[Henry Kolesnik]

I just got the book Sound System Engineering by Don Davis and Carolyn
Davis and it has 2 short sections on sound decades. Defined as:
HF/LF=10**1=1 decade HF is highest freq and LF is lowest freq
or
HF/LF=10**x decades
or
(Ln HF - Ln LF)/Ln 10 = x decades
An example for a span of 30 to 15,000 Hz
x decades = (Ln 15,000 - Ln 30)/Ln 10 = 2.7 decades
Ok, I think I understand the math but what is a practical application of
a sound decade?
I couldn't find anything in Wikipedia or Goog..

tnx

--

73
Hank WD5JFR

[Scott Dorsey]

Henry Kolesnik wrote:
I just got the book Sound System Engineering by Don Davis and Carolyn
Davis and it has 2 short sections on sound decades. Defined as:
HF/LF=10**1=1 decade HF is highest freq and LF is lowest freq
or
HF/LF=10**x decades
or
(Ln HF - Ln LF)/Ln 10 = x decades
An example for a span of 30 to 15,000 Hz
x decades = (Ln 15,000 - Ln 30)/Ln 10 = 2.7 decades
Ok, I think I understand the math but what is a practical application of
a sound decade?

It's just a convenient way of thinking about pitch changes. Sometimes
it can make the math easier to think about decades instead of octaves,
especially if you are plotting responses on conventional log-log paper
where there is a cycle every tenfold increase.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Henry Kolesnik]

I understand the octave which relates to the doubling of pitch, cent
which is 1/1200 of an octave. Cents are used in tuning and the good
ears can tell a diference of about 3 cents. My imagination seems
limited in understanding what's so great about plotting on log-log
paper. And why is the knowledge of sound decades so limited? From 30
to 15,000 Hz there's 2.7 decades, 8.97 octaves and too many cents. The
book points out that all visible light is one octave making the audible
spectrum much more complicated if you look at it from that standpoint.
In a multiband shortwave radio each band is approximately an octave
because that's all a LC circuit can span.

--

73
Hank WD5JFR

"Scott Dorsey" wrote in message
...
Henry Kolesnik wrote:
I just got the book Sound System Engineering by Don Davis and Carolyn
Davis and it has 2 short sections on sound decades. Defined as:
HF/LF=10**1=1 decade HF is highest freq and LF is lowest freq
or
HF/LF=10**x decades
or
(Ln HF - Ln LF)/Ln 10 = x decades
An example for a span of 30 to 15,000 Hz
x decades = (Ln 15,000 - Ln 30)/Ln 10 = 2.7 decades
Ok, I think I understand the math but what is a practical application
of
a sound decade?

It's just a convenient way of thinking about pitch changes. Sometimes
it can make the math easier to think about decades instead of octaves,
especially if you are plotting responses on conventional log-log paper
where there is a cycle every tenfold increase.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Peter Larsen]

Henry Kolesnik wrote:

what's so great about plotting on log-log paper.

It correlates well to perception, as does writing the comment after what
is commented on. Usenet is very different from corporate email, as is
sound from RF.

Hank WD5JFR


Regards

Peter Larsen

[Scott Dorsey]

Henry Kolesnik wrote:
I understand the octave which relates to the doubling of pitch, cent
which is 1/1200 of an octave. Cents are used in tuning and the good
ears can tell a diference of about 3 cents. My imagination seems
limited in understanding what's so great about plotting on log-log
paper.

Usually it's because log-log paper is what there is, and it's what
people have always used for plotting responses. Look at a product
data sheet and that's how you'll see it plotted.

And why is the knowledge of sound decades so limited? From 30
to 15,000 Hz there's 2.7 decades, 8.97 octaves and too many cents. The
book points out that all visible light is one octave making the audible
spectrum much more complicated if you look at it from that standpoint.

Mostly it's because one of those things that engineers came up with
because it makes the arithmetic easier to do, and today with pocket
calculators people don't really care so much about making everything
fall into nice easy numbers any more.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Don Pearce]

On Sat, 24 Feb 2007 16:37:46 GMT, "Henry Kolesnik"
wrote:

I understand the octave which relates to the doubling of pitch, cent
which is 1/1200 of an octave. Cents are used in tuning and the good
ears can tell a diference of about 3 cents. My imagination seems
limited in understanding what's so great about plotting on log-log
paper. And why is the knowledge of sound decades so limited? From 30
to 15,000 Hz there's 2.7 decades, 8.97 octaves and too many cents. The
book points out that all visible light is one octave making the audible
spectrum much more complicated if you look at it from that standpoint.
In a multiband shortwave radio each band is approximately an octave
because that's all a LC circuit can span.

The octave is a musical convenience - we all recognise it and it is
the natural interval on which all scales are based. Being power based
rather than linear, it is a logarithmic scale. So much is easy enough.

The decade is an artifact, based on the fact that we are a ten-based
species. It has no greater significance than that it fits nicely with
our mathematical norms. It is something we have come to accept as the
natural order, even though it is no such thing.

