Is vinyl high-end?
jim
As a newcomer to this group I have been reading it with interest over
the past few months and I have to admit I am staggered by what is
posted here:
The number of posts concerning turntables and vinyl!
How can something that consists of scraping rock through plastic to
reproduce sounds that have numerous clicks and pops, limited
frequency reponse, wow and flutter (not to mention little tolerance
to people walking past) be considered as "high-end"?
The actual ability of vinyl to reproduce sounds across the audible
spectrum is pretty poor. To compensate for this the bottom end has to
be
boosted and compression has to be applied to make it sound halfway
decent. Now I do not have a problem if you like listening to music
with lots of compression - but can it *really* be considered
"high-end"?
If I turned up here and started a discussion about this top value
three-in-one portable sound system with 1000 watts of power
(pmpo)that I
got for $300 you would all tell me to go away, so could someone
*please* explain:
Why is vinyl considered high-end?
Jim
P.S. This is a genuine post and not trolling for a big argument. I do
not understand why vinyl is considered high-end and would welcome
some explanations.
Harry Lavo
news:9r2ad...@enews1.newsguy.com...
> Folks I need enlightenment.
>
> As a newcomer to this group I have been reading it with interest over
> the past few months and I have to admit I am staggered by what is
> posted here:
>
> The number of posts concerning turntables and vinyl!
>
<ship>
>
> Why is vinyl considered high-end?
>
> Jim
>
> P.S. This is a genuine post and not trolling for a big argument. I do
> not understand why vinyl is considered high-end and would welcome
> some explanations.
And this is not a snide remark. One listen is worth a thousand words and
caveats. Find a local audio club or dealer who has a really top quality
vinyl rig set up, and spend and hour or two listening. Then you will
understand.
One giveaway: if your experience is only with a phono setup that is prone to
"floor bounce" then you are dealing with a system setup that cannot possibly
yield good sound. So I am assuming that you have never really heard a
near-SOTA vinyl system, circa CY2001.
On such a system, the negatives are minimized and vinyl's virtues are
maximized....the scientists here will call it euphonic distortion, and it
may be, but those of us who love vinyl feel that at this level of
reproduction it often sounds more lifelike that the vast majority of cd
material out there. Another aspect is that there is much older vinyl
floating around that is not on CD and is available for bargain prices. Much
is in good shape, and even that that isn't can often be restored with proper
cleaning and preservative treatment.
Give it a try, and then see if you still have the same questions.
Harry
You Know Who ~
You appear to have some misconceptions about vinyl.
In the four channel mode (yes, there was such a format) it was shown that
the LP could reach a frequency response of well above 30cps.
There was no compression. There was an equalization of the signal which any
decent phono preamp could deal with and restore to a very flat frequency
response. (by the way, the often used dithering, which is deliberately
adding noise to digital recordings could be attacked in the same way,
though, like equalization, it is not something that interferes.
Analog is a continuous signal while digital is still a sampling medium, and
many of us believe it doesn't sample often enough, especially in the high
frequencies.
The bottom line is that I have some music on both CD and LP. In most (not
all) cases, the LP sounds better.
--
You know who~
TAV http://www.audiophilevoice.com/
Cats and audio links: http://you_know_who.home.att.net/
----------------------------------------------------------------
"There are two means of refuge from the miseries of life:
music and cats." -- Albert Schweitzer
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
"jim" <j...@tasmail.com> wrote in message
news:9r2ad...@enews1.newsguy.com...
Chris Johnson
> How can something that consists of scraping rock through plastic to
> reproduce sounds that have numerous clicks and pops, limited
> frequency reponse, wow and flutter (not to mention little tolerance
> to people walking past) be considered as "high-end"?
Clicks and pops are intermittent sounds. The frequency response isn't
as limited as you might think- wow is a mechanical problem dealt with by
inspired design or just plain mass- flutter's more an issue with tape
players, such as cassette players- footsteps are more a problem for
mid-end gear rather than the cheapest or most expensive- but none of this
addresses the real question.
My opinion is that it's a question of linearity in the midrange. It
takes quite a bit of digital trickery to get a comparable amount of
linearity out of digital- the motions of the rock scraping through plastic
are inherently a linear process.
There are high levels of additional noise produced- but they are either
intermittent, or strongly frequency dependent, such as rumble. For years
this has all been lumped under 'noise', but it's more common today to
examine spectral content, and it's unusual for an analog/mechanical
process to produce perfectly uniform noise, or even the moderately uniform
noise floor of digital truncation.
Honestly, it's not as bizarre as you think. People still use
reel-to-reel tape professionally in audio work, too.
Chris Johnson
Stewart Pinkerton
>How can something that consists of scraping rock through plastic to
>reproduce sounds that have numerous clicks and pops, limited
>frequency reponse, wow and flutter (not to mention little tolerance
>to people walking past) be considered as "high-end"?
Because top-quality vinyl rigs are *very* expensive (top cartridges
alone cost $2-6,000), which puts then well into the aspirational
range. Additionally, getting the very best results from this precision
electr0-mechanical process requires great attention to detail and a
lot of 'tweaking'. This qualifies vinyl as high-end, regardless of
absolute sound quality.
>The actual ability of vinyl to reproduce sounds across the audible
>spectrum is pretty poor. To compensate for this the bottom end has to
>be boosted and compression has to be applied to make it sound halfway
>decent. Now I do not have a problem if you like listening to music
>with lots of compression - but can it *really* be considered
>"high-end"?
Not in my opinion, but no doubt you'll hear a *lot* of opposing
opinion in this arena! :-)
--
Stewart Pinkerton | Music is art, audio is engineering
Stewart Pinkerton
>In article <9r2ad...@enews1.newsguy.com>, jim <j...@tasmail.com> wrote:
>> How can something that consists of scraping rock through plastic to
>> reproduce sounds that have numerous clicks and pops, limited
>> frequency reponse, wow and flutter (not to mention little tolerance
>> to people walking past) be considered as "high-end"?
>
> Clicks and pops are intermittent sounds. The frequency response isn't
>as limited as you might think- wow is a mechanical problem dealt with by
>inspired design or just plain mass- flutter's more an issue with tape
>players, such as cassette players- footsteps are more a problem for
>mid-end gear rather than the cheapest or most expensive- but none of this
>addresses the real question.
They are however real effects which would seem to eliminate vinyl as a
contender in any true high-fidelity sound system - as distinct from
that strange term, 'high-end'.
> My opinion is that it's a question of linearity in the midrange. It
>takes quite a bit of digital trickery to get a comparable amount of
>linearity out of digital- the motions of the rock scraping through plastic
>are inherently a linear process.
It takes no 'trickery' at all to get a linearity from digital which is
at least ten times *better* than that of vinyl in the midrange, and
over a *hundred* times better at the frequency extremes. If you think
that vinyl is 'inherently linear' at any frequency, perhaps you should
take some basic engineering courses, as this is demonstrably untrue,
from surface roughness at one end of the dynamic range to elastic
limits at the other.
> There are high levels of additional noise produced- but they are either
>intermittent, or strongly frequency dependent, such as rumble. For years
>this has all been lumped under 'noise', but it's more common today to
>examine spectral content, and it's unusual for an analog/mechanical
>process to produce perfectly uniform noise, or even the moderately uniform
>noise floor of digital truncation.
It remains tha case however, that the noise floor of vinyl is 10-20
times higher than that of 16/44 digital, in relation to peak recorded
level, even for the very best vinyl.
> Honestly, it's not as bizarre as you think. People still use
>reel-to-reel tape professionally in audio work, too.
What has this to do with vinyl?
Arny Krueger
news:9r2ad...@enews1.newsguy.com...
> Folks I need enlightenment.
> Why is vinyl considered high-end?
I think you've made a critical error in trying to relate the
technical performance of vinyl to the phrase "high end".
"High End" is a marketing term. For example, "Rolls Royce" is a high
end brand of automobile. However few if any automotive authorities
suggest that a Rolls Royce automobile is a technological wonder in
terms of raw performance or the automotive technology it embodies.
The car is a out-accelerated, out-handled, out -stopped, and out-fuel
economized by any number of cars with vastly lower price tags.
"High end" relates to price tag and status.
Vinyl itself is neither "high end" or "mid fi" or "low end". However
it remains possible for there to be high end vinyl players. However,
the majority of vinyl playback equipment being sold at this time is
designed to be used by "turntablists" who work in dance clubs and at
parties in private homes. Most new titles being released in the vinyl
format are directed towards the dance market, not the high end of
audio.
Howard Ferstler
> "High End" is a marketing term. For example, "Rolls Royce" is a high
> end brand of automobile. However few if any automotive authorities
> suggest that a Rolls Royce automobile is a technological wonder in
> terms of raw performance or the automotive technology it embodies.
> The car is a out-accelerated, out-handled, out -stopped, and out-fuel
> economized by any number of cars with vastly lower price tags.
>
> "High end" relates to price tag and status.
>
> Vinyl itself is neither "high end" or "mid fi" or "low end". However
> it remains possible for there to be high end vinyl players. However,
> the majority of vinyl playback equipment being sold at this time is
> designed to be used by "turntablists" who work in dance clubs and at
> parties in private homes. Most new titles being released in the vinyl
> format are directed towards the dance market, not the high end of
> audio.
Arny's point is well taken.
High end relates more to a state of mind than anything else,
since you can get excellent sound from a system that is not
normally considered high end. I think that if two-channel
audio is your bag, you can get sound from vinyl that is
adequate to be called high end, provided that you do not
mind the surface noise. And of course, if you purchase a
player that costs a bundle and resembles fine machine art,
you obviously have a high-end device.
Howard Ferstler
Stewart Pinkerton
>You appear to have some misconceptions about vinyl.
Actually, he seems to have nailed it pretty well........
>In the four channel mode (yes, there was such a format) it was shown that
>the LP could reach a frequency response of well above 30cps.
Indeed yes, about a thousand times higher! :-)
However, this signal was down by 20-30dB from the main signal, and
didn't last long once that old rock got scraping in those grooves...
>There was no compression.
Yes, there was, also peak limiting.
> There was an equalization of the signal which any
>decent phono preamp could deal with and restore to a very flat frequency
>response. (by the way, the often used dithering, which is deliberately
>adding noise to digital recordings could be attacked in the same way,
>though, like equalization, it is not something that interferes.
Excuse me? Dither is an *essential* part of the A/D process, and
purely beneficial to sound quality. It has no similarity whatever to
RIAA equalisation.
>Analog is a continuous signal while digital is still a sampling medium, and
>many of us believe it doesn't sample often enough, especially in the high
>frequencies.
Analogue may be a continuopus process, but it is one with very poor
resolution...............
BTW, why don't you consider 192k sampling (as used in two-channel
DVD-A) to be adequate, since it offers a bandwidth much wider than
that available from *any* analogue master tape?
>The bottom line is that I have some music on both CD and LP. In most (not
>all) cases, the LP sounds better.
To you, perhaps. My experience is exactly opposite.
Paul van der Hulst
>
> Folks I need enlightenment.
>
snip
>
> The actual ability of vinyl to reproduce sounds across the audible
> spectrum is pretty poor. To compensate for this the bottom end has to
> be
>
> boosted and compression has to be applied to make it sound halfway
> decent. Now I do not have a problem if you like listening to music
> with lots of compression - but can it *really* be considered
> "high-end"?
>
Over-all frequency response of vinyl is flat, if you use the phono
input sockets like you're supposed to.
On many CDs the same thing is done on the high end: pre-recording
amplification of high frequencies, neutralised by attenuation after
DA-conversion. It's called 'emphasis'
No points there I'm afraid
Paul
Joseph Oberlander
>
> Hi
>
> You appear to have some misconceptions about vinyl.
>
> In the four channel mode (yes, there was such a format) it was shown that
> the LP could reach a frequency response of well above 30cps.
>
> There was no compression. There was an equalization of the signal which any
> decent phono preamp could deal with and restore to a very flat frequency
> response. (by the way, the often used dithering, which is deliberately
> adding noise to digital recordings could be attacked in the same way,
> though, like equalization, it is not something that interferes.
>
> Analog is a continuous signal while digital is still a sampling medium, and
> many of us believe it doesn't sample often enough, especially in the high
> frequencies.
>
> The bottom line is that I have some music on both CD and LP. In most (not
> all) cases, the LP sounds better.
I agree. Good analog recordings do better justice. There is a
reason studios use high-speed tape several decades after it was
introduced. Analog in, analog out.
A good record player and stylus will sound incredible.
If you have good ears, you can hear some cd players generating the
tone in steps - this usually requires a stable tone like a pipe organ
or flute. You can hear the harmonics aligning differently than they
do in real life. It bothers me because I played woodwinds for ten
years and know how they sound - and CD just doesn't have RL sparkle
at higher frequencies because it takes incredibly fast sampling rates
to capture upper-range harmonics and CD is just not quite fast enough
for most insturmental music. Frequency and dynamic range isn't
everything. Neither is the absolute lack of a few clicks and pops.
Sonic accuracy to the original signal is.
I can understand - the technology is over ten years old and they had
to make serious compromises back then based upon the available
technology.
If you spend time around samplers you know this syndrome - the
waveforms are plainly jagged on a scope compared to a pure analog
signal. Most of the time, though, you don't get a tone all by itself
in music and adapt to how a CD sounds over the years.
Now, MP3s can thankfully sample at a high enough rate to do it
better, so it may well be the next step and come closer.
Arny Krueger
news:9r2liv$1u8$1...@bourbaki.localdomain...
> Hi
>
> You appear to have some misconceptions about vinyl.
> In the four channel mode (yes, there was such a format) it was shown that
> the LP could reach a frequency response of well above 30cps.
The proper number was more like 30 KHz. The given number is off by a
factor of 1,000.
> There was no compression.
Compression was optional and frequently used.
> There was an equalization of the signal which any
> decent phono preamp could deal with and restore to a very flat frequency
> response.
It is very rare to find a vinyl record/playback facility that was
flat within 1.0 dB over the range where the human ear is most
sensitive to differences in frequency response. One major issue that
was only infrequently addressed was the fact that the frequency
response of phono cartridges is dependent on the capacitive load
provided by the arm, the cables and the phono preamp. A few preamps
were built that provided means for adjusting this, but far fewer
audiophiles ever set these up properly. BTW for the last decade or so
that I listened to vinyl, I addressed this issue very faithfully, and
was rewarded with a great deal of listening pleasure.
In contrast, a CD player can be flat within 0.1 dB over the same
range of frequencies. It is well known that the human ear is almost
completely insensitive to even total elimination of frequencies in
music that are above 16 KHz. You can hear this for yourself at
http://www.pcabx.com/technical/low_pass/index.htm . Therefore claims
that vinyl is obviously superior because it has response above 16 KHz
are poorly-informed.
> (by the way, the often used dithering, which is deliberately
> adding noise to digital recordings could be attacked in the same way,
> though, like equalization, it is not something that interferes.
The difference is that dither is applied at very low levels, say 90
or more dB below peak levels. Performance at these levels is
generally moot, because it is masked by noise primarily in the
recording studio, but also to a lesser degree in the playback room.
Equalization affects all levels, and therefore equalization errors
are far more audible.
Until 1979 there was no exact solution to the design of RIAA
equalization networks, which hindered the accuracy of all playback
systems designed before that date.
> Analog is a continuous signal while digital is still a sampling medium, and
> many of us believe it doesn't sample often enough, especially in the high
> frequencies.
I think that having various beliefs is a lot of fun, but when
theories can be verified experimentally, they should be. It's easy
to confirm or deny whether or not 16 bit, 44 KHz sampling has audible
artifacts, and it has been repeatedly shown that critical listeners
cannot detect the interposition of a good quality job of 16 bit, 44
KHz coding. An early example of this can be found at
http://users.htdconnect.com/~djcarlst/abx_digi.htm . The same issue
can be investigated in a contemporary, personal context at
http://www.pcabx.com/technical/sample_rates/index.htm .
> The bottom line is that I have some music on both CD and LP. In most (not
> all) cases, the LP sounds better.
Of course, personal preference is sacrosanct and need not be related
to rhyme or reason. It just is.
Arny Krueger
news:9r2m52$29u$1...@bourbaki.localdomain...
> In article <9r2ad...@enews1.newsguy.com>, jim <j...@tasmail.com>
wrote:
> > How can something that consists of scraping rock through plastic to
> > reproduce sounds that have numerous clicks and pops, limited
> > frequency response, wow and flutter (not to mention little tolerance
> > to people walking past) be considered as "high-end"?
Aside from the issues of price and status that I cover in another
post, the performance issues you raise can be addressed more
directly.
