Why 24/96 sampling isn't necessarily better-sounding than 24/44 sampling
Arny Krueger
meeting in New York (Thanks Scott!), I found a mention of the following
paper that sheds some light on other listening tests that have shown that,
all other things being equal, 24/96 sampling isn't necessarily
better-sounding than 24/44 sampling.
Audio Engineering Society Convention Paper 5876: Perceptual Discrimination
between Musical Sounds with and without Very High Frequency Components
This paper describes the test methodology and the results of a series of
listening tests performed by researchers at NHK Science & Technical
Research Laboratories, Tokyo, Japan. These tests compared the playback of
recordings with and without audio signals above 21 KHz.
19 different musical selections and one synthetic sound were used:
1 "Satsuma-Biwa" "Satsuma-Biwa"
2 Litha Drums, Bass, Pf (Jazz Piano Trio)
3 Meditation Vn, Pf
4 Romanian Folk Dances Vn, Pf
5 Intermezzo de "Carmen" Fl, Pf
6 Beethoven: Sym. No.9 4th Mov. Picc
7 Bach: Suite for Vc No.2 - Prelude Sax
8 Bach: Suite for Vc No.6 - Prelude Sax
9 Piece en forme de Habanera Sax, Pf
10 Partie Sax, Pf, Perc
11 Sednalo Bulgarian Chorus (SACD ARHS-1002)
12 TihViatar Bulgarian Chorus (SACD ARHS-1002)
13 Meditation+White Noise Vn, Pf, High frequency band consists of only white
noise.
14 Airs Valagues Fl, Pf
15 Tchaikovski: Sym. No.6 3rd Mov. Full Orchestra
16 Doralice Vo, Gt (Bossa Nova)
17 Chega de Sauadade Vo, Gt, Pf, Perc (Bossa Nova)
18 tiny rose Vo, Pf, Gt, Fl, Perc ("the birds")
19 butterfly Vo, Pf, Gt, Perc ("the birds")
20 Autumn Leaves Drums, Bass, Pf (Jazz Piano Trio)
Notably 2 SACD selections were used.
"First, 36 subjects evaluated 20 kinds of stimulus, and each stimulus was
evaluated 40 times in total. The results showed no significant difference
among the sound stimuli, but that the correct response rate for three sound
stimuli was close to the significance probability (5% level). It is
concluded that one subject attained a 75% correct response rate which
constituted a significant difference. In order to make a strict statistical
test, we conducted a supplementary test with this subject who had attained
the best correct answer rate in the first test. This subject evaluated six
kinds of sound stimulus, and then evaluated each sound stimulus 20 times. As
a result, no significant difference was found among the sound stimuli, and
so this subject could not discriminate between these sound stimuli with and
without very high frequency components."
In other words, of 36 listeners, only one listener scored substantially
better than random guessing, and when retested, he could not duplicate his
earlier results. This indicates that his results were due to luck. A study
of statistics and actual experience suggests that with a group of 36
listeners, it is pretty much certain that one or more listeners will get
good scores due to luck, and that they won't be able to duplicate those
results when re-tested.
So, you can flip pennies or compare 24/44 to 24/96 and get pretty much the
same results, provided you hold all other relevant variables equal.
WillStG
of program at the various sampling rates, summed through an analog mixing bus.
Maybe I can't tell the difference between 44.1 and 48k sampling with two
tracks, but I have been able to tell between when working with a Sony 3348 on a
daily basis when you have a lot of tracks up.
Will Miho
NY Music & TV Audio Guy
Fox News/Fox & Friends
"The large print giveth and the small print taketh away..." Tom Waits
> "Arny Krueger" ar...@hotpop.com
Will Miho
NY Music & TV Audio Guy
Off the Morning Show! & sleepin' In... / Fox News
"The large print giveth and the small print taketh away..." Tom Waits
Sugarite
> meeting in New York (Thanks Scott!), I found a mention of the following
> paper that sheds some light on other listening tests that have shown that,
> all other things being equal, 24/96 sampling isn't necessarily
> better-sounding than 24/44 sampling.
blah blah blah, no mention of sound system/environment of testing, wouldn't
convince anyone anyway.
Sample rate is just one issue. If there's a significant comprimise anywhere
else in the sound system then many advantages like higher sample rates are
effectively negated.
A native 96kHz sample rate recording with suitable gear is unquestionably a
more precise representation of the sound information than a recording using
the same gear set to 44.1kHz. If that advantage is not exploited then it's
not the fault of the higher sample rate, and is certainly not grounds for
dismissing higher sample rates as a progressive step in sound reproduction.
However the gap is narrowed by the fact that equipment capable of higher
sample rates generally has a better digital clock, and renders better
results at lower sample rates than gear with a clock that can only sustain
the rate being played. You may have noticed how good DVD players are at
playing CD's.
Frankly I doubt any consumer sound system but that of a real enthusiast can
effectively pass on the benefits of content at a 96kHz sample rate.
However, the advantages of recording a processing audio at 96kHz during
engineering can certainly be appreciated through any reasonable consumer
stereo.
Bob Cain
Arny Krueger wrote:
>
> This paper describes the test methodology and the results of a series of
> listening tests performed by researchers at NHK Science & Technical
> Research Laboratories, Tokyo, Japan. These tests compared the playback of
> recordings with and without audio signals above 21 KHz.
Arny, could you tell us what the reproduction chain was?
Bob
--
"Things should be described as simply as possible, but no
simpler."
A. Einstein
anthony.gosnell
> These tests compared the playback of
> recordings with and without audio signals above 21 KHz.
Who cares about what is happening above 20K?
The critical difference is what happens at frequencies you can hear.
At 11 Khz you have only 4 samples per cycle using 44.1Khz sampling frequency
but nearly 9 samples per cycle using 96Khz.
Anthony Gosnell
Mike Rivers
> Who cares about what is happening above 20K?
> The critical difference is what happens at frequencies you can hear.
There are those who believe that this is not the whole story. What's
above the range of hearing affects what we hear. If it's missing, we
hear something different from if it's there.
I don't know of any studies which have actually proved this. It would
be very difficult to prove conclusively, so either you accept it or
you don't. It's not a big deal for most people, and it's probably most
important to those who feel that if they can't sell you on higher
sample rates, they won't sell you anything.
> At 11 Khz you have only 4 samples per cycle using 44.1Khz sampling frequency
> but nearly 9 samples per cycle using 96Khz.
And this means? If you're suggesting that the 11 kHz sine wave will
be more accurately reproduced from 9 samples than from four, you're
wrong, provided that all other rules of sampling have been followed.
What's different is that with 96 kHz sampling, stuff at 40 kHz that
happens to be present will be preserved. There are microphones and
loudspeakers which are capable of capturing and reproducing
frequencies above 20 kHz, so the holes are being filled in if you
choose to spend the bucks.
--
I'm really Mike Rivers - (mri...@d-and-d.com)
However, until the spam goes away or Hell freezes over,
lots of IP addresses are blocked from this system. If
you e-mail me and it bounces, use your secret decoder ring
and reach me here: double-m-eleven-double-zero at yahoo
KikeG
> A native 96kHz sample rate recording with suitable gear is unquestionably a
> more precise representation of the sound information than a recording using
> the same gear set to 44.1kHz.
"More precise"? Apart from allowing recording frequencies from 22 KHz
to 48 KHz, why should it be more precise? Intuitively it may seem so,
but it's not.
Arny Krueger
news:3FB553D0...@arcanemethods.com
> Arny Krueger wrote:
>>
>> This paper describes the test methodology and the results of a
>> series of listening tests performed by researchers at NHK Science &
>> Technical Research Laboratories, Tokyo, Japan. These tests compared
>> the playback of recordings with and without audio signals above 21
>> KHz.
> Arny, could you tell us what the reproduction chain was?
Unfortunately, the test system is only described with a diagram, and of
course this is a text-only forum. However,
I'll try to crib a few captions:
DAW SADiE ATEMIS Cool Edit Pro
D/A Dcs 954
Master Clock dcs 992
Controller Laguna Hills SYSTEM 1000E
Amp. SONY FA777ES
Super Tweeter PIONEER PT-R9
Power Supply Accuphase PS-1200V
Speaker B&W Nautilus 801
Amp. Marantz PA02
I get the impression that there were two separate, independent reproduction
chains, one for < 21 KHz and one for > 21 KHz. This was no doubt done to
minimize intermodulation. I suspect they did the 21 KHz filtering with Cool
Edit Pro and used Cool Edit's multitrack facilites to handle the playback.
I'm a little confused because I'm under the impression that the Sadie Atemis
workstation is Mac-based however it does exchange data with PCs.
The > 21 KHz reproduction chain used a DCS 954 DAC, a Sony FA 777ES amp, and
a Pioneer PT-R9 super tweeter.
The < 21 KHz reproduction chain used a DCS 954 DAC, a Marantz PA02 amp, and
a B&W Nautilus 801 speaker system. The DCS 992 handled clocking for both the
low and high frequency DACs.
It was also stated that the listening room conformed to IEC recommendation
BS 1116-1 which is very stringent. For example, BS 1116-1 sates that under
no circumstances should the background noise exceed NR 15.
Steve Jorgensen
I can't tell if you are agreeing or disagreeing with Arny. If you're saying
that 44.1 has an inaccurate representation of higher frequency information,
sure, but that inaccuracy shows up mainly as overtones above 20K that's
removed by the Nyquist filter. The rest is what's called quantization noise,
and dithering takes care of that at the expense of a tiny amount of white
noise too small to bother about.
initialsBB
> "It is
> concluded that one subject attained a 75% correct response rate which
> constituted a significant difference. In order to make a strict statistical
> test, we conducted a supplementary test with this subject who had attained
> the best correct answer rate in the first test.
Why is it necessary to conduct a supplemental test only with the
subject who scored the highest? Shouldn't the researchers have
retested all subjects equally? If he had one "lucky" run wouldn't it
have been equally possible that another of the subjects had an
"unlucky" run and would have scored higher in subsequent tests?
Bob Cain
Arny Krueger wrote:
>
> The > 21 KHz reproduction chain used a DCS 954 DAC, a Sony FA 777ES amp, and
> a Pioneer PT-R9 super tweeter.
> The < 21 KHz reproduction chain used a DCS 954 DAC, a Marantz PA02 amp, and
> a B&W Nautilus 801 speaker system. The DCS 992 handled clocking for both the
> low and high frequency DACs.
Hrmmph. If they used different repro chains I don't see how
any valid conclusions at all about intrinic differences due
to sample rate can be drawn from the listening tests.