So why use log/log paper? The frequency axis is log because that is
how music and out appreciation of pitch works. The vertical scale is
log (dB0 because that is how our appreciation of loudness works -
roughly. You could use a linear vertical scale, but then if you
encompassed the loudest sounds, the small detail would be invisible.
So a log decibel scale makes sense.

Why decade paper? Because we have ten fingers.

d

--
Pearce Consulting
http://www.pearce.uk.com

[Henry Kolesnik]

Hey guys, I'm a graduate engineer and have done probably hundreds of
plots on all kinds of log papers. I'm retired and still have a nice
little stack of K&E blanks, just in case I need them. And for the wise
guy, all I want to know is what is a practical application of using
decade plots for sound. Give one example that shows how a decade plot
is more useful than some other plot. Decade info on Goog is scarcer
than hen's teeth so it must be a big secret or am I missing something
about this artifact?

--

73
Hank WD5JFR

"Don Pearce" wrote in message
...
On Sat, 24 Feb 2007 16:37:46 GMT, "Henry Kolesnik"
wrote:

I understand the octave which relates to the doubling of pitch, cent
which is 1/1200 of an octave. Cents are used in tuning and the good
ears can tell a diference of about 3 cents. My imagination seems
limited in understanding what's so great about plotting on log-log
paper. And why is the knowledge of sound decades so limited? From
30
to 15,000 Hz there's 2.7 decades, 8.97 octaves and too many cents.
The
book points out that all visible light is one octave making the
audible
spectrum much more complicated if you look at it from that standpoint.
In a multiband shortwave radio each band is approximately an octave
because that's all a LC circuit can span.

The octave is a musical convenience - we all recognise it and it is
the natural interval on which all scales are based. Being power based
rather than linear, it is a logarithmic scale. So much is easy enough.

The decade is an artifact, based on the fact that we are a ten-based
species. It has no greater significance than that it fits nicely with
our mathematical norms. It is something we have come to accept as the
natural order, even though it is no such thing.

So why use log/log paper? The frequency axis is log because that is
how music and out appreciation of pitch works. The vertical scale is
log (dB0 because that is how our appreciation of loudness works -
roughly. You could use a linear vertical scale, but then if you
encompassed the loudest sounds, the small detail would be invisible.
So a log decibel scale makes sense.

Why decade paper? Because we have ten fingers.

d

--
Pearce Consulting
http://www.pearce.uk.com

[Don Pearce]

You see, you have to actually read what I wrote, then you'll
understand.

And please don't top post (like this). It screws the threading, and
the stuff you are replying to disappears into your sig area.

d

On Sat, 24 Feb 2007 17:20:10 GMT, "Henry Kolesnik"
wrote:

Hey guys, I'm a graduate engineer and have done probably hundreds of
plots on all kinds of log papers. I'm retired and still have a nice
little stack of K&E blanks, just in case I need them. And for the wise
guy, all I want to know is what is a practical application of using
decade plots for sound. Give one example that shows how a decade plot
is more useful than some other plot. Decade info on Goog is scarcer
than hen's teeth so it must be a big secret or am I missing something
about this artifact?

--

73
Hank WD5JFR

"Don Pearce" wrote in message
...
On Sat, 24 Feb 2007 16:37:46 GMT, "Henry Kolesnik"
wrote:

I understand the octave which relates to the doubling of pitch, cent
which is 1/1200 of an octave. Cents are used in tuning and the good
ears can tell a diference of about 3 cents. My imagination seems
limited in understanding what's so great about plotting on log-log
paper. And why is the knowledge of sound decades so limited? From
30
to 15,000 Hz there's 2.7 decades, 8.97 octaves and too many cents.
The
book points out that all visible light is one octave making the
audible
spectrum much more complicated if you look at it from that standpoint.
In a multiband shortwave radio each band is approximately an octave
because that's all a LC circuit can span.

The octave is a musical convenience - we all recognise it and it is
the natural interval on which all scales are based. Being power based
rather than linear, it is a logarithmic scale. So much is easy enough.

The decade is an artifact, based on the fact that we are a ten-based
species. It has no greater significance than that it fits nicely with
our mathematical norms. It is something we have come to accept as the
natural order, even though it is no such thing.

So why use log/log paper? The frequency axis is log because that is
how music and out appreciation of pitch works. The vertical scale is
log (dB0 because that is how our appreciation of loudness works -
roughly. You could use a linear vertical scale, but then if you
encompassed the loudest sounds, the small detail would be invisible.
So a log decibel scale makes sense.

Why decade paper? Because we have ten fingers.

d

--
Pearce Consulting
http://www.pearce.uk.com

--
Pearce Consulting
http://www.pearce.uk.com

[Scott Dorsey]

Henry Kolesnik wrote:
Hey guys, I'm a graduate engineer and have done probably hundreds of
plots on all kinds of log papers. I'm retired and still have a nice
little stack of K&E blanks, just in case I need them. And for the wise
guy, all I want to know is what is a practical application of using
decade plots for sound. Give one example that shows how a decade plot
is more useful than some other plot. Decade info on Goog is scarcer
than hen's teeth so it must be a big secret or am I missing something
about this artifact?

