Spectrum Analysis Of Contemporary Rap Songs in Audacity

[ChrisCoaster]

Anyone possessing the current version of Audacity has the capacity to
analyze the "energy distribution" of any song loaded into this
editor. It will average the content of up to a 23 sec sample of the
song. It will appear as a profile, from left to right, of all
frequencies of sound in a recording, with sub-bass on the left, bass,
mid, upper mid in the middle, and highs to the right.

I loaded and analyzed several hip-hop tracks from the last 12 years,
including "So Into You"(Fabolous), "6 Foot 7 Foot"(Lil' Wayne), and
"Start It Up"(Lloyd Banks). One of them was a mp3, the other two
ripped from CD.

The spectra returned from all these songs all show a very pronounced
LF hump centered between 50-70 Hertz, about 15-20dB higher than the
rest of the spectrograph! The 256kbps mp3 showed a slight roll off at
the high end - over 12kHz - compared to the CD rips, but since my
hearing also drops off above that point I couldn't hear the difference
if I wanted too!

In Audacity, I utilized the EQ plug-in that comes with it to "scoop"
out a complimentary curve from about 30Hz up to 150Hz, and re-ran the
analysis over the exact same 23 seconds, and the hump was considerably
flattened compared to before, now only about 10dB higher than the rest
of the curve.

I also found I was able to apply the Amplify plug in and raise the
overall volume of the song higher than it had been ripped at(!),
because of the headroom I gained by scooping out some of those
excessive lows.

How did it sound?

Well, the first impression was that the song had more vitality, more
rhythmic "snap" to it. The highs were clearer, even on the mp3. The
vocals were more FOCUSED - and I could clearly understand without a
doubt what the rappers were saying! Compared to the original
production, it now sounded as though Lil' Wayne or Jay-Z were standing
in my living room!

I then applied some low EQ in Windows media Player, and the bottom
energy was once again there, but without the muddiness of the track as
produced.

My question to the pros on R.A.P: I understand why certain genres of
music are "bassed up" the way they are, but is it accurate, sound, and
proper engineering practice to do this?

Do the recording or post-engineers realize they could be causing
damage to consumer systems lacking the power to handle these boosted
up lows? And isn't it better to produce a relatively balanced
recording spectrally speaking, and to let the end user/consumer add
bass to their preference?

I already know how I feel on this. How about you?

-CC

[Scott Dorsey]

ChrisCoaster > wrote:
>
>My question to the pros on R.A.P: I understand why certain genres of
>music are "bassed up" the way they are, but is it accurate, sound, and
>proper engineering practice to do this?

It's not accurate, but it's what the customer wants. It's interesting that
you're seeing a peak maybe an octave lower than I would have expected. Most
rock stuff has exaggerated kick but somewhat higher, so that it can be heard
on cheap speakers without much low end.

>Do the recording or post-engineers realize they could be causing
>damage to consumer systems lacking the power to handle these boosted
>up lows? And isn't it better to produce a relatively balanced
>recording spectrally speaking, and to let the end user/consumer add
>bass to their preference?

The customer is already playing those recordings with the pumped-up bass on
equipment with the bass turned up, the loudness contour turned on, and with
narrowband "subwoofers" tuned somewhere in the 70-100 Hz region to make a big
thump.

>I already know how I feel on this. How about you?

I can't stand it, but more importantly I don't really understand it. I
still cut LPs, but I won't cut rap and hip hop stuff because I don't have a
good feel for why it's supposed to sound like that and therefore how it really
should sound.

But it's the current fashion, a lot of people like it, and sooner or later it
will go away to be replaced with something else.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[PStamler]