Thanks for the info.
Arny Krueger
news:3fb621b5$0$64...@hades.is.co.za
> "Arny Krueger" <ar...@hotpop.com> wrote
>> These tests compared the playback of
>> recordings with and without audio signals above 21 KHz.
> Who cares about what is happening above 20K?
Good question.
> The critical difference is what happens at frequencies you can hear.
Only if there is a meaningful difference.
> At 11 Khz you have only 4 samples per cycle using 44.1Khz sampling
> frequency but nearly 9 samples per cycle using 96Khz.
What difference would you expect this to make?
Digital theory and practice say that it takes only slightly over 2 samples
per cycle to get as good of a sampling job of a sine wave as you can
imagine. So 4 samples per cycle is overkill and 9 samples per cycle is gross
overkill. IOW, there's no meaningful difference in accuracy.
anthony.gosnell
> live...@metroweb.nospam.co.za writes:
> > At 11 Khz you have only 4 samples per cycle using 44.1Khz sampling
frequency
> > but nearly 9 samples per cycle using 96Khz.
>
> And this means? If you're suggesting that the 11 kHz sine wave will
> be more accurately reproduced from 9 samples than from four, you're
> wrong, provided that all other rules of sampling have been followed.
Since when did music consist only of pure sine waves?
--
Anthony Gosnell
to reply remove nospam.
Chris Hornbeck
wrote:
>Digital theory and practice say that it takes only slightly over 2 samples
>per cycle to get as good of a sampling job of a sine wave as you can
>imagine. So 4 samples per cycle is overkill and 9 samples per cycle is gross
>overkill. IOW, there's no meaningful difference in accuracy.
Isn't this only true for perfect, non-quantized samples?
Chris Hornbeck
new email address
"That is my theory, and what it is too."
Anne Elk
Arny Krueger
news:t3tdrvs7442l3tddi...@4ax.com
> On Sat, 15 Nov 2003 19:24:04 -0500, "Arny Krueger" <ar...@hotpop.com>
> wrote:
>
>> Digital theory and practice say that it takes only slightly over 2
>> samples per cycle to get as good of a sampling job of a sine wave as
>> you can imagine. So 4 samples per cycle is overkill and 9 samples
>> per cycle is gross overkill. IOW, there's no meaningful difference
>> in accuracy.
>
> Isn't this only true for perfect, non-quantized samples?
Please explain what you mean by that.
Arny Krueger
> "Mike Rivers" <mri...@d-and-d.com> wrote
>> live...@metroweb.nospam.co.za writes:
>> "anthony.gosnell" <live...@metroweb.nospam.co.za> wrote in message
>> news:3fb621b5$0$64...@hades.is.co.za
>>> At 11 Khz you have only 4 samples per cycle using 44.1Khz sampling
>>> frequency but nearly 9 samples per cycle using 96Khz.
>
>> And this means? If you're suggesting that the 11 kHz sine wave will
>> be more accurately reproduced from 9 samples than from four, you're
>> wrong, provided that all other rules of sampling have been followed.
> Since when did music consist only of pure sine waves?
Anthony, since you are the author of the example based on sine waves, that
would be your question to answer.
Chris Hornbeck
wrote:
>"Chris Hornbeck" <chrishornbe...@att.net> wrote in message
>news:t3tdrvs7442l3tddi...@4ax.com
>> On Sat, 15 Nov 2003 19:24:04 -0500, "Arny Krueger" <ar...@hotpop.com>
>> wrote:
>>
>>> Digital theory and practice say that it takes only slightly over 2
>>> samples per cycle to get as good of a sampling job of a sine wave as
>>> you can imagine. So 4 samples per cycle is overkill and 9 samples
>>> per cycle is gross overkill. IOW, there's no meaningful difference
>>> in accuracy.
>>
>> Isn't this only true for perfect, non-quantized samples?
>
>Please explain what you mean by that.
Sorry, I guess that is about as clear as mud. How about:
Isn't this only true for samples of infinite wordsize?
Arny Krueger
Only if you want absolutely perfect results!
If you have finite word size then you have finite SNR. The finite SNR
creates ambiguities in how precisely the sine wave has been measured.
Remixer
(up to the Nyquist frequency of 1/2 the sample rate) you need some pretty
severe reconstruction filters. These real-world filters have audibly
destructive effects on audio in the passband and, more significantly, on
impulse response. A perfectly reconstructed sine wave and good sound are
far from the same thing. Using double and quadruple sample rates moves
filter artifacts further away from the audible frequency range. That is the
advantage of over-sampling, rather than a frequency response up to 50 or
100kHz.
Don Pearce
wrote:
It is standard practice. The fact is that in any statistical test
there will be results that stand out from the others. It is important
to look at these results to see if they really are different, or
merely the result of normal statistical clumping. If the odd results
are of no significance, you will see "reversion to mean". IE, on a
re-test they will tend to drop back into the average of results.
In "real" statistics, for example when looking at sickness associated
with an industry, when a cluster of cases is observed, the only
statistically valid thing to do with those cases is to note them, then
ignore them. Only if a wholly new cluster can subsequently be
associated with the same location is it statistically worth pursuing
the matter.
Clusters happen.
d
_____________________________
anthony.gosnell
Arny, I didn't say anything about sine waves. I just said "At 11 Khz", you
were the one who assumed that this must of course be a sine wave, and so I
challenged your assumption.
With just four samples per cycle you can reproduce that frequency but you
don't really stand a chance at getting the shape right.
Scott Dorsey
>Since when did music consist only of pure sine waves?
All waveforms can be decomposed into pure sound waves.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Arny Krueger
news:huDtb.179030$pT1....@twister.nyc.rr.com
> A large part of the problem is that to materialize these perfect sine
> waves (up to the Nyquist frequency of 1/2 the sample rate) you need
> some pretty severe reconstruction filters.
A major advance came about a decade ago when these filters were finally
moved into the digital domain with great success. It took about ten years of
fits and starts to get things REALLY right.
> These real-world filters have audibly destructive effects on audio in the
passband
They may or they may not have audibly destructive effects. For example, I've
tested fine brick-wall filters that have less phase shift up to 20 KHz and
beyond, than a fine power amp. I've posted links to the phase/amplitude test
results here many times.
> and, more significantly, on impulse response.
The bottom line is your claim that "These real-world filters have audibly
destructive effects on audio in the passband". IME you missed an important
hedge-word, namely the word *can*. The correct statement is: "These
real-world filters can have audibly destructive effects on audio in the
passband". That means they may or they may not have audibly destructive
effects.
In the end it all comes down to listening. The only valid way to listen for
potentially audible artifacts in good converters is the canonical
level-matched, time-synchronized, bias-controlled listening test. However
there are many perfectly acceptable ways to do good listening tests as long
as these three requirements are paid attention to.
For three years I've posted *everything* it takes to do a good listening
tests of a number of converters at www.pcabx.com, except a DAW and a good
monitoring system. If you're in this game, then you have those two remaining
ingredients. If you try the PCABX web site listening tests you will find
that some converters audibly trash sound quality in just one pass, and
others don't make *any* audible changes after 10 or 20 passes. AFAIK, nobody
who has ever tried the same thing by other reasonable means has produced
results that significantly differ.
> A perfectly reconstructed sine
> wave and good sound are far from the same thing.
Agreed. But we've had at least a few very good converters for years, They
don't make any audible changes to even the most complex, demanding musical
sounds. Let your ears (and just your ears) be your guide!
> Using double and
> quadruple sample rates moves filter artifacts further away from the
> audible frequency range.
If a good 44.1 KHz converter has no audible artifacts with demanding and
complex sounds, even after the conversion process is repeated 10 or 20 or
even 40 times, what audible artifacts are we talking about anyway?
> That is the advantage of over-sampling,
> rather than a frequency response up to 50 or 100kHz.
If it's not broke, don't fix it!
Arny Krueger
> "Arny Krueger" <ar...@hotpop.com> wrote in message
> news:8OudnfzktM3...@comcast.com...
>> "anthony.gosnell" <live...@metroweb.nospam.co.za> wrote in message
>> news:3fb6f3e7$0$64...@hades.is.co.za
>>
>>> "Mike Rivers" <mri...@d-and-d.com> wrote
>>>> live...@metroweb.nospam.co.za writes:
>>
>>>> "anthony.gosnell" <live...@metroweb.nospam.co.za> wrote in message
>>>> news:3fb621b5$0$64...@hades.is.co.za
>>
>>>>> At 11 Khz you have only 4 samples per cycle using 44.1Khz sampling
>>>>> frequency but nearly 9 samples per cycle using 96Khz.
>>>
>>>> And this means? If you're suggesting that the 11 kHz sine wave
>>>> will be more accurately reproduced from 9 samples than from four,
>>>> you're wrong, provided that all other rules of sampling have been
>>>> followed.
>>
>>> Since when did music consist only of pure sine waves?
>>
>> Anthony, since you are the author of the example based on sine
>> waves, that would be your question to answer.
>
> Arny, I didn't say anything about sine waves. I just said "At 11
> Khz", you were the one who assumed that this must of course be a sine
> wave, and so I challenged your assumption.
The 11 KHz number is yours. In a 44 KHz sampled system at 11 KHz and above
there are nothing but sine waves and combinations thereof.
When you say "cycle" Anthony, you are limiting your discussion to periodic
waves. Any periodic wave can always be thought of as being a linear
combination of sine waves. Any wave that has been brick-wall filtered at 22
KHz contains no sine wave components that are above 22 KHz.
A sine wave at any frequency can be fully characterized by its frequency,
amplitude and phase. It takes slightly more than two data points to fully
determine the frequency, amplitude and phase of a sine wave.
These are mathematical theorums and corolaries that are over 170 years old
and have stood the test of time. When I say "slightly more than 2" I'm
referring to 2 plus a mathematical delta, the smallest amount that can be
conceived of.
Therefore, slightly more than two sine waves per 22 KHz cycle is sufficient
to fully characterize any wave that has been brick-wall filtered at 22 Khz.
> With just four samples per cycle you can reproduce that frequency but
> you don't really stand a chance at getting the shape right.
As soon as *any* signal is brick-wall filtered at 22 KHz, there is no shape
that can't be gotten right with a tiny bit more than four samples per cycle
at 11 KHz or above.
For more information please read the rec.audio.pro faq, particularly
Question 5.12 "How can a 44.1 kHz sampling rate be enough to record all the
harmonics of music?"
mike rogers
> listening tests performed by researchers at NHK Science & Technical
> Research Laboratories, Tokyo, Japan. These tests compared the playback of
> recordings with and without audio signals above 21 KHz.