Look at the data sheet of any product. You'll see a decade chart.

It would correlate better with the musical scale if we used 16-cycle
or 8-cycle plots. But we don't, mostly because a century and a half ago
when Helmholtz started plotting responses there wasn't convenient paper
available.

If we had 16 fingers, it would be be an even better fit, and also it
would be a lot easier to play the Rachmaninov second piano concerto too.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Roy W. Rising]

"Henry Kolesnik" wrote:
I just got the book Sound System Engineering by Don Davis and Carolyn
Davis and it has 2 short sections on sound decades. Defined as:
HF/LF=10**1=1 decade HF is highest freq and LF is lowest freq
or
HF/LF=10**x decades
or
(Ln HF - Ln LF)/Ln 10 = x decades
An example for a span of 30 to 15,000 Hz
x decades = (Ln 15,000 - Ln 30)/Ln 10 = 2.7 decades
Ok, I think I understand the math but what is a practical application of
a sound decade?
I couldn't find anything in Wikipedia or Goog..

tnx

Why logs (decades)? Just try plotting 20- 20KHz linearly! Or, 1 microvolt
to 100 volts! (That's the typical range of levels in sound.)

--
~ Roy
"If you notice the sound, it's wrong!"

[Henry Kolesnik]

"Roy W. Rising" wrote in message
...
"Henry Kolesnik" wrote:
I just got the book Sound System Engineering by Don Davis and Carolyn
Davis and it has 2 short sections on sound decades. Defined as:
HF/LF=10**1=1 decade HF is highest freq and LF is lowest freq
or
HF/LF=10**x decades
or
(Ln HF - Ln LF)/Ln 10 = x decades
An example for a span of 30 to 15,000 Hz
x decades = (Ln 15,000 - Ln 30)/Ln 10 = 2.7 decades
Ok, I think I understand the math but what is a practical application
of
a sound decade?
I couldn't find anything in Wikipedia or Goog..

tnx

Why logs (decades)? Just try plotting 20- 20KHz linearly! Or, 1
microvolt
to 100 volts! (That's the typical range of levels in sound.)

--
~ Roy
"If you notice the sound, it's wrong!"
Thanks, it has sunk in...
Hank

[Karl Uppiano]

If we had 16 fingers, it would be be an even better fit, and also it
would be a lot easier to play the Rachmaninov second piano concerto too.

Ha! If we had 16 fingers, Rachmaninov could have written harder concerti.

[Mr.T]

"Scott Dorsey" wrote in message
...
Henry Kolesnik wrote:
My imagination seems
limited in understanding what's so great about plotting on log-log
paper.

Usually it's because log-log paper is what there is, and it's what
people have always used for plotting responses. Look at a product
data sheet and that's how you'll see it plotted.

Which is decidedly NOT because that was all the paper they had!
It's because they chose log-log to best suit the required purpose, just as
everyone else does.
Surely you understand the reason is because of the logarithmic response of
the human auditory system to both frequency and SPL?


Mostly it's because one of those things that engineers came up with
because it makes the arithmetic easier to do, and today with pocket
calculators people don't really care so much about making everything
fall into nice easy numbers any more.

Ignore my question above, obviously you don't.

MrT.

[Mr.T]

"Henry Kolesnik" wrote in message
. ..
Hey guys, I'm a graduate engineer and have done probably hundreds of
plots on all kinds of log papers. I'm retired and still have a nice
little stack of K&E blanks, just in case I need them. And for the wise
guy, all I want to know is what is a practical application of using
decade plots for sound. Give one example that shows how a decade plot
is more useful than some other plot. Decade info on Goog is scarcer
than hen's teeth so it must be a big secret or am I missing something
about this artifact?

So you still don't understand that we usually use a tens based number system
simply for convenience.
Feel free to use a sexagesimal system if you prefer. Of course getting the
rest of the world to change is not so easy.

MrT.

[Mr.T]

"Roy W. Rising" wrote in message
...
Why logs (decades)? Just try plotting 20- 20KHz linearly! Or, 1
microvolt
to 100 volts! (That's the typical range of levels in sound.)

Which is also backwards thinking. We only chose break points of 20-20kHz,
100Volts etc. because we had ALREADY chosen to use a decimal number system!

MrT.

[Scott Dorsey]

Mr.T MrT@home wrote:
"Scott Dorsey" wrote in message
...
Henry Kolesnik wrote:
My imagination seems
limited in understanding what's so great about plotting on log-log
paper.

Usually it's because log-log paper is what there is, and it's what
people have always used for plotting responses. Look at a product
data sheet and that's how you'll see it plotted.

Which is decidedly NOT because that was all the paper they had!
It's because they chose log-log to best suit the required purpose, just as
everyone else does.

