Reply
 
Thread Tools Display Modes
  #1   Report Post  
Mark
 
Posts: n/a
Default 20hz to 20Khz , yea right!

In article ,
says...
I have had no luck in measuring speaker response at low frequency's,


Then you may not be using the right tools.



Almost any speaker system sold today claims 20Hz to 20Kz response yet this
is so far from the truth I cannot understand how they dare claim such
figures.

Because they (or you) are not providing all the information.

It isn't enough to say 20-20k, you must also include (for
the measurement to have Any validity) what the +/- dB range
is across the band. If you want a Truly accurate spec, it
will also include at what power level the measurement was
made, ie "maximum rated power" or "1 watt, 1 meter".

Most consumer and car speakers on the market fudge at least
one, if not all three of these qualifiers. Why? To take
gullible fools' money.


I have yet to find ANY speaker system using an 8 inch driver that has the
ability to produce 40hz or lower frequencys as they all have fallen so far
down in output level as to be useless.


See above. Also, there are significant physics issues
involved in trying to get an 8" driver to reproduce low
frequencies at any significant SPL. They simply can't move
enough air.


How do you measure the responese at 20Hz?, I have tested a few speakers and
basically I cannot get accurate (repeatable) data at much below 60Hz, even
using a borrowed shure KSM141 stereo pair in omnidirectional gives vague
results


First off, you should Never use a stereo mic (or any stereo
set up) for measurement purposes. You are guaranteed to get
inaccurate results. Secondly, the mic is only part of the
equipment. What else are you using?

I have come to the conclusion that you are
really measuring air displacement at anything under 30Hz


Ummm. You're measuring air displacement at Any frequency.
That's what audible sound IS.

and a flat panel
with a transducer is the only repeatable method of measuring the output,


Balderdash.

--
Mark

The truth as I perceive it to be.
Your perception may be different.

Triple Z is spam control.
  #2   Report Post  
Mark
 
Posts: n/a
Default 20hz to 20Khz , yea right!

In article ,
says...
I have had no luck in measuring speaker response at low frequency's,


Then you may not be using the right tools.



Almost any speaker system sold today claims 20Hz to 20Kz response yet this
is so far from the truth I cannot understand how they dare claim such
figures.

Because they (or you) are not providing all the information.

It isn't enough to say 20-20k, you must also include (for
the measurement to have Any validity) what the +/- dB range
is across the band. If you want a Truly accurate spec, it
will also include at what power level the measurement was
made, ie "maximum rated power" or "1 watt, 1 meter".

Most consumer and car speakers on the market fudge at least
one, if not all three of these qualifiers. Why? To take
gullible fools' money.


I have yet to find ANY speaker system using an 8 inch driver that has the
ability to produce 40hz or lower frequencys as they all have fallen so far
down in output level as to be useless.


See above. Also, there are significant physics issues
involved in trying to get an 8" driver to reproduce low
frequencies at any significant SPL. They simply can't move
enough air.


How do you measure the responese at 20Hz?, I have tested a few speakers and
basically I cannot get accurate (repeatable) data at much below 60Hz, even
using a borrowed shure KSM141 stereo pair in omnidirectional gives vague
results


First off, you should Never use a stereo mic (or any stereo
set up) for measurement purposes. You are guaranteed to get
inaccurate results. Secondly, the mic is only part of the
equipment. What else are you using?

I have come to the conclusion that you are
really measuring air displacement at anything under 30Hz


Ummm. You're measuring air displacement at Any frequency.
That's what audible sound IS.

and a flat panel
with a transducer is the only repeatable method of measuring the output,


Balderdash.

--
Mark

The truth as I perceive it to be.
Your perception may be different.

Triple Z is spam control.
  #3   Report Post  
Ethan Winer
 
Posts: n/a
Default 20hz to 20Khz , yea right!

Flash,

Besides all that Mark said, you can't measure loudspeakers by playing tones
in a typical room in a typical house. If you take it outside, far away from
ambient noises, you'll have half a shot at it.

--Ethan


  #4   Report Post  
Ethan Winer
 
Posts: n/a
Default 20hz to 20Khz , yea right!

Flash,

Besides all that Mark said, you can't measure loudspeakers by playing tones
in a typical room in a typical house. If you take it outside, far away from
ambient noises, you'll have half a shot at it.

--Ethan


  #5   Report Post  
Mark
 
Posts: n/a
Default 20hz to 20Khz , yea right!

In article ,
says...

"Mark" wrote in message
th.net...
In article ,
says...
I have had no luck in measuring speaker response at low frequency's,


Then you may not be using the right tools.


What tools do you use?


What is needed is a measurement mic that is calibrated to
the device being used for the test measurements.

I've personally used a number of different set-ups, from a
number of different manufacturers. My favorite for best
bang for the buck is the Audio Control Industrial SA 3052.
There are plenty of other devices that will do a respectable
job as well.

First off, you should Never use a stereo mic (or any stereo
set up) for measurement purposes. You are guaranteed to get
inaccurate results. Secondly, the mic is only part of the
equipment. What else are you using?


Never?

Yes, never. If what you're interested in is the frequency
response from a single driver or cabinet, you want to make
that measurement from one point in space.

By using a stereo setup I can collect two data streams rather than
one, I can compare the streams and quickly analyse microphone placement and
any other abnormalities.


Standard measurement is done at 1 meter, on axis. Period.
(Unless, of course, you're creating a polar chart.) Any
abnormalities that exist are a result of either poor
driver/box design, or space-induced aberrations.

And speaking of which, I note in a different response of
this thread that you put nominal weight on the influence of
the space in which you're testing. I would suggest that
Ethan is dead-on in his comment, _particularly_ when the
concern is low-frequency response. By testing in a confined
space that is Not truly anechoic, you're guaranteed to
introduce abnormalities in the response curve.

I have access to / a collection of equipment, some built by HP other stuff
designed and manufactured for the BBC or others. All the HP gear uses HPIB
which is not what is the most convenient, The A2D - D2A card in the PC is a
much quicker and generates the same results (But much quicker) I have a
number of amplifers, but typically use a Class A 2n3055 based 60 watt mono
amp, It has acceptable characteristics for the type of measurement that I
usually do, 20Hz to 20Khz on / off axis at 1w / 1m / 1khz. Microphones is
the biggest issue as I have not yet found anything that I am really happy
with. I can use the Shures, I have some clone shure that are very flat from
60hz to 20Khz and cost like $60 each! (vs 3 grand for the shures) (I can
compensate in software for the low end droop or use a opamp based corrector
that was build before I started with the PC based A2D card. I have a nice
transducer (Probably from a bridge or Crane) that works from a few Hz.


None of which matters if the measurement tool and the
microphone are not calibrated _To Each Other_. Otherwise
you're fooling yourself into thinking you're getting
accurate measurements.

and a flat panel
with a transducer is the only repeatable method of measuring the output,


Balderdash.


Now thats a technical response!,


Alright, here's the longer response. _Any_ calibrated
mic/tool combination designed for the job will accurately
measure to 20 Hz (and lower) within its own tolerance.

Ok, Can some one tell me what mics / transducers I need to be using to get
accurate low frequency measurements. Not just 'use your ears' as this is far
too subjective!

As I indicated, it isn't just a matter of choosing a mic.
If the mic and measurement tool aren't calibrated to each
other, than accuracy is, at best, a crap shoot.

--
Mark

The truth as I perceive it to be.
Your perception may be different.

Triple Z is spam control.


  #6   Report Post  
Mark
 
Posts: n/a
Default 20hz to 20Khz , yea right!

In article ,
says...

"Mark" wrote in message
th.net...
In article ,
says...
I have had no luck in measuring speaker response at low frequency's,


Then you may not be using the right tools.


What tools do you use?


What is needed is a measurement mic that is calibrated to
the device being used for the test measurements.

I've personally used a number of different set-ups, from a
number of different manufacturers. My favorite for best
bang for the buck is the Audio Control Industrial SA 3052.
There are plenty of other devices that will do a respectable
job as well.

First off, you should Never use a stereo mic (or any stereo
set up) for measurement purposes. You are guaranteed to get
inaccurate results. Secondly, the mic is only part of the
equipment. What else are you using?


Never?

Yes, never. If what you're interested in is the frequency
response from a single driver or cabinet, you want to make
that measurement from one point in space.

By using a stereo setup I can collect two data streams rather than
one, I can compare the streams and quickly analyse microphone placement and
any other abnormalities.


Standard measurement is done at 1 meter, on axis. Period.
(Unless, of course, you're creating a polar chart.) Any
abnormalities that exist are a result of either poor
driver/box design, or space-induced aberrations.

And speaking of which, I note in a different response of
this thread that you put nominal weight on the influence of
the space in which you're testing. I would suggest that
Ethan is dead-on in his comment, _particularly_ when the
concern is low-frequency response. By testing in a confined
space that is Not truly anechoic, you're guaranteed to
introduce abnormalities in the response curve.

I have access to / a collection of equipment, some built by HP other stuff
designed and manufactured for the BBC or others. All the HP gear uses HPIB
which is not what is the most convenient, The A2D - D2A card in the PC is a
much quicker and generates the same results (But much quicker) I have a
number of amplifers, but typically use a Class A 2n3055 based 60 watt mono
amp, It has acceptable characteristics for the type of measurement that I
usually do, 20Hz to 20Khz on / off axis at 1w / 1m / 1khz. Microphones is
the biggest issue as I have not yet found anything that I am really happy
with. I can use the Shures, I have some clone shure that are very flat from
60hz to 20Khz and cost like $60 each! (vs 3 grand for the shures) (I can
compensate in software for the low end droop or use a opamp based corrector
that was build before I started with the PC based A2D card. I have a nice
transducer (Probably from a bridge or Crane) that works from a few Hz.


None of which matters if the measurement tool and the
microphone are not calibrated _To Each Other_. Otherwise
you're fooling yourself into thinking you're getting
accurate measurements.

and a flat panel
with a transducer is the only repeatable method of measuring the output,


Balderdash.


Now thats a technical response!,


Alright, here's the longer response. _Any_ calibrated
mic/tool combination designed for the job will accurately
measure to 20 Hz (and lower) within its own tolerance.