> Clicks and pops are intermittent sounds. The frequency response isn't
> as limited as you might think-
The frequency response of vinyl is difficult or impossible to control
to the same degree as is commonly done in digital audio. It is a
reasonable goal, if true high fidelity is the objective, to maintain
amplitude response within 0.1 dB over the range which the ear is most
sensitive, which runs from about 100 Hz to 10 KHz. Not only is it
difficult or impossible to maintain this tolerance with vinyl, there
are other forms of amplitude distortion that can intrude on vinyl
playback. One of them is the fact that the actual sensititivity of a
cartridge can vary by a dB or more as small imperfections in the
flatness and eccentricity of the record cause the cartridge to vary
its distance from the groove.
>wow is a mechanical problem dealt with by inspired design or just
>plain mass.
Once the rotation of the turntable is stabilized by mechanical
design, there are still wow and flutter problems that remain because
records are rarely if ever perfectly mechanically flat. It is very
difficult to make a record sit perfectly flat once it has been stored
for a while. It is one of the ironies of audio that vinyl advocates
rant about jitter in digital audio when measurements show that jitter
(FM distortion) is not only far greater in magnitude, but generally
also more concentrated in frequency ranges where the ear is most
sensitive, in vinyl.
>- flutter's more an issue with tape players, such as cassette
> players-
One of the things that sold me on digital was the elimination of
flutter from my favorite piano pieces.
> footsteps are more a problem for mid-end gear rather than the
> cheapest or most expensive-
Without the use of elaborate seismic supports, it is difficult or
impossible to have what in modern terms would be a full-range audio
system incorporating vinyl. If you have a high-performance subwoofer
and want to play vinyl you either just add a low-frequency roll-off
for playing vinyl, or provide some means for further isolating the
turntable from the low frequency sound that the subwoofer can
provide.
>but none of this addresses the real question.
> My opinion is that it's a question of linearity in the midrange.
Measurements show that in terms of linearity in the midrange, vinyl
has very substandard performance.
> It takes quite a bit of digital trickery to get a comparable amount of
> linearity out of digital- the motions of the rock scraping through plastic
> are inherently a linear process.
Technically speaking, this is exactly NOT the case.
> There are high levels of additional noise produced- but they are either
> intermittent, or strongly frequency dependent, such as rumble.
Yup, and there is nothing like a good subwoofer to show that vinyl
adds a wealth of low frequency noise to whatever is recorded that
way. Furthermore tone arms used to play vinyl typically have very
pronounced resonance in the 5 to 30 Hz range which make good bass
performance almost impossible. These resonances can even influence
sound quality in the lower midrange.
> For years
> this has all been lumped under 'noise', but it's more common today to
> examine spectral content, and it's unusual for an analog/mechanical
> process to produce perfectly uniform noise, or even the moderately uniform
> noise floor of digital truncation.
The noise floor of vinyl is so high, particularly at low frequencies,
that it in no way compares to the noise floor provided by digital
audio.
> Honestly, it's not as bizarre as you think. People still use
> reel-to-reel tape professionally in audio work, too.
The charm of analog tape is that its sensitivity and frequency
response vary strongly with amplitude. If you look at the frequency
response curves of an analog tape recorder taken say every 10 dB, you
find that almost every one is different, and most of the upper ones
aren't even a uniform distance apart. This provides a very euphonic
means for reducing dynamic range, and a media that is very forgiving
of operator errors.
Keith Garratt
"jim" <j...@tasmail.com> wrote in message
news:9r2ad...@enews1.newsguy.com...
> Folks I need enlightenment.
>
> As a newcomer to this group I have been reading it with interest over
> the past few months and I have to admit I am staggered by what is
> posted here:
>
> The number of posts concerning turntables and vinyl!
Well, this should tell you that a fair number of people represented
in RAHE are passionate about vinyl - one way or the other.
> How can something that consists of scraping rock through plastic to
> reproduce sounds that have numerous clicks and pops, limited
> frequency reponse, wow and flutter (not to mention little tolerance
> to people walking past) be considered as "high-end"?
I agree. Makes about as much sense as throwing steak onto a grid over
an open fire or baking an apple pie with real apples.
> The actual ability of vinyl to reproduce sounds across the audible
> spectrum is pretty poor.
If this isn't just a wind-up, I'll just say that it is patently
obvious you haven't heard any decent vinyl.....
To compensate for this the bottom end has to
> be boosted and compression has to be applied to make it sound halfway
> decent. Now I do not have a problem if you like listening to music
> with lots of compression - but can it *really* be considered
> "high-end"?
"Vinyl boosted and compressed"? Ooh, I don't think so - merely
equalised to reduce stylus excursion which, as was pointed out in
another post, is entirely restored by the phono stage to a ruler-flat
frequency response line even by inexpensive phono stages.
> If I turned up here and started a discussion about this top value
> three-in-one portable sound system with 1000 watts of power
> (pmpo)that I
>
> got for $300 you would all tell me to go away, so could someone
> *please* explain:
No "portable sound system" could be considered "hi-fi" in my book -
my personal belief is that it isn't really true "hi-fi" if you can
lift it, let alone carry it about!!!
> Why is vinyl considered high-end?
"High-end"? What does this really mean? Surely it is an abbreviation
for the "high end of the price scale" for hi-fi equipment? Vinyl, per
se, cannot be considered "high-end" any more than CDs, Radio, Tape
etc. could automatically be called "high-end" (or not). What price 7"
vinyl 45s being played for hours on end, on a Dansette-type
"portable" record player, by millions of kids worldwide, a decade or
two ago? - Never was hi-fi, let alone high-end, but it was most
definitely vinyl.
> Jim
>
> P.S. This is a genuine post and not trolling for a big argument.
Yes Jim, this peeves me somewhat. - For some wacky reason so many
people in this group seem to consider analogue and digital music to
be mutually exclusive - much like the valves/SS arguments.
>I do not understand why vinyl is considered high-end and would welcome some
>explanations.
All I would say to this, having read the posts in this thread, is
that I fully endorse the view "give it a try" that has been expressed
elsewhere. Only then you will know for yourself. No amount of
argument or explanation can actually give you the experience or the
enlightenment you seek.
(Just my two penn'orth.)
Regards,
Keith G.
jj, DBT thug and skeptical philalethist
Chris Johnson <jinx...@sover.net> wrote:
> My opinion is that it's a question of linearity in the midrange.
Sigh, sorry to be contradicting you all the time, but that
is not supported by any evidence.
Linearity is a well-known technical term, and mean-squared
error, also known as SNR, is its complete measure.
>It
>takes quite a bit of digital trickery to get a comparable amount of
>linearity out of digital- the motions of the rock scraping through plastic
>are inherently a linear process.
Digital has much higher SNR than vinyl at any frequency below the
cutoff of the anti-alaising/imaging filters. This is something that
is a question of measurement, and it has been measured to death.
> There are high levels of additional noise produced- but they are either
>intermittent, or strongly frequency dependent, such as rumble. For years
>this has all been lumped under 'noise', but it's more common today to
>examine spectral content, and it's unusual for an analog/mechanical
>process to produce perfectly uniform noise, or even the moderately uniform
>noise floor of digital truncation.
And at least some people find the higher noise level to enhance
their listening experience. Nothing wrong with that.
--
Copyright j...@research.att.com 2001, all rights reserved, except transmission
by USENET and like facilities granted. This notice must be included. Any
use by a provider charging in any way for the IP represented in and by this
article and any inclusion in print or other media are specifically prohibited.
jj, DBT thug and skeptical philalethist
You Know Who ~ <you_kn...@att.net> wrote:
>There was no compression.
Whoa, there, nearly all vinyl has some level compression!
>Analog is a continuous signal while digital is still a sampling medium, and
>many of us believe it doesn't sample often enough, especially in the high
>frequencies.
This, as they say, is a belief you get to have, but one that
lacks any concrete evidence.
jj, DBT thug and skeptical philalethist
>reproduce sounds that have numerous clicks and pops, limited
>frequency reponse, wow and flutter (not to mention little tolerance
>to people walking past) be considered as "high-end"?
How about because the distortions so induced have been shown
to improve many people's perceptions of the limited perceptual
space available with 2 channel audio?
If you're going to suggest that ultimate analytical accuracy
is what is required in a 2-channel system, well, I'm just going
to
WATCH!
Chris Johnson
thug and skeptical philalethist) wrote:
> In article <9r2m52$29u$1...@bourbaki.localdomain>,
> Chris Johnson <jinx...@sover.net> wrote:
> > My opinion is that it's a question of linearity in the midrange.
> Sigh, sorry to be contradicting you all the time, but that
> is not supported by any evidence.
You know, this warms my heart :) I _know_ you're going to be
contradicting me all the time- I'd figured that out by now. It's cheering
to hear that you're sorry about it: this tells me that my efforts to
continue 'flagging' my opinion as such and not pursuing arguments too far,
are working. It suggests I'm coming across as less hostile and annoying
than I might otherwise be :)
> >It takes quite a bit of digital trickery to get a comparable amount of
> >linearity out of digital- the motions of the rock scraping through plastic
> >are inherently a linear process.
>
> Digital has much higher SNR than vinyl at any frequency below the
> cutoff of the anti-alaising/imaging filters. This is something that
> is a question of measurement, and it has been measured to death.
Certainly. In fact, if I'm not mistaken, even the very first CDs, which
are now widely accepted to have sounded awful, did manage to substantially
outperform vinyl for signal-to-noise ratio.
Since a great deal of vinyl out there sounds _substantially_ better,
more lifelike, and more convincingly real than the first, crude attempts
at CDs, given even passably acceptable vinyl playback equipment, I would
ask, why is this? That's the question I'm trying to answer. If my answers
aren't suitable, then they're not suitable. The question remains, and the
answer must be something else.
Note that I'm not talking about modern CD playback- that makes things
more difficult and I myself have tried to make CD rival or beat vinyl,
possibly with success. I'm talking about the old stuff, the first CDs and
the first, no-oversampling, players. Even that generated SNR superior to
records, yet the sound was appalling and many years passed before it got
good. I don't believe for one second that it's due to magic or good vibes-
in my way I am as hardcore-scientific-rationalist as you. There is a
technical reason for this disparity, and that is what good vinyl still
offers. Hence, my speculations.
Chris Johnson
Chris Johnson
<ar...@hotpop.com> wrote:
>> In the four channel mode (yes, there was such a format) it was shown that
>> the LP could reach a frequency response of well above 30cps.
>
> The proper number was more like 30 KHz. The given number is off by a
> factor of 1,000.
I'm keeping records of exactly who pounced on this obvious typo
for the sake of correcting the person publically ;)
Personally, I think the ideal response would have been:
>> the LP could reach a frequency response of well above 30cps.
'yes' ;)
Chris Johnson
Jeff Connelly
> reproduce sounds that have numerous clicks and pops, limited
> frequency reponse, wow and flutter (not to mention little tolerance
> to people walking past) be considered as "high-end"?
"Scraping rock through plastic"? Well, when you put it that way, how can
shining light on holes in metal result in "high-end" sound? ALL musical
REPRODUCTION systems have pros and cons. None have ever been invented that
actually create real live music (unless you consider something like the
Bosendorfer Reproducing Piano, but that is not "audio" in the accepted sense.
It's a real piano.)
> Why is vinyl considered high-end?
This is a meaningless question. This is like asking "Why are speakers
considered high-end?" or "Why are CDs considered high-end?".
It's not the medium per se - it's the reproductive quality of whatever element
in the reproductive chain you're talking about. Let's take a "really good"
system playing a "really good" vinyl record. Then let's take a "really bad"
system playing a "really bad" CD. Which is going to sound better? Sorry for
asking obvious questions, but you get the point.
The answer to your question is "Well, it's not, necessarily." High-end vinyl
reproduction is considered high-end. Crappy vinyl reproduction is considered
crap.
A record playback system is usually harder to get "really good" than a CD
system, simply because it's more complicated. But some are much simpler than
others, and still good. I don't get any clicks or pops on my "good" records
and I certainly don't feel the frequency response is limited. I get no
discernible wow or flutter. On the other hand, many less expensive CD players
can sound harsh and grainy. I'll say this - a $150 record player (including
arm and cartridge) is probably going to sound worse than a $150 CD player,
especially if it's setup carelessly. However, if it's setup correctly, it
might well sound better *in certain respects* than a "weak" CD player. The CD
will probably have it beat in dynamics and extension of frequency response.
On the other hand, I've heard $3000 record players that outperform $3000 CD
players. It depends a lot on the recording of course. There are some bad CD
recordings out there. There are some recordings on vinyl that are simply
higher quality than their CD counterparts (some of the Classic Records
releases, for example, beat the generic version CDs.) But all things
considered, for anyone who has to ask this question about vinyl, I would say
they'd probably be better served with a decent CD player. There are certainly
some "high-end" CD recordings out there too.
Alex Eisenhut
>
Ridiculous, isn't it? Almost as crazy as using a rock IN plastic to
reproduce audio? Like a transistor or an IC? Hmm?
Rob Gold
(top cartridges
>alone cost $2-6,000), which puts then well into the aspirational
>range. Additionally, getting the very best results from this precision
>electr0-mechanical process requires great attention to detail and a
>lot of 'tweaking'. This qualifies vinyl as high-end, regardless of
>absolute sound quality.
~~~~~~~~~~~~~~~~~~~~~~~~~~
$2,000-$6,000? Hey, Stewart, what magazines are you smoking? My Denon DL103D
cart cost me US$175 last year. Together with a Teres turntable and modified
Rega arm, my all-in cost is under $1,800, including shipping. Oh, add $24 for
my DIY wall-mount shelf. You could spend more on an LP playback system, but
I'm having trouble imagining what sonic benefits you might get (well, OK,
getting rid of LP's many obvious problems would be nice, but why do the results
have to sound like CD's?).
Rob Gold
N.P. -- Stravinsky's Apollon Musagette on Argo LP.
Loiskelly1
>REPRODUCTION systems have pros and cons. None have ever been invented that
>actually create real live music (unless you consider something like the
>Bosendorfer Reproducing Piano, but that is not "audio" in the accepted sense.
>It's a real piano.)
An excellent example of which is presented on Telarc CD-80489, Rachmaninoff: A
Window in Time.
As far as the analog/digital debate is concerned, one should consider that
neurons either fire or do not fire, making at least some portion of the process
of perception digital, in that discrete signals are being registered by the
brain. Of course, the mechanisms that precede neurological interpretation are
analog in nature- the basilar membrane, for one. Now why exactly can't we hear
a tone of say 1MHz? It is because there exists no position on the basilar
membrane corresponding to such a frequency. Now this may be an outlandish
number that I have offered, but the point is that as long as there is SOME
limitation to the frequency range (ie: bandwidth) that can be picked up by the
human auditory system, then there MUST be a corresponding bit rate
(mathematically determined via Shannon's information theory) that will provide
sufficient information capacity. In other words, if the desired bandwidth and
dynamic range are known, then the capacity required of the system (in bits per
second) CAN be calculated. This is the same whether we are talking about "high
end" audio or telecommunications, where a common transmission scheme (OC48)
routinely passes 32,256 voice channels multiplexed over a 2.4 Giga bps line.
Now how, exactly, does this relate to digital vs. vinyl in audio? It means
that digital, if not perfect, can at least be perfectible. It also means that
whether 16/44.1 offers up sufficient dynamic range and bandwidth is certainly
open to debate, but whether a digital scheme can be constructed that will
accommodate a given set of parameters is not.
Loiskelly1
It sounds positively space aged compared to the little bits of carbon jumping
around in the mouthpiece of a telephone. Let us never forget how well served
we were by analog technology. I am every bit as impressed by the cumulative
body of science that has brought us radio transmission as virtually anything
since.
Now playing- My guitar wants to kill your momma. F. Zappa
Stewart Pinkerton
>jim wrote:
>>
>> Folks I need enlightenment.
>>
>snip
>>
>> The actual ability of vinyl to reproduce sounds across the audible
>> spectrum is pretty poor. To compensate for this the bottom end has to
>> be
>>
>> boosted and compression has to be applied to make it sound halfway
>> decent. Now I do not have a problem if you like listening to music
>> with lots of compression - but can it *really* be considered
>> "high-end"?
>>
>
>Over-all frequency response of vinyl is flat, if you use the phono
>input sockets like you're supposed to.
Flat to within a dB or two from 50Hz to 15kHz, perhaps. CD is flat to
better than 0.1dB from 20Hz to 20kHz.
>On many CDs the same thing is done on the high end: pre-recording
>amplification of high frequencies, neutralised by attenuation after
>DA-conversion. It's called 'emphasis'
On many CDs made before 1985, perhaps, but that technique hasn't been
used for more than ten years.
>No points there I'm afraid
Be afraid, be very afraid..........
Stewart Pinkerton
>In article <9r2ad...@enews1.newsguy.com>, jim <j...@tasmail.com> wrote:
>> How can something that consists of scraping rock through plastic to
>> reproduce sounds that have numerous clicks and pops, limited
>> frequency reponse, wow and flutter (not to mention little tolerance
>> to people walking past) be considered as "high-end"?