Ok, has everyone lost the plot here or is it just me? So, what this
test says is that basically music sounds the same regardless of
whether or not it has everything above 21khz removed. So what the hell
has this got to do with 24/48 or 24/96 recording? Nothing.
Here is the simple computer programmers explanation of sampling rates.
Analog sound is converted to digital. In order to do this it needs to
be stored as bits. The more bits we can use to re-create the analog
wavesform, the better the sound.
So, the analog waveform is converted to digital in the A>D process and
expressed as a series of numbers. The waveform is sampled at the
sample rate, like 48000 times a second or 96000 times a second. Each
of these samples is then stored as a number. The the precision of this
number is determined by bit size. A 16 bit number is significantly
smaller and therefore less precise than a 24 bit number.
So, in a nutshell. Moving from 16 bit to 24 bit, we have 8 extra bits
per sample to represent the analog wave which is a massive gain.
Moving from 48khz to 96khz we simply double the number of bits we
have. Which is why people will easily notice 16 bit vs 24 bit and less
so 48khz vs 96khz.
Nothing to do with the peak human hearing frequency of 20khz as
expressed in the start of this thread.
Don Pearce
wrote:
Clearly being a simple programmer isn't sufficient. Moving from 16 to
24 bits improves matters if - and only if - the noise floor of the
original analogue signal is below that of the 16 bit dither signal.
The chances of that happening in any real recording are vanishingly
close to zero. In virtually any scenario encountered in real life, 16
bits record just as high a quality as 24. There may well be special
demo recordings that don't obey this rule of thumb.
As regards sampling rate, the situation is perhaps not quite as clear.
Certainly it is possible to find microphones that are pretty flat up
to 20kHz, and have useful output above. Whether that makes any audible
difference to a recording is debatable. Certainly recording above
20kHz will result in the capture of some pretty major untreated
resonances from any microphone - manufacturers aren't quite as fussy
about flatness up there.
If you accept that 20kHz represents a useful upper limit to human
hearing, and there is nothing significant above, then 44.1 is every
bit as good as 48 or 96. 44.1 captures *everything* up to 20kHz with
no exceptions.
d
_____________________________
Mike Rivers
> Arny, I didn't say anything about sine waves. I just said "At 11 Khz", you
> were the one who assumed that this must of course be a sine wave, and so I
> challenged your assumption.
If you said "11 kHz" you specified a single frequency, and that means
a sine wave. If it's not a sine wave, it contains other frequencies
higher than the fundamental. If you're talking about a 100 Hz square
wave, then the reproduction of the 11 kHz component of that waveform
with 44.1 kHz sampling will be pretty accurate, but not perfect. If
you're talking about putting an 11 kHz square wave in to a 44.1 kHz
sampling system, you'll get an 11 kHz sine wave out because that's all
that will be left after filtering out all the frequences that aren't
allowed into or out of the system for the given sample rate.
> With just four samples per cycle you can reproduce that frequency but you
> don't really stand a chance at getting the shape right.
Sorry, but this is completely incorrect. You really need to read up on
recongnized sampling theory before you make your own proclamations.
Scott Dorsey
>> This paper describes the test methodology and the results of a series of
>> listening tests performed by researchers at NHK Science & Technical
>> Research Laboratories, Tokyo, Japan. These tests compared the playback of
>> recordings with and without audio signals above 21 KHz.
>
>Ok, has everyone lost the plot here or is it just me? So, what this
>test says is that basically music sounds the same regardless of
>whether or not it has everything above 21khz removed. So what the hell
>has this got to do with 24/48 or 24/96 recording? Nothing.
It does, in that the only thing that the higher sample rate buys you is
the ultrasonic response.
>Here is the simple computer programmers explanation of sampling rates.
>Analog sound is converted to digital. In order to do this it needs to
>be stored as bits. The more bits we can use to re-create the analog
>wavesform, the better the sound.
Not necessarily, no. I can find all kinds of ways to waste data.
>So, the analog waveform is converted to digital in the A>D process and
>expressed as a series of numbers. The waveform is sampled at the
>sample rate, like 48000 times a second or 96000 times a second. Each
>of these samples is then stored as a number. The the precision of this
>number is determined by bit size. A 16 bit number is significantly
>smaller and therefore less precise than a 24 bit number.
Right.
>So, in a nutshell. Moving from 16 bit to 24 bit, we have 8 extra bits
>per sample to represent the analog wave which is a massive gain.
Not really. It gives you more dynamic range, which is often wasted
anyway. 96 dB is an awful lot.
>Moving from 48khz to 96khz we simply double the number of bits we
>have. Which is why people will easily notice 16 bit vs 24 bit and less
>so 48khz vs 96khz.
You might want to go back and read a good description of elementary
sampling theory. I think Gabe references one in the FAQ.
Scott Dorsey
>
> With just four samples per cycle you can reproduce that frequency but you
> don't really stand a chance at getting the shape right.
No. Look up "Shannon's Sampling Theorem" in the encyclopedia.
KikeG
> Why is it necessary to conduct a supplemental test only with the
> subject who scored the highest?
To rule out he got good results just by chance.
> If he had one "lucky" run wouldn't it
> have been equally possible that another of the subjects had an
> "unlucky" run and would have scored higher in subsequent tests?
If a subject can really hear a difference, then "bad luck" has no
sense. On the other side, it's possible that you can't hear a
difference, but have good luck and pass the test.
Luke Kaven
> "Sugarite" <nob...@home.com> wrote
>
> > A native 96kHz sample rate recording with suitable gear is unquestionably a
> > more precise representation of the sound information than a recording using
> > the same gear set to 44.1kHz.
>
> "More precise"? Apart from allowing recording frequencies from 22 KHz
> to 48 KHz, why should it be more precise? Intuitively it may seem so,
> but it's not.
There is an increase in the information content. The hard question is
whether and how that additional information might be exploited. That
is an empirical question that will more than likely only be answered
with subsequent discoveries in things such as downsampling
technologies over the coming years (and years, and years). Looking
back in recent history, one can see that downsampling technology has
been improved, so there is at least precedent for projecting such a
claim. The one justification I can see for recording at higher
sample rates is for archival purposes, which involves betting on the
future. That doesn't seem like an unreasonable bet.
Luke
initialsBB
> It is standard practice. The fact is that in any statistical test
> there will be results that stand out from the others. It is important
> to look at these results to see if they really are different, or
> merely the result of normal statistical clumping.
Ah, I understand now. Thanks for the explanation.
S O'Neill
> Ok, has everyone lost the plot here or is it just me? So, what this
> test says is that basically music sounds the same regardless of
> whether or not it has everything above 21khz removed. So what the hell
> has this got to do with 24/48 or 24/96 recording? Nothing.
If you take the study as gospel, it means that any difference between Fs
= 48 kHz and Fs = 96 kHz or even Fs = 1 THz is inaudible, therefore
those higher sample rates are a waste of money, disk space, and CPU time.
If you don't agree with the study, then there may be value in those
higher rates.
So actually, it has everything to do with those sample rates' necessity
in recording.
Arny Krueger
news:772f2e86.0311...@posting.google.com
>> This paper describes the test methodology and the results of a
>> series of listening tests performed by researchers at NHK Science &
>> Technical Research Laboratories, Tokyo, Japan. These tests compared
>> the playback of recordings with and without audio signals above 21
>> KHz.
> Ok, has everyone lost the plot here or is it just me? So, what this
> test says is that basically music sounds the same regardless of
> whether or not it has everything above 21khz removed. So what the hell
> has this got to do with 24/48 or 24/96 recording? Nothing.
Huh?
24/44 coding removes *everything* above 22 KHz. 24/96, 24/192 and 24/384
coding doesn't. They move the cut-off points to 48, 96, and 192 KHz
respectively.
> Here is the simple computer programmers explanation of sampling rates.
> Analog sound is converted to digital. In order to do this it needs to
> be stored as bits. The more bits we can use to re-create the analog
> waveform, the better the sound.
True only if the law of diminishing returns has been repealed. Furthermore
there are two different and independent ways to use more bits to code audio
signals. You can use more samples or you can use larger samples or both. All
three options use more bits, but they have differing consequences.
> So, the analog waveform is converted to digital in the A>D process and
> expressed as a series of numbers. The waveform is sampled at the
> sample rate, like 48000 times a second or 96000 times a second. Each
> of these samples is then stored as a number.
The major benefit of this option is that the high frequency cutoff gets
moved up in the frequency domain as the sample rate goes up. However it's
not a sure thing that moving the high frequency cutoff up indefinitely
provides improved sound quality. The law of diminishing returns has to start
rearing its ugly head at some point.
> The precision of this
> number is determined by bit size. A 16 bit number is significantly
> smaller and therefore less precise than a 24 bit number.
The major benefit of increased sample size is that the noise floor gets
moved down in the amplitude domain as the samples get larger. However it's
not a sure thing that moving the noise floor down indefinitely provides
improved sound quality. The law of diminishing returns has to start rearing
its ugly head at some point. This paper says nothing about this issue one
way or the other.
> So, in a nutshell. Moving from 16 bit to 24 bit, we have 8 extra bits
> per sample to represent the analog wave which is a massive gain.
The benefit is a reduced noise floor, or if you will higher resolution.
However at some point the digital noise floor moves under the analog noise
floor and further improvements are moot.
> Moving from 48khz to 96khz we simply double the number of bits we
> have.
The benefit of an increased sample rate is an increased high frequency
bandpass, or if you will an increased high frequency cutoff point. However
the human ear is well known to lose accuracy and sensitivity above as little
as 4 KHz. At some point so much accuracy and sensitivity is lost that
further improvements the high frequency cutoff point become moot.
This paper is about investigations into the benefits of moving the cutoff
point well beyond 21 KHz. The investigations showed zero benefit for
increasing the cutoff point beyond 21 KHz. This approximately corresponds to
the real-world situation with a 44 KHz sample rate.
The question the paper addresses is whether or not increasing the sample
rate above 44 KHz (e.g. 96 KHz) has any audible benefits. It struggled
diligently with the question and found no benefits to the major effect of
increasing the sample rate substantially above 44 KHz.
>Which is why people will easily notice 16 bit vs. 24 bit and less
> so 48khz vs. 96khz.
In actuality neither change is very easily noticed. If you have a DAW with
24/96 converters and a monitoring system you respect, you can investigate
this with your own ears by downloading files from
http://www.pbcabx.com/technical/sample_rates/index.htm and listening to
them.