Ok, Can some one tell me what mics / transducers I need to be using to get
accurate low frequency measurements. Not just 'use your ears' as this is far
too subjective!

As I indicated, it isn't just a matter of choosing a mic.
If the mic and measurement tool aren't calibrated to each
other, than accuracy is, at best, a crap shoot.

--
Mark

The truth as I perceive it to be.
Your perception may be different.

Triple Z is spam control.
  #7   Report Post  
Goofball_star_dot_etal
 
Posts: n/a
Default 20hz to 20Khz , yea right!

On Sun, 23 Nov 2003 02:44:43 -0800, "The Flash" wrote:

I have had no luck in measuring speaker response at low frequency's, I can
measure from ~60Hz to 20Khz with near 100% repeatability (How accurate is
questionable)

I have questioned a number of speaker builders and a couple of companies
that 'tune' speakers the answers given on what they measure with and how the
do the tests give serious rise to the claims at low frequency.

Almost any speaker system sold today claims 20Hz to 20Kz response yet this
is so far from the truth I cannot understand how they dare claim such
figures.

I have yet to find ANY speaker system using an 8 inch driver that has the
ability to produce 40hz or lower frequencys as they all have fallen so far
down in output level as to be useless.


How do you measure the responese at 20Hz?, I have tested a few speakers and
basically I cannot get accurate (repeatable) data at much below 60Hz, even
using a borrowed shure KSM141 stereo pair in omnidirectional gives vague
results ($3000 for the mic's!) I have come to the conclusion that you are
really measuring air displacement at anything under 30Hz and a flat panel
with a transducer is the only repeatable method of measuring the output, how
one calibrates said device is open for discussion.

(Oh one firm told me that they use a laser to measure the low frequency of
their speakers, check this out for novel! They place a small piece of
reflective foil on the base driver, and shine a laser beam on it, they then
apply signal and measure via 'laser' the deflection, Also they place a
passive radiator 1 meter infront of the driver unit and use the same method
to measure its deflection.

The apply a 'correction factor' and the produce the frequency response data
(company builds very expensive car and home audio subs!)






I thought up my own method for eliminating the effect of the room on
bass measurement (although, no doubt it has been done before). The
main feature is to place the (pressure) microphone within the speaker
and to compensate by 12 dB/oct.

Normally I use a MLS and filter out higher frequencies, say above
100Hz, with a brick wall FIR filter using Coooledit. The 12 dB/oct
compensation is also done in a similar way, either to the input MLS
or to the output recording. This sort of pre-filtering can give a
large increase to the signal to noise of the meaasurement.

There must be some restrictions on the accuracy of this method but
the only one I can think off at the moment is that the wavelength must
be large compared to the internal dimensions of the speaker. The
meaasurement includes the effect of any port and of the enclosure
flexing.

There is probably an assumption that the air is compressed
adiabatically. The pressures involved are, of course, small if the mic
is not to be overloaded.

  #8   Report Post  
Goofball_star_dot_etal
 
Posts: n/a
Default 20hz to 20Khz , yea right!

On Sun, 23 Nov 2003 02:44:43 -0800, "The Flash" wrote:

I have had no luck in measuring speaker response at low frequency's, I can
measure from ~60Hz to 20Khz with near 100% repeatability (How accurate is
questionable)

I have questioned a number of speaker builders and a couple of companies
that 'tune' speakers the answers given on what they measure with and how the
do the tests give serious rise to the claims at low frequency.

Almost any speaker system sold today claims 20Hz to 20Kz response yet this
is so far from the truth I cannot understand how they dare claim such
figures.

I have yet to find ANY speaker system using an 8 inch driver that has the
ability to produce 40hz or lower frequencys as they all have fallen so far
down in output level as to be useless.


How do you measure the responese at 20Hz?, I have tested a few speakers and
basically I cannot get accurate (repeatable) data at much below 60Hz, even
using a borrowed shure KSM141 stereo pair in omnidirectional gives vague
results ($3000 for the mic's!) I have come to the conclusion that you are
really measuring air displacement at anything under 30Hz and a flat panel
with a transducer is the only repeatable method of measuring the output, how
one calibrates said device is open for discussion.

(Oh one firm told me that they use a laser to measure the low frequency of
their speakers, check this out for novel! They place a small piece of
reflective foil on the base driver, and shine a laser beam on it, they then
apply signal and measure via 'laser' the deflection, Also they place a
passive radiator 1 meter infront of the driver unit and use the same method
to measure its deflection.

The apply a 'correction factor' and the produce the frequency response data
(company builds very expensive car and home audio subs!)






I thought up my own method for eliminating the effect of the room on
bass measurement (although, no doubt it has been done before). The
main feature is to place the (pressure) microphone within the speaker
and to compensate by 12 dB/oct.

Normally I use a MLS and filter out higher frequencies, say above
100Hz, with a brick wall FIR filter using Coooledit. The 12 dB/oct
compensation is also done in a similar way, either to the input MLS
or to the output recording. This sort of pre-filtering can give a
large increase to the signal to noise of the meaasurement.

There must be some restrictions on the accuracy of this method but
the only one I can think off at the moment is that the wavelength must
be large compared to the internal dimensions of the speaker. The
meaasurement includes the effect of any port and of the enclosure
flexing.

There is probably an assumption that the air is compressed
adiabatically. The pressures involved are, of course, small if the mic
is not to be overloaded.

  #9   Report Post  
gregs
 
Posts: n/a
Default 20hz to 20Khz , yea right!

In article , "The Flash" wrote:
I have had no luck in measuring speaker response at low frequency's, I can
measure from ~60Hz to 20Khz with near 100% repeatability (How accurate is
questionable)

I have questioned a number of speaker builders and a couple of companies
that 'tune' speakers the answers given on what they measure with and how the
do the tests give serious rise to the claims at low frequency.

Almost any speaker system sold today claims 20Hz to 20Kz response yet this
is so far from the truth I cannot understand how they dare claim such
figures.

I have yet to find ANY speaker system using an 8 inch driver that has the
ability to produce 40hz or lower frequencys as they all have fallen so far
down in output level as to be useless.


Well its does produce it, no?


How do you measure the responese at 20Hz?, I have tested a few speakers and
basically I cannot get accurate (repeatable) data at much below 60Hz, even
using a borrowed shure KSM141 stereo pair in omnidirectional gives vague
results ($3000 for the mic's!) I have come to the conclusion that you are
really measuring air displacement at anything under 30Hz and a flat panel
with a transducer is the only repeatable method of measuring the output, how
one calibrates said device is open for discussion.

(Oh one firm told me that they use a laser to measure the low frequency of
their speakers, check this out for novel! They place a small piece of
reflective foil on the base driver, and shine a laser beam on it, they then
apply signal and measure via 'laser' the deflection, Also they place a
passive radiator 1 meter infront of the driver unit and use the same method
to measure its deflection.

The apply a 'correction factor' and the produce the frequency response data
(company builds very expensive car and home audio subs!)


I guess you could compute response measuring the deflection of a sealed box driver.
The passive technique seems like it would work, but it doesn't seem better than
simply using a microphone in front of the box on a flat plate.
If you measure the deflection of a passive diaphram its deflection properties must be fully
understood to intrepret anything.

greg
  #10   Report Post  
gregs
 
Posts: n/a
Default 20hz to 20Khz , yea right!

In article , "The Flash" wrote:
I have had no luck in measuring speaker response at low frequency's, I can
measure from ~60Hz to 20Khz with near 100% repeatability (How accurate is
questionable)

I have questioned a number of speaker builders and a couple of companies
that 'tune' speakers the answers given on what they measure with and how the
do the tests give serious rise to the claims at low frequency.

Almost any speaker system sold today claims 20Hz to 20Kz response yet this
is so far from the truth I cannot understand how they dare claim such
figures.

I have yet to find ANY speaker system using an 8 inch driver that has the
ability to produce 40hz or lower frequencys as they all have fallen so far
down in output level as to be useless.


Well its does produce it, no?


How do you measure the responese at 20Hz?, I have tested a few speakers and
basically I cannot get accurate (repeatable) data at much below 60Hz, even
using a borrowed shure KSM141 stereo pair in omnidirectional gives vague
results ($3000 for the mic's!) I have come to the conclusion that you are
really measuring air displacement at anything under 30Hz and a flat panel
with a transducer is the only repeatable method of measuring the output, how
one calibrates said device is open for discussion.

(Oh one firm told me that they use a laser to measure the low frequency of
their speakers, check this out for novel! They place a small piece of
reflective foil on the base driver, and shine a laser beam on it, they then
apply signal and measure via 'laser' the deflection, Also they place a
passive radiator 1 meter infront of the driver unit and use the same method
to measure its deflection.

The apply a 'correction factor' and the produce the frequency response data
(company builds very expensive car and home audio subs!)


I guess you could compute response measuring the deflection of a sealed box driver.
The passive technique seems like it would work, but it doesn't seem better than
simply using a microphone in front of the box on a flat plate.
If you measure the deflection of a passive diaphram its deflection properties must be fully
understood to intrepret anything.

greg


  #11   Report Post  
Dick Pierce
 
Posts: n/a
Default 20hz to 20Khz , yea right!

"The Flash" wrote in message ...
I have had no luck in measuring speaker response at low frequency's, I can
measure from ~60Hz to 20Khz with near 100% repeatability (How accurate is
questionable)

I have questioned a number of speaker builders and a couple of companies
that 'tune' speakers the answers given on what they measure with and how the
do the tests give serious rise to the claims at low frequency.

Almost any speaker system sold today claims 20Hz to 20Kz response yet this
is so far from the truth I cannot understand how they dare claim such
figures.


Almost ANY speaker system? You haven't looked at many speaker systems
if that's your claim.

Further, a specification such as "20 Hz to 20 kHz" is, in an of itself,
pretty damned meaningless. You have left out a crucial portion of the
specification, the tolerance of the amplitude response within those
limits.

I have yet to find ANY speaker system using an 8 inch driver that has the
ability to produce 40hz or lower frequencys as they all have fallen so far
down in output level as to be useless.


Determined how?