Yes, but why 10-cycle? You'd think 8-cycle or 16 cycle would be
more appropriate for the purpose, except that they don't make it...

Surely you understand the reason is because of the logarithmic response of
the human auditory system to both frequency and SPL?

Sure, but we tend to think of frequencies in powers of two rather than ten.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Mr.T]

"Scott Dorsey" wrote in message
...
Yes, but why 10-cycle? You'd think 8-cycle or 16 cycle would be
more appropriate for the purpose, except that they don't make it...

For some reason the decimal system is prefered over the octal and
hexadecimal in the general community. Something to do with what they were
first taught at school I think. Even with the introduction of computers
making Hex slightly more well known, I don't think that is likely to change
anytime soon.

Sure, but we tend to think of frequencies in powers of two rather than
ten.

Being logarithmic to power of two, suits that already. Line groupings of two
would just be silly. Getting people to change to an octal number system
would be useful though, I will admit.

Good luck with that.

MrT.

[Roy W. Rising]

"Mr.T" MrT@home wrote:
"Roy W. Rising" wrote in message
...
Why logs (decades)? Just try plotting 20- 20KHz linearly! Or, 1
microvolt to 100 volts! (That's the typical range of levels in sound.)

Which is also backwards thinking. We only chose break points of 20-20kHz,
100Volts etc. because we had ALREADY chosen to use a decimal number
system!

MrT.

Your other responses were interestingly confused. This one is somewhat
upside-down. Logarithmic representations include natural logs that do not
use the base 10. While they more closely resemble our natural perceptions,
their math is unfamiliar.

The break point of 20 Hz is conveniently close to 16 Hz, the lowest
fundamental in acoustical music. 20 KHz is in the neighborhood of the
highest frequency humans can conciously identify. These nearby decimal
locations remain convenient approximations.

Log 10 based scale compression gives us a convenient display, but many
ten-fingered humans haven't the foggiest understanding of logs, hence this
thread. Many of the same primates continue to insist on a system of
weights and measures that conspicuously avoids decimal simplicity.
Velocities measured in furlongs per fortnight or miles per hour and
pressure measured in pounds per square inch deny the *assumption* "we had
ALREADY chosen to use a decimal number system!" ;-)

--
~ Roy
"If you notice the sound, it's wrong!"

[Mr.T]

"Roy W. Rising" wrote in message
...
Which is also backwards thinking. We only chose break points of
20-20kHz,
100Volts etc. because we had ALREADY chosen to use a decimal number
system!

Your other responses were interestingly confused.

Sorry that you think so.

This one is somewhat upside-down.
Logarithmic representations include natural logs that do not
use the base 10. While they more closely resemble our natural
perceptions,
their math is unfamiliar.

Exactly, log(e) is used where appropriate in the scientific world. It's use
would be of limited value if you are trying to reach the general population.

The break point of 20 Hz is conveniently close to 16 Hz, the lowest
fundamental in acoustical music. 20 KHz is in the neighborhood of the
highest frequency humans can conciously identify. These nearby decimal
locations remain convenient approximations.

Exactly. That's what I said, they were chosen as suitably appropriate
DECIMAL numbers for convenience. They did NOT precede the decimal system.

Many of the same primates continue to insist on a system of
weights and measures that conspicuously avoids decimal simplicity.
Velocities measured in furlongs per fortnight or miles per hour and
pressure measured in pounds per square inch deny the *assumption* "we had
ALREADY chosen to use a decimal number system!" ;-)

Yes humans are irrational. However sometimes there are good reasons for
things the way they are. I doubt we will ever have decimal time for example,
but who knows, if we ever colonise another planet things might change.
Angular measurements in degrees, radians etc. have their own benefits too.
Many such examples.

But there is a good reason why *most* countries have gone Metric anyway. And
*most* have decimal currency.

MrT.

[Roy W. Rising]

"Mr.T" MrT@home wrote:
"Roy W. Rising" wrote in message
...
Which is also backwards thinking. We only chose break points of
20-20kHz,
100Volts etc. because we had ALREADY chosen to use a decimal number
system!

Your other responses were interestingly confused.

Sorry that you think so.

This one is somewhat upside-down.
Logarithmic representations include natural logs that do not
use the base 10. While they more closely resemble our natural
perceptions,
their math is unfamiliar.

Exactly, log(e) is used where appropriate in the scientific world. It's
use would be of limited value if you are trying to reach the general
population.

The break point of 20 Hz is conveniently close to 16 Hz, the lowest
fundamental in acoustical music. 20 KHz is in the neighborhood of the
highest frequency humans can conciously identify. These nearby decimal
locations remain convenient approximations.

Exactly. That's what I said, they were chosen as suitably appropriate
DECIMAL numbers for convenience. They did NOT precede the decimal system.