> Nothing to do with the peak human hearing frequency of 20khz as
> expressed in the start of this thread.
I think I've explained why this is not a correct statement several different
ways.
Carey Carlan
news:772f2e86.0311...@posting.google.com:
> So, in a nutshell. Moving from 16 bit to 24 bit, we have 8 extra bits
> per sample to represent the analog wave which is a massive gain.
> Moving from 48khz to 96khz we simply double the number of bits we
> have. Which is why people will easily notice 16 bit vs 24 bit and less
> so 48khz vs 96khz.
>
> Nothing to do with the peak human hearing frequency of 20khz as
> expressed in the start of this thread.
I agreed with you until they showed me the math, which says:
It only takes two samples per cycle over time to define a 20 kHz sine wave.
If it's not a perfect sine wave (if you introduce more bits to put ripples
or bumps in the wave) then you are adding components higher than 20 kHz to
your simple sine wave. To get a more complex waveform, you must add a very
high frequency component or lots of somewhat higher frequency components or
both. In either case, the human ear is not supposed to be able to hear it.
I know my ear can't with any of the gear I own.
Jay - atldigi
Dorsey) wrote:
> >So, in a nutshell. Moving from 16 bit to 24 bit, we have 8 extra bits
> >per sample to represent the analog wave which is a massive gain.
>
> Not really. It gives you more dynamic range, which is often wasted
> anyway. 96 dB is an awful lot.
>
> >Moving from 48khz to 96khz we simply double the number of bits we
> >have. Which is why people will easily notice 16 bit vs 24 bit and less
> >so 48khz vs 96khz.
>
> You might want to go back and read a good description of elementary
> sampling theory. I think Gabe references one in the FAQ.
> --scott
Scott, please visit the thread entitled:
"16 bit vs 24 bit, 44.1khz vs 48 khz <-- please explain".
The 16 bit 24 bit subject is going on there as well with the same issues
and it may be helpful for you to make a similar post in that thread. I'm
getting lonely over there. Thanks!
--
Jay Frigoletto
Mastersuite
Los Angeles
promastering.com
Jay - atldigi
<live...@metroweb.nospam.co.za> wrote:
Since forever. What do you think overtones are? A fundamental and a
series of overtones can be broken down essentially into a bunch of sine
waves.
Jay - atldigi
> > A 16 bit number is significantly
> > smaller and therefore less precise than a 24 bit number.
>
> Right.
>
> >So, in a nutshell. Moving from 16 bit to 24 bit, we have 8 extra bits
> >per sample to represent the analog wave which is a massive gain.
>
> Not really. It gives you more dynamic range, which is often wasted
> anyway. 96 dB is an awful lot.
The 24 bit number is more precise than the 16 bit. True enough. What
that means in audio, however, is that the 24 bit word can describe
smaller values than the 16 bit word, thus signals that are lower in
level. The 16 bit number is already describing 96 dB of dynamic range
just fine. If you want to carry the precision further and capture
signals that are lower, say to -144 dB, then 24 bits is your ticket.
mike rogers
>
> Clearly being a simple programmer isn't sufficient. Moving from 16 to
> 24 bits improves matters if - and only if - the noise floor of the
> original analogue signal is below that of the 16 bit dither signal.
> The chances of that happening in any real recording are vanishingly
> close to zero. In virtually any scenario encountered in real life, 16
> bits record just as high a quality as 24. There may well be special
> demo recordings that don't obey this rule of thumb.
>
> As regards sampling rate, the situation is perhaps not quite as clear.
> Certainly it is possible to find microphones that are pretty flat up
> to 20kHz, and have useful output above. Whether that makes any audible
> difference to a recording is debatable. Certainly recording above
> 20kHz will result in the capture of some pretty major untreated
> resonances from any microphone - manufacturers aren't quite as fussy
> about flatness up there.
>
> If you accept that 20kHz represents a useful upper limit to human
> hearing, and there is nothing significant above, then 44.1 is every
> bit as good as 48 or 96. 44.1 captures *everything* up to 20kHz with
> no exceptions.
>
I agree with you regarding the noise floor on 16 bit recording. Your
last paragraph is complete bollocks though. I do understand sampling,
having worked on software in this area. This is why you are wrong:
Say we take a analog wave cycling at a fequency of 10khz or 10,000
times per second. We then sample that at 10khz. This means that for
every 1 second of waveform time we take 10,000 samples to see what the
amplitude of the waves is.
From this information the computer can try to calculate what the wave
actually "looked" like and reconstruct it in the D>A process. But the
analog world does not work in samples and there are actually an
infinite number of possible sample points in a 1 second, 10khz wave.
So, when we sample the same wave at 20khz, we now have a much more
accurate representation of the orignal wave form as we have measured
the amplitude in double the number of places so the wave recontruction
is more faithful to the original. To truly represent an analog
waveform, you would have to sample at infinite number of KHZ, which is
obviously rediculous. At some point probably 96khz or a little bit
above, no-one would be able to tell the difference.
To prove my point, try this:
Record an analog signal onto a PC, a higher pitched signal is better,
at a low sample rate, like 8khz. Play the sample back through a
spectral analyser and you will see frequencies above 8khz have been
captured. According to what you are saying, it should be impossible to
record audio frequencies higher than your sample rate. This is not
true. True that the higher frequencies will not sound good as you will
get a very poor representation of the higher frequency waveform, but
they are still there.
Bob Cain
What's too often forgotton is that a signal of finite
length, like a song or a single drum hit, requires a whole
bunch of them. An infinite number in fact. Conversely,
anything that can be decomposed into a finite number of them
must be infinitely long and repeat with a frequency equal to
that of the lowest sin wave.
Bob
--
"Things should be described as simply as possible, but no
simpler."
A. Einstein
Philip Perkins
Philip Perkins
Jay - atldigi
mikero...@hotmail.com (mike rogers) wrote:
> Say we take a analog wave cycling at a fequency of 10khz or 10,000
> times per second. We then sample that at 10khz. This means that for
> every 1 second of waveform time we take 10,000 samples to see what the
> amplitude of the waves is.
You won't be able to sample a 10k wave with a 10k sample rate. Do a
search for Shannon and Nyquist. You need to sample at double the highest
frequency you want to capture, theoretically - assuming an infinite
response filter which can't exist. In real world practice you need a bit
more than double.
> From this information the computer can try to calculate what the wave
> actually "looked" like and reconstruct it in the D>A process. But the
> analog world does not work in samples and there are actually an
> infinite number of possible sample points in a 1 second, 10khz wave.
> So, when we sample the same wave at 20khz, we now have a much more
> accurate representation of the orignal wave form as we have measured
> the amplitude in double the number of places so the wave recontruction
> is more faithful to the original. To truly represent an analog
> waveform, you would have to sample at infinite number of KHZ, which is
> obviously rediculous. At some point probably 96khz or a little bit
> above, no-one would be able to tell the difference.
>
> To prove my point, try this:
>
> Record an analog signal onto a PC, a higher pitched signal is better,
> at a low sample rate, like 8khz. Play the sample back through a
> spectral analyser and you will see frequencies above 8khz have been
> captured. According to what you are saying, it should be impossible to
> record audio frequencies higher than your sample rate. This is not
> true. True that the higher frequencies will not sound good as you will
> get a very poor representation of the higher frequency waveform, but
> they are still there.
Um... I don't even know where to begin with all of this... So I won't.
Lurkers in search of learning beware!
Les Cargill
>
> Jay - atldigi wrote:
> >
> > In article <3fb6f3e7$0$64...@hades.is.co.za>, "anthony.gosnell"
> > <live...@metroweb.nospam.co.za> wrote:
> >
> > > "Mike Rivers" <mri...@d-and-d.com> wrote
> > > > live...@metroweb.nospam.co.za writes:
> > > > > At 11 Khz you have only 4 samples per cycle using 44.1Khz sampling
> > > frequency
> > > > > but nearly 9 samples per cycle using 96Khz.
> > > >
> > > > And this means? If you're suggesting that the 11 kHz sine wave will
> > > > be more accurately reproduced from 9 samples than from four, you're
> > > > wrong, provided that all other rules of sampling have been followed.
> > >
> > > Since when did music consist only of pure sine waves?
> > >
> >
> > Since forever. What do you think overtones are? A fundamental and a
> > series of overtones can be broken down essentially into a bunch of sine
> > waves.
>
> What's too often forgotton is that a signal of finite
> length, like a song or a single drum hit, requires a whole
> bunch of them. An infinite number in fact.
Not if it's bandlimited.
> Conversely,
> anything that can be decomposed into a finite number of them
> must be infinitely long and repeat with a frequency equal to
> that of the lowest sin wave.
>
> Bob
> --
>
> "Things should be described as simply as possible, but no
> simpler."
>
> A. Einstein
--
Les Cargill
mike rogers
>
> It only takes two samples per cycle over time to define a 20 kHz sine wave.
>
That's my point. It takes 2 samples per cycle to define the high and
low peak of a 20khz sine wave. Therefore 40khz. But as you say, if the
wave is not a perfect sine, then the more times you sample the
waveform the more accurately the wave is re-created in the D>A
process.
Steve Jorgensen
>Don Pearce <comp...@nonsense.com> wrote in message news:<lu2frvk2qbf1hlg5f...@4ax.com>...
...
>I agree with you regarding the noise floor on 16 bit recording. Your
>last paragraph is complete bollocks though. I do understand sampling,
>having worked on software in this area. This is why you are wrong:
I hate to put it this way, but perhaps, you don't understand as much as you
think you do.
>Say we take a analog wave cycling at a fequency of 10khz or 10,000
>times per second. We then sample that at 10khz. This means that for
>every 1 second of waveform time we take 10,000 samples to see what the
>amplitude of the waves is.
>
>From this information the computer can try to calculate what the wave
>actually "looked" like and reconstruct it in the D>A process. But the
>analog world does not work in samples and there are actually an
>infinite number of possible sample points in a 1 second, 10khz wave.
>So, when we sample the same wave at 20khz, we now have a much more
>accurate representation of the orignal wave form as we have measured
>the amplitude in double the number of places so the wave recontruction
>is more faithful to the original. To truly represent an analog
>waveform, you would have to sample at infinite number of KHZ, which is
>obviously rediculous. At some point probably 96khz or a little bit
>above, no-one would be able to tell the difference.