How do you measure the responese at 20Hz?,


You measure it with microphones and ancillary equipment that is up
to the task. For myself, I have a number of Bruel & Kjaer, ACO,
GR and pther microphones that have verified flat response to well
below 20 Hz. Some of the B&K 1/2" capsules, for example, are within
+-1/2 dB from approximately 3 Hz to 30 kHz and above.

The remainder of the measurement and analysis chain has similar
properties: the primary measurement chain is DC coupled, for example.

Secondly, the size of the venue required for accurate measurements
is inversely proportional to the frequency you need to measure. Even
using techniques such as gated or windowed measurement, the distance
to the first reflection surface is a prime determinant of how low
you can measure. You want to measure 20 Hz accurately? Then you need
to find a room where the distance between the speaker/microphone and
the NEAREST surface is a minimum of 25 feet.

I have tested a few speakers and
basically I cannot get accurate (repeatable) data at much below 60Hz, even
using a borrowed shure KSM141 stereo pair in omnidirectional gives vague
results ($3000 for the mic's!)


$3000 for microphones that were NEVER designed to be used as measurement
microphones, ESPECIALLY at low frequencies. These are recording
microphones, NOT measurement microphones.

I have come to the conclusion that you are
really measuring air displacement at anything under 30Hz and a flat panel
with a transducer is the only repeatable method of measuring the output,


That may be the conclusion you came to, but that conclusion just happens
to be quite wrong. The physical stimulus that the ear responds to as
sound are periodic pressure variations of a sufficient amplitude and
within certain frequency limits. That's it. As long as a device can
detect these pressure variations, it can be used to measure sound.

The problem with you big flat panel method is that it assumes, quite
incorrectly, that the imnpinging waves are planar: unless you are VERY
far away from the speaker, such the wavefronts are no longer spherical,
it isn't going to work. The smaller the diaphgragm of the microphone,
the less it is affected by such a problem. That's one reason why
measurement microphones have very small diaphragms: they are essentially
point transducers over a wide range of frequencies.

how one calibrates said device is open for discussion.


One calibrates it by throwing it in the nearest landfill and going
out an learning the proper ways of measuring acoustic phenomenon.

(Oh one firm told me that they use a laser to measure the low frequency of
their speakers, check this out for novel! They place a small piece of
reflective foil on the base driver, and shine a laser beam on it, they then
apply signal and measure via 'laser' the deflection,


Well, gee golly, since it can be shown on physical first principles
that the requirement for a constant sound pressure level (that would
mean flat frequency response) from a piston radiator is simply a
displacement which goes as the reciprocal of the square of frequency,
then if you know the displacement, which you can measure with a pretty
high degree of accuracy, then you can, over the piston range of the
driver, DIRECTLY and UNAMBIGUOUSLY determine the total acoustic power
as:

Pa = p/(2 pi c) * (Sd w^2 X)^2

where

p = density of air, typ. 1.18 kg/m^3
c = velocity of sound, typ 343 m/s
Sd = emissive area of the diaphragm in m^2
w = radian frequency
X = displacement of the diaphragm, in m.

the method is HARDLY novel at all, as it is well understood and utilized
in the field. If provides, for example, a means of measuring acoustical
power output without the confounding innaccuracies of microphones, rooms
and such, though the microphone innaccuracies are not a problem if you
use proper microphones to begin with.

More to your notion that it is "novel," you might want to modify that
opinion when you discover the technique is described in nearly every
text on acoustics.

Also they place a
passive radiator 1 meter infront of the driver unit and use the same method
to measure its deflection.


Well, not quite, I believe you are completely misinderstanding what they
told you. I would suggest you look up "reciprocity methods."

The apply a 'correction factor' and the produce the frequency response data
(company builds very expensive car and home audio subs!)


Hardly any correction factor needed: simply understand what's going on
and that you are simply relying on the physical first principles of
sound.

What's wrong with that?

Compare that to using recording microphones whose measurement capabilities
are entirely unknown, in a room of unknown characteristics, using unknown
poorly calibrated and undoubtedly poorly controlled techniques by someone
who has little or know experience in measurement and acoustics...

I'd not bet good money on getting ANY reliable data out of the latter.
  #12   Report Post  
Dick Pierce
 
Posts: n/a
Default 20hz to 20Khz , yea right!

"The Flash" wrote in message ...
I have had no luck in measuring speaker response at low frequency's, I can
measure from ~60Hz to 20Khz with near 100% repeatability (How accurate is
questionable)

I have questioned a number of speaker builders and a couple of companies
that 'tune' speakers the answers given on what they measure with and how the
do the tests give serious rise to the claims at low frequency.

Almost any speaker system sold today claims 20Hz to 20Kz response yet this
is so far from the truth I cannot understand how they dare claim such
figures.


Almost ANY speaker system? You haven't looked at many speaker systems
if that's your claim.

Further, a specification such as "20 Hz to 20 kHz" is, in an of itself,
pretty damned meaningless. You have left out a crucial portion of the
specification, the tolerance of the amplitude response within those
limits.

I have yet to find ANY speaker system using an 8 inch driver that has the
ability to produce 40hz or lower frequencys as they all have fallen so far
down in output level as to be useless.


Determined how?

How do you measure the responese at 20Hz?,


You measure it with microphones and ancillary equipment that is up
to the task. For myself, I have a number of Bruel & Kjaer, ACO,
GR and pther microphones that have verified flat response to well
below 20 Hz. Some of the B&K 1/2" capsules, for example, are within
+-1/2 dB from approximately 3 Hz to 30 kHz and above.

The remainder of the measurement and analysis chain has similar
properties: the primary measurement chain is DC coupled, for example.

Secondly, the size of the venue required for accurate measurements
is inversely proportional to the frequency you need to measure. Even
using techniques such as gated or windowed measurement, the distance
to the first reflection surface is a prime determinant of how low
you can measure. You want to measure 20 Hz accurately? Then you need
to find a room where the distance between the speaker/microphone and
the NEAREST surface is a minimum of 25 feet.

I have tested a few speakers and
basically I cannot get accurate (repeatable) data at much below 60Hz, even
using a borrowed shure KSM141 stereo pair in omnidirectional gives vague
results ($3000 for the mic's!)


$3000 for microphones that were NEVER designed to be used as measurement
microphones, ESPECIALLY at low frequencies. These are recording
microphones, NOT measurement microphones.

I have come to the conclusion that you are
really measuring air displacement at anything under 30Hz and a flat panel
with a transducer is the only repeatable method of measuring the output,


That may be the conclusion you came to, but that conclusion just happens
to be quite wrong. The physical stimulus that the ear responds to as
sound are periodic pressure variations of a sufficient amplitude and
within certain frequency limits. That's it. As long as a device can
detect these pressure variations, it can be used to measure sound.

The problem with you big flat panel method is that it assumes, quite
incorrectly, that the imnpinging waves are planar: unless you are VERY
far away from the speaker, such the wavefronts are no longer spherical,
it isn't going to work. The smaller the diaphgragm of the microphone,
the less it is affected by such a problem. That's one reason why
measurement microphones have very small diaphragms: they are essentially
point transducers over a wide range of frequencies.

how one calibrates said device is open for discussion.


One calibrates it by throwing it in the nearest landfill and going
out an learning the proper ways of measuring acoustic phenomenon.

(Oh one firm told me that they use a laser to measure the low frequency of
their speakers, check this out for novel! They place a small piece of
reflective foil on the base driver, and shine a laser beam on it, they then
apply signal and measure via 'laser' the deflection,


Well, gee golly, since it can be shown on physical first principles
that the requirement for a constant sound pressure level (that would
mean flat frequency response) from a piston radiator is simply a
displacement which goes as the reciprocal of the square of frequency,
then if you know the displacement, which you can measure with a pretty
high degree of accuracy, then you can, over the piston range of the
driver, DIRECTLY and UNAMBIGUOUSLY determine the total acoustic power
as:

Pa = p/(2 pi c) * (Sd w^2 X)^2

where

p = density of air, typ. 1.18 kg/m^3
c = velocity of sound, typ 343 m/s
Sd = emissive area of the diaphragm in m^2
w = radian frequency
X = displacement of the diaphragm, in m.

the method is HARDLY novel at all, as it is well understood and utilized
in the field. If provides, for example, a means of measuring acoustical
power output without the confounding innaccuracies of microphones, rooms
and such, though the microphone innaccuracies are not a problem if you
use proper microphones to begin with.

More to your notion that it is "novel," you might want to modify that
opinion when you discover the technique is described in nearly every
text on acoustics.

Also they place a
passive radiator 1 meter infront of the driver unit and use the same method
to measure its deflection.


Well, not quite, I believe you are completely misinderstanding what they
told you. I would suggest you look up "reciprocity methods."

The apply a 'correction factor' and the produce the frequency response data
(company builds very expensive car and home audio subs!)


Hardly any correction factor needed: simply understand what's going on
and that you are simply relying on the physical first principles of
sound.

What's wrong with that?

Compare that to using recording microphones whose measurement capabilities
are entirely unknown, in a room of unknown characteristics, using unknown
poorly calibrated and undoubtedly poorly controlled techniques by someone
who has little or know experience in measurement and acoustics...

I'd not bet good money on getting ANY reliable data out of the latter.
  #13   Report Post  
malcolm
 
Posts: n/a
Default 20hz to 20Khz , yea right!


"The Flash" wrote in message
...

"Mark" wrote in message
th.net...
In article ,
says...
I have had no luck in measuring speaker response at low frequency's,


Then you may not be using the right tools.


What tools do you use?

First off, you should Never use a stereo mic (or any stereo
set up) for measurement purposes. You are guaranteed to get
inaccurate results. Secondly, the mic is only part of the
equipment. What else are you using?


Never? By using a stereo setup I can collect two data streams rather than
one, I can compare the streams and quickly analyse microphone placement

and
any other abnormalities.

I would dearly love to sample 8 or more soundpoints at realtime but this

is
beyond my gear.