On Apr 12, 10:16*am, (Scott Dorsey) wrote:
> ChrisCoaster > wrote:
>
> >My question to the pros on R.A.P: *I understand why certain genres of
> >music are "bassed up" the way they are, but is it accurate, sound, and
> >proper engineering practice to do this?
>
> It's not accurate, but it's what the customer wants. *It's interesting that
> you're seeing a peak maybe an octave lower than I would have expected. *Most
> rock stuff has exaggerated kick but somewhat higher, so that it can be heard
> on cheap speakers without much low end.
>
> >Do the recording or post-engineers realize they could be causing
> >damage to consumer systems lacking the power to handle these boosted
> >up lows? * And isn't it better to produce a relatively balanced
> >recording spectrally speaking, and to let the end user/consumer add
> >bass to their preference?
>
> The customer is already playing those recordings with the pumped-up bass on
> equipment with the bass turned up, the loudness contour turned on, and with
> narrowband "subwoofers" tuned somewhere in the 70-100 Hz region to make a big
> thump.
>
> >I already know how I feel on this. *How about you?
>
> I can't stand it, but more importantly I don't really understand it. *I
> still cut LPs, but I won't cut rap and hip hop stuff because I don't have a
> good feel for why it's supposed to sound like that and therefore how it really
> should sound. *
>
> But it's the current fashion, a lot of people like it, and sooner or later it
> will go away to be replaced with something else.

I'd add that the songs are mixed this way, to a great extent, for the
effect they have at a dance club, rather than in home listening or
earbuds.

Peace,
Paul

[alex]

Il 12/04/2011 1.50, ChrisCoaster ha scritto:
> My question to the pros on R.A.P: I understand why certain genres of
> music are "bassed up" the way they are, but is it accurate, sound, and
> proper engineering practice to do this?
>
> Do the recording or post-engineers realize they could be causing
> damage to consumer systems lacking the power to handle these boosted
> up lows? And isn't it better to produce a relatively balanced
> recording spectrally speaking, and to let the end user/consumer add
> bass to their preference?
>
> I already know how I feel on this. How about you?

wow RAP on R.A.P.... :-)
Measuring that way doesn't figure out how really the energy is
distributed. The only things you know is that the "bass" peak is the
most common frequency used across the whole 23 sec sample, so
statistically will show you a peak. Sadly this says almost nothing about
the frequency balance of the song.
Bassy sounding music doesn't really always mean that the very low
frequency are boosted, harmonics help the perception and the same will
do some kind of ambience. Frequencies around 40Hz are not well
reproduced by most audio consumer systems which attenuate the
reproduction not causing damage to speaker unless you drive the amp into
clipping, which is dangerous. Amplifiers and speakers are designed to
handle a very big dynamic range without problems. The bad habits to
overlimit CDs in order to obtain "louder" sounding songs can be
dangerous because the average level (most critical for the system) is
much higher, with respect to peaks, leading to a more "stressed" system,
prone to distortion and drivers overheating.

[ChrisCoaster]

On Apr 12, 2:03*pm, alex > wrote:
> Il 12/04/2011 1.50, ChrisCoaster ha scritto:
>
> > My question to the pros on R.A.P: *I understand why certain genres of
> > music are "bassed up" the way they are, but is it accurate, sound, and
> > proper engineering practice to do this?
>
> > Do the recording or post-engineers realize they could be causing
> > damage to consumer systems lacking the power to handle these boosted
> > up lows? * And isn't it better to produce a relatively balanced
> > recording spectrally speaking, and to let the end user/consumer add
> > bass to their preference?
>
> > I already know how I feel on this. *How about you?
>
> wow RAP on R.A.P.... :-)
> Measuring that way doesn't figure out how really the energy is
> distributed. The only things you know is that the "bass" peak is the
> most common frequency used across the whole 23 sec sample, so
> statistically will show you a peak. Sadly this says almost nothing about
> the frequency balance of the song.
> Bassy sounding music doesn't really always mean that the very low
> frequency are boosted, harmonics help the perception and the same will
> do some kind of ambience. Frequencies around 40Hz are not well
> reproduced by most audio consumer systems which attenuate the
> reproduction not causing damage to speaker unless you drive the amp into
> clipping, which is dangerous. Amplifiers and speakers are designed to
> handle a very big dynamic range without problems. The bad habits to
> overlimit CDs in order to obtain "louder" sounding songs can be
> dangerous because the average level (most critical for the system) is
> much higher, with respect to peaks, leading to a more "stressed" system,
> prone to distortion and drivers overheating.
_____________________

Well, I'll put this way: Nearly any post-1995 rap song I've analyzed
has a MOUNTAIN of lows centered around the area I mentioned, and this
dome peters out only above 300Hz! Down at 20Hz it's still up +5dB.