What the heck doeas "accurate" mean at this point? An accurate representation
of the wave's true shape as it differs from a pure sine wave? If so, these
would amount to higher frequency overtones that are filtered out before A/D to
prevent aliasing, and filtered out during D/A to reverse the effects of
quantization. In short, if a sound is near the top of the human hearing
range, it doesn't matter how "accurate" its representation is except with
respect to plain frequency and amplitude. Its "shape" is irrelevant because
we can't here that.
>
>To prove my point, try this:
>
>Record an analog signal onto a PC, a higher pitched signal is better,
>at a low sample rate, like 8khz. Play the sample back through a
>spectral analyser and you will see frequencies above 8khz have been
>captured. According to what you are saying, it should be impossible to
>record audio frequencies higher than your sample rate. This is not
>true. True that the higher frequencies will not sound good as you will
>get a very poor representation of the higher frequency waveform, but
>they are still there.
If you're seeing frequencies above 4 KHz, then the spectum analyzer is bad.
It should be seeing that the sample rate is 8KHz, and filtering out everything
above 4KHz. Frequency content above that is quantization noise due to the
square shapes of the individual samples. Of course, if the A/D failes to
filter out frequencies above 4KHz, then higher frequency content in the input
can show up as lower frequency aliasing kind of like ring modulation with the
sampling frequency.
Scott Dorsey
>
>That's my point. It takes 2 samples per cycle to define the high and
>low peak of a 20khz sine wave. Therefore 40khz. But as you say, if the
>wave is not a perfect sine, then the more times you sample the
>waveform the more accurately the wave is re-created in the D>A
>process.
If the wave is not a perfect sine, then it has components that are
higher frequency than 20KHz.
Any arbitrary waveform can be decomposed to a sum of sine waves of
varying amplitude, phase, and frequency. (Fourier and Heaviside).
Bob Cain
Les Cargill wrote:
> >
> > What's too often forgotton is that a signal of finite
> > length, like a song or a single drum hit, requires a whole
> > bunch of them. An infinite number in fact.
>
> Not if it's bandlimited.
No finite number of sin waves (which are defined as infinite
in duration by Fourier theory) can sum to a signal that goes
to zero and stays there nor comes from an infinite past of
zero and suddenly becomes non-zero. Any finite number of
sin waves will sum to a periodic signal that repeats forever
in both directions.
S O'Neill
> Don Pearce <comp...@nonsense.com> wrote
>>Clearly being a simple programmer isn't sufficient. Moving from 16 to
This is the D/A converor's job, whose input it purely binary. A
programmer who deals with device drivers is intimately familiar with
these. The computer never needs to deal with the "analogness" of it,
except to never screw it up.
But the
> analog world does not work in samples and there are actually an
> infinite number of possible sample points in a 1 second, 10khz wave.
> So, when we sample the same wave at 20khz, we now have a much more
> accurate representation of the orignal wave form as we have measured
> the amplitude in double the number of places so the wave recontruction
> is more faithful to the original. To truly represent an analog
> waveform, you would have to sample at infinite number of KHZ, which is
> obviously rediculous. At some point probably 96khz or a little bit
> above, no-one would be able to tell the difference.
Right, the debate rages on about all that....
> To prove my point, try this:
>
> Record an analog signal onto a PC, a higher pitched signal is better,
> at a low sample rate, like 8khz. Play the sample back through a
> spectral analyser and you will see frequencies above 8khz have been
> captured. According to what you are saying, it should be impossible to
> record audio frequencies higher than your sample rate. This is not
> true. True that the higher frequencies will not sound good as you will
> get a very poor representation of the higher frequency waveform, but
> they are still there.
Of course, you'll get a lot of frequencies if you don't perform
anti-alias filtering before sampling.
Jay - atldigi
Dorsey) wrote:
Listen to Scott. This is an important concept. The sine example is just
a starting point. When you start to think of complex sounds as several
sines combined, then it starts to make more sense. If you add together
several sines and look at a scope you'll see the shape of the wave
changing and getting more complex. When you have a complex wave, any
wiggles that are very small actually represent signal that is higher in
frequency. Since the bandwidth of digital audio is indeed limited by
Nyquist (1/2 the sample rate), any of those little squiggles that are
small enough to represent frequencies that are higher than Nyquist will
be lost. This in no way impacts the detail or accuracy of the waveform
withing the bandwidth of the system, i.e. below Nyquist. In practice
there are filter issues and some other arcane details that influence
what you hear, but you need to understand the basic truths before you
can move on to the other details. If not, later conclusions will be
erroneous as they stem from an erroneous premise.
Don Pearce
wrote:
>Don Pearce <comp...@nonsense.com> wrote in message news:<lu2frvk2qbf1hlg5f...@4ax.com>...
No, I'm afraid you do not understand sampling. Signals below the
Nyquist limit (half the sampling rate) are "perfectly" reproduced by
the sampling process. There are no exceptions to this. Your demand to
sample infinitely finely shows that you don't understand the process.
>To prove my point, try this:
>
>Record an analog signal onto a PC, a higher pitched signal is better,
>at a low sample rate, like 8khz. Play the sample back through a
>spectral analyser and you will see frequencies above 8khz have been
>captured. According to what you are saying, it should be impossible to
>record audio frequencies higher than your sample rate. This is not
>true. True that the higher frequencies will not sound good as you will
>get a very poor representation of the higher frequency waveform, but
>they are still there.
You are talking about aliasing. I have written a paper on this if you
are interested. Visit my web site at www.pearce.uk.com and you can
read it. It will explain all his to you (as well as the
amplitude-related side of things with the paper on dither).
Incidentally I have designed a modulation analyser that sampled at
2GS/s (2 thousand million samples per second) and performed perfectly
up to 40GHz. It used alias responses in a controlled manner to do
this.
d
_____________________________
Justin Ulysses Morse
> If you accept that 20kHz represents a useful upper limit to human
> hearing, and there is nothing significant above, then 44.1 is every
> bit as good as 48 or 96. 44.1 captures *everything* up to 20kHz with
> no exceptions.
Well, there is one exception and it's the one that apparently makes the
difference: The in-band ripple in the anti-aliasing filter will
"except" a portion of certain frequency intervals. It would appear
that "perfect sound forever" is more likely when these artefacts are
scaled up out of the range of human hearing along with the corner
frequency.
ulysses
Justin Ulysses Morse
> That's my point. It takes 2 samples per cycle to define the high and
> low peak of a 20khz sine wave. Therefore 40khz. But as you say, if the
> wave is not a perfect sine, then the more times you sample the
> waveform the more accurately the wave is re-created in the D>A
> process.
But if you filter out everything over 20kHz, then a 20kHz wave has no
choice but to be a perfect sine wave. Any imperfection in it is
nothing more than higher-frequency content. If you want to represent
that content, you need a higher sampling rate. If you don't, you
don't.
ulysses
Justin Ulysses Morse
> No finite number of sin waves (which are defined as infinite
> in duration by Fourier theory) can sum to a signal that goes
> to zero and stays there nor comes from an infinite past of
> zero and suddenly becomes non-zero. Any finite number of
> sin waves will sum to a periodic signal that repeats forever
> in both directions.
So what?
I think I understand the theoretical implications of what you're
saying, but I don't understand why it's practically relevant. Which is
not to say I think it isn't relevant. So my question is neither
judgemental nor rhetorical. I want to learn: What's your point?
ulysses
Don Pearce
Inadequacies in implementation are yet another matter...
d
_____________________________
Les Cargill
>
> Les Cargill wrote:
> > >
> > > What's too often forgotton is that a signal of finite
> > > length, like a song or a single drum hit, requires a whole
> > > bunch of them. An infinite number in fact.
> >
> > Not if it's bandlimited.
>
> No finite number of sin waves (which are defined as infinite
> in duration by Fourier theory) can sum to a signal that goes
> to zero and stays there nor comes from an infinite past of
> zero
Meaning where t is undefined?
> and suddenly becomes non-zero. Any finite number of
> sin waves will sum to a periodic signal that repeats forever
> in both directions.
>
But PCM is not so much a Fourier thing as it is a
Nyquist Theorem thing. The FT in Nyquist-land is a DFT, not
a continuous FT.
I am pretty sure that PCM 44.1/16 can represent a drum hit
properly, so long as all the constraints on domain and range are
met (IOW no overs).
Did I miss your point?
> Bob
> --
>
> "Things should be described as simply as possible, but no
> simpler."
>
> A. Einstein
--
Les Cargill
Arny Krueger
news:3677d4b3.03111...@posting.google.com
> The Sadie "ARTEMIS" is PC/Windows based.
>
Thanks. It seems to be considered obsolete for sales purposes, and therefore
I couldn't find out much about it, online.
mike rogers
> > If the wave is not a perfect sine, then it has components that are
> > higher frequency than 20KHz.
> >
> > Any arbitrary waveform can be decomposed to a sum of sine waves of
> > varying amplitude, phase, and frequency. (Fourier and Heaviside).
> > --scott
>
> Listen to Scott. This is an important concept. The sine example is just
> a starting point. When you start to think of complex sounds as several
> sines combined, then it starts to make more sense. If you add together
> several sines and look at a scope you'll see the shape of the wave
> changing and getting more complex. When you have a complex wave, any
> wiggles that are very small actually represent signal that is higher in
> frequency.
Jay, I think that I am agreeing with you and Scott. A waveform is not
a simple sine which, if you know the peak high and low amplitudes the
wave form is easy to reconstruct. But, as you say, when a wave is a
complex and made up of several sines then higher frequency sampling is
required to capture the nuances.
Arny Krueger
> Jay, I think that I am agreeing with you and Scott. A waveform is not
> a simple sine which, if you know the peak high and low amplitudes the
> wave form is easy to reconstruct. But, as you say, when a wave is a
> complex and made up of several sines then higher frequency sampling is
> required to capture the nuances.
That's not how it works. It doesn't take any more samples to properly
capture a complex wave as a simple one, given that both waves have the same
bandpass.
IOW if you have a signal whose highest frequency content is say 20 KHz, add
what you will at lower frequencies, and it still doesn't take any more
samples to properly capture the wave.
IOW it takes no more samples per second to properly capture 20 KHz + 15 KHz
+ 15.001 Hz + 5 KHz + 1 KHz + unbelievable amounts of other stuff below 20
KHz, as it takes to properly capture a single 20 KHz wave.
anthony.gosnell
> But if you filter out everything over 20kHz, then a 20kHz wave has no
> choice but to be a perfect sine wave.
What about a perfect sawtooth or a perfect square?