I have access to / a collection of equipment, some built by HP other

stuff
designed and manufactured for the BBC or others. All the HP gear uses HPIB
which is not what is the most convenient, The A2D - D2A card in the PC is

a
much quicker and generates the same results (But much quicker) I have a
number of amplifers, but typically use a Class A 2n3055 based 60 watt mono
amp, It has acceptable characteristics for the type of measurement that I
usually do, 20Hz to 20Khz on / off axis at 1w / 1m / 1khz. Microphones is
the biggest issue as I have not yet found anything that I am really happy
with. I can use the Shures, I have some clone shure that are very flat

from
60hz to 20Khz and cost like $60 each! (vs 3 grand for the shures) (I can
compensate in software for the low end droop or use a opamp based

corrector
that was build before I started with the PC based A2D card. I have a nice
transducer (Probably from a bridge or Crane) that works from a few Hz.

I don't compensate for impedance/frequency change as this would results
that would never be seen in the real world, so a crappy crossover can make

a
good driver look bad.

and a flat panel
with a transducer is the only repeatable method of measuring the

output,

Balderdash.


Now thats a technical response!, Ok what I have found is that it seem

almost
all manuafctures fiddle the figures to make them look good, here is a
classic.

http://www.aperionaudio.com/products...onse_graph.jpg

From the specs page they claim a 60Hz to 20Khz response! Yet their graph
shows the speaker is more than 10dB down in output at 60Hz from its output
at ~200Hz and again at the high end its ouput looks to be ~8dB down at

20Khz
from 15Khz

Ok, Can some one tell me what mics / transducers I need to be using to get
accurate low frequency measurements. Not just 'use your ears' as this is

far
too subjective!




try these people, they sell interesting FAST dataloggers,
and if you want multichannel, use a reed relay or solid state switch box to
multiplex your 'Mics' ,
as long as the source is the same for each run you then have multichannel.
regards malcolm



  #14   Report Post  
malcolm
 
Posts: n/a
Default 20hz to 20Khz , yea right!


"The Flash" wrote in message
...

"Mark" wrote in message
th.net...
In article ,
says...
I have had no luck in measuring speaker response at low frequency's,


Then you may not be using the right tools.


What tools do you use?

First off, you should Never use a stereo mic (or any stereo
set up) for measurement purposes. You are guaranteed to get
inaccurate results. Secondly, the mic is only part of the
equipment. What else are you using?


Never? By using a stereo setup I can collect two data streams rather than
one, I can compare the streams and quickly analyse microphone placement

and
any other abnormalities.

I would dearly love to sample 8 or more soundpoints at realtime but this

is
beyond my gear.

I have access to / a collection of equipment, some built by HP other

stuff
designed and manufactured for the BBC or others. All the HP gear uses HPIB
which is not what is the most convenient, The A2D - D2A card in the PC is

a
much quicker and generates the same results (But much quicker) I have a
number of amplifers, but typically use a Class A 2n3055 based 60 watt mono
amp, It has acceptable characteristics for the type of measurement that I
usually do, 20Hz to 20Khz on / off axis at 1w / 1m / 1khz. Microphones is
the biggest issue as I have not yet found anything that I am really happy
with. I can use the Shures, I have some clone shure that are very flat

from
60hz to 20Khz and cost like $60 each! (vs 3 grand for the shures) (I can
compensate in software for the low end droop or use a opamp based

corrector
that was build before I started with the PC based A2D card. I have a nice
transducer (Probably from a bridge or Crane) that works from a few Hz.

I don't compensate for impedance/frequency change as this would results
that would never be seen in the real world, so a crappy crossover can make

a
good driver look bad.

and a flat panel
with a transducer is the only repeatable method of measuring the

output,

Balderdash.


Now thats a technical response!, Ok what I have found is that it seem

almost
all manuafctures fiddle the figures to make them look good, here is a
classic.

http://www.aperionaudio.com/products...onse_graph.jpg

From the specs page they claim a 60Hz to 20Khz response! Yet their graph
shows the speaker is more than 10dB down in output at 60Hz from its output
at ~200Hz and again at the high end its ouput looks to be ~8dB down at

20Khz
from 15Khz

Ok, Can some one tell me what mics / transducers I need to be using to get
accurate low frequency measurements. Not just 'use your ears' as this is

far
too subjective!




try these people, they sell interesting FAST dataloggers,
and if you want multichannel, use a reed relay or solid state switch box to
multiplex your 'Mics' ,
as long as the source is the same for each run you then have multichannel.
regards malcolm



  #15   Report Post  
malcolm
 
Posts: n/a
Default 20hz to 20Khz , yea right!

try these people, they sell interesting FAST dataloggers,
and if you want multichannel, use a reed relay or solid state switch box

to
multiplex your 'Mics' ,
as long as the source is the same for each run you then have multichannel.
regards malcolm




oops http://www.picotech.com/data.html




  #16   Report Post  
malcolm
 
Posts: n/a
Default 20hz to 20Khz , yea right!

try these people, they sell interesting FAST dataloggers,
and if you want multichannel, use a reed relay or solid state switch box

to
multiplex your 'Mics' ,
as long as the source is the same for each run you then have multichannel.
regards malcolm




oops http://www.picotech.com/data.html


  #17   Report Post  
Tony Pearce
 
Posts: n/a
Default 20hz to 20Khz , yea right!


"The Flash" wrote in message
...
Uh, I used to test in a anechoic chamber at the local university but don't
now,


Most anechoic chambers are useless below 100 Hz as well.

I also tested other areas and found the following. You can test fine in
you living room. Background noiselevel is typically 5OdB at home. You

place
a mat of acoustic foam on the floor, put yout mics up (1m away) place 4

more
sheets of acoustic foam round the speaker and mic's and final sheet on top

These knock 10dB off background noise


Not at 20 Hz they don't.

and stop echo, run up 1 watt at 1khz
gives about 90dB in most case, this is 50dB above background


???.
50dB BG from 90 dB SPL gives 40 dB.

and you can run
a sweep test and get results the same as in an anechoic chamber.
Again I think this is all smoke and mirrors stuff so that its 'too hard'

to
do yourself.


Sure you can do that, and your measurement uncertainty is ?????

TonyP.




  #18   Report Post  
Tony Pearce
 
Posts: n/a
Default 20hz to 20Khz , yea right!


"The Flash" wrote in message
...
Uh, I used to test in a anechoic chamber at the local university but don't
now,


Most anechoic chambers are useless below 100 Hz as well.

I also tested other areas and found the following. You can test fine in
you living room. Background noiselevel is typically 5OdB at home. You

place
a mat of acoustic foam on the floor, put yout mics up (1m away) place 4

more
sheets of acoustic foam round the speaker and mic's and final sheet on top

These knock 10dB off background noise


Not at 20 Hz they don't.

and stop echo, run up 1 watt at 1khz
gives about 90dB in most case, this is 50dB above background


???.
50dB BG from 90 dB SPL gives 40 dB.

and you can run
a sweep test and get results the same as in an anechoic chamber.
Again I think this is all smoke and mirrors stuff so that its 'too hard'

to
do yourself.


Sure you can do that, and your measurement uncertainty is ?????

TonyP.




  #19   Report Post  
Goofball_star_dot_etal
 
Posts: n/a
Default 20hz to 20Khz , yea right!

On Sun, 23 Nov 2003 12:29:42 -0800, "The Flash" wrote:

What type of / brand of mic are you using?



I was not trying to make a very accurate measurement so I just used a
cheap Panasonic WM-60AT price £2
http://rocky.digikey.com/WebLib/Pana...ata/WM-60A.pdf
http://www.digikey.com/digihome.html
  #20   Report Post  
Goofball_star_dot_etal
 
Posts: n/a
Default 20hz to 20Khz , yea right!

On Sun, 23 Nov 2003 12:29:42 -0800, "The Flash" wrote:

What type of / brand of mic are you using?



I was not trying to make a very accurate measurement so I just used a
cheap Panasonic WM-60AT price £2
http://rocky.digikey.com/WebLib/Pana...ata/WM-60A.pdf
http://www.digikey.com/digihome.html


  #21   Report Post  
The Flash
 
Posts: n/a
Default 20hz to 20Khz , yea right!

I have had no luck in measuring speaker response at low frequency's, I can
measure from ~60Hz to 20Khz with near 100% repeatability (How accurate is
questionable)

I have questioned a number of speaker builders and a couple of companies
that 'tune' speakers the answers given on what they measure with and how the
do the tests give serious rise to the claims at low frequency.

Almost any speaker system sold today claims 20Hz to 20Kz response yet this
is so far from the truth I cannot understand how they dare claim such
figures.

I have yet to find ANY speaker system using an 8 inch driver that has the
ability to produce 40hz or lower frequencys as they all have fallen so far
down in output level as to be useless.


How do you measure the responese at 20Hz?, I have tested a few speakers and
basically I cannot get accurate (repeatable) data at much below 60Hz, even
using a borrowed shure KSM141 stereo pair in omnidirectional gives vague
results ($3000 for the mic's!) I have come to the conclusion that you are
really measuring air displacement at anything under 30Hz and a flat panel
with a transducer is the only repeatable method of measuring the output, how
one calibrates said device is open for discussion.

(Oh one firm told me that they use a laser to measure the low frequency of
their speakers, check this out for novel! They place a small piece of
reflective foil on the base driver, and shine a laser beam on it, they then
apply signal and measure via 'laser' the deflection, Also they place a
passive radiator 1 meter infront of the driver unit and use the same method
to measure its deflection.

The apply a 'correction factor' and the produce the frequency response data
(company builds very expensive car and home audio subs!)





  #22   Report Post  
The Flash
 
Posts: n/a
Default 20hz to 20Khz , yea right!


"Mark" wrote in message
th.net...
In article ,
says...
I have had no luck in measuring speaker response at low frequency's,


Then you may not be using the right tools.


What tools do you use?

First off, you should Never use a stereo mic (or any stereo
set up) for measurement purposes. You are guaranteed to get
inaccurate results. Secondly, the mic is only part of the
equipment. What else are you using?


Never? By using a stereo setup I can collect two data streams rather than
one, I can compare the streams and quickly analyse microphone placement and
any other abnormalities.

I would dearly love to sample 8 or more soundpoints at realtime but this is
beyond my gear.