Many of the same primates continue to insist on a system of
weights and measures that conspicuously avoids decimal simplicity.
Velocities measured in furlongs per fortnight or miles per hour and
pressure measured in pounds per square inch deny the *assumption* "we
had ALREADY chosen to use a decimal number system!" ;-)

Yes humans are irrational. However sometimes there are good reasons for
things the way they are. I doubt we will ever have decimal time for
example, but who knows, if we ever colonise another planet things might
change.
Angular measurements in degrees, radians etc. have their own benefits
too. Many such examples.

But there is a good reason why *most* countries have gone Metric anyway.
And *most* have decimal currency.

MrT.

Excelent answers and examples! Regarding decimal time, some companies log
their employees' presence in 6 minute increments ... tenths of the hour.

--
~ Roy
"If you notice the sound, it's wrong!"

[RD Jones]

"Henry Kolesnik" wrote:

Ok, I think I understand the math but what is a practical application of
a sound decade?

Crossovers, ie: the woofer is the first decade, mids are the second
and the tweeter is the third.

Multiband compression (3 band) is another that falls into 3 neat
decades for processing, if you choose the bands in a method
that seems most logical to me.

rd

r.a.o removed from xpost reply

[Sander deWaal]

"Mr.T" MrT@home said:


"Henry Kolesnik" wrote in message
...
Hey guys, I'm a graduate engineer and have done probably hundreds of
plots on all kinds of log papers. I'm retired and still have a nice
little stack of K&E blanks, just in case I need them. And for the wise
guy, all I want to know is what is a practical application of using
decade plots for sound. Give one example that shows how a decade plot
is more useful than some other plot. Decade info on Goog is scarcer
than hen's teeth so it must be a big secret or am I missing something
about this artifact?


So you still don't understand that we usually use a tens based number system
simply for convenience.
Feel free to use a sexagesimal system if you prefer. Of course getting the
rest of the world to change is not so easy.



Then start using the SI (metric) system first.
Measuring distances in feet, yards etc. is archaic. ;-)

--

- Maggies are an addiction for life. -

[Mike Rivers]

On Feb 24, 10:36 am, "Henry Kolesnik"
wrote:
I just got the book Sound System Engineering by Don Davis and Carolyn
Davis and it has 2 short sections on sound decades.

I've been able to make absolutely no sense of the replies (relevant to
your question) here, so I got out my copy of Sound System Engineering,
checked the index for "decade" and found nothing. In what section or
chapter (by title, please, not page number or chapter number) did you
find this? My edition is very old (the hand calculator pictured is an
HP-35) so it may not be in mine, but I'm willing to look for it if I
knew where to look. Your question, and the answers, may make more
sense in context than just looking at a formula.

[Henry Kolesnik]

"Mike Rivers" wrote in message
ups.com...
On Feb 24, 10:36 am, "Henry Kolesnik"
wrote:
I just got the book Sound System Engineering by Don Davis and Carolyn
Davis and it has 2 short sections on sound decades.

I've been able to make absolutely no sense of the replies (relevant to
your question) here, so I got out my copy of Sound System Engineering,
checked the index for "decade" and found nothing. In what section or
chapter (by title, please, not page number or chapter number) did you
find this? My edition is very old (the hand calculator pictured is an
HP-35) so it may not be in mine, but I'm willing to look for it if I
knew where to look. Your question, and the answers, may make more
sense in context than just looking at a formula.



Mike
My book is the Second Edition FIRST PRINTING 1987 and decades are in
the index. In The Contents at the first of the book we see Section 2
Mathematics for Audio Systems, Decade Calibration page 26 and Section 3
Using the Decibel, Calculating the Number of Decades in a Frequency Span
page 57..
I accept and agree that it is easier and much more convenient to plot on
log-log paper any values that have a wide range and log base 10 is fine.
I also find it very practical to say some pitch or frequency depending
on whether we're talking about sound or electrical signals that the
range is 2 or 3 octaves. If we're discussing audio and I say starting
at 20Hz every one knows 2 octaves is 80Hz. Or if I say 880Hz but one
octave lower you go to 440Hz or 0.3010 decade in either case. So
decades aren't of much practical use as far as I can see. Or take the
span 20 to 20Khz is 3 decades is easy to do mentally and easy to
understand whereas that same span is not quite 10 octaves, not too
practical, but failrly easy to calculate on HP 11C.
RDJones mentioned; "Crossovers, ie: the woofer is the first decade, mids
are the second
and the tweeter is the third." For speakers that puts woofers producing
20 to 200Hz, midrange producing 200 to 2000Hz and tweeters 2000 to
20,000Hz. This is my example to be used only for illustration because
you never want any crossover in the 2000Hz neighborhood because it can
degrade the human voice.
From another book, Audio Cyclopedia by Tremaine it shows crossovers at
450hz and 5000Hz for a 3 way and 200, 1000, 3500, & 10KHz for a 5 way.
It should be noted that the plots are on semilog paper but the filter
rolloff is specd at 6dB per octave.
In the Audio Engineering Handbook by Benson plots are on semilog paper
and rolloff in dB per octave.
Neither of the above books cover decades. I'm starting to think that
decades are only used because we plot on log rule graph paper and that's
because we don't have octave ruled graph paper. Their practical use is
limited.
Out of curiosity and because I've been shook up by sub woofers in
theaters and now some cars can vibrate mine, what is the frequwncy range
for these sub woofers and what is the rolloff, octaves or decades,
however they are specd.
tnx

--

73
Hank WD5JFR

[Roy W. Rising]

"RD Jones" wrote:
"Henry Kolesnik" wrote:

Ok, I think I understand the math but what is a practical application
of a sound decade?