According to your logic sampling at 44k works because anything which isn't a
sine wave should be converted to a sine wave before you sample it so that it
can be reproduced correctly. (as a sine wave of course)
What's happened here is that you have defined analog theory in terms of
digital sampling theory.
Of course digital sampling reproduces analog exactly if you define analog
only as what can be reproduced digitally.
--
Anthony Gosnell
to reply remove nospam.
anthony.gosnell
> For more information please read the rec.audio.pro faq,
Where are these?
Isn't this something which should be posted once a week for the benefit of
newbies?
Arny Krueger
news:3fb8cc68$0$64...@hades.is.co.za
> "Arny Krueger" <ar...@hotpop.com> wrote
>> For more information please read the rec.audio.pro faq,
>
> Where are these?
> Isn't this something which should be posted once a week for the
> benefit of newbies?
Try google, it will get you to the HTML version of the rec.audio.pro FAQ
version in flash.
Carey Carlan
news:3fb8cc68$0$64...@hades.is.co.za:
> "Arny Krueger" <ar...@hotpop.com> wrote
>> For more information please read the rec.audio.pro faq,
>
> Where are these?
> Isn't this something which should be posted once a week for the
> benefit of newbies?
It is often posted and can be found with a quick google search. Saving you
that 10-second effort:
Arny Krueger
> "Justin Ulysses Morse"
>> But if you filter out everything over 20kHz, then a 20kHz wave has no
>> choice but to be a perfect sine wave.
>
> What about a perfect sawtooth or a perfect square?
Filtering these waves with a 20 KHz brick wall filter ahem, profoundly
changes them. Doesn't matter whether you implement the filter in the digital
or analog domain.
> According to your logic sampling at 44k works because anything which
> isn't a sine wave should be converted to a sine wave before you
> sample it so that it can be reproduced correctly. (as a sine wave of
> course)
That's very close to what actually happens. If you 22 KHz brick wall filter
an 8 KHz (or higher through 22 KHz) square wave you end up with an 8 KHz (or
higher) sine wave, period.
> What's happened here is that you have defined analog theory in terms
> of digital sampling theory.
Nope, the Fourier math involved is equally applicable to the analog and
digital domains.
> Of course digital sampling reproduces analog exactly if you define
> analog only as what can be reproduced digitally.
Nope, all you have to do is define your analog signal to be something that
is band-limited to some frequency below the Nyquist frequency.
I learned Fourier math in 1965 when digital audio was in its infancy. At
that time we thought of *everything* as being in the analog domain. The
basic math and its practical applications worked great then, and work great
today. We just see a lot more practical applications of Fourier math today,
because doing arithmetic is so much faster and cheaper. Most of what we see
today is in the digital domain of course, but the basic concepts apply as
well as ever.
My point is that it is a great mistake to cop a plea on the grounds that
this stuff only works in the digital domain. It started out when for all
intents and purposes, analog was all we had.
Fourier did his basic work in the late 1600's, and Nyquist and Shannon did
their basic work in the late 1920s and early 1930s. Digital audio as we now
know it was just about equally practically improbable in either case.
Carey Carlan
news:3fb8cc5c$0$64...@hades.is.co.za:
> "Justin Ulysses Morse"
>> But if you filter out everything over 20kHz, then a 20kHz wave has no
>> choice but to be a perfect sine wave.
> What about a perfect sawtooth or a perfect square?
Perfect sawtooth and square waves don't exist in the real world. In order
to "turn the corner" instantly, with a sharp edge, you need an infinite
frequency component.
OTOH, rounding the corners to the "radius" of a 20 kHz tone should be
inaudible to the human ear.
> According to your logic sampling at 44k works because anything which
> isn't a sine wave should be converted to a sine wave before you sample
> it so that it can be reproduced correctly. (as a sine wave of course)
Because any component that can't be represented as a sine wave at 20 kHz
should be inaudible.
Mike Rivers
> You won't be able to sample a 10k wave with a 10k sample rate. Do a
> search for Shannon and Nyquist.
Sure you can, but you'll get a bogus result, and that's what he's
shown. So yeah, you can prove that it's possible to violate the rules
of the sampling theorem and get back something different than what you
put in.
> Um... I don't even know where to begin with all of this... So I won't.
> Lurkers in search of learning beware!
Good advice. I think this guy has more theories than Nyquist and
Shannon combined.
--
I'm really Mike Rivers - (mri...@d-and-d.com)
However, until the spam goes away or Hell freezes over,
lots of IP addresses are blocked from this system. If
you e-mail me and it bounces, use your secret decoder ring
and reach me here: double-m-eleven-double-zero at yahoo
Mike Rivers
> What about a perfect sawtooth or a perfect square?
>
> According to your logic sampling at 44k works because anything which isn't a
> sine wave should be converted to a sine wave before you sample it so that it
> can be reproduced correctly. (as a sine wave of course)
A perfect sawtooth contains frequencies higher than the fundamental.
The thing you have to understand about the principle that any complex
waveform can be expressed as the sum of a number of single frequencies
(sine waves) is that not all of those sine waves are the same
amplitude. In the case of a sawtooth, the amplitude of the harmonics
diminishes fairly rapidly, so that after the first few, they don't
contribute a whole lot to the wave shape. So given that not all
frequencies will be sampled (since some will be above the Nyquist
"half the sample rate" frequency) it won't matter much. IN THEORY, it
won't be perfectly reproduced, however in practice, the difference
between before and after sampling will be pretty small. How small
depends on the sample rate and to a certain extent, the resolution.
It's important to know where to draw the line. Generally this is
dictated by the pocketbook for most people.
A square wave is a special case since it toggles between two known
amplitudes. If you knew that you would be sampling nothing but square
waves, you could use a one-bit sample, not bother to filter the input
or the output, and it would be correct. However, to use conventional
sampling of a square wave, you need to be able to reproduce several
harmonics above the fundamental frequency in order to get an output
that resembles a square wave. The closer the fundamental gets to half
the sample rate, the more the reproduction looks like a sine wave.
Actually, since a square wave is composed of the fundamental and a
series of odd (only) harmonics, once the fundamental is above 1/3 the
sample rate, if you obey the law, the reproduction will look like a
sine wave.
Roger W. Norman
--
Roger W. Norman
SirMusic Studio
Purchase your copy of the Fifth of RAP CD set at www.recaudiopro.net.
See how far $20 really goes.
"Arny Krueger" <ar...@hotpop.com> wrote in message
news:AqWdnRNLJN2...@comcast.com...
> "Bob Cain" <arc...@arcanemethods.com> wrote in message
> news:3FB553D0...@arcanemethods.com
> > Arny Krueger wrote:
> >>
> >> This paper describes the test methodology and the results of a
> >> series of listening tests performed by researchers at NHK Science &
> >> Technical Research Laboratories, Tokyo, Japan. These tests compared
> >> the playback of recordings with and without audio signals above 21
> >> KHz.
>
> > Arny, could you tell us what the reproduction chain was?
>
> Unfortunately, the test system is only described with a diagram, and of
> course this is a text-only forum. However,
> I'll try to crib a few captions:
>
> DAW SADiE ATEMIS Cool Edit Pro
> D/A Dcs 954
> Master Clock dcs 992
> Controller Laguna Hills SYSTEM 1000E
> Amp. SONY FA777ES
> Super Tweeter PIONEER PT-R9
> Power Supply Accuphase PS-1200V
> Speaker B&W Nautilus 801
> Amp. Marantz PA02
>
> I get the impression that there were two separate, independent
reproduction
> chains, one for < 21 KHz and one for > 21 KHz. This was no doubt done to
> minimize intermodulation. I suspect they did the 21 KHz filtering with
Cool
> Edit Pro and used Cool Edit's multitrack facilites to handle the playback.
> I'm a little confused because I'm under the impression that the Sadie
Atemis
> workstation is Mac-based however it does exchange data with PCs.
>
> The > 21 KHz reproduction chain used a DCS 954 DAC, a Sony FA 777ES amp,
and
> a Pioneer PT-R9 super tweeter.
> The < 21 KHz reproduction chain used a DCS 954 DAC, a Marantz PA02 amp,
and
> a B&W Nautilus 801 speaker system. The DCS 992 handled clocking for both
the
> low and high frequency DACs.
>
> It was also stated that the listening room conformed to IEC recommendation
> BS 1116-1 which is very stringent. For example, BS 1116-1 sates that under
> no circumstances should the background noise exceed NR 15.
>
>
>
Roger W. Norman
--
Roger W. Norman
SirMusic Studio
Purchase your copy of the Fifth of RAP CD set at www.recaudiopro.net.
See how far $20 really goes.
"anthony.gosnell" <live...@metroweb.nospam.co.za> wrote in message
news:3fb8cc68$0$64...@hades.is.co.za...
Bob Cain
For the purpose of this group, none I suppose. There are
subtle but important consequences in the world of DSP where
I often encounter an unawareness or oversight of the fact
and correcting it has become pretty much a knee jerk.
Bob Cain
Les Cargill wrote:
>
> Bob Cain wrote:
> >
> > Les Cargill wrote:
> > > >
> > > > What's too often forgotton is that a signal of finite
> > > > length, like a song or a single drum hit, requires a whole
> > > > bunch of them. An infinite number in fact.
> > >
> > > Not if it's bandlimited.
> >
> > No finite number of sin waves (which are defined as infinite
> > in duration by Fourier theory) can sum to a signal that goes
> > to zero and stays there nor comes from an infinite past of
> > zero
>
> Meaning where t is undefined?
No, where it extends to the infinite. There are no ideally
bandlimited signals of finite duration.
>
> > and suddenly becomes non-zero. Any finite number of
> > sin waves will sum to a periodic signal that repeats forever
> > in both directions.
> >
>
> But PCM is not so much a Fourier thing as it is a
> Nyquist Theorem thing. The FT in Nyquist-land is a DFT, not
> a continuous FT.
FWIW, the DFT has the same property.
>
> I am pretty sure that PCM 44.1/16 can represent a drum hit
> properly, so long as all the constraints on domain and range are
> met (IOW no overs).
Yes, close enough for all practical purposes.
>
> Did I miss your point?
The point was pedantic, and as I responded to ulysses, it
isn't particularly signifigant for the purposes of this
group.
anthony.gosnell
> >> For more information please read the rec.audio.pro faq,
> >
> > Where are these?
> > Isn't this something which should be posted once a week for the
> > benefit of newbies?
>
> Try google, it will get you to the HTML version of the rec.audio.pro FAQ
> version in flash.