I have access to / a collection of equipment, some built by HP other stuff
designed and manufactured for the BBC or others. All the HP gear uses HPIB
which is not what is the most convenient, The A2D - D2A card in the PC is a
much quicker and generates the same results (But much quicker) I have a
number of amplifers, but typically use a Class A 2n3055 based 60 watt mono
amp, It has acceptable characteristics for the type of measurement that I
usually do, 20Hz to 20Khz on / off axis at 1w / 1m / 1khz. Microphones is
the biggest issue as I have not yet found anything that I am really happy
with. I can use the Shures, I have some clone shure that are very flat from
60hz to 20Khz and cost like $60 each! (vs 3 grand for the shures) (I can
compensate in software for the low end droop or use a opamp based corrector
that was build before I started with the PC based A2D card. I have a nice
transducer (Probably from a bridge or Crane) that works from a few Hz.

I don't compensate for impedance/frequency change as this would results
that would never be seen in the real world, so a crappy crossover can make a
good driver look bad.

and a flat panel
with a transducer is the only repeatable method of measuring the output,


Balderdash.


Now thats a technical response!, Ok what I have found is that it seem almost
all manuafctures fiddle the figures to make them look good, here is a
classic.

http://www.aperionaudio.com/products...onse_graph.jpg

From the specs page they claim a 60Hz to 20Khz response! Yet their graph
shows the speaker is more than 10dB down in output at 60Hz from its output
at ~200Hz and again at the high end its ouput looks to be ~8dB down at 20Khz
from 15Khz

Ok, Can some one tell me what mics / transducers I need to be using to get
accurate low frequency measurements. Not just 'use your ears' as this is far
too subjective!



  #23   Report Post  
The Flash
 
Posts: n/a
Default 20hz to 20Khz , yea right!


"Mark" wrote in message
th.net...
In article ,
says...
I have had no luck in measuring speaker response at low frequency's,


Then you may not be using the right tools.


What tools do you use?

First off, you should Never use a stereo mic (or any stereo
set up) for measurement purposes. You are guaranteed to get
inaccurate results. Secondly, the mic is only part of the
equipment. What else are you using?


Never? By using a stereo setup I can collect two data streams rather than
one, I can compare the streams and quickly analyse microphone placement and
any other abnormalities.

I would dearly love to sample 8 or more soundpoints at realtime but this is
beyond my gear.

I have access to / a collection of equipment, some built by HP other stuff
designed and manufactured for the BBC or others. All the HP gear uses HPIB
which is not what is the most convenient, The A2D - D2A card in the PC is a
much quicker and generates the same results (But much quicker) I have a
number of amplifers, but typically use a Class A 2n3055 based 60 watt mono
amp, It has acceptable characteristics for the type of measurement that I
usually do, 20Hz to 20Khz on / off axis at 1w / 1m / 1khz. Microphones is
the biggest issue as I have not yet found anything that I am really happy
with. I can use the Shures, I have some clone shure that are very flat from
60hz to 20Khz and cost like $60 each! (vs 3 grand for the shures) (I can
compensate in software for the low end droop or use a opamp based corrector
that was build before I started with the PC based A2D card. I have a nice
transducer (Probably from a bridge or Crane) that works from a few Hz.

I don't compensate for impedance/frequency change as this would results
that would never be seen in the real world, so a crappy crossover can make a
good driver look bad.

and a flat panel
with a transducer is the only repeatable method of measuring the output,


Balderdash.


Now thats a technical response!, Ok what I have found is that it seem almost
all manuafctures fiddle the figures to make them look good, here is a
classic.

http://www.aperionaudio.com/products...onse_graph.jpg

From the specs page they claim a 60Hz to 20Khz response! Yet their graph
shows the speaker is more than 10dB down in output at 60Hz from its output
at ~200Hz and again at the high end its ouput looks to be ~8dB down at 20Khz
from 15Khz

Ok, Can some one tell me what mics / transducers I need to be using to get
accurate low frequency measurements. Not just 'use your ears' as this is far
too subjective!



  #24   Report Post  
The Flash
 
Posts: n/a
Default 20hz to 20Khz , yea right!

Uh, I used to test in a anechoic chamber at the local university but don't
now, I also tested other areas and found the following. You can test fine in
you living room. Background noiselevel is typically 5OdB at home. You place
a mat of acoustic foam on the floor, put yout mics up (1m away) place 4 more
sheets of acoustic foam round the speaker and mic's and final sheet on top

These knock 10dB off background noise and stop echo, run up 1 watt at 1khz
gives about 90dB in most case, this is 50dB above background and you can run
a sweep test and get results the same as in an anechoic chamber. (were not
interested in its frequency response or efficency at 40 dB any way!)

Again I think this is all smoke and mirrors stuff so that its 'too hard' to
do yourself.

None of us live inside a anechoic chamber and speakers sound awful when
played inside one (no natural room reverberation)

Alot of this smacks of the car makes 'economy mpg' by driving at 56Mph round
an oval test circut. Something that we all do....


"Ethan Winer" ethanw at ethanwiner dot com wrote in message
...
Flash,

Besides all that Mark said, you can't measure loudspeakers by playing

tones
in a typical room in a typical house. If you take it outside, far away

from
ambient noises, you'll have half a shot at it.

--Ethan




  #25   Report Post  
The Flash
 
Posts: n/a
Default 20hz to 20Khz , yea right!

Uh, I used to test in a anechoic chamber at the local university but don't
now, I also tested other areas and found the following. You can test fine in
you living room. Background noiselevel is typically 5OdB at home. You place
a mat of acoustic foam on the floor, put yout mics up (1m away) place 4 more
sheets of acoustic foam round the speaker and mic's and final sheet on top

These knock 10dB off background noise and stop echo, run up 1 watt at 1khz
gives about 90dB in most case, this is 50dB above background and you can run
a sweep test and get results the same as in an anechoic chamber. (were not
interested in its frequency response or efficency at 40 dB any way!)

Again I think this is all smoke and mirrors stuff so that its 'too hard' to
do yourself.

None of us live inside a anechoic chamber and speakers sound awful when
played inside one (no natural room reverberation)

Alot of this smacks of the car makes 'economy mpg' by driving at 56Mph round
an oval test circut. Something that we all do....


"Ethan Winer" ethanw at ethanwiner dot com wrote in message
...
Flash,

Besides all that Mark said, you can't measure loudspeakers by playing

tones
in a typical room in a typical house. If you take it outside, far away

from
ambient noises, you'll have half a shot at it.

--Ethan






  #26   Report Post  
Dick Pierce
 
Posts: n/a
Default 20hz to 20Khz , yea right!

"The Flash" wrote in message ...
I have had no luck in measuring speaker response at low frequency's, I

can
measure from ~60Hz to 20Khz with near 100% repeatability (How accurate

is
questionable)



You measure it with microphones and ancillary equipment that is up
to the task. For myself, I have a number of Bruel & Kjaer, ACO,
GR and pther microphones that have verified flat response to well
below 20 Hz. Some of the B&K 1/2" capsules, for example, are within
+-1/2 dB from approximately 3 Hz to 30 kHz and above.


Can you give a recommend model of Bruel that you like?


Bruel & Kjaer makes a wide variety of laboratory microphones that
are intended for different purposes. They range in size from 1"
diameter to 1/8", there are microphones intended for pressure,
random incidence and such, carrier models for extremely low
frequency and so on. I might usggest a visit to the B&K, ACO
pacific and other websites for more information (q.v., bkhome.com).

As to what Bruel & Kjaer microphone I "like," well, no insult
intended, but it isn't a matter of what I "like," it's a matter
of what is best suited for the task at hand based on technical
considerations. I have a number of 4133 and 4134 that I use for common
frequency response measurements.

Please note that the price of these microphone is NOT cheap. You will
find that by the time you are done with the impdeance converter/
preamplifier, powersupply and capsule, you're already up in the
$2000 range.

I have come to the conclusion that you are
really measuring air displacement at anything under 30Hz and a flat

panel
with a transducer is the only repeatable method of measuring the output,


One calibrates it by throwing it in the nearest landfill and going
out an learning the proper ways of measuring acoustic phenomenon.


You comments are most welcome but not vey helpful, I seem to be getting the
feeling you are giving me the 'I've be doing it for decade and got really
expensive gear, so my methods, results and opinions are gospel - do it my
way or be burned at the stake as a heratic'


Hold on there, bucko, YOU'RE the one who came riding into town
yelling at everyone about how all this 20-20 kHz is bogus and
declared the whole world of speaker measurement bogus and then
declared a VERY well understood measurement paradigm as "novel."

One starts to wonder if perhaps you are one of the 'Experts' that 'corrects'
the measurments to what you 'know' is correct. Your attitude in your
responses seems to show this blinded mentality.


Sir, you make assumptions out of tyhin air. Never once did I
say ANYTHING about correcting measurements, I talked about
making measurements correctly. I suggest it's time you get
down off your mighty steed and start learning instead of
holding forth on topics about which you state quite clearly
you know little about.

(Oh one firm told me that they use a laser to measure the low frequency

of
their speakers, check this out for novel! They place a small piece of
reflective foil on the base driver, and shine a laser beam on it, they

then
apply signal and measure via 'laser' the deflection,


Well, gee golly, since it can be shown on physical first principles
that the requirement for a constant sound pressure level (that would
mean flat frequency response) from a piston radiator is simply a
displacement which goes as the reciprocal of the square of frequency,
then if you know the displacement, which you can measure with a pretty
high degree of accuracy, then you can, over the piston range of the
driver, DIRECTLY and UNAMBIGUOUSLY determine the total acoustic power
as:

Pa = p/(2 pi c) * (Sd w^2 X)^2

where

p = density of air, typ. 1.18 kg/m^3
c = velocity of sound, typ 343 m/s
Sd = emissive area of the diaphragm in m^2
w = radian frequency
X = displacement of the diaphragm, in m.

the method is HARDLY novel at all, as it is well understood and utilized
in the field. If provides, for example, a means of measuring acoustical
power output without the confounding innaccuracies of microphones, rooms
and such, though the microphone innaccuracies are not a problem if you
use proper microphones to begin with.