>
> Clicks and pops are intermittent sounds. The frequency response isn't
>as limited as you might think- wow is a mechanical problem dealt with by
>inspired design or just plain mass- flutter's more an issue with tape
>players, such as cassette players- footsteps are more a problem for
>mid-end gear rather than the cheapest or most expensive- but none of this
>addresses the real question.
They are however real effects which would seem to eliminate vinyl as a
contender in any true high-fidelity sound system - as distinct from
that strange term, 'high-end'.
> My opinion is that it's a question of linearity in the midrange. It
>takes quite a bit of digital trickery to get a comparable amount of
>linearity out of digital- the motions of the rock scraping through plastic
>are inherently a linear process.
It takes no 'trickery' at all to get a linearity from digital which is
at least ten times *better* than that of vinyl in the midrange, and
over a *hundred* times better at the frequency extremes. If you think
that vinyl is 'inherently linear' at any frequency, you need to revise
your thinking, as this is demonstrably untrue, from surface roughness
at one end of the dynamic range to elastic limits at the other.
> There are high levels of additional noise produced- but they are either
>intermittent, or strongly frequency dependent, such as rumble. For years
>this has all been lumped under 'noise', but it's more common today to
>examine spectral content, and it's unusual for an analog/mechanical
>process to produce perfectly uniform noise, or even the moderately uniform
>noise floor of digital truncation.
It remains the case however, that the noise floor of vinyl is 10-20
times higher than that of 16/44 digital, in relation to peak recorded
level, even for the very best vinyl.
> Honestly, it's not as bizarre as you think. People still use
>reel-to-reel tape professionally in audio work, too.
What has this to do with vinyl?
--
Paul van der Hulst
>
> Paul van der Hulst <pa...@home.nl> writes:
>
> >Over-all frequency response of vinyl is flat, if you use the phono
> >input sockets like you're supposed to.
>
> Flat to within a dB or two from 50Hz to 15kHz, perhaps. CD is flat to
> better than 0.1dB from 20Hz to 20kHz.
Ofcourse you understand very well that I'm saying there is no delibirate
boost or compression of the whole low-frequency range. In those days
they were really trying to do it properly. No nitpicking _please_.
> >On many CDs the same thing is done on the high end: pre-recording
> >amplification of high frequencies, neutralised by attenuation after
> >DA-conversion. It's called 'emphasis'
>
> On many CDs made before 1985, perhaps, but that technique hasn't been
> used for more than ten years.
>
> >No points there I'm afraid
>
> Be afraid, be very afraid..........
>
Maybe you're right, I don't know much about current recording practices.
The point is still: no deliberate boost or compression of some frequency
range for either CD or LP
PS Disclaimer: Being an engineer with only limited hearing experience, I
know what I'm talking about technically, but I have absolutely _no_
opinion about which _sounds_ better. Both have their pro's and con's.
Maybe even more important: both have their supporters and opponents. I'd
rather not get stuck in between.
Knowing what I know, I know that I know very little. (Try to be creative
if you want to get me on this one.) Understanding the theoretical
details just isn't enough if you cannot translate them into qualitative
or even quantitative terms; audio perception theory just isn't very
sophisticated at the moment. Even a quantitative term such as THD is
only significant to a certain extent (one frequency or whole frequency
range etc. ). First we have to figure out which measurements really tell
us something about the audio quality. An example: for their new SACD/DVD
player, the engineers at PHILIPS made 15 modifications to their design
based purely on listening tests. Measurements revealed no differences at
all. (I know, PHILIPS isn't exactly a high-end company, but they do know
how to measure. Just try to get my point.)
If you don't know what and how to measure audio, how can you say
anything significant?
Instead of _fighting_ over which is best you better go to work and try
to figure out what really is the point of good sound reproduction.
PPS This thread is turning into a flame war (maybe my fault, I should
have known). Bye!
Paul
Rob Gold
>>There was no compression.
>
>Whoa, there, nearly all vinyl has some level compression!
~~~~~~~~~~~~~~~~~~
Hiya, JJ!
Correct, most real world vinyl LP's use compression, rolloff at the
frequency extremes and (often) summed-mono for the bass. Of course,
most real world CD's do exactly the same (without the summed-mono).
In neither format is this really necessary, but is done for reasons
as varied as marketing (to sound "good" on cheap walkmen), practical
issues (playing time limitations for vinyl offering true bass
extension), and sheer orneriness (not many folks *really* want to
hear what a trumpet playing fortissimo would sound like in their
living rooms).
There are two separate arguments here: the sorry commercial realities
of either storage/playback media, and the potential of each when
fully and sensitively implemented. I certainly don't want the idiots
who run record companies and radio stations messing with my logical
absolutes, so I stick with the latter perspective, which says:
1) The Rube Goldberg contraptions known as vinyl LPs have no right
sounding good at all, yet at their best they do. 2) Theoretically,
CDs *should* sound magnificent, but too often in my experience they
are flat, two-dimensional, uninvolving and lifeless-sounding. Go
figure. 3) For the time being I'll keep building my LP collection on
the cheap, and buy CD's (or whatever) to fulfill my new-music jones.
Rob Gold
N.P. Holst's St. Paul Suite on Lyrita LP
Ron-A...@magmar.freeserve.co.uk
<you_kn...@att.net> wrote:
>
>Analog is a continuous signal while digital is still a sampling medium, and
>many of us believe it doesn't sample often enough, especially in the high
>frequencies.
>
>The bottom line is that I have some music on both CD and LP. In most (not
>all) cases, the LP sounds better.
>
Most analogue LPs actually come from tape recorders that don't have a
"continuous signal" - the bias frequency certainly sets an upper
frequency limit. I think for my Revox the bias frequency is around 190
KHz - but that's from memory.
Whilst I'd agree that LP reply can sound better than a lot of CDs - it
does seem to cost a lot to achieve. For example, my LP player system
costs more than double the CD player, and requires a periodic
expensive cartridge up-date. And these days a lot of preamps are only
line level, so you need to buy another item of kit - plus
interconnects ...
If I was starting out afresh (wiothout an LP collection) I'd invest in
a better all-round system than spread the money out more thinly.
Hope this helps.
Stewart Pinkerton
>Stewart Pinkerton wrote:
>>
>> Paul van der Hulst <pa...@home.nl> writes:
>>
>> >Over-all frequency response of vinyl is flat, if you use the phono
>> >input sockets like you're supposed to.
>>
>> Flat to within a dB or two from 50Hz to 15kHz, perhaps. CD is flat to
>> better than 0.1dB from 20Hz to 20kHz.
>
>Ofcourse you understand very well that I'm saying there is no delibirate
>boost or compression of the whole low-frequency range. In those days
>they were really trying to do it properly. No nitpicking _please_.
Contrary to several statements in this thread however, the vast
majority of vinyl *is* compressed and peak limited, plus it's rolled
off above 15kHz to avoid overheating the cutter, while bass is summed
to mono below 80Hz to preserve groove depth, and rolled off below 40Hz
to maximise playing time.
>> >On many CDs the same thing is done on the high end: pre-recording
>> >amplification of high frequencies, neutralised by attenuation after
>> >DA-conversion. It's called 'emphasis'
>>
>> On many CDs made before 1985, perhaps, but that technique hasn't been
>> used for more than ten years.
>>
>> >No points there I'm afraid
>>
>> Be afraid, be very afraid..........
>>
>
>Maybe you're right, I don't know much about current recording practices.
>The point is still: no deliberate boost or compression of some frequency
>range for either CD or LP
See above.
>Instead of _fighting_ over which is best you better go to work and try
>to figure out what really is the point of good sound reproduction.
Been doing that for over thirty years, still got a long way to go! :-)
Curtis L. Coleman
(Arcam Alpha 9 amp, B&W DM 302 speakers, Velodyne sub) and a Music Hall MMF2
with an AT OC9 cartridge. Good CDs sound good. Well recorded and mastered
CDs sound great. Poor CDs sound flat and lifeless. Good LPs (e.g., the new
Classic Records "Kind of Blue", Loggins & Messina "Full Sail", etc.) sound
superb; unnoticeable surface noise, huge sense of space, coherent sound.
Poor LPs sound flat and lifeless. Comparing the "Kind of Blue" CD I have to
the Classic LP, the LP seems to have a larger and deeper soundstage.
Obviously, I don't have the ultimate in CD playback. But I don't have the
ultimate in LP playback either. I enjoy listening to music on both media. I
think it needs to be stated, however, that one can enjoy listening to LPs
on a sub-$1,000 system. For those of us with signficant numbers of LPs, the
choice is obvious, and people should not be scared away from resurrecting
their vinyl collection by statements that multi-kilobuck playback equipment
is necessary. Get a good belt-drive turntable and a decent or better
cartridge (Goldring, Grado, Benz, Clearaudio, or the AT OC9 even), clean
your old records and enjoy.
What is equally necessary, however, is a good record cleaning system. For
that, I recommend the Disc Doctor system of brushes and cleaning fluids.
Curt Coleman
Rob Gold <rgvi...@aol.com> wrote in message
news:9r6od...@enews4.newsguy.com...
Fill
and get into personal, bitter, reductionist discourses about what it
all means. Don't worry about specs. Worry about how it sounds to
you.Go listen to a good analog system at a store and a good digital
one. Better yet, listen to several. You'll probably prefer one or the
other. You don't listen to specs, you listen to music.
P h i l i p
______________________________
"Whenever a friend succeeds, a little something in me dies."
- Gore Vidal
PETEKOW
I thought that was Franks kid? D. Zappa?
Stewart Pinkerton
>Stewart Pinkerton wrote:>Because top-quality vinyl rigs are *very* expensive
>(top cartridges
>>alone cost $2-6,000), which puts then well into the aspirational
>>range. Additionally, getting the very best results from this precision
>>electro-mechanical process requires great attention to detail and a
>>lot of 'tweaking'. This qualifies vinyl as high-end, regardless of
>>absolute sound quality.
>
>~~~~~~~~~~~~~~~~~~~~~~~~~~
>
>$2,000-$6,000? Hey, Stewart, what magazines are you smoking? My Denon DL103D
>cart cost me US$175 last year.
That otherwise very nice cart (I used one for several years) has a
spherical stylus, and cannot be considered a 'top' cart. You know I'm
talking about Clearaudio, VdH etc.
> Together with a Teres turntable and modified
>Rega arm, my all-in cost is under $1,800, including shipping. Oh, add $24 for
>my DIY wall-mount shelf. You could spend more on an LP playback system, but
>I'm having trouble imagining what sonic benefits you might get (well, OK,
>getting rid of LP's many obvious problems would be nice, but why do the results
>have to sound like CD's?).
Try listening to a Rockport fitted with a Clearaudio Insider, then
tell me it doesn't sound noticeably better than your rig.
Stewart Pinkerton
>You Know Who ~ wrote:
>> The bottom line is that I have some music on both CD and LP. In most (not
>> all) cases, the LP sounds better.
>
>I agree. Good analog recordings do better justice.
In your opinion, of course. I mostly find the opposite to be the case.
>There is a
>reason studios use high-speed tape several decades after it was
>introduced. Analog in, analog out.
Irrelevant, they just like the sound of the soft limiting offered by
analogue tape. You can hear the effect quite clearly on good CDs made
from analogue tapes, although you'd never know on vinyl.......
>A good record player and stylus will sound incredible.
Yup, I don't find the sound very believable, either....... :-)
>If you have good ears, you can hear some cd players generating the
>tone in steps - this usually requires a stable tone like a pipe organ
>or flute. You can hear the harmonics aligning differently than they
>do in real life.
If this is the case, it's because of poor recording - there is
absolutely nothing in CD technology which would cause such an effect.
Vinyl OTOH has the harmonics wandering about all over the place
because of phase shifts.
> It bothers me because I played woodwinds for ten
>years and know how they sound - and CD just doesn't have RL sparkle
>at higher frequencies because it takes incredibly fast sampling rates
>to capture upper-range harmonics and CD is just not quite fast enough
>for most insturmental music.
What, a 20kHz bandwidth is not enough for music? It's wider than
you'll find on 99.9% of LPs.............
>Frequency and dynamic range isn't
>everything. Neither is the absolute lack of a few clicks and pops.
>Sonic accuracy to the original signal is.
>
>I can understand - the technology is over ten years old and they had
>to make serious compromises back then based upon the available
>technology.
Seious compromises? CD captures *everything* that is on any analogue
master tape up to more than 20kHz, which is above the limit of hearing
of the vast majority of adults. How is this a 'serious compromise'?
>If you spend time around samplers you know this syndrome - the
>waveforms are plainly jagged on a scope compared to a pure analog
>signal. Most of the time, though, you don't get a tone all by itself
>in music and adapt to how a CD sounds over the years.
This is utter rubbish! The output of a CD player contains *no*
jaggedness, this is just flat-out WRONG!
>Now, MP3s can thankfully sample at a high enough rate to do it
>better, so it may well be the next step and come closer.
Shame about the heavy compression that throws away 90% of the data...
Denis
reflects the difference between piano and forte) of current CD
format(44 kHz/16bits) is about 20 dB. The acoustic guitar has a
dynamic range of 26 dB. Virtually nothing can be recorded on this
format without any degree of compression.
From other side, vinyl as a recording medium has no limitations on
dynamic and frequency range. These limitations are originated first
from the playback, and second, from the recording equipment. The
playback limitations now are gone together with piezo cartridges!
Modern vinil releases in fact handle absolutely incredible dynamics,
and this is a pure analog recording medium. Even 24 bit digital
formats cannot compete. Need one say more?
Dennis.
jim <j...@tasmail.com> wrote in message news:<9r2ad...@enews1.newsguy.com>...
> Folks I need enlightenment.
>
> As a newcomer to this group I have been reading it with interest over
> the past few months and I have to admit I am staggered by what is
> posted here:
>
> The number of posts concerning turntables and vinyl!
>
> How can something that consists of scraping rock through plastic to
> reproduce sounds that have numerous clicks and pops, limited
> frequency reponse, wow and flutter (not to mention little tolerance
> to people walking past) be considered as "high-end"?
>
> The actual ability of vinyl to reproduce sounds across the audible
> spectrum is pretty poor. To compensate for this the bottom end has to
> be
>
> boosted and compression has to be applied to make it sound halfway
> decent. Now I do not have a problem if you like listening to music
> with lots of compression - but can it *really* be considered
> "high-end"?
>
> If I turned up here and started a discussion about this top value
> three-in-one portable sound system with 1000 watts of power
> (pmpo)that I
>
> got for $300 you would all tell me to go away, so could someone
> *please* explain:
>
> Why is vinyl considered high-end?
>
> Jim
>
> P.S. This is a genuine post and not trolling for a big argument. I do
Steven Sullivan
: us something about the audio quality. An example: for their new SACD/DVD
: player, the engineers at PHILIPS made 15 modifications to their design
: based purely on listening tests. Measurements revealed no differences at
: all.
THis is incredible. IF there are no measurements to correlate to the
differences, how did they know what to change?
--
-S.
The 80's was a really hard thing to put up with. -- Steve Howe
fathom
news:9r6se...@enews2.newsguy.com:
> Shame about the heavy compression that throws away 90% of the
> data...
Using 256 or 320 kb/s encoding is throwing out approx. 75% and
66% of the data, respectively. Music sounds pretty good at those
bitrates.
Also, using lossless encoding it is possible to reduce filesizes
by 45% while retaining 100% of the data. This is where the
future lies.
Graeme Jaye
wrote:
>Most analogue LPs actually come from tape recorders that don't have a
>"continuous signal" - the bias frequency certainly sets an upper
>frequency limit.
Nonsense! Of course an analogue tape recorder provides an analogue
output.
Upper frequency limit is mainly established as a combination of tape
speed and head gap - it has nothing to do with the biasing.
The bias signal (which, incidentally, is sinusoidal and certainly
continuous) is there to push the recorded signal onto the linear part
of the hysteresis curve of the recording medium - ie, rust :-)
Audio Restoration and CD Repair
http://www.personal-cd.com
Steven Sullivan
: 1) The Rube Goldberg contraptions known as vinyl LPs have no right
: sounding good at all, yet at their best they do. 2) Theoretically,
: CDs *should* sound magnificent, but too often in my experience they
: are flat, two-dimensional, uninvolving and lifeless-sounding. Go
: figure. 3) For the time being I'll keep building my LP collection on
: the cheap, and buy CD's (or whatever) to fulfill my new-music jones.
4) By deduction from (2) , if even *one* CD does *not* sound flat,
two-dimensional , uninvolving and lifeless, then the fault cannot be
in the compact disc medium or digital audio per se.