I tried that, but only found endless discussion on whether to have an HTML
version or not and a few links to a site which doesn't exist anymore.
Jay - atldigi
mikero...@hotmail.com (mike rogers) wrote:
The nuances that are the content of higher frequencies than the Nyquist
limit of your current sample rate, yes. It's just what's above 1/2 the
sampling rate that you lose (albeit a little less than 1/2 Fs in
practice due to filters etc.). I wasn't aiming specifically at you -
just reinforcing an important point that often gets lost.
Jay - atldigi
<ar...@hotpop.com> wrote:
> "mike rogers" <mikero...@hotmail.com> wrote in message
> news:772f2e86.03111...@posting.google.com
>
> > Jay, I think that I am agreeing with you and Scott. A waveform is not
> > a simple sine which, if you know the peak high and low amplitudes the
> > wave form is easy to reconstruct. But, as you say, when a wave is a
> > complex and made up of several sines then higher frequency sampling is
> > required to capture the nuances.
>
> That's not how it works. It doesn't take any more samples to properly
> capture a complex wave as a simple one, given that both waves have the
> same bandpass.
>
True - the key point is that those "small nuances" in the waveform
represent higher frequencies, and only audio of frequencies that are
higher than your Nyquist limit (1/2 the sample rate theoretically,
though a little less in practice due to filtering realities) are lost.
So yes, a triangle wave with a fundamental at 20k won't look like a
triangle because the upper harmonics are filtered, but you won't notice
when listening to it. A 1k trianlge will look like a triangle, but even
it won't have any harmonics above Nyquist, and it will still sound just
like you expect it to. It's ONLY the frequency components above Nyquist
that are lost, but within the bandwidth of the system, you gain nada,
zip, zero as far as accuracy or detail by having a higher sampling rate
(filter and noise shaping issues nonwithstanding). Higher smapling rates
allow you to capture higher frequencies but do not improve the accuracy
or detail of what you can capture within the bandwidth of lower sample
rate systems.
Jay - atldigi
> In article <atldigi-530ACB...@news1.news.adelphia.net>
> atl...@aol.com writes:
>
> > You won't be able to sample a 10k wave with a 10k sample rate. Do a
> > search for Shannon and Nyquist.
>
> Sure you can, but you'll get a bogus result, and that's what he's
> shown. So yeah, you can prove that it's possible to violate the rules
> of the sampling theorem and get back something different than what you
> put in.
LOL! I stand corrected! You can indeed do it wrong and get a bad result.
Arny Krueger
You're telling me that you searched google web pages for rec.audio.pro faq
and couldn't find anything?
That's not what happened when I tried it!
Blind Joni
>can move on to the other details. If not, later conclusions will be
>erroneous as they stem from an erroneous premise.
And this NEVER happens in the pro audio world!!!
John A. Chiara
SOS Recording Studio
Live Sound Inc.
Albany, NY
www.sosrecording.net
518-449-1637
Justin Ulysses Morse
> "Justin Ulysses Morse"
> > But if you filter out everything over 20kHz, then a 20kHz wave has no
> > choice but to be a perfect sine wave.
>
> What about a perfect sawtooth or a perfect square?
Do you know what a sawtooth wave is? Or a square wave? It's a sine
wave with a bunch of harmonics added to it. Those harmonics are simply
higher-frequency sine waves. A 10kHz square wave contains a 10kHz sine
wave and a bunch of other sine waves, but those other sine waves are
all 30kHz and higher. There is NO in-band information in a 10kHz
square wave other than the 10kHz sine wave.
> According to your logic sampling at 44k works because anything which isn't a
> sine wave should be converted to a sine wave before you sample it so that it
> can be reproduced correctly. (as a sine wave of course)
No, you can have in-band harmonics of a 100Hz tone, and you've got room
for all kinds of different wave shapes.
> What's happened here is that you have defined analog theory in terms of
> digital sampling theory.
The only "term" that's defining anything in this case is the bandwidth
limitation. Once you understand what this means, you'll see there are
no compromises being made other than bandwidth.
> Of course digital sampling reproduces analog exactly if you define analog
> only as what can be reproduced digitally.
Read up on the subject and then come back and talk to us.
ulysses
Justin Ulysses Morse
> It's important to know where to draw the line. Generally this is
> dictated by the pocketbook for most people.
I would say it's more usually dictated by the limits of human hearing.
People who are willing to spend whatever is necessary for quality audio
still won't be able to hear the ultrasonic content.
ulysses
Garthrr
Morse <ulysse...@rollmusic.com> writes:
>I would say it's more usually dictated by the limits of human hearing.
>People who are willing to spend whatever is necessary for quality audio
>still won't be able to hear the ultrasonic content.
Ulysses,
Whats your take on the claim that some people make that the ultrasonic content,
while inaudible by definition, has an audible effect on the audio within the
passband?
Garth~
"I think the fact that music can come up a wire is a miracle."
Ed Cherney
Justin Ulysses Morse
> Ulysses,
> Whats your take on the claim that some people make that the
> ultrasonic content, while inaudible by definition, has an audible
> effect on the audio within the passband?
My first reaction is to ask: "How?" If this were true, there would be
some mechanism causing it that we would already know about. Maybe
there is. I don't pretend to be expert or knowledgeable on the
subject. However, my feeling based on what I know so far is that the
entirety of the perceived differences between CD and higher sampling
rates is due to artefacts of the digital filters. This is consistent
with the fact that some people don't notice any difference at all while
others hear a "big" difference. The quality of the conversion, and a
host of implementation variables make it a moving target that would be
subtle at best anyway.
Maybe if I ever find myself in a well-controlled listening environment
that contains ultrasonic tweeters, equipment capable of driving them
precisely, and source material that was recorded in a manner to make it
relevant...Maybe I'll learn something new. It's hard enough to get a
decent listening experience to happen without worrying about
ultrasonics though. The instances of an entire record/reproduce chain
being accurate into the ultrasonic range, from musician to storage to
listener, must be so few that the number of people who have experienced
it firsthand could probably fit into a small schoolbus.
ulysses
Mike Rivers
> You're telling me that you searched google web pages for rec.audio.pro faq
> and couldn't find anything?
>
> That's not what happened when I tried it!
I suspect that he couldn't find a message that said in flashing neon
lights: "You can find an FAQ for rec.audio.pro at the web page
http://www.recaudiopro.net/faq/index.html"
It's true that the FAQ document hasn't been updated recently, but
neither have basic principles of acoustics, electronics, physics, and
mathematics. It's not the intent of the FAQ to tell you what you buy
or how to hook it up. That's what we have magazines, newsgroups,
salespeople, and real live paid human consultants.
Garthrr
Morse <ulysse...@rollmusic.com> writes:
>The instances of an entire record/reproduce chain
>being accurate into the ultrasonic range, from musician to storage to
>listener, must be so few that the number of people who have experienced
>it firsthand could probably fit into a small schoolbus.
>
>ulysses
Point taken.
Arny Krueger
news:20031118025019...@mb-m06.aol.com
> In article <171120032330436040%ulysse...@rollmusic.com>, Justin
> Ulysses Morse <ulysse...@rollmusic.com> writes:
>
>> I would say it's more usually dictated by the limits of human
>> hearing. People who are willing to spend whatever is necessary for
>> quality audio still won't be able to hear the ultrasonic content.
>
> Ulysses,
> Whats your take on the claim that some people make that the
> ultrasonic content, while inaudible by definition, has an audible
> effect on the audio within the passband?
It's a testable hypothesis and there are no valid tests that support it.
The world is now full of equipment that people can use to test the
hypothesis for themselves, and many people do try to test it for themselves.
However most people don't do the testing with very much rigor. So the
results might be valid to them, but in the larger context it is hard to give
their test results much global meaning.
I've made (with some help) all the unique pieces required to do rigorous
testing freely available at www.pcabx.com .
Doing rigorous tests isn't that gosh awful much work if you have a good DAW
and monitoring system at your disposal.
*All* the people who've done non-rigorous tests have all the pieces required
for a rigorous test at their disposal.
So why don't more people do their own rigorous tests? It can be done in less
than an hour.
Garthrr
<ar...@hotpop.com> writes:
>*All* the people who've done non-rigorous tests have all the pieces required
>for a rigorous test at their disposal.
>
>So why don't more people do their own rigorous tests? It can be done in less
>than an hour.
Is there perhaps one or two critical mistakes they are making or is it a lot of
things.
Mike Rivers
> > It's important to know where to draw the line. Generally this is
> > dictated by the pocketbook for most people.
>
> I would say it's more usually dictated by the limits of human hearing.
> People who are willing to spend whatever is necessary for quality audio
> still won't be able to hear the ultrasonic content.
But people who are only willing to spend as little as possible to get
sound coming at them from a loudspeaker usually won't hear all they're
able to hear.
Mike Rivers
> Whats your take on the claim that some people make that the ultrasonic content,
> while inaudible by definition, has an audible effect on the audio within the
> passband?
I have read of experiments with golden ears not being able to detect
the presence of mathematically accurate filters cutting things above
20 kHz.
I did sit in on a listening experiment where a singer/guitarist played
into a pair of Sennheiser MKH-800 mics (which have usable response up
to at least 40 kHz. From the mics, the signal chain was through a True
preamp (recognized as unexciting but very clean and accurate) to two
sets of back-to-back Lavry A/D and D/A converters, one set running at
48 kHz, the other set running at 96 kHz. The outputs of the two
converters, as well as the direct preamp output were run into a
Coleman switchbox, accurately level-matched, and played through a
Bryston amplifier and PMC speakers. (yes, factory reps were involved
in the demo - it was set up by a dealer)
I was quite surprised and pleased that I could hear a difference
between the 96 and 48 kHz converter chain, and also between the
converter chain and the direct preamp output. It wasn't dramatic by
any sense, but given the A/B/C environment, it wasn't difficult to
tell which was which. As expected, 96 kHz sounded better than 48 kHz,
and no converters at all sounded best. Everything sounded a little
diffrent than the orignal source, but that could easily be attributed
to the difference in the acoustics of the listening room and the
speakers.
This didn't make me want to go out and buy a set of Lavry converters,
but it did renew my faith in the fact that there really is something
going on that I could hear in one case and not in the other. It would
be interesting to repeat the experiment using ideal low pass filters
rather than having A/D and D/A converters do that job, but part of the
deal was to sell wide-bandwidth goodies.
Ty was there. Maybe he heard something different.
