[apa]

Can anyone explain the difference between these two readings in Adobe
Audition's Amplitude Statistics window?
Thanks, Andy

[Randy Yates]

apa writes:

Can anyone explain the difference between these two readings in Adobe
Audition's Amplitude Statistics window?
Thanks, Andy

If you gun your engine at 10:00:00am for 5 seconds to go from 0 to 60
MPH, you might use 350 horsepower during those 5 seconds. However, if
you then level off at 60 and drive that speed for an hour, your average
over the entire hour may only be something like 60 horsepower.

Power is energy per time. Depending on the extent of the time, you can
consider it "total power" or "average power" over some averaging time.
--
% Randy Yates % "The dreamer, the unwoken fool -
%% Fuquay-Varina, NC % in dreams, no pain will kiss the brow..."
%%% 919-577-9882 %
%%%% % 'Eldorado Overture', *Eldorado*, ELO
http://www.digitalsignallabs.com

[apa]

On Jan 9, 3:53 pm, Randy Yates wrote:
apa writes:
Can anyone explain the difference between these two readings in Adobe
Audition's Amplitude Statistics window?
Thanks, Andy

If you gun your engine at 10:00:00am for 5 seconds to go from 0 to 60
MPH, you might use 350 horsepower during those 5 seconds. However, if
you then level off at 60 and drive that speed for an hour, your average
over the entire hour may only be something like 60 horsepower.

Power is energy per time. Depending on the extent of the time, you can
consider it "total power" or "average power" over some averaging time.
--
% Randy Yates % "The dreamer, the unwoken fool -
%% Fuquay-Varina, NC % in dreams, no pain will kiss the brow..."
%%% 919-577-9882 %
%%%% % 'Eldorado Overture', *Eldorado*, ELOhttp://www.digitalsignallabs.com

Here's the confusion more specifically. Adobe Audition Statistics
window has two different averaged values for the same selected audio
segment (i.e. the same time period). One they call "Average Power RMS"
which would seem straight forward except that they also have a "Total
Power RMS". This value tends to be slightly lower than the "Average
Power RMS" value. I'm wonder what the difference in the averaging is.

[Richard Crowley]

"apa" wrote ...
Here's the confusion more specifically. Adobe Audition Statistics
window has two different averaged values for the same selected audio
segment (i.e. the same time period). One they call "Average Power RMS"
which would seem straight forward except that they also have a "Total
Power RMS". This value tends to be slightly lower than the "Average
Power RMS" value. I'm wonder what the difference in the averaging is.

I would guess that the "Total" number is the RMS average over the entire
timeline. The question is what period the "Average" number is integrated
over?

[[email protected]]

X-no-archive:
On Jan 9, 5:29*pm, "Richard Crowley" wrote:
"apa" wrote ...

Here's the confusion more specifically. Adobe Audition Statistics
window has two different averaged values for the same selected audio
segment (i.e. the same time period). One they call "Average Power RMS"
which would seem straight forward except that they also have a "Total
Power RMS". This value tends to be slightly lower than the "Average
Power RMS" value. *I'm wonder what the difference in the averaging is..

I would guess that the "Total" number is the RMS average over the entire
timeline. The question is what period the "Average" number is integrated
over?

well first off the term "RMS power" is not really valid...

Irms * Vrms = Power average..

But lots of people use the term "RMS power" incorrectly when they
really mean average power.

And the total average power integrated over a time would be units of
energy... maybe thats what they mean..

i.e. if you burned a 60 Watt (power) lightbulb for 1 hour that would
be 60 Watthours of energy.

What are the units that they use for both terms?

And do some experiments, create a tone at -10 dBFS for 1 minute and
see what it reads. Then make it 1 minute one and 1 minute off in a 2
minute "song" and see what it reads.. That is probably the only way
you are going to be able to know for sure what it IS doing..



Mark

[Norbert Hahn[_2_]]

"Richard Crowley" wrote:

"apa" wrote ...
Here's the confusion more specifically. Adobe Audition Statistics
window has two different averaged values for the same selected audio
segment (i.e. the same time period). One they call "Average Power RMS"
which would seem straight forward except that they also have a "Total
Power RMS". This value tends to be slightly lower than the "Average
Power RMS" value. I'm wonder what the difference in the averaging is.

I would guess that the "Total" number is the RMS average over the entire
timeline. The question is what period the "Average" number is integrated
over?

Both values are always calculated over the highlighted part of the song.
In nothing is highlighted the total song is evaluated.
When you click "help" on the statistics windows and then goto "Waveform
Statistics options" you can read:
"Average RMS Power - Shows the average amplitude. This value reflects
perceived loudness."
"Total RMS Power - Represents the total power of the entire selection."