Crossovers, ie: the woofer is the first decade, mids are the second
and the tweeter is the third.

Multiband compression (3 band) is another that falls into 3 neat
decades for processing, if you choose the bands in a method
that seems most logical to me.

rd

r.a.o removed from xpost reply

Let me see if I understand you.
First Decade = 20 Hz to 200 Hz?
Second Decade = 200 Hz to 2000 Hz?
Third Decade = 2000 Hz to 20,000 Hz?

--
~ Roy
"If you notice the sound, it's wrong!"

[Mike Rivers]

On Feb 25, 11:13 am, "Henry Kolesnik"
wrote:

My book is the Second Edition FIRST PRINTING 1987 and decades are in
the index. In The Contents at the first of the book we see Section 2
Mathematics for Audio Systems, Decade Calibration page 26 and Section 3
Using the Decibel, Calculating the Number of Decades in a Frequency Span
page 57..

Oh, you have a new one. Mine is First Edition 8th printing, 1983.
There's a lot about decimals and logs in that section but I didn't see
anything like what you were talking about. Must have been a new idea
they had for a class. I wouldn't think that they'd make such a big
deal over what's obvious by looking at just about any frequency
response graph, but maybe they felt it necessary to explain why we
plot from 20 to 20,000 and not 10 to 100,000.

At first I thought you were talking about a rule of thumb that
suggested that the relationship between the lowest and highest
frequency of whatever you chose to consider the bandwidth should is a
constant for a "balanced" spectrum over the bandwidth, in other words,
to not be bass-heavy or treble-heavy. But that's clearly not it, and I
don't remember what the relationship is. It's interesting, anyway, and
I'm sure that by mentioning it, someone will know what I'm talking
about and state it properly. I love Usenet!

I'm starting to think that
decades are only used because we plot on log rule graph paper and that's
because we don't have octave ruled graph paper.

Now that's something that would confuse everyone, except maybe for
musicians trained in harmonic theory. Cranesong makes an equalizer
where the knobs aren't calibrated in frequency, they're calibrated in
musical notes. I suspect that's an attempt to get people to just turn
the knobs until it sounds right instead of boosting a 2.5 kHz because
someone told them to do it.

Out of curiosity and because I've been shook up by sub woofers in
theaters and now some cars can vibrate mine, what is the frequwncy range
for these sub woofers and what is the rolloff, octaves or decades,
however they are specd.

In theater systems, they're designed to start taking over at around 80
Hz. In cars, they're designed to produce a very small frequency range,
probably in the order of 30-40 Hz. Anything down in that range will
come out of the subwoofer pretty much at its tuned frequency because,
well, it's tuned for that.

[Roy W. Rising]

"Mike Rivers" wrote:

At first I thought you were talking about a rule of thumb that
suggested that the relationship between the lowest and highest
frequency of whatever you chose to consider the bandwidth should is a
constant for a "balanced" spectrum over the bandwidth, in other words,
to not be bass-heavy or treble-heavy. But that's clearly not it, and I
don't remember what the relationship is. It's interesting, anyway, and
I'm sure that by mentioning it, someone will know what I'm talking
about and state it properly. I love Usenet!

The relationship you mention is "Let the product of the LF and HF -3dB
frequencies equal 600,000". Familiar examples are 30-20,000Hz; 50-12,000Hz
and 100-6000Hz.

--
~ Roy
"If you notice the sound, it's wrong!"

[Mike Rivers]

On Feb 25, 1:12 pm, Roy W. Rising wrote:

The relationship you mention is "Let the product of the LF and HF -3dB
frequencies equal 600,000". Familiar examples are 30-20,000Hz;

Oh, well, there goes 50 years of "20 to 20,000 cycles" shot to hell.

[Roy W. Rising]

"Mike Rivers" wrote:
On Feb 25, 1:12 pm, Roy W. Rising wrote:

The relationship you mention is "Let the product of the LF and HF -3dB
frequencies equal 600,000". Familiar examples are 30-20,000Hz;

Oh, well, there goes 50 years of "20 to 20,000 cycles" shot to hell.

There's not much music below 30 Hz, only a few pipe organs. However, I've
always thought in terms of 20 to 30,000 Hz. Consider ... if someone can
differentiate between a perfectly produced 8000 Hz sine wave and square
wave, they're hearing 24,000 Hz! It's another good reason for 96 KHz
sampling.

--
~ Roy
"If you notice the sound, it's wrong!"