Scott Dorsey
Garthrr <gar...@aol.com> wrote:
>In article <171120032330436040%ulysse...@rollmusic.com>, Justin Ulysses
>Morse <ulysse...@rollmusic.com> writes:
>
>>I would say it's more usually dictated by the limits of human hearing.
>>People who are willing to spend whatever is necessary for quality audio
>>still won't be able to hear the ultrasonic content.
>
>Ulysses,
>Whats your take on the claim that some people make that the ultrasonic content,
>while inaudible by definition, has an audible effect on the audio within the
>passband?
This is the ONLY good argument for high sample rate systems.
And it might be a good one or it might not be. I dunno. I can give you
some nice data either way.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Tommi
"Scott Dorsey" <klu...@panix.com> wrote in message
news:bpde01$7uh$1...@panix2.panix.com...
But having an audible effect and having a more pleasing sensation isn't
necessarily the same thing. I posted this already here as a new post, but
I'll still include a link also here, if someone's interested, here are the
results from a test where the effects of ultrasonics were measured:
http://www.yamashirogumi.gr.jp/kumigashira/jnp-hse.pdf
On the other hand, it is indeed possible that ultrasonics may produce
interference tones when they're mixed in air with audible frequencies, at
least at higher sound pressures.
Arny Krueger
news:znr1069154989k@trad
> In article <1JCdndH384t...@comcast.com> ar...@hotpop.com
> writes:
>> You're telling me that you searched google web pages for
>> rec.audio.pro faq and couldn't find anything?
>> That's not what happened when I tried it!
> I suspect that he couldn't find a message that said in flashing neon
> lights: "You can find an FAQ for rec.audio.pro at the web page
> http://www.recaudiopro.net/faq/index.html"
Well, he got to the right place anyhow, I guess.
Ironically, the link is broken at the moment, at least from Comcast.
> It's true that the FAQ document hasn't been updated recently, but
> neither have basic principles of acoustics, electronics, physics, and
> mathematics. It's not the intent of the FAQ to tell you what you buy
> or how to hook it up. That's what we have magazines, newsgroups,
> salespeople, and real live paid human consultants.
I've looked at the FAQ pretty hard and critically from time to time, and it
still looks really good to me.
anthony.gosnell
>
> In article <1JCdndH384t...@comcast.com> ar...@hotpop.com writes:
>
> > You're telling me that you searched google web pages for rec.audio.pro
faq
> > and couldn't find anything?
> >
> > That's not what happened when I tried it!
>
> I suspect that he couldn't find a message that said in flashing neon
> lights: "You can find an FAQ for rec.audio.pro at the web page
> http://www.recaudiopro.net/faq/index.html"
I looked through at least 20 threads and only came up with the old address
and endless discussion about what to include in the faq and if it should be
available on the web etc. I also found a list of the questions, but not the
answers.
My point wasn't so much about not being able to find it but that the address
should be posted say once a week so that newbies can find it easily.
Newbies in my experience don't know how to use google.
> It's true that the FAQ document hasn't been updated recently, but
> neither have basic principles of acoustics, electronics, physics, and
> mathematics. It's not the intent of the FAQ to tell you what you buy
> or how to hook it up. That's what we have magazine
One of the gear shops should start running a customer ratings thing like
Amazon does. The FAQ could point to it for a quick answer on what is the
best mic for under $200 type questions. The gear shop could generate a lot
of sales from a service like that.
Arny Krueger
So how were these tests bias-controlled?
Arny Krueger
> In article <u9ydndcwd8g...@comcast.com>, "Arny Krueger"
> <ar...@hotpop.com> writes:
>
>> *All* the people who've done non-rigorous tests have all the pieces
>> required for a rigorous test at their disposal.
>>
>> So why don't more people do their own rigorous tests? It can be done
>> in less than an hour.
>
> Is there perhaps one or two critical mistakes they are making or is
> it a lot of things.
Listening test rigor is composed of three basic things:
(1) Level matching (+/-) < 0.1 dB
(2) Time synchronization (ideally +/- 2 mSec)
(3) Bias control (Double Blind or equivalent).
Often all three are missing, sometimes just one or two. Miss any of them and
the so-called test is highly questionable, particularly if it develops
positive results. I can show that missing any of these three requirements by
a significant amount will produce positive results for comparisons between
equipment that is absolutely identical. Honoring them will change the
outcome of that listening test comparing identical equipment to be negative.
There are other requirements but they are a little less glaringly important.
See the sidebar "10 Requirements...." posted at www.pcabx.com . The three
basics address false positives, the remaining seven mostly address false
negatives.
Mike Rivers
> I also found a list of the questions, but not the
> answers.
Well, you know that FAQ stands for "Frequently Asked QUESTIONS."
> My point wasn't so much about not being able to find it but that the address
> should be posted say once a week so that newbies can find it easily.
I always wondered why there was even a FAQ for this newsgroup. Both
the questions and the answers are constantly changing, yet the
fundamentals never change. I guess it's there just because people sort
of expect it. There are probalby three of four books (yes, whole
books), reference to which would probably make for a great FAQ for
this newsgroup, but then you know that the next most frequently asked
question would be "Does anyone know a web site where I can read a copy
of Jeff Cooper's 'How to build a studio' book?
> Newbies in my experience don't know how to use google.
They should learn. It's one of the basic skills today. They should
also learn that they should read a newsgroup for a week or so before
posting a question or comment, and get to know who's who, who they're
flaming or insulting, and who's "condescending" to speak to them.
> One of the gear shops should start running a customer ratings thing like
> Amazon does. The FAQ could point to it for a quick answer on what is the
> best mic for under $200 type questions. The gear shop could generate a lot
> of sales from a service like that.
This would imply that the newsgroup approves of those ratings as
legitimate, and I doubt that anyone who knows mics (or compressors, of
preamps, or DAW software) would do that. If you read the newsgroup for
a while (as suggested above) you'll get a sense of who has some
credibility, who's totally incredible, and whose opinions you should
consider but not as gospel. This is why the usual answer to just about
any question is "Go listen for yourself." And you knowm, that's the
right answer.
Arny Krueger
news:uDtub.78$Lo...@reader1.news.jippii.net
> "Scott Dorsey" <klu...@panix.com> wrote in message
> news:bpde01$7uh$1...@panix2.panix.com...
>> In article <20031118025019...@mb-m06.aol.com>,
>> Garthrr <gar...@aol.com> wrote:
>>> In article <171120032330436040%ulysse...@rollmusic.com>, Justin
>>> Ulysses Morse <ulysse...@rollmusic.com> writes:
>>>
>>>> I would say it's more usually dictated by the limits of human
>>>> hearing. People who are willing to spend whatever is necessary for
>>>> quality audio still won't be able to hear the ultrasonic content.
>>>
>>> Ulysses,
>>> Whats your take on the claim that some people make that the
>>> ultrasonic content, while inaudible by definition, has an audible
>>> effect on the audio within the passband?
>>
>> This is the ONLY good argument for high sample rate systems.
This is a hypothesis that is reasonably easy to test given how many modern
mics and speakers have response > 20 KHz, and the fact that such a high
proportion of digital recording equipment is 24/96 capable, i.e., capable of
powerful clean response up to 44 KHz or more.
Here's my shot at testing the hypothesis:
http://www.pcabx.com/technical/sample_rates/index.htm .Try it for yourself!
>> And it might be a good one or it might not be. I dunno. I can give
>> you some nice data either way.
>
> But having an audible effect and having a more pleasing sensation
> isn't necessarily the same thing. I posted this already here as a new
> post, but I'll still include a link also here, if someone's
> interested, here are the results from a test where the effects of
> ultrasonics were measured:
>
> http://www.yamashirogumi.gr.jp/kumigashira/jnp-hse.pdf
Trouble is, this paper's test conditions and conclusions are themselves
highly controversial. For example, figure 1 in that paper shows that the
test setup had appreciable differences in response *below* 20 KHz.
> On the other hand, it is indeed possible that ultrasonics may produce
> interference tones when they're mixed in air with audible
> frequencies, at least at higher sound pressures.
It seems like it is more than a possibility. There are speakers that are
being promoted for use at exhibits, etc. that claim to produce a highly
focused audible sound by means of projecting extremely intense ultrasonic
sounds. I've never experienced them, but they seem to be working for some
people.
However, there's copious evidence that even wide-bandwidth recordings of
music don't often have highly intense ultrasonics. In fact, their ultrasonic
content is often 40 dB and more below mid and low frequency levels.
http://world.std.com/~griesngr/intermod.ppt as one example.
http://www.cco.caltech.edu/~boyk/spectra/spectra.htm for another.
Mike Rivers
> So how were these tests bias-controlled?
No control, just listening wherever people happened to be sitting, or
where they moved around the room. It wasn't a test, it was a
demonstration, and I don't expect it to prove anything other than what
I observed myself. All I claim is that I was able to detect some small
differences.
S O'Neill
> Listening test rigor is composed of three basic things:
>
> (2) Time synchronization (ideally +/- 2 mSec)
I've read in other groups that this is the hard part. You have to make the
transition between A and B (or whatever) undetectable. A click, a gap, a
momentary burst of "lightness" or "air" or "warmth" or "grunge" can all be cues
that will expose the fact that a parameter has changed. And apparently this can
occur in the millisecond range or less.
> ...
Arny Krueger
news:bpgbn0$n0a$1...@woodrow.ucdavis.edu
> Arny Krueger wrote:
>> Listening test rigor is composed of three basic things:
>> ...
>> (2) Time synchronization (ideally +/- 2 mSec)
> I've read in other groups that this is the hard part.
It most certainly can be a very hard part.
> You have to
> make the transition between A and B (or whatever) undetectable. A
> click, a gap, a momentary burst of "lightness" or "air" or "warmth"
> or "grunge" can all be cues that will expose the fact that a
> parameter has changed. And apparently this can occur in the
> millisecond range or less.
Whether clicks and pops of reasonable size at the transition point are a
problem is controversial.
I'm of the opinion that a minor click is a good thing, because it bookmarks
the transition point for the ear.
Others are death on them.
There's two ways to avoid clicks, one being to do a quick fade out followed
by a quick fade in, the other to have precise time synch and no switching
transients. The problem with fade-in/fade-out is that the required variable
gain amplifier circuit can be a point where distortion and noise are
introduced into the signal path, or are potentially introduced into the
signal path. I have to make the latter distinction because there's quite a
bit of concern about the DBT test environment masking differences that would
be otherwise audible.
More significant to me is the fact that if the one of the alternatives leads
or lags the other by too much, one can detect an echo effect and reliably
identify the unknowns, even if they are otherwise identical.