HTH
Norbert

[Randy Yates]

Norbert Hahn writes:

"Richard Crowley" wrote:

"apa" wrote ...
Here's the confusion more specifically. Adobe Audition Statistics
window has two different averaged values for the same selected audio
segment (i.e. the same time period). One they call "Average Power RMS"
which would seem straight forward except that they also have a "Total
Power RMS". This value tends to be slightly lower than the "Average
Power RMS" value. I'm wonder what the difference in the averaging is.

I would guess that the "Total" number is the RMS average over the entire
timeline. The question is what period the "Average" number is integrated
over?

Both values are always calculated over the highlighted part of the song.
In nothing is highlighted the total song is evaluated.
When you click "help" on the statistics windows and then goto "Waveform
Statistics options" you can read:
"Average RMS Power - Shows the average amplitude. This value reflects
perceived loudness."
"Total RMS Power - Represents the total power of the entire
selection."

Norbert,

What do you mean by "average amplitude?"

Power is joules / second. volts are joules / coulomb. amps are coulombs
/ second. v * i = joules / second, or power. If you integrate v * i over
time, you get the total joules over that time. If you then divide the
total joules by the time period T, you get power back again.

--Randy

PS: I'm assuming that you mean "Average Power" and "Total Power," i.e.,
that the "RMS" is a misnomer. Volts can be RMS, not power.
--
% Randy Yates % "Maybe one day I'll feel her cold embrace,
%% Fuquay-Varina, NC % and kiss her interface,
%%% 919-577-9882 % til then, I'll leave her alone."
%%%% % 'Yours Truly, 2095', *Time*, ELO
http://www.digitalsignallabs.com

[Norbert Hahn[_2_]]

wrote:

X-no-archive:
On Jan 9, 5:29*pm, "Richard Crowley" wrote:
"apa" wrote ...

Here's the confusion more specifically. Adobe Audition Statistics
window has two different averaged values for the same selected audio
segment (i.e. the same time period). One they call "Average Power RMS"
which would seem straight forward except that they also have a "Total
Power RMS". This value tends to be slightly lower than the "Average
Power RMS" value. *I'm wonder what the difference in the averaging is.

I would guess that the "Total" number is the RMS average over the entire
timeline. The question is what period the "Average" number is integrated
over?

well first off the term "RMS power" is not really valid...

Irms * Vrms = Power average..

That's right in the real world. In software dealing with samples most
likely coming from a mic that converts air pressure changes into
voltage...

You may change the windows size that Auditions uses for calculation
the average values. It defaults to 10 ms - at least in my version 1.5.
One of the advantages of Audition is its online help - every parameter
which can be configured is explaned along with its side effects. You
may press F1 any time and the html browser comes up and shows help.

And do some experiments, create a tone at -10 dBFS for 1 minute and
see what it reads. Then make it 1 minute one and 1 minute off in a 2
minute "song" and see what it reads.. That is probably the only way
you are going to be able to know for sure what it IS doing..

That's always a good idea.

Norbert

[Randy Yates]

writes:
[...]
And the total average power integrated over a time would be units of
energy...

Hi Mark,

That doesn't make sense. You don't integrate total average power - it's
already integrated, i.e., it's just a constant.

Perhaps what you meant was "the instantaneous power integrated over a
time would be units of energy," to which I would agree.
--
% Randy Yates % "Ticket to the moon, flight leaves here today
%% Fuquay-Varina, NC % from Satellite 2"
%%% 919-577-9882 % 'Ticket To The Moon'
%%%% % *Time*, Electric Light Orchestra
http://www.digitalsignallabs.com

[Don Pearce[_2_]]

On Fri, 09 Jan 2009 15:53:16 -0500, Randy Yates
wrote:

apa writes:

Can anyone explain the difference between these two readings in Adobe
Audition's Amplitude Statistics window?
Thanks, Andy

If you gun your engine at 10:00:00am for 5 seconds to go from 0 to 60
MPH, you might use 350 horsepower during those 5 seconds. However, if
you then level off at 60 and drive that speed for an hour, your average
over the entire hour may only be something like 60 horsepower.

Power is energy per time. Depending on the extent of the time, you can
consider it "total power" or "average power" over some averaging time.

I did an analysis and explanation of this phenomenon a couple of years
ago. It actually deals with the FFT window in Audition, but it applies
equally to the statistics window, which is based on the FFT.

I think it explains what is going on.

http://81.174.169.10/odds/fftoddity

d

[Mike Rivers]

On Jan 9, 12:36 pm, apa wrote:
Can anyone explain the difference between these two readings in Adobe
Audition's Amplitude Statistics window?

I think the whole concept is bogus, the kind of thing that a
programmer might imagine is useful to an audio engineer. There are no
amps or watts in digital audio (it's only voltage levels that are
digitized) so the program has no clue what the power of anything is,
at least not without some calibration.