[Karl Uppiano]

"Mike Rivers" wrote in message
ups.com...
On Feb 25, 1:12 pm, Roy W. Rising wrote:

The relationship you mention is "Let the product of the LF and HF -3dB
frequencies equal 600,000". Familiar examples are 30-20,000Hz;

Oh, well, there goes 50 years of "20 to 20,000 cycles" shot to hell.

Some old textbooks I've read give the human hearing range "30 cps to 15,000
cps" [sic]. I think it was gradually normalized to 20Hz to 20KHz during the
1960s.

Of course, either one is just a rule of thumb. Actual human sensitivity
varies. You can "hear" frequencies significantly lower than 20Hz if they are
loud enough. Ditto for ultrasonic, although I think there is more variation
at the high end.

In my youth, I used to easily hear TV flyback transformers at 15,750Hz. I
could walk into any department store, and tell immediately if they sold TVs,
because I could hear the sets whining away in the distance like a bunch of
cicadas. I'm lucky if I can hear 5KHz today.

[Karl Uppiano]

"Roy W. Rising" wrote in message
...
"Mike Rivers" wrote:
On Feb 25, 1:12 pm, Roy W. Rising wrote:

The relationship you mention is "Let the product of the LF and HF -3dB
frequencies equal 600,000". Familiar examples are 30-20,000Hz;

Oh, well, there goes 50 years of "20 to 20,000 cycles" shot to hell.

There's not much music below 30 Hz, only a few pipe organs. However, I've
always thought in terms of 20 to 30,000 Hz. Consider ... if someone can
differentiate between a perfectly produced 8000 Hz sine wave and square
wave, they're hearing 24,000 Hz! It's another good reason for 96 KHz
sampling.

Can someone do that? If they can, it would be the *only* good reason for
96KHz sampling that I can think of, unless you're selling bandwidth. :-)

Keep in mind that, for the test to be valid, the *fundamental* (8KHz)
frequency component must be at exactly the same level for the square wave
and the sine wave, or the difference in loudness will provide a cue.

I did this test on myself many years ago, and IIRC the math, the square wave
peak level needed to be 0.707 (-3dB) relative to the sinewave peak level. If
they were the same peak level you *could* hear the difference, due to the
3dB loudness mismatch, not the harmonics -- but at 8KHz, it wasn't always
completely obvious to my ear what the real difference was.

[Mike Rivers]

On Feb 25, 2:07 pm, Roy W. Rising wrote:

There's not much music below 30 Hz, only a few pipe organs.

No, but there's lots of advertising below 30 Hz.

However, I've
always thought in terms of 20 to 30,000 Hz. Consider ... if someone can
differentiate between a perfectly produced 8000 Hz sine wave and square
wave, they're hearing 24,000 Hz! It's another good reason for 96 KHz
sampling.

I'm not young enough to know anyone who can differential between an 8
kHz sine wave and an 8 kHz square wave. In fact, a lot of people I
know can't hear 8 kHz at all, and some of them are pretty young.

I haven't heard much from James Boyk lately (was it you who mentioned
hiim in another thread recently?) but he's long been a proponent of
recording and reproducing frequencies above 20 kHz. While he doesn't
have a lot of scientific evidence that peope can hear a difference
when response is extended beyond 20 kHz, he does have some
measurements showing that common musicial instruments produce
frequencies about 20 kHz. The Audio Impressions (or Expressions -
one's a sample library, the other is a school, I think) people push
their 96 kHz and 192 kHz pipe organ samples because they say that part
of what we hear is the difference frequency when the pipes radiate
frequencies above 20 kHz. They admit that there isn't a whole lot of
it, but in the interest of greater accuracy, they offer the samples
for those who want to use them

[Scott Dorsey]

Mike Rivers wrote:

I'm not young enough to know anyone who can differential between an 8
kHz sine wave and an 8 kHz square wave. In fact, a lot of people I
know can't hear 8 kHz at all, and some of them are pretty young.

I could make the distinction when I was a kid, but even today I can still
tell the difference between a 6 KHz sine and square. Is that close enough?

I haven't heard much from James Boyk lately (was it you who mentioned
hiim in another thread recently?) but he's long been a proponent of
recording and reproducing frequencies above 20 kHz. While he doesn't
have a lot of scientific evidence that peope can hear a difference
when response is extended beyond 20 kHz, he does have some
measurements showing that common musicial instruments produce
frequencies about 20 kHz. The Audio Impressions (or Expressions -
one's a sample library, the other is a school, I think) people push
their 96 kHz and 192 kHz pipe organ samples because they say that part
of what we hear is the difference frequency when the pipes radiate
frequencies above 20 kHz. They admit that there isn't a whole lot of
it, but in the interest of greater accuracy, they offer the samples
for those who want to use them

The basic philosophy is that even though pure sounds above some point
are not audible, the added harmonics change our perception of sounds
below that point. I've seen no solid arguments for or against this
philosophy.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Roy W. Rising]

(Scott Dorsey) wrote:
Mike Rivers wrote:

I'm not young enough to know anyone who can differential between an 8
kHz sine wave and an 8 kHz square wave. In fact, a lot of people I
know can't hear 8 kHz at all, and some of them are pretty young.