Analyzing R&B from the 80s on back and this hump is far less obvious.
Analyzing ANYTHING from the 70's and back and the spectrum shows up
far more balanced than the artists I mentioned in my openng post.

As far as what PStamler said about mixing for clubs, I'd rather post-
produce spectrally flat, and let the club eq their system to market
the type of sound that will fill their dance floor and move drinks
across their bar.

-CC

[Scott Dorsey]

ChrisCoaster > wrote:
>As far as what PStamler said about mixing for clubs, I'd rather post-
>produce spectrally flat, and let the club eq their system to market
>the type of sound that will fill their dance floor and move drinks
>across their bar.

"I must have maximum bass at all frequencies."
-- Lee "Scratch" Perry

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Peter Larsen[_3_]]

ChrisCoaster wrote:

> Anyone possessing the current version of Audacity has the capacity to
> analyze the "energy distribution" of any song loaded into this
> editor. It will average the content of up to a 23 sec sample of the
> song. It will appear as a profile, from left to right, of all
> frequencies of sound in a recording, with sub-bass on the left, bass,
> mid, upper mid in the middle, and highs to the right.

Even old Cool Edit 96 did it better in as much as it could average over the
length of any file it could load. It is however important for such an
analysis to make sense to come to grips with what constitutes "linear", pink
noise is not linear compared to real world acoustic sound sources and white
noise most certainly isn't. My finding is that the best approach is to add a
suitable eq prior to analyzing with a display that displays pink or white as
linear.

Kind regards

Peter Larsen

[ChrisCoaster]

On Apr 21, 4:59*pm, "Peter Larsen" > wrote:
> ChrisCoaster wrote:
> > Anyone possessing the current version of Audacity has the capacity to
> > analyze the "energy distribution" of any song loaded into this
> > editor. *It will average the content of up to a 23 sec sample of the
> > song. *It will appear as a profile, from left to right, of all
> > frequencies of sound in a recording, with sub-bass on the left, bass,
> > mid, upper mid in the middle, and highs to the right.
>
> Even old Cool Edit 96 did it better in as much as it could average over the
> length of any file it could load. It is however important for such an
> analysis to make sense to come to grips with what constitutes "linear", pink
> noise is not linear compared to real world acoustic sound sources and white
> noise most certainly isn't. My finding is that the best approach is to add a
> suitable eq prior to analyzing with a display that displays pink or white as
> linear.
>
> * *Kind regards
>
> * *Peter Larsen
__________________
I honestly do not believe Audacity is LYING. The sub-100Hz hump is
there! I saw it in Cool Edit in real time(as the song was playing)
and in Audacity as an average of a 23sec sample snapshot. And
scooping out a complimentary amount of LF content with the EQ app in
either yielded postive results - the song just sounded better. This
does not undermine what you are saying - I'm just reporting my (visual
and aural) observations.

-CC

[Scott Dorsey]

ChrisCoaster > wrote:
>I honestly do not believe Audacity is LYING. The sub-100Hz hump is
>there! I saw it in Cool Edit in real time(as the song was playing)
>and in Audacity as an average of a 23sec sample snapshot. And
>scooping out a complimentary amount of LF content with the EQ app in
>either yielded postive results - the song just sounded better. This
>does not undermine what you are saying - I'm just reporting my (visual
>and aural) observations.

It's not lying, but if you change the window configuration on the FFT it
will look different. The bottom end on an FFT is always problematic.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Mark]

> It's not lying, but if you change the window configuration on the FFT it
> will look different. *The bottom end on an FFT is always problematic.
> --scott
> --
> "C'est un Nagra. C'est suisse, et tres, tres precis."

NTrack Studio also has a nice FFT spectrum analysis feature that
overlays the EQ settings visually.

Mark

[Peter Larsen[_3_]]

ChrisCoaster wrote:

[audacity unable to analyze more than 23 seconds of audio]

> I honestly do not believe Audacity is LYING.

Look here, if you want to analyze an entire CD then do it.

> The sub-100Hz hump is there!

No contest, but when you analyze only 23 seconds rather than the entire CD
it is by your assertion that you have selected 23 representative seconds,
when you analyze the lot then that is how it is. Google my old posts - I
wrote quite much about this some years ago. Could have added to it because I
have gotten wiser since, but didn't because it didn't get published and
because some writer in some context appeared to be just slightly possibly
paraphrasing some other of the same subject matter without credit or
reference.