I suppose that with a calibrated microphone that puts out a given
voltage proportional to acoustical power it might be meaningful, but
then the program would have to know something about the microphone and
recording chain, which it doesn't.

These RMS "statistics" are just numbers. If you find them helpful in
deciding something that you can't decide by listening, like if one
tune is perceived as louder, softer, or about the same as the adjacent
tune in collection, then by all means make whatever use you can of it.
Otherwise, don't fret about it.

[[email protected]]

On Jan 9, 7:47*pm, Randy Yates wrote:
writes:
[...]
And the total average power integrated over a time would be units of
energy... *

Hi Mark,

That doesn't make sense. You don't integrate total average power - it's
already integrated, i.e., it's just a constant.

Perhaps what you meant was "the instantaneous power integrated over a
time would be units of energy," to which I would agree.
--
% *Randy Yates * * * * * * * * *% "Ticket to the moon, flight leaves here today

Yep, you are correct...
thanks

Mark

[Peter Larsen[_3_]]

Richard Crowley wrote:

"apa" wrote ...

Here's the confusion more specifically. Adobe Audition Statistics
window has two different averaged values for the same selected audio
segment (i.e. the same time period). One they call "Average Power
RMS" which would seem straight forward except that they also have a
"Total Power RMS". This value tends to be slightly lower than the
"Average Power RMS" value. I'm wonder what the difference in the
averaging is.

I would guess that the "Total" number is the RMS average over the
entire timeline. The question is what period the "Average" number is
integrated over?

Assuming some slight amount of logic in the wordings Total RMS is that, ie.
the entire piece. Averaging is done with a time window with adjustable
width. Default is 50 ms, I always use 300 ms. My vague perception is that
the "average" is then the average of the sliding average.

My choice when aligning files numerically is to always use the average, and
it is very rare that it needs adjusting because it "sounds wrong". What
could void it would be a file with either repetive close transients or a
lack of bass. It is a bit more work than using the automated process, but it
seems to be less work overall because less after-adjustments or re-doings
are likely to be needed.

Kind regards

Peter Larsen

[Peter Larsen[_3_]]

Randy Yates wrote:
Norbert Hahn writes:

"Richard Crowley" wrote:

"apa" wrote ...
Here's the confusion more specifically. Adobe Audition Statistics
window has two different averaged values for the same selected
audio segment (i.e. the same time period). One they call "Average
Power RMS" which would seem straight forward except that they also
have a "Total Power RMS". This value tends to be slightly lower
than the "Average Power RMS" value. I'm wonder what the
difference in the averaging is.

I would guess that the "Total" number is the RMS average over the
entire timeline. The question is what period the "Average" number
is integrated over?

Both values are always calculated over the highlighted part of the
song. In nothing is highlighted the total song is evaluated.
When you click "help" on the statistics windows and then goto
"Waveform Statistics options" you can read:
"Average RMS Power - Shows the average amplitude. This value reflects
perceived loudness."
"Total RMS Power - Represents the total power of the entire
selection."

Norbert,

What do you mean by "average amplitude?"

Power is joules / second. volts are joules / coulomb. amps are
coulombs / second. v * i = joules / second, or power. If you
integrate v * i over time, you get the total joules over that time.
If you then divide the total joules by the time period T, you get
power back again.

--Randy

PS: I'm assuming that you mean "Average Power" and "Total Power,"
i.e., that the "RMS" is a misnomer. Volts can be RMS, not power.

[Peter Larsen[_3_]]

Randy Yates wrote:

Norbert,

What do you mean by "average amplitude?"

Power is joules / second. volts are joules / coulomb. amps are
coulombs / second. v * i = joules / second, or power. If you
integrate v * i over time, you get the total joules over that time.
If you then divide the total joules by the time period T, you get
power back again.

Randy, this is about what a specific software program does, ie. about what
the designer of it means.

--Randy

Kind regards

Peter Larsen

[Peter Larsen[_3_]]

Mike Rivers wrote:

On Jan 9, 12:36 pm, apa wrote:

Can anyone explain the difference between these two readings in Adobe
Audition's Amplitude Statistics window?

I think the whole concept is bogus, the kind of thing that a
programmer might imagine is useful to an audio engineer.

And it happens to so be, bogus or not, it works and is certainly useful for
getting channel balance right to the hundreth of a dB. You may not be able
to hear so small differences without practice .... O;-) ... but it keeps the
boring brain half happy.


Kind regards

Peter Larsen

[William Sommerwerck]

I've waited to chime in on this...

An RMS measurement manipulates data in a particular way. (It's the square
root of the average of the sum of the squares.) It stands alone. The idea of
a "total" or "average" RMS measurement strikes me as odd, and likely
meaningless.

The only valid explanation is going to come from the software's publisher.