I do undestand the theory, and I expect it would work well if the speaker
was moving in piston motion. (An some of the car sub woofers with ridged
aluminium cones would no doubt at low frequency)


As ALL speaker cones do in the piston region of operation, below
a frequency which is roughly determined when the wavelength reaches
the circumference of the driver. For an 8" driver, that frequency
is roughly 600 Hz, for example.

However typical stereo speakers using either doped paper or pp cones won't,
as you know the LF breakup happens quite early on these type of drivers due
to the tradeoff of trying to give a wide response range with only 2 or 3
drives in most cases and stiff cone suspension.


My humble suggestion, sir, is that you're experience in this area
is not congruent with the rather large amount of data that is known
and has been known for quite some time in this realm. The operation
of drivers operating in the piston region has been very well studied,
not by car audio or home audio speaker makers, most of whom,
ESPECIALLY those in the car audio business, are clueless and work
more on legenbd, rumor and unfounded assumption, but by the likes of
MacClachlan, Beranek, Thiele, SMall, and many others. Some of these
researchers did their work 60 years ago, and developed information
that much of the speaker industry has yet to clue themselves in on.

SHould you want a comprehensive set of references, I can start
supplying such should you desire.

More to your notion that it is "novel," you might want to modify that
opinion when you discover the technique is described in nearly every
text on acoustics.


-Noted and yes I have seen it mention and was taught such things, but it was
always theory and I had not heard of it done on audio speakers (however I do
know it is done on building structures to measure characteristic)


Well, again, you seem to be unfamiliar with the vast amount of
literature on the topic. Indeed, far back in the 1970's, there is
an extensive amount of literature in the technical press describing
the characterisation of loudspeakers via such techniques. Anyone
familiar with the theory is likely to be aware of these articles.

I do know that laser scanning of drive to measure the breakup is done, but
this was not what they were doing (So they said)

Compare that to using recording microphones whose measurement capabilities
are entirely unknown, in a room of unknown characteristics, using unknown
poorly calibrated and undoubtedly poorly controlled techniques by someone
who has little or know experience in measurement and acoustics...

I'd not bet good money on getting ANY reliable data out of the latter.


Well I have to say getting reliable data from 'experts' is not very
successful!


Not when you come charging in challenging a tpoic with which you
are unfamiliar. You made some smeasurements and you ASSUMED without
ANY supporting evidence that your methods have any validity.

The apparently very narrow topic of measurement of acoustical
phenomenon is, in fact, a very large, complex and quite fascinating
one. You've not even touched on 1% of it. It's not something you're
going to learn about via an Usenet article or two. You've not even
touched on the issues of boundary conditions and nearfield vs farfield
and diffraction effects, an entire sub-topic on suitable stimuli,
of MLS vs impulse vs gated sine vs swept sine vs narrow band noise
vs wide band noise vs averaged narrow-band analysis and so one and
so forth. None of that.

Despite that, you seem to have the attitude that you have just
discovered some vast conspiracy or fraud, indeed the very title
of your post states that. Yet, you clearly state that you know
little or nothing about the basic principles involved.

You claim to "understand the theory" that in the pistonic region
of operation, there is a direct causal link between driver excursion,
frequency and sound pressure level. If you are "familiar" with it as
you claim, you coudn't possibly find the method "novel." It's
a fundamental property that is taight in the very early stages
of acoustical principles. All due respect, you don't even use the
lexicon of the subject in a way that suggests you have any familiarity
with the topic.

You want some information? Fine, ask the questions. Don't come
charging in like some expert, which you are not. The experts here
are happy to answer your questions. Accusations such as "20hz to
20Khz, yea right!" are going to be dismissed as uninformed rantings
that they are.
  #27   Report Post  
Dick Pierce
 
Posts: n/a
Default 20hz to 20Khz , yea right!

"The Flash" wrote in message ...
I have had no luck in measuring speaker response at low frequency's, I

can
measure from ~60Hz to 20Khz with near 100% repeatability (How accurate

is
questionable)



You measure it with microphones and ancillary equipment that is up
to the task. For myself, I have a number of Bruel & Kjaer, ACO,
GR and pther microphones that have verified flat response to well
below 20 Hz. Some of the B&K 1/2" capsules, for example, are within
+-1/2 dB from approximately 3 Hz to 30 kHz and above.


Can you give a recommend model of Bruel that you like?


Bruel & Kjaer makes a wide variety of laboratory microphones that
are intended for different purposes. They range in size from 1"
diameter to 1/8", there are microphones intended for pressure,
random incidence and such, carrier models for extremely low
frequency and so on. I might usggest a visit to the B&K, ACO
pacific and other websites for more information (q.v., bkhome.com).

As to what Bruel & Kjaer microphone I "like," well, no insult
intended, but it isn't a matter of what I "like," it's a matter
of what is best suited for the task at hand based on technical
considerations. I have a number of 4133 and 4134 that I use for common
frequency response measurements.

Please note that the price of these microphone is NOT cheap. You will
find that by the time you are done with the impdeance converter/
preamplifier, powersupply and capsule, you're already up in the
$2000 range.

I have come to the conclusion that you are
really measuring air displacement at anything under 30Hz and a flat

panel
with a transducer is the only repeatable method of measuring the output,


One calibrates it by throwing it in the nearest landfill and going
out an learning the proper ways of measuring acoustic phenomenon.


You comments are most welcome but not vey helpful, I seem to be getting the
feeling you are giving me the 'I've be doing it for decade and got really
expensive gear, so my methods, results and opinions are gospel - do it my
way or be burned at the stake as a heratic'


Hold on there, bucko, YOU'RE the one who came riding into town
yelling at everyone about how all this 20-20 kHz is bogus and
declared the whole world of speaker measurement bogus and then
declared a VERY well understood measurement paradigm as "novel."

One starts to wonder if perhaps you are one of the 'Experts' that 'corrects'
the measurments to what you 'know' is correct. Your attitude in your
responses seems to show this blinded mentality.


Sir, you make assumptions out of tyhin air. Never once did I
say ANYTHING about correcting measurements, I talked about
making measurements correctly. I suggest it's time you get
down off your mighty steed and start learning instead of
holding forth on topics about which you state quite clearly
you know little about.

(Oh one firm told me that they use a laser to measure the low frequency

of
their speakers, check this out for novel! They place a small piece of
reflective foil on the base driver, and shine a laser beam on it, they

then
apply signal and measure via 'laser' the deflection,


Well, gee golly, since it can be shown on physical first principles
that the requirement for a constant sound pressure level (that would
mean flat frequency response) from a piston radiator is simply a
displacement which goes as the reciprocal of the square of frequency,
then if you know the displacement, which you can measure with a pretty
high degree of accuracy, then you can, over the piston range of the
driver, DIRECTLY and UNAMBIGUOUSLY determine the total acoustic power
as:

Pa = p/(2 pi c) * (Sd w^2 X)^2

where

p = density of air, typ. 1.18 kg/m^3
c = velocity of sound, typ 343 m/s
Sd = emissive area of the diaphragm in m^2
w = radian frequency
X = displacement of the diaphragm, in m.

the method is HARDLY novel at all, as it is well understood and utilized
in the field. If provides, for example, a means of measuring acoustical
power output without the confounding innaccuracies of microphones, rooms
and such, though the microphone innaccuracies are not a problem if you
use proper microphones to begin with.


I do undestand the theory, and I expect it would work well if the speaker
was moving in piston motion. (An some of the car sub woofers with ridged
aluminium cones would no doubt at low frequency)


As ALL speaker cones do in the piston region of operation, below
a frequency which is roughly determined when the wavelength reaches
the circumference of the driver. For an 8" driver, that frequency
is roughly 600 Hz, for example.

However typical stereo speakers using either doped paper or pp cones won't,
as you know the LF breakup happens quite early on these type of drivers due
to the tradeoff of trying to give a wide response range with only 2 or 3
drives in most cases and stiff cone suspension.


My humble suggestion, sir, is that you're experience in this area
is not congruent with the rather large amount of data that is known
and has been known for quite some time in this realm. The operation
of drivers operating in the piston region has been very well studied,
not by car audio or home audio speaker makers, most of whom,
ESPECIALLY those in the car audio business, are clueless and work
more on legenbd, rumor and unfounded assumption, but by the likes of
MacClachlan, Beranek, Thiele, SMall, and many others. Some of these
researchers did their work 60 years ago, and developed information
that much of the speaker industry has yet to clue themselves in on.

SHould you want a comprehensive set of references, I can start
supplying such should you desire.

More to your notion that it is "novel," you might want to modify that
opinion when you discover the technique is described in nearly every
text on acoustics.


-Noted and yes I have seen it mention and was taught such things, but it was
always theory and I had not heard of it done on audio speakers (however I do
know it is done on building structures to measure characteristic)


Well, again, you seem to be unfamiliar with the vast amount of
literature on the topic. Indeed, far back in the 1970's, there is
an extensive amount of literature in the technical press describing
the characterisation of loudspeakers via such techniques. Anyone
familiar with the theory is likely to be aware of these articles.

I do know that laser scanning of drive to measure the breakup is done, but
this was not what they were doing (So they said)

Compare that to using recording microphones whose measurement capabilities
are entirely unknown, in a room of unknown characteristics, using unknown
poorly calibrated and undoubtedly poorly controlled techniques by someone
who has little or know experience in measurement and acoustics...

I'd not bet good money on getting ANY reliable data out of the latter.


Well I have to say getting reliable data from 'experts' is not very
successful!


Not when you come charging in challenging a tpoic with which you
are unfamiliar. You made some smeasurements and you ASSUMED without
ANY supporting evidence that your methods have any validity.

The apparently very narrow topic of measurement of acoustical
phenomenon is, in fact, a very large, complex and quite fascinating
one. You've not even touched on 1% of it. It's not something you're
going to learn about via an Usenet article or two. You've not even
touched on the issues of boundary conditions and nearfield vs farfield
and diffraction effects, an entire sub-topic on suitable stimuli,
of MLS vs impulse vs gated sine vs swept sine vs narrow band noise
vs wide band noise vs averaged narrow-band analysis and so one and
so forth. None of that.

