The Real Farfield Measurement Distance

[jmb]

As a general rule, software and people in general speak of the accurate farfield measurement distance being a function of the distance between the measurement microphone and source and the nearest reflectie surface. Others indicate that though this provides a reflection free window, true farfield cannot me measured unless the measurement microphone is at least 2 wavelengths away from the sound source of the lowest frequency of interest. This would put a damper in the accuracy of full range (that is, for the low frequency) ground plane measurements. I was curious to hear the thoughts of any informed engineers.

I did some tests and seem to get some consistent results ignoring the wavelength limited distance requirements, but still, my scientific curiosity got the best of me.

Thanks, Jay

[billfitzmaurice]

Re: The Real Farfiedl Measurement Distance

2 wavelengths throughout the pass band isn't practical for a subwoofer, but that's OK as you don't have to be that far away to be in the farfield, because the farfield distance is determined by the size of the radiating plane. The current de facto standard for measuring pro-sound subs as practiced by Danley Sound Labs is 100 watts at 10 meters ground plane, which not only ensures a farfield measurement but also gives a result far more real world than 1 watt.

[Sydney]

Re: The Real Farfiedl Measurement Distance

Jay - you may find some of this of interest:
http://www.etcinc.us/tech/nl043_far_field_criteria.pdf
The Klippel web site had at one time details for "stitching" near and far field measurements

[jmb]

Re: The Real Farfiedl Measurement Distance

Thanks, Sydney. I had seen that and Klippel's write up, as well. It is just that I am attempting to grasp the mathematical understanding of the why. This paper helps to some degree, but I can't seem to get my hands around how you can have an accurate read being within a wavelength of the source.

Bill, that ground plane distance is impressive and would certainly require a fairly quiet environment though a 100 dB source would mitigate some if this issue. It would be even more difficult to get a reflection free area to really get down to the real low frequencies if you were that far away. I think that for most of us DIY'ers, having access to a facility like this might be a challenge.

Jay

[Sydney]

Re: The Real Farfiedl Measurement Distance

You may have seen a distinction draw between the geometric near field and the acoustic near field?

but I can't seem to get my hands around how you can have an accurate read being within a wavelength of the source.

Multiple measurements on-axis or Multiple positional measurements in a defined matrix that are summed and averaged are approaches.

[ryanbouma]

Re: The Real Farfiedl Measurement Distance

Interesting discussion.

If measuring at 4 watts and 2m in the ground plane, I wonder how accurate that would be. I could find an area that good. 10m is tough though.

[Sydney]

Re: The Real Farfiedl Measurement Distance

Ryan: Do you tilt the cab forward?
Jay: Just to be clear -
Is the goal to measure on-axis FR of a single driver in a "theoretical" infinite baffle?
A driver or driver(s) in a cab?
Or something else?

[ryanbouma]

Re: The Real Farfiedl Measurement Distance

Nope. Do you think that's necessary? I'm only interested in <200hz if doing ground plane.

[Sydney]

Re: The Real Farfiedl Measurement Distance

IMO: Probably not if you are after woofer output.

I could find an area that good. 10m is tough though.

A speaker pit for flush half space measurement isn't an option for many either.

[billfitzmaurice]

Re: The Real Farfiedl Measurement Distance

Bill, that ground plane distance is impressive and would certainly require a fairly quiet environment though a 100 dB source would mitigate some if this issue. It would be even more difficult to get a reflection free area to really get down to the real low frequencies if you were that far away. I think that for most of us DIY'ers, having access to a facility like this might be a challenge.

Danley uses their parking lot. I use my back yard, but since I only have a standard quarter-acre plot, and neighbors, I measure at 4 meters on average.

[Pete Basel]

Re: The Real Farfiedl Measurement Distance

This is Don Keele's near field measurement AES paper and I think it provides the analysis that you're looking for. He uses terminology that looks to be in agreement with Beranek's classic text on Acoustics IIRC - it has been a while since I read that chapter:
http://www.xlrtechs.com/dbkeele.com/...ld%20Paper.pdf

More papers:
http://techtalk.parts-express.com/sh...d.php?t=230830

[jmb]

Re: The Real Farfiedl Measurement Distance

Hi Pete,
Keele's discussion really focusses on Nearfield and does not really explain the farfield distance rationale from what I can tell.

Sydney, the utility to me and I suspect to most would be to be able to measure a boxed speaker (i.e. measurement of both individual drivers and later the combined unit) full range (outdoors, ground plane) or semi anechoic (indoors or outdoors).

I did several measurements with a gated measurement (so it cut out the bottom 250 Hz or so and I see the high frequency issues as one gets too close to the boxed unit. As one moves away, you can see the standard 6 dB drop per doubling of distance. So, like the paper that you reference, Sydney, there is the HF issue of when Farfield truly begins and it is certainly frequency dependent.

...but what about low frequency? We can get nearfield measurements and then model the diffraction effects and fit it to our farfield measurements; but can we truly get farfield measurements on low frequency signals without being several wavelengths away from the source? It does not seem so from what I have read. Historically, we have used (as the lowest reliable frequency), the time window based upon the direct measurement until the initiation of the geometric calculation of the nearest reflection - but this does not take into account that a 300 Hz signal is over a meter long and would therefore require a measurement at least a couple of meters away to truly capture a far field signal.

This is what I am trying to wrap my mind around. I guess the real issue here is, perhaps we have been splicing our anechoic measurements too low? Perhaps we should be attempting to splice our signals closer to 500-600 Hz (nearfield allowing)?

Jay

[Sydney]

Re: The Real Farfield Measurement Distance

Jay:
I feel like I telling you what you already know here.
Practically speaking - far field is where the energy drops inversely (-6db ). Closer in is the interference field.
Of course small wavelength high frequency energy is subject to attenuation by humidity ( which can be noted and accounted for ).
I believe you know the reason for using windowing, and what would need to happen to have the window "opened longer" so to speak.
There are numerous approaches to near field measurement methods, like other testing methodology in other domains they have limitations or compromises depending on what is being quantified.

[billfitzmaurice]

Re: The Real Farfiedl Measurement Distance

this does not take into account that a 300 Hz signal is over a meter long and would therefore require a measurement at least a couple of meters away to truly capture a far field signal.

You're confusing the issue. There are two 'nearfields'. One is 'very close', as opposed to 'far'. The other is where a source differentiates from point to line source radiation characteristics. That definition relates to the size of the radiating plane versus frequency plus distance, not distance alone. Yes, a 300Hz signal is a meter long, but the source must be approximately three times that in height to be a line source. So if you're measuring a ten foot high speaker and want to do so within the farfield zone you need to be well away. With a two foot high speaker a few meters will do.

[johnk...]

Re: The Real Farfield Measurement Distance

I think you guys may be looking at this incorrectly. Here is a figure of the near to far field transition for a flat, circular piston source:



Far field is where the response drops 6dB per doubling of distance. Notice that regardless of wave length, as indicated by ka, (ka = 2Pi x a / wavelength where a is the radius of the source), the response is pretty much dropping by 6dB/doubling at distances of r/a = 3 or 4. Small ka indicates low frequency, long wave length, relative to the source size. Notice that it is r/a, distance from source divided by source radius, that matters, not distance divided by wave length. For a 12" woofer, regardless of frequency, the far field transition is complete by about 2 feet (r/a = 4). To be super cautious, use r/a = 10, or about 5 feet. For a 6" midrange the far field is at about 1 ft of a little more.


The size of the baffle should probably be figured in, but for the most part 1 M measurements are far field for a single source.