> -CC

Kind regards

Peter Larsen

[ChrisCoaster]

On Apr 22, 9:32*am, (Scott Dorsey) wrote:
> ChrisCoaster > wrote:
> >I honestly do not believe Audacity is LYING. *The sub-100Hz hump is
> >there! *I saw it in Cool Edit in real time(as the song was playing)
> >and in Audacity as an average of a 23sec sample snapshot. *And
> >scooping out a complimentary amount of LF content *with the EQ app in
> >either yielded postive results - the song just sounded better. *This
> >does not undermine what you are saying - I'm just reporting my (visual
> >and aural) observations.
>
> It's not lying, but if you change the window configuration on the FFT it
> will look different. *The bottom end on an FFT is always problematic.
> --scott
> --
> "C'est un Nagra. C'est suisse, et tres, tres precis."
_______________
FFT? Remember, I'm "acronym illiterate"!
I was 20 years old before I realized what FBI and NASCAR stood
for. :)
-CC

[Scott Dorsey]

ChrisCoaster > wrote:
>On Apr 22, 9:32=A0am, (Scott Dorsey) wrote:
>> ChrisCoaster > wrote:
>> >I honestly do not believe Audacity is LYING. =A0The sub-100Hz hump is
>> >there! =A0I saw it in Cool Edit in real time(as the song was playing)
>> >and in Audacity as an average of a 23sec sample snapshot. =A0And
>> >scooping out a complimentary amount of LF content =A0with the EQ app in
>> >either yielded postive results - the song just sounded better. =A0This
>> >does not undermine what you are saying - I'm just reporting my (visual
>> >and aural) observations.
>>
>> It's not lying, but if you change the window configuration on the FFT it
>> will look different. =A0The bottom end on an FFT is always problematic.
>
>FFT? Remember, I'm "acronym illiterate"!
>I was 20 years old before I realized what FBI and NASCAR stood
>for. :)

The Cooley-Tukey Fast Fourier Transform algorithm is what is being used to
turn that short sample into the frequency domain. FFTs are very powerful,
but they are also very flexible and they have to be configured properly for
the material you're looking at.

The FFT basically looks across the dataset, makes up a set of different "bins"
for different frequencies, and then decides how much of the original signal
gets dropped into the individual bins. Since the bins are a constant number
of hertz wide, the low frequency bins have a much larger portion of an octave
than the upper ones. This means if you care about low frequencies in an FFT
display you MUST set a very large number of bins. It also means you have to
be very careful about the window configuration (which defines the borders
between the bins and how you handle marginal cases).

The FFT data looks nice but it's not always as meaningful as you might expect
and by fiddling with the FFT configuration you can get all kinds of weird
artifacts in the bottom octaves.

My guess is that the hump you're seeing is probably a little bit higher than
you think it is, and that some of the stuff way down at the bottom is a
measurement artifact, but play with the settings and see for yourself. You
should be able to add bins without the shape of the curve changing much...
if it does, keep adding more until it stops.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Mark]

*
>
> The FFT data looks nice but it's not always as meaningful as you might expect
> and by fiddling with the FFT configuration you can get all kinds of weird
> artifacts in the bottom octaves.
>
>
I use a 4096 FFT with Blackman window and a lot of material seems to
on average, have an overall flat power curve below 2 kHz or so and a
decrease above that as we all know most music will do.

I think looking at an analyzer is like getting used to a set of
speakers, after a while listening/looking at a lot of material, you
get to know what typical material looks/sounds like and when something
comes along that deviates from that, you can see and hear it.

I will the SA as a guide for setting levels of bass while mixing, I'll
set it so its flat with the rest of the spectrum, just like most other
pre-recorded material looks.

(I'm not a pro at this and don't pretend to be one)

I don't know about rap music, i don't listen to it.

Mark

[rakman]

On Apr 12, 12:50*am, ChrisCoaster > wrote:

> My question to the pros on R.A.P: *I understand why certain genres of
> music are "bassed up" the way they are, but is it accurate, sound, and
> proper engineering practice to do this?