Richard D Pierce
Denis <denis_af...@mail.ru> wrote:
>The real undistorted dynamic range of RMS values(which really
>reflects the difference between piano and forte) of current CD
>format(44 kHz/16bits) is about 20 dB. The acoustic guitar has a
>dynamic range of 26 dB. Virtually nothing can be recorded on this
>format without any degree of compression.
And where on this good earth did you come up with THIS piece of
utter techno-rubbish? I hope you didn't pay for this
informati because, even free, it's a bad deal, because it is
totally incorrect.
Having, in fact measured innumerable systems, I have yet to see
a 16 bit 44 kHz system that had less than about 86 dB dynamic
range, quite a few that attained the teoreitical 93 or so dB
with dither, and ALL of them with dither achieve narrow-band
dynamic ranges well in excess of 96 dB.
From other side, vinyl as a recording medium has no limitations on
>dynamic and frequency range.
Thus, you are claiming that an LP has infinite dynamic range and
infinite bandwidth, right?
If that's youre claim, then it requires that LP system use
infinite energy and must exist for infinite time.
Clearly, this is absurd, as are your claims of dynmaic range and
frequency response. These absurd requirements simply follow
logically from your absurd claims.
>These limitations are originated first
>from the playback, and second, from the recording equipment. The
>playback limitations now are gone together with piezo cartridges!
>Modern vinil releases in fact handle absolutely incredible dynamics,
>and this is a pure analog recording medium. Even 24 bit digital
>formats cannot compete. Need one say more?
Yes. Your assertions are nonsense and are unsupportable.
Where's the obilgatory smiley, because I can't believe anyone
would post such technically incorrect with a straight face.
--
| Dick Pierce |
| Professional Audio Development |
| 1-781/826-4953 Voice and FAX |
| DPi...@world.std.com |
Howard Ferstler
>
> The real undistorted dynamic range of RMS values(which really
> reflects the difference between piano and forte) of current CD
> format(44 kHz/16bits) is about 20 dB.
Where on earth did you get this information?
> From other side, vinyl as a recording medium has no limitations on
> dynamic and frequency range.
Are you saying that if the mastering technician adjusts the
cutter that makes the disc so that the surface noise will be
inaudible during quiet passages that the disc will have
grooves that will be trackable by even the best phono
cartridge? I don't think so.
The fact is that if a disc is cut so that the loudest
passages are trackable, the surface noise will still be
audible, and considerably louder than what you find with the
CD. That takes care of your dynamic-range argument. As for
frequency range, how many LP records have significant
response to beyond 15 kHz. Most engineers do not let their
cutters work up above that point, because it would possibly
cause damage. Besides, there is no musical reason to go much
higher, anyway.
> These limitations are originated first
> from the playback, and second, from the recording equipment. The
> playback limitations now are gone together with piezo cartridges!
> Modern vinil releases in fact handle absolutely incredible dynamics,
> and this is a pure analog recording medium.
Just what do you think allows an LP with a S/N ratio of 60
dB to satisfactorily handle orchestral program sources with
dynamic ranges in excess of 70 dB?
> Even 24 bit digital
> formats cannot compete. Need one say more?
Well, yes they should, since you are dead wrong in your
comments about the dynamic range of the LP and the CD.
Howard Ferstler
Chris Johnson
ste...@pinkertons.fsnet.co.uk wrote:
> Seious compromises? CD captures *everything* that is on any analogue
> master tape up to more than 20kHz, which is above the limit of hearing
> of the vast majority of adults. How is this a 'serious compromise'?
This is an absurd claim. Talk to some mastering engineers before
mouthing off in this manner... you've got no idea how foolish that sounds
to someone who actually has the job of doing the capturing.
It does not. It's hard work to capture _most_ of the openness, ease and
tonal richness of even 15 ips tape, and the professional standard is 30
ips half-inch two-track tape. I question if you've ever heard _or_
measured this.
Perhaps you were referring to the Philips cassette?
Chris Johnson
Richard D Pierce
wrote:
>Most analogue LPs actually come from tape recorders that don't have a
>"continuous signal" - the bias frequency certainly sets an upper
>frequency limit.
Uh, no. If that were true, all we would need to do is change the
bias oscillator frequency and we would change the high frequency
limit. And that CERTAINLY is not the case at all.
What it comes down to, and this is hard for a lot of analog-only
proponents to swallow because it is non-intuitive, is that the
output of a discrete time-sampled system is just as smooth and
continuous as the output of a continuous-time analog system of
the same bandwidth. To put it in more practical terms: if your
tape recorder has the same bandwidth as your CD player, the CD
player will be just as continuous as the tape recorder. If your
LP has the same bandwidth as you CD, the output from CD will be
just as smooth and continuous as the output from the LP,
REGARDLESS of the way in which the signal was originally
recorded.
Contrary to much of the bunko surround these topics, some of
which we have seen repeated YET AGAIN by respondants to this
thread, the output of a CD DOES NOT have discrete stairsteps,
the output of a LP DOES NOT have more resolution in time, and
all that.
You may like one over the other for whatever reasons you want,
and no one will argue with your choice. WHat WILL be argued with
is spouting such nonsense as "stairsteps from CD's" and "the
infinite resolution of analog" and "the phase distortion of the
RIAA curve" and any other such examples of uninformed technical
nonsense.
Stewart Pinkerton
>ste...@pinkertons.fsnet.co.uk (Stewart Pinkerton) wrote in
>news:9r6se...@enews2.newsguy.com:
>
>> Shame about the heavy compression that throws away 90% of the
>> data...
>
>Using 256 or 320 kb/s encoding is throwing out approx. 75% and
>66% of the data, respectively. Music sounds pretty good at those
>bitrates.
Shame that the *vast* majority of MP3 files are 128kb/s.........
>Also, using lossless encoding it is possible to reduce filesizes
>by 45% while retaining 100% of the data. This is where the
>future lies.
However, this (MLP) has no relevance to MP3.
Richard D Pierce
Chris Johnson <jinx...@sover.net> wrote:
>In article <9r6se...@enews2.newsguy.com>,
>ste...@pinkertons.fsnet.co.uk wrote:
>
>> Seious compromises? CD captures *everything* that is on any analogue
>> master tape up to more than 20kHz, which is above the limit of hearing
>> of the vast majority of adults. How is this a 'serious compromise'?
>
> This is an absurd claim. Talk to some mastering engineers before
>mouthing off in this manner... you've got no idea how foolish that sounds
>to someone who actually has the job of doing the capturing.
Well, I, for one, HAVE and DO talk to a number of mastering
engineers all the time. And quite a number of them flatly
diagree with your mischaracterizations of their opinions and
flatly state the contrary of what you are claiming. I would only
refer you, for example, to the late Mr. Wiener, who mastered
some of the finest classical recordings in ANY medium, who
clearly contradicts what you claim.
>
> It does not. It's hard work capture _most_ of the openness, ease and
>tonal richness of even 15 ips tape, and the professional standard is 30
>ips half-inch two-track tape. I question if you've ever heard _or_
>measured this.
You may question it all you want, but I and quite a few others
have seen instances where self-proclaimed astute audiophiles
have been unable to distiguish the direct feed from such a tape
recorder from a 44.1/16 bit encode/decode process, using
equipment of their choice to listen.
I thus, in light of these result, question whether YOU'VE heard
to many mastering engineers or heard and measured the results
you claim.
Chris Johnson
ste...@pinkertons.fsnet.co.uk wrote:
> Contrary to several statements in this thread however, the vast
> majority of vinyl *is* compressed and peak limited, plus it's rolled
> off above 15kHz to avoid overheating the cutter
OK, quick question for Stew here: what order of filter would that
be, Stewart? Do you understand the concept of filter slope? Given a
broadband input of 10-100Khz, how many db down would you figure 60K
would be- infinite?
> while bass is summed
> to mono below 80Hz to preserve groove depth, and rolled off below 40Hz
> to maximise playing time.
By the same token, the vast majority of CDs from the last 5-10
years or so are produced through Pro Tools' undithered 16 bit output
buss, and do you hear us analog fans complaining?
Well- yes, of course you do. Because that sounds pretty awful and
has actually given Pro Tools itself a somewhat unjustified bad name
among pro audio types. But my point is, most audiophiles either don't
get the records like that, or get 'em and eventually sell them or
throw them away. In particular, the audiophile warhorse vinyl records
do not have these characteristics. You're talking about teen pop
disposable records- you're talking about a lot of records which
audiophiles would run screaming from.
I could as easily give a list of CDs which have been horribly
maimed in sound for the sake of loudness- did you know that the big
secret weapon from a couple years ago up to now has been digital
distortion, nothing less than brutally clipping the waveform AFTER
hard peak limiting, so that it becomes intentionally flat-topped and
puts out fierce crunchy distortion?
One advantage of vinyl is that you _can't_ do that to it: you
can't peaklimit that hard without busting the needle. Instead of
fighting to defend the dubious honor of the CD, maybe you'd better
keep an eye out for what's being done to it, because it's been a
steadily worsening progression- and the direction it's leading will
have any schmuck off the street insisting that vinyl has more dynamic
range than CD in level-matched comparisons, simply because the CD is
_destroyed_ with incredibly aggressive peak limiting until it has no
dynamics at all, nothing but 'loud'.
That is what lurks out there waiting to exact karmic vengeance on
your generalizations of LP bass-summing/rolling-off habits. In all
seriousness- don't even go there. ;)
Chris Johnson
Chris Johnson
<sull...@gwis2.circ.gwu.edu> wrote:
> Rob Gold <rgvi...@aol.com> wrote:
> : 1) The Rube Goldberg contraptions known as vinyl LPs have no right
> : sounding good at all, yet at their best they do. 2) Theoretically,
> : CDs *should* sound magnificent, but too often in my experience they
> : are flat, two-dimensional, uninvolving and lifeless-sounding. Go
> : figure. 3) For the time being I'll keep building my LP collection on
> : the cheap, and buy CD's (or whatever) to fulfill my new-music jones.
>
> 4) By deduction from (2) , if even *one* CD does *not* sound flat,
> two-dimensional , uninvolving and lifeless, then the fault cannot be
> in the compact disc medium or digital audio per se.
Worth commenting on- the same is true for analog, of course. If
even one vinyl record (a direct-to-disc extravaganza?) substantially
outperforms the usually dismal technical performance cited, then...
To me, the interesting question becomes: what needs to be done, in
each format, to accomplish this- in particular, what can make each
format reach towards the strong suit of the other?
For vinyl and analog in general, keeping a simple signal path
seems to be the proper way- and a lot of the goofy audiophile
practices come into play in, I think, a _cumulative_ fashion: if none
of them make a hearable difference but all of them put together
amount to a hearable difference, that's a win. The first thing to go
is frequency extension and noise level.
For digital, the situation becomes very interesting- cumulative
mathematical distortions take on extreme importance, and the first
thing to go is soundstage depth and image dimensionality. There's a
good litmus test: Bob Katz of Digital Domain invented what's called a
'bitscope', which simply links bits of digital audio to LEDs: for
instance, running CD audio through a 24-bit bitscope, only the top 16
bits will light (anything over half volume will make the top bit
light, by the way). If you haven't heard of a bitscope, you're not
doing professional work in digital sound mastering (at least not on
the Internet!) and if you can't see why one would be needed, you are
_way_ too trusting. One Yamaha mixer Bob tested sent only 20 bits of
information out the SPDIF output, and 24 out the AES/PDU output.
Another DAW exhibited a veiled quality that was eventually traced to
a one-bit level offset on positive wavefronts only as the signal
crossed zero- literally, crossover distortion in digital alone, due
to a programming mistake! <http://www.digido.com/morebits.html>.
Having mastery of this sort of thing is what's needed to bring out
the maximum soundstage and dimensionality of CDs, and the usual
trappings of audiophilia are no help at all...
It's an evolving situation, and neither 'side' are standing still.
For every Andy Rockport on the analog end, making quarter-ton
turntables, there's a Bob Katz pushing digital designers to use
48-bit floating point for the internal math of digital audio
workstations, minimizing the inevitable mathematical errors.
The only difference is in the character of the error when these
things aren't done. Rob's done a good job of describing what bad or
mediocre digital sounds like. This is the digital equivalent of a
plastic turntable from Radio Shack, and the problems are known and
understood (though sometimes hotly denied by people who want to
insist all digital is perfect- and they really should be more
grateful). As both analog and digital approach perfection, they grow
closer to each other in sound- from opposite extremes. The trick is
to not try to improve either by emulating the FLAWS of the other.
Chris Johnson
Loiskelly1
>Now playing- My guitar wants to kill your momma. F. Zappa<< >>
>
>I thought that was Franks kid? D. Zappa?
>
D. Zappa did indeed record a cover of the aforementioned tune, which
first appeared on daddy's album "Weasels Ripped My Flesh".
Now playing-
Catacombae Sepulchrum Romanum, from Jean Guillou's infamous organ
transcription of Mussorgsky's Pictures, on Dorian.
Now there's something you can't hear on vinyl!!!
jj, DBT thug and skeptical philalethist
Graeme Jaye <graem...@iname.com> wrote:
>On Wed, 24 Oct 2001 16:34:42 GMT, Ron-A...@magmar.freeserve.co.uk
>wrote:
>>Most analogue LPs actually come from tape recorders that don't have a
>>"continuous signal" - the bias frequency certainly sets an upper
>>frequency limit.
>Nonsense! Of course an analogue tape recorder provides an analogue
>output.
Really now, this is relevant how? The statemetn made by
Ron Ardell is very simple, the bias frequency sets an
upper limit.
DO you have any evidence otherwise? Do you know why stereo FM
tuners have "subcarrier traps", and the like?
He's right, you know.
>Upper frequency limit is mainly established as a combination of tape
>speed and head gap - it has nothing to do with the biasing.
I see, and if I put in a signal at 90% of the biasing frequency,
you assert I won't get anything out at 10% of the bias frequency?
Really?
Try it!
--
Copyright j...@research.att.com 2001, all rights reserved, except transmission
by USENET and like facilities granted. This notice must be included. Any
use by a provider charging in any way for the IP represented in and by this
article and any inclusion in print or other media are specifically prohibited.
N. Thornton
music is available on vinyl, from the 1950s to the present day. Far
more than on CD. And most of us already have lots of vinyl, and want
to hear it!
So if you want to enjoy your music, you need vinyl. With CD you won't
hear it at all, as it just wont play properly when you jam the record
into the CD player ;)
This has to be one big reason to keep using vinyl.
The first requirement of a high end system is that it can play your
music. If it can't, it's not high end in my book, its just
pretentious.
Vinyl has many technical shortcomings, yet surprisingly it still seems
to sound superb some of the time. For some reason all those sources of
distortion just don't seem to be ones my ears are too worried about -
except for the clicks and pops. I have watched a record do some very
non textbook things and still sound pretty good. I have found that
pitch accuracy isn't the most important thing after all.
Those of us who've put the work into setting up a decent turntable
know that it is *possible* to get excellent results from vinyl. But
most decks people own are pretty awful. Vinyl has such a poor
reputation simply because it is so easy to get it wrong, and quite
difficult to get right. Add all those scratchy abused records, and its
no wonder people usually go for CD.
BTW CD is *far* more complex than vinyl, but the complexities are all
hidden from the user. With vinyl you have to play with the small
number of complexities yourself.
Vinyl is very primitive technology, yet for some reason it just seems
to shine on.
I have masses of vinyl I bought second hand, of music that is simply
not available on CD.
Stewart Pinkerton
>In article <9r6se...@enews2.newsguy.com>,
>ste...@pinkertons.fsnet.co.uk wrote:
>
>> Seious compromises? CD captures *everything* that is on any analogue
>> master tape up to more than 20kHz, which is above the limit of hearing
>> of the vast majority of adults. How is this a 'serious compromise'?
>
> This is an absurd claim. Talk to some mastering engineers before
>mouthing off in this manner... you've got no idea how foolish that sounds
>to someone who actually has the job of doing the capturing.
I have done, often. Please specify *exactly* what is *not* captured.
Try to avoid artistic hand-waving, and get down to specifics.
> It does not. It's hard work to capture _most_ of the openness, ease and
>tonal richness of even 15 ips tape, and the professional standard is 30
>ips half-inch two-track tape. I question if you've ever heard _or_
>measured this.
I certainly have, for many years. Care to provide some numbers to back
up your rhetoric?
> Perhaps you were referring to the Philips cassette?
No, 30 ips 2-track 1/2 inch masters. Do you understand anything about
the real limits of such a medium?
Loiskelly1
>reflects the difference between piano and forte) of current CD
>format(44 kHz/16bits) is about 20 dB. The acoustic guitar has a
>dynamic range of 26 dB. Virtually nothing can be recorded on this
>format without any degree of compression.
Earlier in this thread, someone, quite accidently I am sure, claimed
that vinyl could extend way out to 30cps. Even though he clearly
meant 30 kHz, he was promptly ripped a second sphincter.