[[email protected]]

x-no-archive:

On Jan 13, 9:00*am, "William Sommerwerck"
wrote:
I've waited to chime in on this...

An RMS measurement manipulates data in a particular way. (It's the square
root of the average of the sum of the squares.) It stands alone. The idea of
a "total" or "average" RMS measurement strikes me as odd, and likely
meaningless.

The only valid explanation is going to come from the software's publisher..

that's one way to get an answer

the other way is to run some tests on the software with some defined
signals to see what it does...

Mark

[Randy Yates]

"Peter Larsen" writes:

Randy Yates wrote:

Norbert,

What do you mean by "average amplitude?"

Power is joules / second. volts are joules / coulomb. amps are
coulombs / second. v * i = joules / second, or power. If you
integrate v * i over time, you get the total joules over that time.
If you then divide the total joules by the time period T, you get
power back again.

Randy, this is about what a specific software program does, ie. about what
the designer of it means.

Agreed, and the thread you interrupted is about the operations the
designer might have implemented, given the common meanings of power and
averaging.
--
% Randy Yates % "Though you ride on the wheels of tomorrow,
%% Fuquay-Varina, NC % you still wander the fields of your
%%% 919-577-9882 % sorrow."
%%%% % '21st Century Man', *Time*, ELO
http://www.digitalsignallabs.com

[Peter Larsen[_3_]]

Randy Yates wrote:

Agreed, and the thread you interrupted is about the operations the
designer might have implemented, given the common meanings of power
and averaging.

I also *did* comment on those Randy ... O;-) ... this software-designer has
more of a musical or musicians approach to this aspect of it than a
technical one.

Kind regards

Peter Larsen

[Norbert Hahn[_2_]]

"William Sommerwerck" wrote:

I've waited to chime in on this...

An RMS measurement manipulates data in a particular way. (It's the square
root of the average of the sum of the squares.) It stands alone.

I compared the results of RMS values obtained with Audition 1.5 to
those obtained with SoX (http://sox.sourceforge.net/) and found that
Audition's "Total RMS power" equals "Scaled by rms" returned by SoX
after translating the fractional number to dB. Example of calling SoX:

C:\Program Files\Util\SoX\sox "Song201.wav" -n stat -rms

The idea of
a "total" or "average" RMS measurement strikes me as odd, and likely
meaningless.

According to the help file Audition's "Average RMS power" tries to
resample the perceived loudness it *should* apply some weighing factor
because low and high frequencies give a different loudness than mid
range frequencies do. Mid range contributes most to perceived
loudness. I checked this assumption with 44 Hz, 440 Hz, and 14000 Hz
sine wave @ -6 dB and always got the same values for both Average RMS
power and Total RMS power - unfortunatley.

Wavegain, however, does filter the files to approximate the perceived
loudness:
Analyzing...

Gain | Peak | Scale | New Peak | Track
--------------------------------------------
+6.00 dB | 16422 | 2.00 | 32765 | C:\temp\44.wav
-8.81 dB | 16422 | 0.36 | 5956 | C:\temp\440.wav
+6.00 dB | 16415 | 2.00 | 32765 | C:\temp\14000.wav

WaveGain Processing completed normally

MP3gain is said to do the same processing.

Norbert

[Peter Larsen[_3_]]

Norbert Hahn wrote:

According to the help file Audition's "Average RMS power" tries to
resample the perceived loudness it *should* apply some weighing factor
because low and high frequencies give a different loudness than mid
range frequencies do.

That concept per se is as broken as the concept of non-adjustable loudness
in an amplifier because playback level is unknown.-

Mid range contributes most to perceived
loudness.

The helpfile may quite possibly be ill applied or broken, there is a
frequency compensation mode for the averaging in the group normalize tool,
but not in the edit view statistics tool. The group normalize statistics
seem to also make some other asumptions since the results seemed to differ
from the statistics tool in the edit view window last time I was there. I
have some average loudness guidelines that work well for me as "house
standard" and they are based on the statistics tool in the edit view, ain't
seen no reason to use the group normalize tool since it appears to me to
make some undisclosed asumptions and weightings.

Kind regards

Peter Larsen

I checked this assumption with 44 Hz, 440 Hz, and 14000 Hz
sine wave @ -6 dB and always got the same values for both Average RMS
power and Total RMS power - unfortunatley.

Wavegain, however, does filter the files to approximate the perceived
loudness:
Analyzing...

Gain | Peak | Scale | New Peak | Track
--------------------------------------------
+6.00 dB | 16422 | 2.00 | 32765 | C:\temp\44.wav
-8.81 dB | 16422 | 0.36 | 5956 | C:\temp\440.wav
+6.00 dB | 16415 | 2.00 | 32765 | C:\temp\14000.wav

WaveGain Processing completed normally

MP3gain is said to do the same processing.

Norbert