It takes skill to get the sub bass sounding heavy, loud,
tight and controlled whilst keeping the entire mix loud
and clear, so it's absolutely "proper engineering practice"
to do so.

[ChrisCoaster]

On Apr 22, 10:43*am, (Scott Dorsey) wrote:

>
> My guess is that the hump you're seeing is probably a little bit higher than
> you think it is, and that some of the stuff way down at the bottom is a
> measurement artifact, but play with the settings and see for yourself. *You
> should be able to add bins without the shape of the curve changing much....
> if it does, keep adding more until it stops.
> --scott
>
> --
> "C'est un Nagra. C'est suisse, et tres, tres precis."- Hide quoted text -
>
> - Show quoted text -
______________________

Hump higher in amplitude or frequency? Frequncy wise the peak is no
lower than 50Hz, and is typically between 60-80. I have seen a
gradual uptick down to 10-20Hz and just throw that away in my mind.
LOL! But it's definitely there and at the frequencies I just
mentioned.

I'm beginning to feel like I'm being made out to be a nut here, but
that's fine - I use my own eyes and ears to make my observation.
Nothing like good old empirical observation! No one here is obliged
to believe me.

-CC

[ChrisCoaster]

On Apr 22, 6:59*pm, rakman > wrote:

> It takes skill to get the sub bass sounding heavy, loud,
> tight and controlled whilst keeping the entire mix loud
> and clear, so it's absolutely "proper engineering practice"
> to do so.
_______________
Well well, we have here a man of the cloth, so to speak. Whazzup my
man? Some of the bottom on these tracks makes my speakers sound like
they're farting, yo!

-CC

[rakman]

On Apr 23, 3:10*am, ChrisCoaster > wrote:
> On Apr 22, 6:59*pm, rakman > wrote:
>
> > It takes skill to get the sub bass sounding heavy, loud,
> > tight and controlled whilst keeping the entire mix loud
> > and clear, so it's absolutely "proper engineering practice"
> > to do so.
>
> _______________
> Well well, we have here a man of the cloth, so to speak. *Whazzup my
> man? *Some of the bottom on these tracks makes my speakers sound like
> they're farting, yo!
>
> -CC

Yeah.
Bangladesh's 808 on 6 foot 7 foot sounds like it's gone
through an overdrive plugin, ie deliberate distortion.
(UK DnB producers have been doing this for years).
But it would have been possible to say... distort only the
top end and keep the bass end cleaner.

I think some electro house tracks have the bass going
through a multiband distortion plugin, making it fill
the entire frequency spectrum. Like the Kesha
Cirkut remix.

A low sub without any top end definition is hard to hear
on normal systems, so the it's common to have a short
click sound in a higher frequency range for attack and
definition. And/or layer two kicks one sub and one
high end attack kick.

The sparser the track the easier it is to fit in
a big 808. If you have a fully arranged track
with lots of instruments there just might not
be enough room for it.

It's also easier to fit in a big long 808 on a slow tempo
track like dirty south, dubstep, slowjam etc,
than it is on a higher tempo house/electro track.

One faster track that has a mega powerful and
fairly clean snappy 808 is boom boom pow but it's a
sparse track with very few instruments and they're
switching between/layering at least 3-4 kick sounds, the
sub 808 with relatively long decay, a snappy shorter 808,
a "knock" electronic kick with 600Hz boost and a short
high freq kick as well.

[Peter Larsen[_3_]]

ChrisCoaster wrote:

> On Apr 22, 6:59 pm, rakman > wrote:

>> It takes skill to get the sub bass sounding heavy, loud,
>> tight and controlled whilst keeping the entire mix loud
>> and clear, so it's absolutely "proper engineering practice"
>> to do so.

> Well well, we have here a man of the cloth, so to speak. Whazzup my
> man? Some of the bottom on these tracks makes my speakers sound like
> they're farting, yo!

Yo, man, yo, yo speakers are wrong, yo. Get some dual 12" in large tubes for
truck use, put them in the corners and be happy.