Accordingly, you may want to retract/rethink/reconsider the above.
Each bit added to the word length gives about 6 dB of dynamic range
(20 times the base 10 logarithm of 2, or 6.02). A 24 bit word length
provides a dynamic range of better than 144 dB, which not only covers
the region from the threshold of audibility to the painfully
intolerable, but will even accommodate the needs of that blaring,
raucous, stentorian, and earsplitting instrument of torture, the
dread acoustic guitar.
"Sarcasm- I've never been very good at it, but I hear it can be
terribly effective."
-Slartibartfast
Stewart Pinkerton
>The real undistorted dynamic range of RMS values(which really
>reflects the difference between piano and forte) of current CD
>format(44 kHz/16bits) is about 20 dB. The acoustic guitar has a
>dynamic range of 26 dB. Virtually nothing can be recorded on this
>format without any degree of compression.
I don't know where you got this weird stuff from, but the reality is
that the *average* CD has 90-93dB of available dynamic range.
From other side, vinyl as a recording medium has no limitations on
>dynamic and frequency range.
The dynamic range from surface noise of the finest virgin vinyl to the
'hottest' peak cutting level of 'killer' Telarcs is around 80dB in the
midrange, and significantly lower at the frequency extremes.
Frequency range is *in practice* limited to around 40Hz at the low end
(to allow more than ten minutes per side), and 15kHz or so at the top
end (to stop the cutter head melting!). A *tiny* minority of
'audiophile' vinyl has exceeded this, with resulting playing times as
low as 8 minutes a side on some direct-cut Sheffields, and half-speed
mastering on the defunct MFSL label.
No available vinyl however can exceed 20Hz-30kHz or 80dB dynamic
range, which are hardly examples of 'no limitations'.
> These limitations are originated first
>from the playback, and second, from the recording equipment. The
>playback limitations now are gone together with piezo cartridges!
>Modern vinil releases in fact handle absolutely incredible dynamics,
>and this is a pure analog recording medium. Even 24 bit digital
>formats cannot compete. Need one say more?
One might say something accurate, which would be nice..........
Joseph Oberlander
>
> Denis wrote:
> >
> > The real undistorted dynamic range of RMS values(which really
> > reflects the difference between piano and forte) of current CD
> > format(44 kHz/16bits) is about 20 dB.
>
> Where on earth did you get this information?
>
> > From other side, vinyl as a recording medium has no limitations on
> > dynamic and frequency range.
>
> Are you saying that if the mastering technician adjusts the
> cutter that makes the disc so that the surface noise will be
> inaudible during quiet passages that the disc will have
> grooves that will be trackable by even the best phono
> cartridge? I don't think so.
>
> The fact is that if a disc is cut so that the loudest
> passages are trackable, the surface noise will still be
> audible, and considerably louder than what you find with the
> CD. That takes care of your dynamic-range argument. As for
> frequency range, how many LP records have significant
> response to beyond 15 kHz. Most engineers do not let their
> cutters work up above that point, because it would possibly
> cause damage. Besides, there is no musical reason to go much
> higher, anyway.
So, given that:
- Sounds less than 20db are usually not present in recorded music.
- Sound over 80db are not usually present either.
- 40hz is pretty much the limit for musical recordings.
- 15khz is pretty much the top of most music for both formats.
- People can't hear the difference between .1 db and 1db...
I see it as a 6 of 1 type argument...
Pro:
CDs do have a benefit from the point of durability and ease of
manufacture. Like MP3s, they sound acceptable with even
very inexpensive equipment and sound roughly the same in all
players. This is great for marketing and use as a format for
music sales.
Con:
Cd's suffer greatly from poor engineering and mastering and are
often "slapped onto the disc" as if proper attention isn't
required(the "digital solves everything" view)
They also have a problem of a too low sampling rate.
My bet is on DVDs for music in a few years.
They can encode MP3s with no loss/compression and 512 bit+
encoding and still fit over 80 minutes of music plus extras on the
video tracks. Nice.
5-channel audio would also be a reality - not the compressed
junk on movies, but real 5-channel recordings of symphonies and
such at 512 bit or better.
Joseph Oberlander
> Seious compromises? CD captures *everything* that is on any analogue
> master tape up to more than 20kHz, which is above the limit of hearing
> of the vast majority of adults. How is this a 'serious compromise'?
You need to do some research about sampling rates. CD can do 20Khz,
but the sampling suffers. Remeber - this is a 15+yr old
technology now and when they made the standard, it was for practical
(marketing and legngth) reasons and not audio quality.
The sampling quality of high-frequency harmonics makes a dramatic
impact on subtle things like imaging and whether or not it sounds
somewhat like a live performance.
> >If you spend time around samplers you know this syndrome - the
> >waveforms are plainly jagged on a scope compared to a pure analog
> >signal. Most of the time, though, you don't get a tone all by itself
> >in music and adapt to how a CD sounds over the years.
>
> This is utter rubbish! The output of a CD player contains *no*
> jaggedness, this is just flat-out WRONG!
Then you need to spend some time with a real Oscilloscope.
> >Now, MP3s can thankfully sample at a high enough rate to do it
> >better, so it may well be the next step and come closer.
>
> Shame about the heavy compression that throws away 90% of the data...
You can record the signal with no compression if you want. What you
gain is a huge increase in sampling quality. Of course, the file
tends to be 20+ megs a minute...(grin)
jj, DBT thug and skeptical philalethist
Richard D Pierce <DPi...@world.std.com> wrote:
>If that were true, all we would need to do is change the
>bias oscillator frequency and we would change the high frequency
>limit. And that CERTAINLY is not the case at all.
It's also true that if you get anywhere near a
bias frequency input, you'll get, due to nonlinearity,
something hideously close to aliasing.
However, increasing the bias frequency ALONE won't do a
thing for the HF response of the average tape deck, indeed.
David Collins
jinx...@sover.net (Chris Johnson) wrote:
> For digital, the situation becomes very interesting- cumulative
> mathematical distortions take on extreme importance, and the first
> thing to go is soundstage depth and image dimensionality. There's a
> good litmus test: Bob Katz of Digital Domain invented what's called a
> 'bitscope', which simply links bits of digital audio to LEDs:
Sorry Chris, but Bob didn't "invent" this. There have been logic
analyzers for many moons. Also, it does'nt use a LED's as a display,
you have to use an oscilloscope triggered from an AES receiver... There
are however, a number of other products that will display the number of
active bits, either on a LED ladder, or on a computer monitor.
DC
fathom
> tgr...@maig.com (fathom) writes:
>
>>ste...@pinkertons.fsnet.co.uk (Stewart Pinkerton) wrote in
>>news:9r6se...@enews2.newsguy.com:
>>
>>> Shame about the heavy compression that throws away 90% of
>>> the data...
>>
>>Using 256 or 320 kb/s encoding is throwing out approx. 75%
>>and 66% of the data, respectively. Music sounds pretty good
>>at those bitrates.
>
> Shame that the *vast* majority of MP3 files are
> 128kb/s.........
I disagree. Maybe this was true 2 years ago, but most people now
use 160 or 192 as a minimum. I did a random sampling of 8 tracks
from one of the mp3 mewsgroups and came up with:
2@128
2@160
3@192
1@224
That's an average of 172kb/s.
Try it yourself and see what you come up with. It'll be
significantly higher than 128.
>>Also, using lossless encoding it is possible to reduce
>>filesizes by 45% while retaining 100% of the data. This is
>>where the future lies.
>
> However, this (MLP) has no relevance to MP3.
I wasn't referring to Meridian's Lossless scheme used for DVD-A.
I use Monkey's Audio format (stupid name, excellent codec).
IMO, there should be more "buzz" in the audiophile community for
lossless, as it removes all concerns about the sound quality of
mp3 or other formats. Isn't that why mp3 is rejected out-of-hand
by most audiophiles?
From the website:
"Monkey’s Audio is a fast and easy way to compress digital music.
Unlike traditional methods such as mp3, ogg, or lqt that
permanently discard quality to save space, Monkey’s Audio only
makes perfect, bit-for-bit copies of your music. That means it
always sounds perfect exactly the same as the original. Even
though the sound is perfect, it still saves a lot of space.
(think of it as a beefed-up Winzip for your music) The other
great thing is that you can always decompress your Monkey's Audio
files back to the exact, original files. That way, you'll never
have to recopy your CD collection to switch formats, and you'll
always be able to recreate the original music CD if something
ever happens to yours."
To keep this in line with the thread "Is vinyl high-end?",
lossless is an IDEAL way to keep a pristine copy of treasured
vinyl. I have a lot of vinyl stored this way. The Mobile
Fidelity Beatles Collection (on vinyl), digitized and losslessly
encoded, sounds incredible! And I don't have to clean the files
each time I want to hear them :)
Hard drives get cheaper and bigger every year. Lossless is the
future.
Harry Lavo
news:9r6sq9$mt9$1...@bourbaki.localdomain...
> Rob Gold's experience mirrors mine. I have a Rotel RCD 945BX CD player
> (Arcam Alpha 9 amp, B&W DM 302 speakers, Velodyne sub) and a Music Hall
MMF2
> with an AT OC9 cartridge. Good CDs sound good. Well recorded and mastered
> CDs sound great. Poor CDs sound flat and lifeless. Good LPs (e.g., the new
> Classic Records "Kind of Blue", Loggins & Messina "Full Sail", etc.) sound
> superb; unnoticeable surface noise, huge sense of space, coherent sound.
> Poor LPs sound flat and lifeless. Comparing the "Kind of Blue" CD I have
to
> the Classic LP, the LP seems to have a larger and deeper soundstage.
>
> Obviously, I don't have the ultimate in CD playback. But I don't have the
> ultimate in LP playback either. I enjoy listening to music on both media.
I
> think it needs to be stated, however, that one can enjoy listening to LPs
> on a sub-$1,000 system. For those of us with signficant numbers of LPs,
the
> choice is obvious, and people should not be scared away from resurrecting
> their vinyl collection by statements that multi-kilobuck playback
equipment
> is necessary. Get a good belt-drive turntable and a decent or better
> cartridge (Goldring, Grado, Benz, Clearaudio, or the AT OC9 even), clean
> your old records and enjoy.
>
> What is equally necessary, however, is a good record cleaning system. For
> that, I recommend the Disc Doctor system of brushes and cleaning fluids.
>
> Curt Coleman
To paraphrase Steven Sullivan in another post:
"If only one record compared to the equivalent CD (both superb recordings
and masterings) sounds better, than ipso facto the problems associated with
vinyl cannot be attributable to the medium".
Two different experiences, two different conclusions. How interesting.
Stewart Pinkerton
>In article <9r737...@enews3.newsguy.com>, Steven Sullivan
><sull...@gwis2.circ.gwu.edu> wrote:
>> 4) By deduction from (2) , if even *one* CD does *not* sound flat,
>> two-dimensional , uninvolving and lifeless, then the fault cannot be
>> in the compact disc medium or digital audio per se.
>
> Worth commenting on- the same is true for analog, of course. If
>even one vinyl record (a direct-to-disc extravaganza?) substantially
>outperforms the usually dismal technical performance cited, then...
However, they don't, *except* in frequency response, and then only
marginally.
> To me, the interesting question becomes: what needs to be done, in
>each format, to accomplish this- in particular, what can make each
>format reach towards the strong suit of the other?
>
> For vinyl and analog in general, keeping a simple signal path
>seems to be the proper way- and a lot of the goofy audiophile
>practices come into play in, I think, a _cumulative_ fashion: if none
>of them make a hearable difference but all of them put together
>amount to a hearable difference, that's a win. The first thing to go
>is frequency extension and noise level.
Please explain how you're going to get below 20Hz with vinyl, or above
80dB dynamic range.
> For digital, the situation becomes very interesting- cumulative
>mathematical distortions take on extreme importance, and the first
>thing to go is soundstage depth and image dimensionality. There's a
>good litmus test: Bob Katz of Digital Domain invented what's called a
>'bitscope', which simply links bits of digital audio to LEDs:
Um, no he didn't, but logic analysers are indeed useful devices.
> for
>instance, running CD audio through a 24-bit bitscope, only the top 16
>bits will light (anything over half volume will make the top bit
>light, by the way). If you haven't heard of a bitscope, you're not
>doing professional work in digital sound mastering (at least not on
>the Internet!) and if you can't see why one would be needed, you are
>_way_ too trusting. One Yamaha mixer Bob tested sent only 20 bits of
>information out the SPDIF output, and 24 out the AES/PDU output.
You might want to check out the Red Book standards for these signals,
if as you imply you are doing 'professional' work in digital sound
mastering.........
>Another DAW exhibited a veiled quality that was eventually traced to
>a one-bit level offset on positive wavefronts only as the signal
>crossed zero- literally, crossover distortion in digital alone, due
>to a programming mistake! <http://www.digido.com/morebits.html>.
>Having mastery of this sort of thing is what's needed to bring out
>the maximum soundstage and dimensionality of CDs, and the usual
>trappings of audiophilia are no help at all...
There exists no technology which cannot be badly implemented.......
> It's an evolving situation, and neither 'side' are standing still.
>For every Andy Rockport on the analog end, making quarter-ton
>turntables, there's a Bob Katz pushing digital designers to use
>48-bit floating point for the internal math of digital audio
>workstations, minimizing the inevitable mathematical errors.
Unfortunately, this is unlikely to improve the analogue input/output
side of DAWs, which are still woefully short of 24-bit capability.
> The only difference is in the character of the error when these
>things aren't done. Rob's done a good job of describing what bad or
>mediocre digital sounds like. This is the digital equivalent of a
>plastic turntable from Radio Shack, and the problems are known and
>understood (though sometimes hotly denied by people who want to
>insist all digital is perfect- and they really should be more
>grateful). As both analog and digital approach perfection, they grow
>closer to each other in sound- from opposite extremes. The trick is
>to not try to improve either by emulating the FLAWS of the other.
The trick is not to pretend that vinyl is other than *grossly* flawed
from a technical standpoint. The other trick is not to pretend that
SACD and DVD-A are *significantly* different in sound quality from the
best of CD (probably represented by JVC XRCD2).
Stewart Pinkerton
>ste...@pinkertons.fsnet.co.uk (Stewart Pinkerton) wrote in
>news:9r7cp9$2bf$1...@bourbaki.localdomain:
>
>> tgr...@maig.com (fathom) writes:
>>
>>>ste...@pinkertons.fsnet.co.uk (Stewart Pinkerton) wrote in
>>>news:9r6se...@enews2.newsguy.com:
>>>
>>>> Shame about the heavy compression that throws away 90% of
>>>> the data...
>>>
>>>Using 256 or 320 kb/s encoding is throwing out approx. 75%
>>>and 66% of the data, respectively. Music sounds pretty good
>>>at those bitrates.
>>
>> Shame that the *vast* majority of MP3 files are
>> 128kb/s.........
>
>I disagree. Maybe this was true 2 years ago, but most people now
>use 160 or 192 as a minimum. I did a random sampling of 8 tracks
>from one of the mp3 mewsgroups and came up with:
>
>2@128
>2@160
>3@192
>1@224
>
>That's an average of 172kb/s.
>
>Try it yourself and see what you come up with. It'll be
>significantly higher than 128.
OK, such is progress, but we're still talking about greater than 5:1
lossy compression (i.e. throwing away >80% of the data), which is
*never* going to be as good as leaving the data in one piece.
>>>Also, using lossless encoding it is possible to reduce
>>>filesizes by 45% while retaining 100% of the data. This is
>>>where the future lies.
>>
>> However, this (MLP) has no relevance to MP3.
>
>I wasn't referring to Meridian's Lossless scheme used for DVD-A.
>I use Monkey's Audio format (stupid name, excellent codec).
Fine, same irrelevance. Also, this is *not* where the future lies,
it's just a minor stopgap until storage capacity is adequate to remove
*any* form of compression. All this argy-bargy about DD vs DTS will
vanish when 'ultra DVD' provides DVD-A quality on movie soundtracks.
>IMO, there should be more "buzz" in the audiophile community for
>lossless, as it removes all concerns about the sound quality of
>mp3 or other formats. Isn't that why mp3 is rejected out-of-hand
>by most audiophiles?
Yes, but note that you get less than 2:1 compression, so it's hardly a
major breakthrough in audio quality.
You could get 24/96 on a 60-minute CD by using this technique, which
would be nice, but isn't going to happen.
>To keep this in line with the thread "Is vinyl high-end?",
>lossless is an IDEAL way to keep a pristine copy of treasured
>vinyl. I have a lot of vinyl stored this way. The Mobile
>Fidelity Beatles Collection (on vinyl), digitized and losslessly
>encoded, sounds incredible! And I don't have to clean the files
>each time I want to hear them :)
>Hard drives get cheaper and bigger every year. Lossless is the
>future.