Despite that, you seem to have the attitude that you have just
discovered some vast conspiracy or fraud, indeed the very title
of your post states that. Yet, you clearly state that you know
little or nothing about the basic principles involved.

You claim to "understand the theory" that in the pistonic region
of operation, there is a direct causal link between driver excursion,
frequency and sound pressure level. If you are "familiar" with it as
you claim, you coudn't possibly find the method "novel." It's
a fundamental property that is taight in the very early stages
of acoustical principles. All due respect, you don't even use the
lexicon of the subject in a way that suggests you have any familiarity
with the topic.

You want some information? Fine, ask the questions. Don't come
charging in like some expert, which you are not. The experts here
are happy to answer your questions. Accusations such as "20hz to
20Khz, yea right!" are going to be dismissed as uninformed rantings
that they are.
  #28   Report Post  
Victor
 
Posts: n/a
Default 20hz to 20Khz , yea right!

"Goofball_star_dot_etal" wrote:
I was not trying to make a very accurate measurement so I just used a
cheap Panasonic WM-60AT price £2

----------------------------------------

That's almost exactly what Behringer uses in ECM8000...


  #29   Report Post  
Victor
 
Posts: n/a
Default 20hz to 20Khz , yea right!

"Goofball_star_dot_etal" wrote:
I was not trying to make a very accurate measurement so I just used a
cheap Panasonic WM-60AT price £2

----------------------------------------

That's almost exactly what Behringer uses in ECM8000...


  #30   Report Post  
Ethan Winer
 
Posts: n/a
Default 20hz to 20Khz , yea right!

Flash,

Tony hit the salient high points, and I'll add this:

I think this is all smoke and mirrors stuff so that its 'too hard' to do

yourself.

Not smoke and mirrors, acoustic interference. That's the real issue. And
that's why even 1/12th octave pink noise tests are useless to obtain a true
response within a room.

None of us live inside a anechoic chamber


Yes, but that's an entirely different issue. You said your stated goal is to
measure the *speaker's* response. Well, if you meausre it in a room you're
measuring far more of the *room's* response.

--Ethan




  #31   Report Post  
Ethan Winer
 
Posts: n/a
Default 20hz to 20Khz , yea right!

Flash,

Tony hit the salient high points, and I'll add this:

I think this is all smoke and mirrors stuff so that its 'too hard' to do

yourself.

Not smoke and mirrors, acoustic interference. That's the real issue. And
that's why even 1/12th octave pink noise tests are useless to obtain a true
response within a room.

None of us live inside a anechoic chamber


Yes, but that's an entirely different issue. You said your stated goal is to
measure the *speaker's* response. Well, if you meausre it in a room you're
measuring far more of the *room's* response.

--Ethan


  #32   Report Post  
Ian
 
Posts: n/a
Default 20hz to 20Khz , yea right!


"The Flash" wrote in message
...

I have had no luck in measuring speaker response at low frequency's, I

can
measure from ~60Hz to 20Khz with near 100% repeatability (How accurate

is
questionable)



You measure it with microphones and ancillary equipment that is up
to the task. For myself, I have a number of Bruel & Kjaer, ACO,
GR and pther microphones that have verified flat response to well
below 20 Hz. Some of the B&K 1/2" capsules, for example, are within
+-1/2 dB from approximately 3 Hz to 30 kHz and above.


Can you give a recommend model of Bruel that you like?

Summerised I have tested in a number of different situation and with a
collection of mics and such. what I have found is that it is easy to get a
mike with a flat response from ~60Hz to ~20Kz, it seems that below this is
where a specialist product is required.

The remainder of the measurement and analysis chain has similar
properties: the primary measurement chain is DC coupled, for example.

Secondly, the size of the venue required for accurate measurements
is inversely proportional to the frequency you need to measure. Even
using techniques such as gated or windowed measurement, the distance
to the first reflection surface is a prime determinant of how low
you can measure. You want to measure 20 Hz accurately? Then you need
to find a room where the distance between the speaker/microphone and
the NEAREST surface is a minimum of 25 feet.

I have tested a few speakers and
basically I cannot get accurate (repeatable) data at much below 60Hz,

even
using a borrowed shure KSM141 stereo pair in omnidirectional gives

vague
results ($3000 for the mic's!)


$3000 for microphones that were NEVER designed to be used as measurement
microphones, ESPECIALLY at low frequencies. These are recording
microphones, NOT measurement microphones.

I have come to the conclusion that you are
really measuring air displacement at anything under 30Hz and a flat

panel
with a transducer is the only repeatable method of measuring the

output,

That may be the conclusion you came to, but that conclusion just happens
to be quite wrong. The physical stimulus that the ear responds to as
sound are periodic pressure variations of a sufficient amplitude and
within certain frequency limits. That's it. As long as a device can
detect these pressure variations, it can be used to measure sound.

The problem with you big flat panel method is that it assumes, quite
incorrectly, that the imnpinging waves are planar: unless you are VERY
far away from the speaker, such the wavefronts are no longer spherical,
it isn't going to work. The smaller the diaphgragm of the microphone,
the less it is affected by such a problem. That's one reason why
measurement microphones have very small diaphragms: they are essentially
point transducers over a wide range of frequencies.

how one calibrates said device is open for discussion.


snip good stuff from Dick

I do undestand the theory, and I expect it would work well if the speaker
was moving in piston motion. (An some of the car sub woofers with ridged
aluminium cones would no doubt at low frequency)

However typical stereo speakers using either doped paper or pp cones

won't,
as you know the LF breakup happens quite early on these type of drivers

due
to the tradeoff of trying to give a wide response range with only 2 or 3
drives in most cases and stiff cone suspension.

more snip

Compare that to using recording microphones whose measurement

capabilities
are entirely unknown, in a room of unknown characteristics, using

unknown
poorly calibrated and undoubtedly poorly controlled techniques by

someone
who has little or know experience in measurement and acoustics...

I'd not bet good money on getting ANY reliable data out of the latter.


Well I have to say getting reliable data from 'experts' is not very
successful!


I think Dick gave you good information.

The B&K microphones are very expensive. Someone earlier posted a
suggestion for a Panasonic capsule. These are available calibrated.
The Behringer ECM8000 is more expensive, but still relatively cheap (£30).
Either solution is far better than you have been using.

Dick mentions the problem of reflections when doing in-room response.
At low frequencies this can be dealt with using near field techniques,
valid up to where the diaphragm reaches its piston limit. This is often
several hundred Hz, allowing you to "stitch together" responses from
different regions.

Your assertion that it is not possible to get low frequencies using an
8 inch driver turns out not to be correct. I have a pair where the LF
is done using 8 inch drivers in a bandpass enclosure. Those have a
measured response which is 1dB down at 30Hz, and an overall
response at 3dB down of 27Hz to 90Hz.

Regards
Ian


  #33   Report Post  
Ian
 
Posts: n/a
Default 20hz to 20Khz , yea right!


"The Flash" wrote in message
...

I have had no luck in measuring speaker response at low frequency's, I

can
measure from ~60Hz to 20Khz with near 100% repeatability (How accurate

is
questionable)



You measure it with microphones and ancillary equipment that is up
to the task. For myself, I have a number of Bruel & Kjaer, ACO,
GR and pther microphones that have verified flat response to well
below 20 Hz. Some of the B&K 1/2" capsules, for example, are within
+-1/2 dB from approximately 3 Hz to 30 kHz and above.


Can you give a recommend model of Bruel that you like?

Summerised I have tested in a number of different situation and with a
collection of mics and such. what I have found is that it is easy to get a
mike with a flat response from ~60Hz to ~20Kz, it seems that below this is
where a specialist product is required.

The remainder of the measurement and analysis chain has similar
properties: the primary measurement chain is DC coupled, for example.

Secondly, the size of the venue required for accurate measurements
is inversely proportional to the frequency you need to measure. Even
using techniques such as gated or windowed measurement, the distance
to the first reflection surface is a prime determinant of how low
you can measure. You want to measure 20 Hz accurately? Then you need
to find a room where the distance between the speaker/microphone and
the NEAREST surface is a minimum of 25 feet.

I have tested a few speakers and
basically I cannot get accurate (repeatable) data at much below 60Hz,

even
using a borrowed shure KSM141 stereo pair in omnidirectional gives

vague
results ($3000 for the mic's!)


$3000 for microphones that were NEVER designed to be used as measurement
microphones, ESPECIALLY at low frequencies. These are recording
microphones, NOT measurement microphones.

I have come to the conclusion that you are
really measuring air displacement at anything under 30Hz and a flat

panel
with a transducer is the only repeatable method of measuring the

output,

That may be the conclusion you came to, but that conclusion just happens
to be quite wrong. The physical stimulus that the ear responds to as
sound are periodic pressure variations of a sufficient amplitude and
within certain frequency limits. That's it. As long as a device can
detect these pressure variations, it can be used to measure sound.

The problem with you big flat panel method is that it assumes, quite
incorrectly, that the imnpinging waves are planar: unless you are VERY
far away from the speaker, such the wavefronts are no longer spherical,
it isn't going to work. The smaller the diaphgragm of the microphone,
the less it is affected by such a problem. That's one reason why
measurement microphones have very small diaphragms: they are essentially
point transducers over a wide range of frequencies.

how one calibrates said device is open for discussion.


snip good stuff from Dick

I do undestand the theory, and I expect it would work well if the speaker
was moving in piston motion. (An some of the car sub woofers with ridged
aluminium cones would no doubt at low frequency)

However typical stereo speakers using either doped paper or pp cones

won't,
as you know the LF breakup happens quite early on these type of drivers

due
to the tradeoff of trying to give a wide response range with only 2 or 3
drives in most cases and stiff cone suspension.

more snip

Compare that to using recording microphones whose measurement

capabilities
are entirely unknown, in a room of unknown characteristics, using

unknown
poorly calibrated and undoubtedly poorly controlled techniques by

someone
who has little or know experience in measurement and acoustics...

I'd not bet good money on getting ANY reliable data out of the latter.


Well I have to say getting reliable data from 'experts' is not very
successful!


I think Dick gave you good information.

The B&K microphones are very expensive. Someone earlier posted a
suggestion for a Panasonic capsule. These are available calibrated.
The Behringer ECM8000 is more expensive, but still relatively cheap (£30).
Either solution is far better than you have been using.