> -CC

Kind regards

Peter Larsen

[Mike Rivers]

On 4/22/2011 10:43 AM, Scott Dorsey wrote:

> The FFT basically looks across the dataset, makes up a set of different "bins"
> for different frequencies, and then decides how much of the original signal
> gets dropped into the individual bins. Since the bins are a constant number
> of hertz wide, the low frequency bins have a much larger portion of an octave
> than the upper ones. This means if you care about low frequencies in an FFT
> display you MUST set a very large number of bins. It also means you have to
> be very careful about the window configuration (which defines the borders
> between the bins and how you handle marginal cases).

I was playing around with various spectrum display tools the
other day. I have a Focusrite VRM Box here for review, a
gadget that's supposed to let you hear, over headphones,
what your mix would sound like through different speakers in
assorted listening environments - kind of like a modern
digital/headphone counterpart to Auratones. I hear obvious
differences in frequency balance, but mostly I hear room
reverberation and problems caused by rooms, not speakers.

I was looking for a way to show this graphically, hoping
that by playing pink noise through the box and looking at
the spectrum coming out I could somehow correlate what I'm
hearing with a plot. I tried fiddling around with number of
bins and smoothing algorithms, I even tried a waterfall
plot, and none of it seemed to make sense to me. I guess
this is at least partially why acoustic designers use FFT
analysis to prove out their work, not to tell them what they
need to do.



--
"Today's production equipment is IT based and cannot be
operated without a passing knowledge of computing, although
it seems that it can be operated without a passing knowledge
of audio." - John Watkinson

http://mikeriversaudio.wordpress.com - useful and
interesting audio stuff

[ChrisCoaster]

On Apr 23, 6:25*am, "Peter Larsen" > wrote:
> ChrisCoaster wrote:
> > On Apr 22, 6:59 pm, rakman > wrote:
> >> It takes skill to get the sub bass sounding heavy, loud,
> >> tight and controlled whilst keeping the entire mix loud
> >> and clear, so it's absolutely "proper engineering practice"
> >> to do so.
> > Well well, we have here a man of the cloth, so to speak. *Whazzup my
> > man? *Some of the bottom on these tracks makes my speakers sound like
> > they're farting, yo!
>
> Yo, man, yo, yo speakers are wrong, yo. Get some dual 12" in large tubes for
> truck use, put them in the corners and be happy.
>
> > -CC
>
> * Kind regards
>
> * Peter Larsen
_______________
LOL mah man!!

[Scott Dorsey]

rakman > wrote:
>
>A low sub without any top end definition is hard to hear
>on normal systems, so the it's common to have a short
>click sound in a higher frequency range for attack and
>definition. And/or layer two kicks one sub and one
>high end attack kick.

This is something that folks have known since the early sixties. Add
distortion and the kick and electric bass will come across on a 3"
transistor radio speaker.

>The sparser the track the easier it is to fit in
>a big 808. If you have a fully arranged track
>with lots of instruments there just might not
>be enough room for it.

And, when you add a lot of harmonics, you get a track that takes up
a whole lot of space in the mix, just like that. That's the bad end
of the deal.

>One faster track that has a mega powerful and
>fairly clean snappy 808 is boom boom pow but it's a
>sparse track with very few instruments and they're
>switching between/layering at least 3-4 kick sounds, the
>sub 808 with relatively long decay, a snappy shorter 808,
>a "knock" electronic kick with 600Hz boost and a short
>high freq kick as well.

Bingo.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[High Plains Thumper[_2_]]

Mark wrote:
> Scott Dorsey wrote:
>
>> The FFT data looks nice but it's not always as meaningful as you might
>> expect and by fiddling with the FFT configuration you can get all kinds
>> of weird artifacts in the bottom octaves.
>
> I use a 4096 FFT with Blackman window and a lot of material seems to on
> average, have an overall flat power curve below 2 kHz or so and a
> decrease above that as we all know most music will do.
>
> I think looking at an analyzer is like getting used to a set of
> speakers, after a while listening/looking at a lot of material, you get
> to know what typical material looks/sounds like and when something comes
> along that deviates from that, you can see and hear it.
>
> I will the SA as a guide for setting levels of bass while mixing, I'll
> set it so its flat with the rest of the spectrum, just like most other
> pre-recorded material looks.
>
> (I'm not a pro at this and don't pretend to be one)
>
> I don't know about rap music, i don't listen to it.

I enjoyed reading Scott's explanation of the use of the Fast Fourier
Transform (FFT) series and applications to audio.