Actually, since mass storage gets bigger and cheaper every year, *any*
form of compression is rapidly becoming unnecessary. Ten years from
now, you'll be able to get 6 channels of 24/192 on a smart card.
Stewart Pinkerton
>Stewart Pinkerton wrote:
>
>> Seious compromises? CD captures *everything* that is on any analogue
>> master tape up to more than 20kHz, which is above the limit of hearing
>> of the vast majority of adults. How is this a 'serious compromise'?
>
>You need to do some research about sampling rates. CD can do 20Khz,
>but the sampling suffers.
How does the sampling 'suffer'?
> Remeber - this is a 15+yr old
>technology now and when they made the standard, it was for practical
>(marketing and legngth) reasons and not audio quality.
The standard was however set at a level which well exceeds the
capability of vinyl.
>The sampling quality of high-frequency harmonics makes a dramatic
>impact on subtle things like imaging and whether or not it sounds
>somewhat like a live performance.
Oh, really? How? What do you mean by 'high-frequency harmonics'?
>> >If you spend time around samplers you know this syndrome - the
>> >waveforms are plainly jagged on a scope compared to a pure analog
>> >signal. Most of the time, though, you don't get a tone all by itself
>> >in music and adapt to how a CD sounds over the years.
>>
>> This is utter rubbish! The output of a CD player contains *no*
>> jaggedness, this is just flat-out WRONG!
>
>Then you need to spend some time with a real Oscilloscope.
I have done, about thirty years. You need to spend some time with a
real DAC.............
>> >Now, MP3s can thankfully sample at a high enough rate to do it
>> >better, so it may well be the next step and come closer.
>>
>> Shame about the heavy compression that throws away 90% of the data...
>
>You can record the signal with no compression if you want. What you
>gain is a huge increase in sampling quality. Of course, the file
>tends to be 20+ megs a minute...(grin)
Quite so. Now, back here in the real world of <256kb/sec...........
Arny Krueger
news:9r7cv...@enews4.newsguy.com...
> For digital, the situation becomes very interesting- cumulative
> mathematical distortions take on extreme importance, and the first
> thing to go is soundstage depth and image dimensionality. There's a
> good litmus test: Bob Katz of Digital Domain invented what's called
a
> 'bitscope', which simply links bits of digital audio to LEDs: for
> instance, running CD audio through a 24-bit bitscope, only the top
16
> bits will light (anything over half volume will make the top bit
> light, by the way).
This isn't a universal rule. It's a designer's choice to handle the
low order bits as he wishes. The most common technique is to extend
the low order bits with zeros. In fact, extending them with 1's has
very similar results. Certain CD players (example, Wadia) extend the
16 bit data with a noise source.
> If you haven't heard of a bitscope, you're not
> doing professional work in digital sound mastering (at least not on
> the Internet!) and if you can't see why one would be needed, you
are
> _way_ too trusting.
Bitscopes based on rows of indicator lights are very old tools. I
think I saw the first one during the 60's, and they were old, then.
They largely fell out of favor as digital test equipment. There are
many ways to analyze digital data, and each has its strengths and
weaknesses.
> One Yamaha mixer Bob tested sent only 20 bits of
> information out the SPDIF output, and 24 out the AES/PDU output.
That would be AES/EBU, and again this it not very significant since
AFAIK there is no commercial equipment that converts analog signals
reliably, below the 19 or 20 bit level.
> Another DAW exhibited a veiled quality that was eventually traced
to
> a one-bit level offset on positive wavefronts only as the signal
> crossed zero- literally, crossover distortion in digital alone, due
> to a programming mistake! <http://www.digido.com/morebits.html>.
> Having mastery of this sort of thing is what's needed to bring out
> the maximum soundstage and dimensionality of CDs, and the usual
> trappings of audiophilia are no help at all...
What's ironic is the fact that loss of low order bits is actually
reasonably-tolerated by the ear. I provide a variety of
demonstrations of this at
http://www.cdabx.com/technical/bits44/index.htm and
http://www.pcabx.com/technical/sample_rates/index.htm .
People can listen to these samples, which were quite carefully made,
and reach their own conclusions. The general response seems to be
that below about 14 bits, additional digital precision is moot. This
is probably partially due to the pervasiveness of acoustic noise in
real-world recordings. By means of close-microphone techniques and
careful production methods, some of these samples have more dynamic
range, based on a fairly wide measurement window - 50 mSec) than any
known commercial recording.
Reports of audible artifacts at lower levels seem to invariable trace
to sighted evaluations, which are well-known to be grossly flawed
for the purpose of evaluating subtle effects.
Howard Ferstler
> - 40hz is pretty much the limit for musical recordings.
Actually, I have a large number of CD releases with
substantial musical content below 40 Hz, and even below 30
Hz.
> 5-channel audio would also be a reality - not the compressed
> junk on movies, but real 5-channel recordings of symphonies and
> such at 512 bit or better.
I am a big fan of 5-channel audio with both music and
movies. I do need to footnote your comments and point out
that many movies have very uncompressed sound. Some have
immense dynamic range. Some of the Dolby Digital music
recordings I have (at 448 kbps) are very clean and realistic
sounding.
Howard Ferstler
Harry Lavo
news:9r6v8o$ond$1...@bourbaki.localdomain...
> Paul van der Hulst <pa...@home.nl> wrote:
> : us something about the audio quality. An example: for their new SACD/DVD
> : player, the engineers at PHILIPS made 15 modifications to their design
> : based purely on listening tests. Measurements revealed no differences at
> : all.
>
> THis is incredible. IF there are no measurements to correlate to the
> differences, how did they know what to change?
Perhaps they used their intuition and then listened carefully to
empirical experiments?
Richard D Pierce
Joseph Oberlander <josephob...@earthlink.net> wrote:
>Stewart Pinkerton wrote:
>
>> Seious compromises? CD captures *everything* that is on any analogue
>> master tape up to more than 20kHz, which is above the limit of hearing
>> of the vast majority of adults. How is this a 'serious compromise'?
>
>You need to do some research about sampling rates. CD can do 20Khz,
>but the sampling suffers. Remeber - this is a 15+yr old
>technology now and when they made the standard, it was for practical
>(marketing and legngth) reasons and not audio quality.
False, false, false, false.
This nonsense has been debunked time and time and time again and
yet we see it here again.
>The sampling quality of high-frequency harmonics makes a dramatic
>impact on subtle things like imaging and whether or not it sounds
>somewhat like a live performance.
Gerat assertion, vigorously stated. Now substantiate it.
And if you are able, the assertions of sampling at 20 kHz made
above are STILL wrong.
>> >If you spend time around samplers you know this syndrome - the
>> >waveforms are plainly jagged on a scope compared to a pure analog
>> >signal. Most of the time, though, you don't get a tone all by itself
>> >in music and adapt to how a CD sounds over the years.
>>
>> This is utter rubbish! The output of a CD player contains *no*
>> jaggedness, this is just flat-out WRONG!
>
>Then you need to spend some time with a real Oscilloscope.
I do, all the time, every day, and they DO NOT show jaggedness
unless that system is BROKEN.
>> >Now, MP3s can thankfully sample at a high enough rate to do it
>> >better, so it may well be the next step and come closer.
>>
>> Shame about the heavy compression that throws away 90% of the data...
>
>You can record the signal with no compression if you want. What you
>gain is a huge increase in sampling quality. Of course, the file
>tends to be 20+ megs a minute...(grin)
Uh, then it ain't MP3
Harry Lavo
news:9r7cv...@enews4.newsguy.com...
I understand that my Otari MTR 12II 30ips machine is down -3db at
28-30khz at 0db reference level, which is a hell of a lot better than
CD. And it is only a "B" level professional recorder.
I understand that my Teac 7030SL 15ips machine is down -2db at 24khz
at 0db reference level on most tape formulations. And it is only a
semi-pro (albeit top quality) machine.
And I understand that my Schoeps small diameter mics in omni mode
extended to 26khz.
And that the new Sennheiser 800 microphone extends to 50khz.
What is your point?
Arny Krueger
news:9r7if...@enews1.newsguy.com...
> Stewart Pinkerton wrote:
> > Serious compromises? CD captures *everything* that is on any analogue
> > master tape up to more than 20kHz, which is above the limit of hearing
> > of the vast majority of adults. How is this a 'serious compromise'?
> You need to do some research about sampling rates. CD can do 20Khz,
> but the sampling suffers. Remember - this is a 15+yr old
> technology now and when they made the standard, it was for practical
> (marketing and length) reasons and not audio quality.
Anybody who wants to do level-matched blind listening tests and has a
windows PC with a good sound card and speakers can investigate this
issue using files downloaded from
http://www.pcabx.com/technical/sample_rates/index.htm and
http://www.pcabx.com/technical/low_pass/index.htm . In the past two
years, thousands of people have downloaded tens of thousands of files
from this source. Anybody who wants to experience these issues for
themselves need only download files and report their results to RAHE.
> The sampling quality of high-frequency harmonics makes a dramatic
> impact on subtle things like imaging and whether or not it sounds
> somewhat like a live performance.
The general experience is that brick wall filtering as low as 16 KHz
and downsampling to sample rates as low as 32 KHz can be difficult or
impossible to detect, even using program material with considerable
energy above those frequencies.
> > >If you spend time around samplers you know this syndrome - the
> > >waveforms are plainly jagged on a scope compared to a pure analog
> > >signal. Most of the time, though, you don't get a tone all by itself
> > >in music and adapt to how a CD sounds over the years.
> > This is utter rubbish! The output of a CD player contains *no*
> > jaggedness, this is just flat-out WRONG!
> Then you need to spend some time with a real Oscilloscope.
Oscilloscopes are often not the best tool to use to view audio
signals. I recommend that interested parties use tools like a good
computer-based digital editing software. Shareware such as that which
can be downloaded from www.syntrillium.com can be used to view and
analyze audio signals at one's leisure, with highly detailed displays
that can be used to view down to the individual sample level.
> > >Now, MP3s can thankfully sample at a high enough rate to do it
> > >better, so it may well be the next step and come closer.
> > Shame about the heavy compression that throws away 90% of the
> > data...
If its the right 90%, detecting the loss can be highly challenging.
If the goal is action and experience, not talk, a variety of musical
samples can be downloaded from
http://www.pcabx.com/product/coder_decoder/index.htm .
> You can record the signal with no compression if you want. What you
> gain is a huge increase in sampling quality. Of course, the file
> tends to be 20+ megs a minute...(grin)
16/44 files run about 11 megs a minute, but 24/96 files run well over
20 megs a minute.
fathom
> Actually, since mass storage gets bigger and cheaper every
> year, *any* form of compression is rapidly becoming
> unnecessary. Ten years from now, you'll be able to get 6
> channels of 24/192 on a smart card.
Maybe, but certain kinds of compression will always be useful.
Hard drive capacities keep doubling, but I don't see computer
users abandoning WinZip and other forms of file compression.
Perhaps lossless technology will mature enough to reduce
filesizes by 75% or more. No matter how cheap storage gets, it
would always be beneficial to fit 4 times as much data in a given
space.
Wouldn't you rather store four 24/192 songs on your smart card
instead of one, as long as the sound is the same? Just a
thought.
You Know Who ~
unlike you, most of us can't talk to the late Mr. Wiener...
--
You Know Who~
"No matter how cynical you get,
it is impossible to keep up." -- Lily Tomlin
-----------------------------------------------------
Cat stuff, Win Help & Purgatory
@ http://You_Know_Who.home.att.net/
`````````````````````````````````````````````
"Richard D Pierce" <DPi...@world.std.com> wrote in message
news:9r7di2$2ri$1...@bourbaki.localdomain...
Alex Eisenhut
> >waveforms are plainly jagged on a scope compared to a pure analog
>
Tube or transistor scope? There's negative feedback in there too! Oh
my goodness!
Arny Krueger
> Howard Ferstler wrote:
> > The fact is that if a disc is cut so that the loudest
> > passages are trackable, the surface noise will still be
> > audible, and considerably louder than what you find with the
> > CD. That takes care of your dynamic-range argument. As for
> > frequency range, how many LP records have significant
> > response to beyond 15 kHz. Most engineers do not let their
> > cutters work up above that point, because it would possibly
> > cause damage. Besides, there is no musical reason to go much
> > higher, anyway.
> - So, given that:
> - Sounds less than 20db are usually not present in recorded music.
This is a limit based on acoustic noise in recording studios, not a
limitation of the 16/44 recording process.
> - Sound over 80db are not usually present either.
80 dB corresponds to a normal average to low playing level for
music. Peak levels from single acoustical instruments can be over 100
dB at the microphone. Orchestras often develop sound levels in the
105 dB range in large rooms full of thousands of people. A pipe organ
can develop sound levels well over 100 dB, even in very large rooms.
> - 40hz is pretty much the limit for musical recordings.
The fundamental frequency lowest string on a 5-string bass guitar is
around 32 Hz. 16' organ pipe ranks are fairly common and the lowest
pipe has a fundamental of 32 Hz.
> - 15khz is pretty much the top of most music for both formats.
This claim has been falsified by a number of sources including
http://www.pcabx.com/technical/sample_rates/index.htm and
http://www.its.caltech.edu/~boyk/spectra/spectra.htm .
> - People can't hear the difference between .1 db and 1db...
People can falsify this claim for themselves by listening to musical
sounds they can download from
http://www.pcabx.com/technical/levels/index.htm . The audible limits
for variations in level is frequency and bandwidth dependent as shown
by http://www.pcavtech.com/soundcards/techtalk/FR/index.htm (chart at
the bottom of page taken from a JAES article. )
> I see it as a 6 of 1 type argument...
I see it as a collection of deconstructed claims.
> Pro:
> CDs do have a benefit from the point of durability and ease of
> manufacture.
They also have superior technical performance in a variety of very
meaningful ways.
> Like MP3s, they sound acceptable with even
> very inexpensive equipment and sound roughly the same in all
> players.
In fact, 16/44 coding done well has been shown to be sonically
transparent in listening tests going back about two decades:
http://users.htdconnect.com/~djcarlst/abx_digi.htm .
> This is great for marketing and use as a format for music sales.
It's also great for people who love the natural sound of music, as
opposed to sonic artifacts.
> Con:
> Cd's suffer greatly from poor engineering and mastering and are
> often "slapped onto the disc" as if proper attention isn't
> required(the "digital solves everything" view)
Every form of commercial media has this problem. CD's and other
digital media are more resistant than analog media to errors at many
stages of processing and production, particularly those that involve
large volume production.
> They also have a problem of a too low sampling rate.
This claim can be falsified by anybody who listens to the samples
freely downloadable from http://www.pcabx.com/technical/index.htm .
> My bet is on DVDs for music in a few years.
My understanding is that many industry insiders think that DVD-A and
SACD are likely to fail as mass-market products.
> They can encode MP3s with no loss/compression and 512 bit+
> encoding and still fit over 80 minutes of music plus extras on the
> video tracks. Nice.
Think A/V. Kids are growing up in an A/V world.
> 5-channel audio would also be a reality - not the compressed
> junk on movies, but real 5-channel recordings of symphonies and
> such at 512 bit or better.
Please explain the use of "512 bit" in this context, as it seems to
make no sense.
David Collins
jinx...@sover.net (Chris Johnson) wrote:
> In article <9r6se...@enews2.newsguy.com>,
> ste...@pinkertons.fsnet.co.uk wrote:
>
> > Seious compromises? CD captures *everything* that is on any analogue
> > master tape up to more than 20kHz, which is above the limit of hearing
> > of the vast majority of adults. How is this a 'serious compromise'?
>
> This is an absurd claim. Talk to some mastering engineers before
> mouthing off in this manner... you've got no idea how foolish that sounds
> to someone who actually has the job of doing the capturing.
>
Sorry Chris, but I compare the raw output of the 1/2" 30ips master
with the A/D/A chain every day and even at 44/16, it's pretty
close. Now maybe I don't have your ears or fancy modified gear, but
around these parts digital sounds pretty good. I have made these
same comparisons with 88/24 as well as the latest DSD, and while
there are some subtle improvements, I share Stewarts' opinion.
DC
Stewart Pinkerton
>So, given that:
>- Sounds less than 20db are usually not present in recorded music.
>- Sound over 80db are not usually present either.
>- 40hz is pretty much the limit for musical recordings.
>- 15khz is pretty much the top of most music for both formats.
>- People can't hear the difference between .1 db and 1db...
>
>I see it as a 6 of 1 type argument...
>
>Pro:
>CDs do have a benefit from the point of durability and ease of
>manufacture. Like MP3s, they sound acceptable with even
>very inexpensive equipment and sound roughly the same in all
>players. This is great for marketing and use as a format for
>music sales.