Dick mentions the problem of reflections when doing in-room response.
At low frequencies this can be dealt with using near field techniques,
valid up to where the diaphragm reaches its piston limit. This is often
several hundred Hz, allowing you to "stitch together" responses from
different regions.

Your assertion that it is not possible to get low frequencies using an
8 inch driver turns out not to be correct. I have a pair where the LF
is done using 8 inch drivers in a bandpass enclosure. Those have a
measured response which is 1dB down at 30Hz, and an overall
response at 3dB down of 27Hz to 90Hz.

Regards
Ian


  #34   Report Post  
Geoff Wood
 
Posts: n/a
Default 20hz to 20Khz , yea right!


"The Flash" wrote in message
...
Uh, I used to test in a anechoic chamber at the local university but don't
now, I also tested other areas and found the following. You can test fine

in
you living room. Background noiselevel is typically 5OdB at home.



Noisae isn't the problem . Speaker measurements in rooms are flawed becuse
of the effects of standing waves.



  #35   Report Post  
Geoff Wood
 
Posts: n/a
Default 20hz to 20Khz , yea right!


"The Flash" wrote in message
...
Uh, I used to test in a anechoic chamber at the local university but don't
now, I also tested other areas and found the following. You can test fine

in
you living room. Background noiselevel is typically 5OdB at home.



Noisae isn't the problem . Speaker measurements in rooms are flawed becuse
of the effects of standing waves.





  #36   Report Post  
Goofball_star_dot_etal
 
Posts: n/a
Default 20hz to 20Khz , yea right!

On Sun, 23 Nov 2003 09:11:31 -0500, "Victor"
wrote:

"Goofball_star_dot_etal" wrote:
I was not trying to make a very accurate measurement so I just used a
cheap Panasonic WM-60AT price £2

----------------------------------------

That's almost exactly what Behringer uses in ECM8000...


I guess I am just lucky when it comes to measurements and boats. . .

  #37   Report Post  
Goofball_star_dot_etal
 
Posts: n/a
Default 20hz to 20Khz , yea right!

On Sun, 23 Nov 2003 09:11:31 -0500, "Victor"
wrote:

"Goofball_star_dot_etal" wrote:
I was not trying to make a very accurate measurement so I just used a
cheap Panasonic WM-60AT price £2

----------------------------------------

That's almost exactly what Behringer uses in ECM8000...


I guess I am just lucky when it comes to measurements and boats. . .

  #38   Report Post  
The Flash
 
Posts: n/a
Default 20hz to 20Khz , yea right!

I thought up my own method for eliminating the effect of the room on
bass measurement (although, no doubt it has been done before). The
main feature is to place the (pressure) microphone within the speaker
and to compensate by 12 dB/oct.

Normally I use a MLS and filter out higher frequencies, say above
100Hz, with a brick wall FIR filter using Coooledit. The 12 dB/oct
compensation is also done in a similar way, either to the input MLS
or to the output recording. This sort of pre-filtering can give a
large increase to the signal to noise of the meaasurement.

There must be some restrictions on the accuracy of this method but
the only one I can think off at the moment is that the wavelength must
be large compared to the internal dimensions of the speaker. The
meaasurement includes the effect of any port and of the enclosure
flexing.

There is probably an assumption that the air is compressed
adiabatically. The pressures involved are, of course, small if the mic
is not to be overloaded.


What type of / brand of mic are you using?



  #39   Report Post  
The Flash
 
Posts: n/a
Default 20hz to 20Khz , yea right!

I thought up my own method for eliminating the effect of the room on
bass measurement (although, no doubt it has been done before). The
main feature is to place the (pressure) microphone within the speaker
and to compensate by 12 dB/oct.

Normally I use a MLS and filter out higher frequencies, say above
100Hz, with a brick wall FIR filter using Coooledit. The 12 dB/oct
compensation is also done in a similar way, either to the input MLS
or to the output recording. This sort of pre-filtering can give a
large increase to the signal to noise of the meaasurement.

There must be some restrictions on the accuracy of this method but
the only one I can think off at the moment is that the wavelength must
be large compared to the internal dimensions of the speaker. The
meaasurement includes the effect of any port and of the enclosure
flexing.

There is probably an assumption that the air is compressed
adiabatically. The pressures involved are, of course, small if the mic
is not to be overloaded.


What type of / brand of mic are you using?



  #40   Report Post  
The Flash
 
Posts: n/a
Default 20hz to 20Khz , yea right!


I have had no luck in measuring speaker response at low frequency's, I

can
measure from ~60Hz to 20Khz with near 100% repeatability (How accurate

is
questionable)



You measure it with microphones and ancillary equipment that is up
to the task. For myself, I have a number of Bruel & Kjaer, ACO,
GR and pther microphones that have verified flat response to well
below 20 Hz. Some of the B&K 1/2" capsules, for example, are within
+-1/2 dB from approximately 3 Hz to 30 kHz and above.


Can you give a recommend model of Bruel that you like?

Summerised I have tested in a number of different situation and with a
collection of mics and such. what I have found is that it is easy to get a
mike with a flat response from ~60Hz to ~20Kz, it seems that below this is
where a specialist product is required.

The remainder of the measurement and analysis chain has similar
properties: the primary measurement chain is DC coupled, for example.

Secondly, the size of the venue required for accurate measurements
is inversely proportional to the frequency you need to measure. Even
using techniques such as gated or windowed measurement, the distance
to the first reflection surface is a prime determinant of how low
you can measure. You want to measure 20 Hz accurately? Then you need
to find a room where the distance between the speaker/microphone and
the NEAREST surface is a minimum of 25 feet.

I have tested a few speakers and
basically I cannot get accurate (repeatable) data at much below 60Hz,

even
using a borrowed shure KSM141 stereo pair in omnidirectional gives vague
results ($3000 for the mic's!)


$3000 for microphones that were NEVER designed to be used as measurement
microphones, ESPECIALLY at low frequencies. These are recording
microphones, NOT measurement microphones.

I have come to the conclusion that you are
really measuring air displacement at anything under 30Hz and a flat

panel
with a transducer is the only repeatable method of measuring the output,


That may be the conclusion you came to, but that conclusion just happens
to be quite wrong. The physical stimulus that the ear responds to as
sound are periodic pressure variations of a sufficient amplitude and
within certain frequency limits. That's it. As long as a device can
detect these pressure variations, it can be used to measure sound.

The problem with you big flat panel method is that it assumes, quite
incorrectly, that the imnpinging waves are planar: unless you are VERY
far away from the speaker, such the wavefronts are no longer spherical,
it isn't going to work. The smaller the diaphgragm of the microphone,
the less it is affected by such a problem. That's one reason why
measurement microphones have very small diaphragms: they are essentially
point transducers over a wide range of frequencies.

how one calibrates said device is open for discussion.


One calibrates it by throwing it in the nearest landfill and going
out an learning the proper ways of measuring acoustic phenomenon.


You comments are most welcome but not vey helpful, I seem to be getting the
feeling you are giving me the 'I've be doing it for decade and got really
expensive gear, so my methods, results and opinions are gospel - do it my
way or be burned at the stake as a heratic'

One starts to wonder if perhaps you are one of the 'Experts' that 'corrects'
the measurments to what you 'know' is correct. Your attitude in your
responses seems to show this blinded mentality.

(Oh one firm told me that they use a laser to measure the low frequency

of
their speakers, check this out for novel! They place a small piece of
reflective foil on the base driver, and shine a laser beam on it, they

then
apply signal and measure via 'laser' the deflection,


Well, gee golly, since it can be shown on physical first principles
that the requirement for a constant sound pressure level (that would
mean flat frequency response) from a piston radiator is simply a
displacement which goes as the reciprocal of the square of frequency,
then if you know the displacement, which you can measure with a pretty
high degree of accuracy, then you can, over the piston range of the
driver, DIRECTLY and UNAMBIGUOUSLY determine the total acoustic power
as:

Pa = p/(2 pi c) * (Sd w^2 X)^2

where

p = density of air, typ. 1.18 kg/m^3
c = velocity of sound, typ 343 m/s
Sd = emissive area of the diaphragm in m^2
w = radian frequency
X = displacement of the diaphragm, in m.

the method is HARDLY novel at all, as it is well understood and utilized
in the field. If provides, for example, a means of measuring acoustical
power output without the confounding innaccuracies of microphones, rooms
and such, though the microphone innaccuracies are not a problem if you
use proper microphones to begin with.


I do undestand the theory, and I expect it would work well if the speaker
was moving in piston motion. (An some of the car sub woofers with ridged
aluminium cones would no doubt at low frequency)

However typical stereo speakers using either doped paper or pp cones won't,
as you know the LF breakup happens quite early on these type of drivers due
to the tradeoff of trying to give a wide response range with only 2 or 3
drives in most cases and stiff cone suspension.

More to your notion that it is "novel," you might want to modify that
opinion when you discover the technique is described in nearly every
text on acoustics.


-Noted and yes I have seen it mention and was taught such things, but it was
always theory and I had not heard of it done on audio speakers (however I do
know it is done on building structures to measure characteristic)

I do know that laser scanning of drive to measure the breakup is done, but
this was not what they were doing (So they said)

Compare that to using recording microphones whose measurement capabilities
are entirely unknown, in a room of unknown characteristics, using unknown
poorly calibrated and undoubtedly poorly controlled techniques by someone
who has little or know experience in measurement and acoustics...

I'd not bet good money on getting ANY reliable data out of the latter.


Well I have to say getting reliable data from 'experts' is not very
successful!


Reply
Thread Tools
Display Modes

Posting Rules

Smilies are On
[IMG] code is On
HTML code is Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
When did home theater take over? chexxon Audio Opinions 305 January 14th 04 11:50 PM
Learning 30 tones from 20hz to 20khz Tom Pro Audio 8 November 22nd 03 07:08 PM


All times are GMT +1. The time now is 09:17 PM.

Powered by: vBulletin
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Copyright ©2004-2024 AudioBanter.com.
The comments are property of their posters.
 

About Us

"It's about Audio and hi-fi"