[OT] I am familiar with that series, having taken advanced numerical
(computer) methods and also did a directed research project ME-499 on
using the FFT for post processing real time data 30 years ago. These were
model rocket thrust curves, considerable amount of data gathered from
short duration firings under several seconds.

My research involved using the computer (in this case an HP-85) to
develop filters to remove noise from these data sets. That included
calculating Fourier coefficients for those filters.

I can see how they can apply to audio as data is data. FFT's can be used
to mathematically mimic just about any filter shape desired, including
band pass, notch, etc. The FFT is a power series, depending on the
"pureness" to form desired, can be taken to the nth degree. Practicalness
depends on the quantity of data and the power of the computer to do the
number crunching.

As with any tool, I imagine there are caveats and it has its place.

--
HPT

[hank alrich]

Richard Webb > wrote:

> Can you say snake oil? I knew you could.

I got a band by that name, sort of. Sierra Snake Oil.

But we're the real thing.

--
shut up and play your guitar * http://hankalrich.com/
http://www.youtube.com/watch?v=NpqXcV9DYAc
http://www.sonicbids.com/HankandShaidriAlrichwithDougHarman

[[email protected]]

On 2011-04-23 (hankalrich) said:
>Richard Webb >
>wrote:
>> Can you say snake oil? I knew you could.
>I got a band by that name, sort of. Sierra Snake Oil.
>But we're the real thing.
<rotfl> Dig it!




Richard webb,

replace anything before at with elspider
ON site audio in the southland: see www.gatasound.com



Great audio is never heard by the average person, but bad
audio is heard by everyone.

[Scott Dorsey]

ChrisCoaster > wrote:
>
>Hump higher in amplitude or frequency? Frequncy wise the peak is no
>lower than 50Hz, and is typically between 60-80. I have seen a
>gradual uptick down to 10-20Hz and just throw that away in my mind.
>LOL! But it's definitely there and at the frequencies I just
>mentioned.

The gradual uptick is a measurement artifact. My guess is that the
peak is actually higher in frequency than you think it is, and that
if you use more bins you will see it move.

>I'm beginning to feel like I'm being made out to be a nut here, but
>that's fine - I use my own eyes and ears to make my observation.
>Nothing like good old empirical observation! No one here is obliged
>to believe me.

No, we believe you. This is what the music is like, it's supposed to
be this way. It's not designed to play on a 3" transistor radio speaker.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Scott Dorsey]

hank alrich > wrote:
>Richard Webb > wrote:
>
>> Can you say snake oil? I knew you could.
>
>I got a band by that name, sort of. Sierra Snake Oil.
>
>But we're the real thing.

Is snake oil like baby oil or motor oil, which is squeezed out of babies
and motors? Or is it like olive oil, which is the oil you rub olives with?

Who oils their snakes anyway?
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[ChrisCoaster]

On Apr 23, 1:39*pm, (Scott Dorsey) wrote:
> ChrisCoaster > wrote:
>
> >Hump higher in amplitude or frequency? *Frequncy wise the peak is no
> >lower than 50Hz, and is typically between 60-80. *I have seen a
> >gradual uptick down to 10-20Hz and just throw that away in my mind.
> >LOL! *But it's definitely there and at the frequencies I just
> >mentioned.
>
> The gradual uptick is a measurement artifact. *My guess is that the
> peak is actually higher in frequency than you think it is, and that
> if you use more bins you will see it move.
>
> >I'm beginning to feel like I'm being made out to be a nut here, but
> >that's fine - I use my own eyes and ears to make my observation.
> >Nothing like good old empirical observation! *No one here is obliged
> >to believe me.
>
> No, we believe you. *This is what the music is like, it's supposed to
> be this way. *It's not designed to play on a 3" transistor radio speaker.

[Richard Webb[_3_]]

Mike Rivers writes:

> I have a Focusrite VRM Box here for review, a
> gadget that's supposed to let you hear, over headphones,
> what your mix would sound like through different speakers in
> assorted listening environments - kind of like a modern
> digital/headphone counterpart to Auratones.