A gross denigration of the sound of CD. Not only does it sound
'acceptable', it sounds *much* better than most MP3 files, and there
reason that most good CD players sound the same is that they are all
as near as dammit extracting everything that's on the disc. The litmus
test is to record an LP on to a CD-R. The CD-R then sounds almost
exactly like the original LP, showing that the 'magic' of lp is simply
'added' artifacts, and that CD is inherently the more transparent
medium.
What, you have a problem with CD sounding good without having to spend
$10k on the replay gear? Interesting 'high end' approach to life.
>Con:
>Cd's suffer greatly from poor engineering and mastering and are
>often "slapped onto the disc" as if proper attention isn't
>required(the "digital solves everything" view)
Also a problem with LP, note.
>They also have a problem of a too low sampling rate.
In your opinion. Mine too, as it happens, but still not a problem when
merely comparing to vinyl.
>My bet is on DVDs for music in a few years.
Technically, agreed. Whether 24/192 is *necessary* is another matter.
I believe that the multi-channel potential is far more important.
>They can encode MP3s with no loss/compression
Pretty good trick, since MP3 is a lossy encoding process.......
> and 512 bit+
>encoding and still fit over 80 minutes of music plus extras on the
>video tracks. Nice.
I think you need to revise your numbers - or are you simply referring
to MP3 at 512 kb/sec bitrate? Still not CD quality, note.
>5-channel audio would also be a reality - not the compressed
>junk on movies, but real 5-channel recordings of symphonies and
>such at 512 bit or better.
6 channels at 24/96 is part of the DVD-A spec, no need for nasty
compression techniques - with apologies to JJ! :-)
jj, DBT thug and skeptical philalethist
Stewart Pinkerton <ste...@pinkertons.fsnet.co.uk> wrote:
>>Try it yourself and see what you come up with. It'll be
>>significantly higher than 128.
>OK, such is progress, but we're still talking about greater than 5:1
>lossy compression (i.e. throwing away >80% of the data), which is
>*never* going to be as good as leaving the data in one piece.
Really, even from a perceptual point of view? Have some
proof on hand?
Try using MPEG-2 AAC at 5:1 compression and see how things
sound.
By the way, Stewart, people have been trumpeting the demise of
coding of all sorts since at least 1976 (and I'm sure before).
In 1977 at the first technical talk I ever gave, a person stood
up to tell me "This ADPCM stuff will never fly, the networks will
be big enough that you don't compress anything within 5 years.
Coding is a dead end."
In 1979 at an AES meeting in LA, when I was discussing 56 and 64 kb/s
"commentary grade coding", someone said "Coding will never be necessary
in the audio field, nobody will accept it and we'll all be able to
send all the bits we want in 5 years".
In 1986 at the AES "Conference on Digital Times", someone stood
up and said, when I'd discussed (briefly) perceptual audio coding,
"Nobody will ever use audio coding, and storage will be so big in
5 years that it won't ever be necessary".
In 1994 a marketeer told me "nobody will ever send music over
the internet, there's no demand".
In 1996 somebody else said: "Really, why bother with coding, we'll
all have broadband connections in 2 years".
There are a whole variety of such anecdotes I could point out,
each one asserting that the network will grow faster than what
we have to put on it. Presently, today, the saturation is worse
than it was in 1976, and the level of coding for a comparible signal,
which was 2:1 in 1976 (for speech) is now 8:1.
Ed Seedhouse
Pinkerton) wrote:
>Joseph Oberlander <josephob...@earthlink.net> writes:
>> It bothers me because I played woodwinds for ten
>>years and know how they sound - and CD just doesn't have RL sparkle
>>at higher frequencies because it takes incredibly fast sampling rates
>>to capture upper-range harmonics and CD is just not quite fast enough
>>for most insturmental music.
>What, a 20kHz bandwidth is not enough for music? It's wider than
>you'll find on 99.9% of LPs.............
The idea that playing an instrument is going to tell you anything much
about how that instrument sounds to the audience when you play it
strikes me as being totally illogical.
I myself used to play flute (badly and a long time ago) and every once
in awhile I got to listen to someone else playing a flute in the
concert hall and I was always struck by how unlike the sound of a
flute in a hall from the audience was to the sound of my flute when I
played it. When you think about it, how could it be otherwise?
I also sing, and I've heard my voice on tape. It is astonishingly
different from the way my voice sounds to me. In fact if you were
looking for an expert to identify my voice I'd say the *last* person
you should ask would be me.
jj, DBT thug and skeptical philalethist
Joseph Oberlander <josephob...@earthlink.net> wrote:
>So, given that:
>- Sounds less than 20db are usually not present in recorded music.
Except that you completely leave out the spectrum of the
recorded sound, and in some critical bands, the level will be
much lower than that.
>- Sound over 80db are not usually present either.
A recent performance I've recorded peaked out in the mid-90's,
and could keep that up for a while, from 6th row center.
>- 40hz is pretty much the limit for musical recordings.
Unless pipe organ or some few instruments are involved, in which
case one can get much lower, the results are mostly sensation,
not audition, but that's perceptable.
>- 15khz is pretty much the top of most music for both formats.
Except for some percussion, some pipe organs (again, although
not a lot of energy) or anything that is (ick!) close-miked,
etc, etc.
>- People can't hear the difference between .1 db and 1db...
Except that they can hear a 1dB level shift and then can't
generally hear a .1dB level shift.
What was the point here?
jj, DBT thug and skeptical philalethist
Joseph Oberlander <josephob...@earthlink.net> wrote:
>You need to do some research about sampling rates.
Excuse me? Who are you addrressing here?
>CD can do 20Khz,
>but the sampling suffers.
Now that's a total, absolute, complete, nonsensical fallacy.
It's the same fallacy that people have been repeateding
over and over and over again, time after time, since the
first time that the world in general got to look at PCM,
and it's FALSE. I don't mean "kinda false" I mean
100.00%, completely wrong.
Any information at 20kHz is preserved to the level of quantization
accuracy, just like any information at any other frequency below
20kHz. Period. This is not a "guess", this is not a "supposition",
this is the basic, provable MATHEMATICS of the situation, and this
basic MATHEMATICS is born out exactly by the performance of PCM
devices worldwide.
>Remeber - this is a 15+yr old
>technology now and when they made the standard, it was for practical
>(marketing and legngth) reasons and not audio quality.
That's partially correct, but your statement about 20khz is still
dead wrong, and completely inaccurage.
>The sampling quality of high-frequency harmonics makes a dramatic
>impact on subtle things like imaging and whether or not it sounds
>somewhat like a live performance.
Do you have any proof of this assertion? I doubt it, but let's
see your proof, if any.
>Then you need to spend some time with a real Oscilloscope.
YOU need to learn how to READ your oscilloscope. I'm talking
tested, verified and easily demonstrated here. If you can see
any "jagged" edges, you have failed to use an antiimaging
filter at the output.
The fact that you have incorrectly designed your system is
not a PCM prolem.
For some graphs that clear this up pretty much completely, please
see http://www.AudioAsylum.com/audio/general/messages/43134.html.
The discussion with it may help, too. DO Be sure to read the
first article so you know what you're looking at. This demonstration
shows clearly that the "jaggedness" comes purely and absolutely
from signals ABOVE half the sampling rate, putting your fallacy
straight to death.
>You can record the signal with no compression if you want. What you
>gain is a huge increase in sampling quality. Of course, the file
>tends to be 20+ megs a minute...(grin)
Horse puckey. Perhaps you need to read how MP3 works. I have read
how MP3 works, after all, I'm one of the people who wrote
the standard, thank you.
Aaron J. Grier
> My bet is on DVDs for music in a few years. They can encode MP3s with
> no loss/compression and 512 bit+ encoding and still fit over 80
> minutes of music plus extras on the video tracks. Nice.
>
> 5-channel audio would also be a reality - not the compressed junk on
> movies, but real 5-channel recordings of symphonies and such at 512
> bit or better.
it appears that you are confusing bit depth to bit rate. the two are
somewhat independent of each other.
a CD has a bit depth of 16 bits per channel. there are two of these 16
bit channels, and each one is sampled at 44.1kHz. doing some simple
arithmetic, we can figure out the bit rate: 2 * 16 * 44100 = 1411200
bits per second.
I assume the "512 bit" MP3 encoding you refer to above is 512kbit/sec,
or 512000 bits/sec.
comparing the bit rates, the 512kbit/sec MP3 must some how get rid of
64% of the data in order to do its job. with MP3 this is a lossy
operation. there is no way to recover the original data from MP3
encoded file.
--
Aaron J. Grier | "Not your ordinary poofy goof." | agr...@poofygoof.com
"No single layer of abstraction can save the world or even make
good coffee." -- Kevin P. Neal
Stewart Pinkerton
>To paraphrase Steven Sullivan in another post:
>
>"If only one record compared to the equivalent CD (both superb recordings
>and masterings) sounds better, than ipso facto the problems associated with
>vinyl cannot be attributable to the medium".
>
>Two different experiences, two different conclusions. How interesting.
You may paraphrase all you choose, but the 'if' still remains very
much a matter of personal preference, rather than absolute fact.
The various real problems of vinyl are indeed inherent to the medium,
only the degree of their annoyance varies with the listener.
Stewart Pinkerton
>In article <9r6roi$m3k$1...@bourbaki.localdomain>,
>ste...@pinkertons.fsnet.co.uk wrote:
>> Contrary to several statements in this thread however, the vast
>> majority of vinyl *is* compressed and peak limited, plus it's rolled
>> off above 15kHz to avoid overheating the cutter
>
> OK, quick question for Stew here: what order of filter would that
>be, Stewart? Do you understand the concept of filter slope? Given a
>broadband input of 10-100Khz, how many db down would you figure 60K
>would be- infinite?
Normally second-order. Well, gee whiz, I may have to think about that
for a while.......Say, could it be 6dB/octave for every order? Yeah, I
think that's it...............
With a 2nd-order minimum-phase slope at 15kHz, I'd expect a 60k signal
to be around 22-24dB down. Your point?
>> while bass is summed
>> to mono below 80Hz to preserve groove depth, and rolled off below 40Hz
>> to maximise playing time.
>
> By the same token, the vast majority of CDs from the last 5-10
>years or so are produced through Pro Tools' undithered 16 bit output
>buss, and do you hear us analog fans complaining?
I believe you will find that statement to be factually challenged.
> Well- yes, of course you do. Because that sounds pretty awful and
>has actually given Pro Tools itself a somewhat unjustified bad name
>among pro audio types.
Anyone actually using an undithered system is deserving of a bad name.
As noted, I believe you're blowing smoke about anything other than
'garage project studios'.
> But my point is, most audiophiles either don't
>get the records like that, or get 'em and eventually sell them or
>throw them away. In particular, the audiophile warhorse vinyl records
>do not have these characteristics. You're talking about teen pop
>disposable records- you're talking about a lot of records which
>audiophiles would run screaming from.
That was also true of '70s vinyl...........
> I could as easily give a list of CDs which have been horribly
>maimed in sound for the sake of loudness- did you know that the big
>secret weapon from a couple years ago up to now has been digital
>distortion, nothing less than brutally clipping the waveform AFTER
>hard peak limiting, so that it becomes intentionally flat-topped and
>puts out fierce crunchy distortion?
Yes, I'm aware of that (no different really to what electric guitar
players have been doing for many decades). What has this to do with
the attainable fidelity of the medium? There is no technology which
cannot be poorly implemented..............
> One advantage of vinyl is that you _can't_ do that to it: you
>can't peaklimit that hard without busting the needle.
You can do it to the master tape *before* it gets to the cutter.
Sheesh!
> Instead of
>fighting to defend the dubious honor of the CD, maybe you'd better
>keep an eye out for what's being done to it, because it's been a
>steadily worsening progression- and the direction it's leading will
>have any schmuck off the street insisting that vinyl has more dynamic
>range than CD in level-matched comparisons, simply because the CD is
>_destroyed_ with incredibly aggressive peak limiting until it has no
>dynamics at all, nothing but 'loud'.
This has nothing to do with CD, but with incompetent record producers.
As you rightly say, who gives a flying fart about the 'quality' of
teen-pop records? Same was true back in Phil Spector's day.......
> That is what lurks out there waiting to exact karmic vengeance on
>your generalizations of LP bass-summing/rolling-off habits. In all
>seriousness- don't even go there. ;)
In all seriousness, please stop blowing smoke about what bad record
producers can do to screw up *any* medium, and talk instead about the
capabilities of the two media when *well* implemented.
Stewart Pinkerton
>In article <9r799k$ved$1...@bourbaki.localdomain>,
>Richard D Pierce <DPi...@world.std.com> wrote:
>>If that were true, all we would need to do is change the
>>bias oscillator frequency and we would change the high frequency
>>limit. And that CERTAINLY is not the case at all.
>
>It's also true that if you get anywhere near a
>bias frequency input, you'll get, due to nonlinearity,
>something hideously close to aliasing.
>
>However, increasing the bias frequency ALONE won't do a
>thing for the HF response of the average tape deck, indeed.
Perhaps we can encapsulate all of the above by noting that tape speed
and head gap width typically set the upper limit on FR, with the bias
frequency set sufficiently high above this point as to avoid the
horrible effects which would occur if you *could* record a signal
within 30kHz of it.
Harry Lavo
news:9r7cu...@enews4.newsguy.com...
> There is one rather major point not yet mentioned. A heck of a lot of
> music is available on vinyl, from the 1950s to the present day. Far
> more than on CD. And most of us already have lots of vinyl, and want
> to hear it!
>
> So if you want to enjoy your music, you need vinyl. With CD you won't
> hear it at all, as it just wont play properly when you jam the record
> into the CD player ;)
>
> This has to be one big reason to keep using vinyl.
>
> The first requirement of a high end system is that it can play your
> music. If it can't, it's not high end in my book, its just
> pretentious.
>
> Vinyl has many technical shortcomings, yet surprisingly it still seems
> to sound superb some of the time. For some reason all those sources of
> distortion just don't seem to be ones my ears are too worried about -
> except for the clicks and pops. I have watched a record do some very
> non textbook things and still sound pretty good. I have found that
> pitch accuracy isn't the most important thing after all.
>
> Those of us who've put the work into setting up a decent turntable
> know that it is *possible* to get excellent results from vinyl. But
> most decks people own are pretty awful. Vinyl has such a poor
> reputation simply because it is so easy to get it wrong, and quite
> difficult to get right. Add all those scratchy abused records, and its
> no wonder people usually go for CD.
>
> BTW CD is *far* more complex than vinyl, but the complexities are all
> hidden from the user. With vinyl you have to play with the small
> number of complexities yourself.
>
> Vinyl is very primitive technology, yet for some reason it just seems
> to shine on.
>
> I have masses of vinyl I bought second hand, of music that is simply
> not available on CD.
Don't you know? Its the piper's lure of euphonic distortion leading
us directly over the cliffs to our death! Children, wake up! Listen
to your parents!
Denis
> Denis wrote:
> >
> > The real undistorted dynamic range of RMS values(which really
> > reflects the difference between piano and forte) of current CD
> > format(44 kHz/16bits) is about 20 dB.
>
> Where on earth did you get this information?
This is simple:
1. The well-known figure of 96 dB reflects the theoretical limit for
the range of INSTANT signal values(the more precise result is even 98
dB). Of course, each bit of quantization adds 6dB of this range if
there are no limitations imposed by real A/D and D/A convertors.
The real-world equipment gets 93 dB average.
This is, again, the range of instant signal values between extreme
positive and negative peaks. This fact simply reflects that there are
2^16=65536 possible discrete values between these extremes.
2. Now we will subtract.
First, 6 dB out for single polarity(from zero to peak) - one gets
93-6=87 dB.
Next, The most significant distortion of analog signal in digital
domain is well-known quantisation noise. The level of this distortion
is defined by least significant bit(16th in our case). The lower the
signal value, the smaller the number of quantization steps between the
negative and positive peaks - the higher this distortion. Moreover,
these high-order distortions are quite noticeable by our ears, and if
a complex signal like music is dealt with, there will appear a broad
spectrum of intermodulation components, which will fill all the
audible range! The reasonable lower limit for the signal level (peak
value!) is therefore at least -40 dB. At this level the THD already
reaches 1% in real-world A/D - D/A chain.(Look at the test results for
any CD player or DAC.)
Here we get 87-40=47dB.
Next, The crest factor, i.e., the ratio between peak and RMS values
for a pure sine wave is 1.41 or 3dB, that of a square wave is 1 or 0
dB, and this ratio for the music signal is at least 3 or 10 dB, even
this signal is heavily processed. The sounds of some acoustical
instruments, for instance, cymbals, may have this crest factor
exceeding 20dB. So, let it be 12 dB.
Here we get 47-12=35dB.
Next, let me again notice the complex nature of the musical signal.
