the far field (or nearly in the far field) of an individual driver.
Now move sideways towards an individual driver and you will enter the near field of an individual driver and experience a "local maximum".
My point being that you could simultaneously be in the far field of an individual driver, but in the near field of the array of multiple drivers in a multi sub configuration.
(In fact in the case of a woofer in each corner of a room and the listener at the centre you are at a maximum distance from all the individual drivers so have no near field effect from an individual driver but you are at the centre of the virtual source that the array forms)
Markus's point is is there any difference between an array of drivers in a room and an array of virtual drivers formed by the reflections in the room boundaries. Do the virtual sources mirrored in the walls form a physically large array which has its own near field and far field characteristic at a macro level over an above the individual drivers ?
Am I making sense ?
Last edited:
You do but then again I've already surrendered to "direct field". This is what I've measured and more important it's what I hear.
Simon
First, I was never talking about a close coupled array of drivers, so let's dismiss that.
If you are saying that the concepts of near and far field get complicated in a small room then I would agree. But then we shouldn't use the terms at all until we have a concrete definition. As I said, the term "near field" only applies to a source not a room so if the two are intimately related then we have no applicable definition of the term - so we should avoid using it.
I have no problem with the term "direct field" used here because that is defined in the context of a room, and it fits Markus situation. What's not to like!
First, I was never talking about a close coupled array of drivers, so let's dismiss that.
If you are saying that the concepts of near and far field get complicated in a small room then I would agree. But then we shouldn't use the terms at all until we have a concrete definition. As I said, the term "near field" only applies to a source not a room so if the two are intimately related then we have no applicable definition of the term - so we should avoid using it.
I have no problem with the term "direct field" used here because that is defined in the context of a room, and it fits Markus situation. What's not to like!
You mean a dipole? Why would a monopole be 180 degrees out of phase behind it?
A dipole is two monopole point sources separated in space by a distance called the "moment". They are out of phase. You will see phase changes even close up, but not in between them. The phase is undefined when the SPL is zero.
A dipole is two monopole point sources separated in space by a distance called the "moment". They are out of phase. You will see phase changes even close up, but not in between them. The phase is undefined when the SPL is zero.
I don't think I made myself clear. As I understand it: a point source is in phase all the way around it. ( at any distance from the source the crest of the wave arives at the same time no matter the angle(I front or behind). Real woofers are not point sources. They don't push the air out in all directions at once, they only push the air forward, which is backwards(180degrees) if your behind the woofer, which should put it out of phase. (Think of a kick drum). Why would a monopole be 180 degrees out of phase behind it?Then at a certain distance( baffle size?) it stars to become in phase. Am I missing something?
Yea, I think that you are. We are talking about wavelengths here that are an order of magnitude greater that the size of the source. You will not see the kinds of things that (I think) you are talking about.
At very high frequencies, there will be a small spot right behind the enclosure that is out of phase. But this is fairly high in frequency and its a fairly small spot.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.