Huh? YOu mean it ain't enough now to tell me I can make a
cheapo microphone sound like a high end Neumann or
something, now they've gotta tell me they can emulate my
favorite monitors in a given room on a set of cans.
<rotflmao!!!>

Can you say snake oil? I knew you could.

HOpe they're paying you for that one Mike!!!

Regards,
Richard
WHo doesn't own an Antares mike modeler or autotune
--
| Remove .my.foot for email
| via Waldo's Place USA Fidonet<->Internet Gateway Site
| Standard disclaimer: The views of this user are strictly his own.

[Mike Rivers]

On 4/23/2011 2:44 PM, Richard Webb wrote:

> Huh? YOu mean it ain't enough now to tell me I can make a
> cheapo microphone sound like a high end Neumann or
> something, now they've gotta tell me they can emulate my
> favorite monitors in a given room on a set of cans.

VRM isn't really for that. They have to have different
examples, and so they have names, most of which (at least
the speakers) we've heard of. If you try to mix while
listening on phones through the VRM processor, even with the
Genelec speakers in a studio control room, you'll do it all
wrong. The idea is just to see how it would sound in other
environments, and perhaps make a change if something
important gets lost.

You hear the right things happen when you go between
headphones and speakers. When I have to mix on headphones, I
know to pan things wider than I think they should be (or I'm
not distressed, when hearing too wide a spread from a stereo
mic setup), and if I'm adding reverb, I add a bit more than
I think sounds right in the phones. When you switch in the
VRM, the stereo width in the phones collapses and you hear
too much ambience.

> Can you say snake oil? I knew you could.
> HOpe they're paying you for that one Mike!!!

Naw, this will be a web site freebie. Nobody pays me to
write any more. I'm too long winded or too honest.



--
"Today's production equipment is IT based and cannot be
operated without a passing knowledge of computing, although
it seems that it can be operated without a passing knowledge
of audio." - John Watkinson

http://mikeriversaudio.wordpress.com - useful and
interesting audio stuff

[Richard Webb[_3_]]

Mike Rivers writes:

On Sat 2039-Apr-23 15:19, Mike Rivers (1:3634/1000) wrote to All: > VRM isn't really for that. They have to have different
> examples, and so they have names, most of which (at least
> the speakers) we've heard of. If you try to mix while
> listening on phones through the VRM processor, even with the
> Genelec speakers in a studio control room, you'll do it all wrong.
> The idea is just to see how it would sound in other
> environments, and perhaps make a change if something
> important gets lost.

Okay, sounds right, but I wouldn't want to base any critical decisions on its output but for a rough idea it might work.

> You hear the right things happen when you go between
> headphones and speakers. When I have to mix on headphones, I know
> to pan things wider than I think they should be (or I'm not
> distressed, when hearing too wide a spread from a stereo mic
> setup), and if I'm adding reverb, I add a bit more than I think
> sounds right in the phones. When you switch in the
> VRM, the stereo width in the phones collapses and you hear
> too much ambience.

YEah that sounds right, I find that true when forced to mix
on phones, verbs require more than you think sounds good for the same effect on speakers, etc. That's why I'd be so
leery.

>> Can you say snake oil? I knew you could.
>> HOpe they're paying you for that one Mike!!!

> Naw, this will be a web site freebie. Nobody pays me to
> write any more. I'm too long winded or too honest.

Okay, hope you didn't have to buy the unit, but I'm sure
that it'll be an honest review that gives folks straight
dope about the unit.



Regards,
Richard
.... Remote audio in the southland: See www.gatasound.com
--
| Remove .my.foot for email
| via Waldo's Place USA Fidonet<->Internet Gateway Site
| Standard disclaimer: The views of this user are strictly his own.

[david gourley[_2_]]

(Scott Dorsey) put forth the notion in...news:iov30j$7lg$1
@panix2.panix.com:

> hank alrich > wrote:
>>Richard Webb > wrote:
>>
>>> Can you say snake oil? I knew you could.
>>
>>I got a band by that name, sort of. Sierra Snake Oil.
>>
>>But we're the real thing.
>
> Is snake oil like baby oil or motor oil, which is squeezed out of babies
> and motors? Or is it like olive oil, which is the oil you rub olives
with?
>
> Who oils their snakes anyway?
> --scott


I guess folks in Morocco would if they want to have a 'buff' adder.

david