I could make the distinction when I was a kid, but even today I can still
tell the difference between a 6 KHz sine and square. Is that close
enough?

--scott

Hearing the difference at 6 KHz means you're hearing 3x6=18 KHz, the first
harmonic present in the square wave. Not bad!

A few years ago I identified the presence of a 20 KHz signal-light tone in
an intercom system. I was surprised.

Further back, I used to be annoyed by a very loud HF signal in the jewelry
department of a local store. There was no reason for the night time motion
sensors to be on when the store was open. I complained to an employee and
he looked at me as if I was planning a heist!

--
~ Roy
"If you notice the sound, it's wrong!"

[Mike Rivers]

On Feb 25, 2:53 pm, (Scott Dorsey) wrote:

I could make the distinction when I was a kid, but even today I can still
tell the difference between a 6 KHz sine and square. Is that close enough?

Close enough for folk music.

The basic philosophy is that even though pure sounds above some point
are not audible, the added harmonics change our perception of sounds
below that point. I've seen no solid arguments for or against this
philosophy.

The Audio Impressons thing is the first argument for high bandwidth
recording is the first argument for it that made sense to me. But that
depends on recording two pipes separately and combining them in
playback. Maybe you get the same thing with two trumpets playing the
same note together in an orchestra. I dunno.

[Mr.T]

"Sander deWaal" wrote in message
...
"Mr.T" MrT@home said:
So you still don't understand that we usually use a tens based number
system
simply for convenience.
Feel free to use a sexagesimal system if you prefer. Of course getting
the
rest of the world to change is not so easy.

Then start using the SI (metric) system first.
Measuring distances in feet, yards etc. is archaic. ;-)

Too right, how does that apply to anything I said?

MrT.

[Mr.T]

"Roy W. Rising" wrote in message
...
The relationship you mention is "Let the product of the LF and HF -3dB
frequencies equal 600,000". Familiar examples are 30-20,000Hz;
50-12,000Hz
and 100-6000Hz.

Shame the more common 20-20,000 doesn't fit then :-)

MrT.

[Mr.T]

"Roy W. Rising" wrote in message
...
There's not much music below 30 Hz, only a few pipe organs.

And piano's, and synth music and ...

However, I've
always thought in terms of 20 to 30,000 Hz. Consider ... if someone can
differentiate between a perfectly produced 8000 Hz sine wave and square
wave, they're hearing 24,000 Hz!

Or IM distortion products, or mismatched levels or ....

It's another good reason for 96 KHz sampling.

Another good reason to require some PROOF.

MrT.

[Mr.T]

"Roy W. Rising" wrote in message
...
"RD Jones" wrote:
"Henry Kolesnik" wrote:

Ok, I think I understand the math but what is a practical application
of a sound decade?

Crossovers, ie: the woofer is the first decade, mids are the second
and the tweeter is the third.

Multiband compression (3 band) is another that falls into 3 neat
decades for processing, if you choose the bands in a method
that seems most logical to me.

Let me see if I understand you.
First Decade = 20 Hz to 200 Hz?
Second Decade = 200 Hz to 2000 Hz?
Third Decade = 2000 Hz to 20,000 Hz?


It's a shame for his theory though that the graphs are almost always grouped
10-100, 100-1000, 1k-10k, and above. :-)

And of course X-overs rarely fall on the 200Hz and 2kHz points either.
Likewise for muti-band compressors.

MrT.

[Arny Krueger]

"Roy W. Rising" wrote in message

"Mike Rivers" wrote:
On Feb 25, 1:12 pm, Roy W. Rising
wrote:

The relationship you mention is "Let the product of the
LF and HF -3dB frequencies equal 600,000". Familiar
examples are 30-20,000Hz;

Oh, well, there goes 50 years of "20 to 20,000 cycles"
shot to hell.

There's not much music below 30 Hz, only a few pipe
organs.

Most of my live recordings have so much content below 30 Hz that I high-pass
them, knowning that they will be primarily played on systems that can't
handle 30Hz effectively, such as OEM car stereos.

However, I've always thought in terms of 20 to
30,000 Hz. Consider ... if someone can differentiate
between a perfectly produced 8000 Hz sine wave and square
wave, they're hearing 24,000 Hz!

AFAIK, this has never happened in a properly run experiment. It takes a
clean system (many tweeters have lots of HF nonlinearity), bias controls,
and intelligent level-matching. In fact 44/16 is an overkill format for
distributing recordings. It has more HF extension and more dynamic range
than can be practically used.

It's another good reason for 96 KHz sampling.

First off, 96 KHz does absolutely zero for one of the items mentioned here -
30Hz.

Secondly, no proof or even reliable suggestive evidence has been provided.
Just speculation that has already been tested and found to be false.