This signal contains a lot of spectral components including those of
very low level. These low-level components determine the fine
structure of signal and are to be preserved. Our ears together with
the hearing analyser in our brain can distinguish the coherent
components which are some 15 dB below the wideband background. The
trained ear of a radio operator can trace the sound of Morse code even
at lower levels. It is the proven fact. Therefore to be sure these
least audible components do not fall below the mentioned -40 dB, we
must get the allowance.
Finally, 35-15=20dB.
Of course, these 20dB is not a hard figure. Depending on the signal
complexity and the use of pre-emphasis, one can add some 6 or
sometimes 10 dB. But no more!
The dither makes just a little help, it reduces the THD figure from
30% at the levels below -90 dB to mere 10%. The distortion figures at
higher levels change a little. Indeed, even this help is necessary to
make the digital distortions less noticeable.
Just 20 or 30 dB is not too bad. For comparison, the RMS dynamic range
of AM broadcasts seldom exceed 10 dB. Such degree of compression is
necessary to get the acceptable transmitter efficiency.
The progress in the sound of CDs over the years of their existence
reflect the gaining experience of recording engineers how to cope with
above limitations.
This experience finally resulted in adoption of higher resolution
formats, first 20bit, and now, 24 bit with higher sampling
frequencies.
At least, why one needs a 144dB dynamic range if it is a true RMS
range?
Theoretically, transition from 16 to 24 bit format can add 48 dB to
above 20 dB figure. This will not so bad, just approaching the dynamic
range of live symphonic orchestra. Indeed, this will be possible only
if the A/D and D/A converters used possess true 24 bit accuracy for
all audio range, i.e., the total conversion error must not exceed 1/2
of 25th bit! This is by now impossible. For example, the jitter even
in the best digital circuits does not allow one to preserve even
16-bit resolution at 20 kHz. Nothing to say about 24 bits. Prof.
Heisenberg rules here! More realistic improvement in RMS dynamic range
for 24-bit format against 16-bit one is about 30 dB, so one gets just
50dB. This is something like the FM broadcast or 7.5 ips analog tape
without compander. Or just like the commercial LPs of 1970s.
Nevertheless, 24-bit format is a major step ahead in digital
recording.
From other side, vast majority of recorded music, rock and pop first,
does not require too large dynamic range on the final medium. So the
present CD format serves well for it's major purpose to be
inexpensive, convenient and reliable mainstream source.
Indeed, there is another music.
For a brief comparison, modern high-end LP releases easely give true
80 dB of RMS dynamic range. The surface noise is an additive one in
contrary to the quantisation distortion, and does not deteriorate the
low-level content of music spectrum. Therefore we should add those 15
dB mentioned above.
And one obtains 80+15=95dB!
Remember, this is true RMS figure! If one wants to get the
peak-to-peak range to make the direct comparison with the digital
formats, one must add 12 db for the mentioned crest factor and 6 dB
for the two polarities.
The final result: 95+12+6=113dB
Serious comments and questions are appreciated.
Denis.
jj, DBT thug and skeptical philalethist
Harry Lavo <harry...@rcn.com> wrote:
>Don't you know? Its the piper's lure of euphonic distortion leading
>us directly
to enjoying the music. What's the problem.
>over the cliffs to our death! Children, wake up! Listen
>to your parents!
Enjoy the music, don't jump!
Richard D Pierce
>ste...@pinkertons.fsnet.co.uk (Stewart Pinkerton) wrote in
>news:9r8gme$qnv$1...@bourbaki.localdomain:
>
>> Actually, since mass storage gets bigger and cheaper every
>> year, *any* form of compression is rapidly becoming
>> unnecessary. Ten years from now, you'll be able to get 6
>> channels of 24/192 on a smart card.
Well, that's not where compression is important: it's really
important in transmission channels, where we are not going to
see the same sorts of size/cost ratio reductions.
>
>Maybe, but certain kinds of compression will always be useful.
>Hard drive capacities keep doubling, but I don't see computer
>users abandoning WinZip and other forms of file compression.
And WinZip and other such compression forms are UTTERLY useless
for compressing audio.
We're not talking about that kind of compressions. It's rare to
see WinZip and other data-priented L-Z based algorithms achioeve
more than a couple of percent reduction, and not uincommon to
see a slight INCREASE in the size of teh compressed file.
>Perhaps lossless technology will mature enough to reduce
>filesizes by 75% or more. No matter how cheap storage gets, it
>would always be beneficial to fit 4 times as much data in a given
>space.
Or 4 times as much data over the same bandwidth channel.
Graeme Jaye
skeptical philalethist) wrote:
>In article <9r737...@enews3.newsguy.com>,
>Graeme Jaye <graem...@iname.com> wrote:
>>On Wed, 24 Oct 2001 16:34:42 GMT, Ron-A...@magmar.freeserve.co.uk
>>wrote:
>
>>>Most analogue LPs actually come from tape recorders that don't have a
>>>"continuous signal" - the bias frequency certainly sets an upper
>>>frequency limit.
>
>>Nonsense! Of course an analogue tape recorder provides an analogue
>>output.
>
>Really now, this is relevant how?
The claim was that tape machines delivered a 'discontinuous' output.
This is nonsense.
>The statemetn made by
>Ron Ardell is very simple, the bias frequency sets an
>upper limit.
No it doesn't.
>DO you have any evidence otherwise?
I need 'evidence' for a basic engineering fundamental - I think not.
How does some 40 odd years in the recording business suit you?
>Do you know why stereo FM
>tuners have "subcarrier traps", and the like?
Yes - I know exactly why. It's to avoid beat frequencies being
generated between the bias and the pilot tone carrier.
>He's right, you know.
He's wrong - but you obviously don't know.
>>Upper frequency limit is mainly established as a combination of tape
>>speed and head gap - it has nothing to do with the biasing.
>
>I see, and if I put in a signal at 90% of the biasing frequency,
>you assert I won't get anything out at 10% of the bias frequency?
If you put a signal in at 90% of the bias frequency it will generate
sum and difference beat frequencies. This will, likely, fall within
the recordable frequency range of the machine and that is what you
will hear.
>Try it!
Try what? Create beats? Been there, done that, many times!!
Wasn't a very desirable effect though :-)
Audio Restoration and CD Repair
http://www.personal-cd.com
jj, DBT thug and skeptical philalethist
Denis <denis_af...@mail.ru> wrote:
>1. The well-known figure of 96 dB reflects the theoretical limit for
>the range of INSTANT signal values(the more precise result is even 98
>dB). Of course, each bit of quantization adds 6dB of this range if
>there are no limitations imposed by real A/D and D/A convertors.
Wrong. You have to dither.
>The real-world equipment gets 93 dB average.
That's because of dither.
>This is, again, the range of instant signal values between extreme
>positive and negative peaks.
No, that's not what it is. It's the noise floor energy dividing the
maximum signal energy.
It's not "instant signal values" it's RMS values, i.e. energy
integrated over time.
>First, 6 dB out for single polarity(from zero to peak) - one gets
>93-6=87 dB.
Fallacy. Sign bit counts. You fail to realize that energy is what
you're dealing with.
>Next, The most significant distortion of analog signal in digital
>domain is well-known quantisation noise.
It's not "distortion", it's noise.
>these high-order distortions are quite noticeable by our ears, and if
>a complex signal like music is dealt with, there will appear a broad
>spectrum of intermodulation components, which will fill all the
>audible range!
Stuff and nonsense. There's a reason nobody ever measured this.
>The reasonable lower limit for the signal level (peak
>value!) is therefore at least -40 dB.
Nice assertion, but you've failed to show any evidence.
>At this level the THD already
>reaches 1% in real-world A/D - D/A chain.(Look at the test results for
>any CD player or DAC.)
And if you go 40 dB down from any LP, you're 10 dB from the noise.
That's 31.6% distortion.
>Here we get 87-40=47dB.
Wrong on two counts. First 40 db is wrong, second, you're subtracting
the wrong thing.
The rest of your analysis is so abjectly flawed that i don't know
where to start.
Joseph Oberlander
>
> I do, all the time, every day, and they DO NOT show jaggedness
> unless that system is BROKEN.
This is largely due to the fact that every poor d/a converter in the chain
adds a little bit of noise/errors. Some engineers are good about getting rid
of this, but many are poorly mastered - and when you combine this with most
converters in consumer-grade CD players, it doesn't look the same as the
original in analog form.
OTOH, as pointed out, moving to 92K solves most of these problems,
and with compression techniques, you can get lossless data at the
higher rate and still fit on a standard CD. What's not to like?
Joseph Oberlander
>
> Joseph Oberlander wrote:
>
> > - 40hz is pretty much the limit for musical recordings.
>
> substantial musical content below 40 Hz, and even below 30
> Hz.
>
> > 5-channel audio would also be a reality - not the compressed
> > junk on movies, but real 5-channel recordings of symphonies and
> > such at 512 bit or better.
>
> movies. I do need to footnote your comments and point out
> that many movies have very uncompressed sound. Some have
> immense dynamic range. Some of the Dolby Digital music
> recordings I have (at 448 kbps) are very clean and realistic
> sounding.
Isn't DVD great? :)
My friend has a simmilar Blue-Man Group DVD. Astounding quality
since they were no longer stuck with 2-channels.
fathom
news:9r9of...@enews2.newsguy.com:
> In article <9r9hl...@enews4.newsguy.com>, fathom
> <tgr...@maig.com> wrote:
>>ste...@pinkertons.fsnet.co.uk (Stewart Pinkerton) wrote in
>>news:9r8gme$qnv$1...@bourbaki.localdomain:
>>
>>> Actually, since mass storage gets bigger and cheaper every
>>> year, *any* form of compression is rapidly becoming
>>> unnecessary. Ten years from now, you'll be able to get 6
>>> channels of 24/192 on a smart card.
>
> Well, that's not where compression is important: it's really
> important in transmission channels, where we are not going to
> see the same sorts of size/cost ratio reductions.
>
>>
>>Maybe, but certain kinds of compression will always be
>>useful. Hard drive capacities keep doubling, but I don't see
>>computer users abandoning WinZip and other forms of file
>>compression.
>
> And WinZip and other such compression forms are UTTERLY
> useless for compressing audio.
> We're not talking about that kind of compressions. It's rare
> to see WinZip and other data-priented L-Z based algorithms
> achioeve more than a couple of percent reduction, and not
> uincommon to see a slight INCREASE in the size of teh
> compressed file.
Yes, I know. I wasn't suggesting WinZip be used for audio; there
are lossless schemes that can compress audio by roughly 50%. I
only used Winzip to illustrate that even with hard drives getting
bigger and cheaper, people still compress their data.
>>Perhaps lossless technology will mature enough to reduce
>>filesizes by 75% or more. No matter how cheap storage gets,
>>it would always be beneficial to fit 4 times as much data in
>>a given space.
>
> Or 4 times as much data over the same bandwidth channel.
Yup. Bandwidth, rather than storage space, is the key reason for
compression.
Joseph Oberlander
(snip)
Note - I said in *general*. Most music sounds accpetable on several
different formats and so you have to look at the pros and cons of the media.
I think CDs and records are just not good enough and there is technology
available that seems to solve all the problems.
Chris Johnson
(Denis) wrote:
> 1. The well-known figure of 96 dB reflects the theoretical limit for
> the range of INSTANT signal values(the more precise result is even 98
> dB). Of course, each bit of quantization adds 6dB of this range if
> there are no limitations imposed by real A/D and D/A convertors.
> The real-world equipment gets 93 dB average.
> This is, again, the range of instant signal values between extreme
> positive and negative peaks. This fact simply reflects that there are
> 2^16=65536 possible discrete values between these extremes.
That's a given- though you're not making any allowance for dither.
Since I use dither, let's say it's 98 db, not 93.
> 2. Now we will subtract.
> First, 6 dB out for single polarity(from zero to peak) - one gets
> 93-6=87 dB.
I'm not quite sure where you're getting this, so I'm not going to give
you this one: 98 db.
> Next, The most significant distortion of analog signal in digital
> domain is well-known quantisation noise. The level of this distortion
> is defined by least significant bit(16th in our case). The lower the
> signal value, the smaller the number of quantization steps between the
> negative and positive peaks - the higher this distortion. Moreover,
> these high-order distortions are quite noticeable by our ears, and if
> a complex signal like music is dealt with, there will appear a broad
> spectrum of intermodulation components, which will fill all the
> audible range! The reasonable lower limit for the signal level (peak
> value!) is therefore at least -40 dB. At this level the THD already
> reaches 1% in real-world A/D - D/A chain.(Look at the test results for
> any CD player or DAC.)
> Here we get 87-40=47dB.
This, I _will_ give you: this is why serious recording and mastering
uses longer wordlengths than 16 bits. As you start compressing and
bringing quiet sounds up, they do get kinda crunchy when you start with 16
bit undithered, and it sounds plausible that -40 db is where you start
getting problems with this, as in THD over 1%, and noticeable tonal
degradation.
So, that's 58 db: the quietest point at which detail is still vaguely
high-resolution.
> Next, The crest factor, i.e., the ratio between peak and RMS values
> for a pure sine wave is 1.41 or 3dB, that of a square wave is 1 or 0
> dB, and this ratio for the music signal is at least 3 or 10 dB, even
> this signal is heavily processed. The sounds of some acoustical
> instruments, for instance, cymbals, may have this crest factor
> exceeding 20dB. So, let it be 12 dB.
> Here we get 47-12=35dB.
This kinda assumes you're comparing to some hypothetical only-RMS
signal, and frankly I don't think it applies here: it applies to people
who are concerned with 'using all the bits' of their recordings, but even
there, the peaks are what bring openness and life to a recording, and you
don't _want_ to squash them. I don't concede that interpretation of the
effective resolution of a digital signal must be a factor of RMS levels,
and I don't grant you this one. We still have 58 db, in my book.
> Next, let me again notice the complex nature of the musical signal.
> This signal contains a lot of spectral components including those of
> very low level. These low-level components determine the fine
> structure of signal and are to be preserved. Our ears together with
> the hearing analyser in our brain can distinguish the coherent
> components which are some 15 dB below the wideband background. The
> trained ear of a radio operator can trace the sound of Morse code even
> at lower levels. It is the proven fact. Therefore to be sure these
> least audible components do not fall below the mentioned -40 dB, we
> must get the allowance.
> Finally, 35-15=20dB.
I'll grant this, but only a little bit- I wouldn't go so far as to say
that counts as 15 db limitation on effective resolution. That said, I
think it's more than 6 db or so- how about 10 db?
So, we then have 48 db usable dynamic range out of CD. Not 20 db.
You've got some right ideas but it's not as bad as all that, you're taking
it too far.
> Of course, these 20dB is not a hard figure. Depending on the signal
> complexity and the use of pre-emphasis, one can add some 6 or
> sometimes 10 dB. But no more!
> The dither makes just a little help, it reduces the THD figure from
> 30% at the levels below -90 dB to mere 10%. The distortion figures at
> higher levels change a little. Indeed, even this help is necessary to
> make the digital distortions less noticeable.
You can say that again- do you know just how much improvement you can
get with specialised dither? You can get upwards of 40 db improvement!
Granted, that's not over the full bandwidth... but still, that's the one
factor that is capable of bringing CDs into the range of seriously good
quality. Maybe it's not always done, maybe it's even a little bit rare,
but you _can_ get upwards of 90 db _usable_ dynamic range out of CD. We
just usually don't: many things can get in the way of that kind of
performance.
> The progress in the sound of CDs over the years of their existence
> reflect the gaining experience of recording engineers how to cope with
> above limitations.
Damn straight.
> This experience finally resulted in adoption of higher resolution
> formats, first 20bit, and now, 24 bit with higher sampling
> frequencies.
Actually, specialised hardware can use 48 bits and more for internal
processing, or floating-point math at 32 bits and up. This stuff is out
there in use now, most notably by the mastering community.
> Theoretically, transition from 16 to 24 bit format can add 48 dB to
> above 20 dB figure. This will not so bad, just approaching the dynamic
So will really high-performance wordlength reduction...
> For a brief comparison, modern high-end LP releases easely give true
> 80 dB of RMS dynamic range. The surface noise is an additive one in
> contrary to the quantisation distortion, and does not deteriorate the
> low-level content of music spectrum. Therefore we should add those 15
> dB mentioned above.
> And one obtains 80+15=95dB!
Actually, I agree with the concept, but disagree strongly that you
therefore add 15 db effective dynamic range. Best you can do is not
subtract it, and your 80 db seems high: how about 80 db at best, and not
subtracting much for the 'nondestructive' type of noise? No way do you add
the noise, it only subtracts. Let's call it 12 db of noise, which makes
the high-end LP more like 68 db, with a very different character to the
noise floor.
> Remember, this is true RMS figure! If one wants to get the
> peak-to-peak range to make the direct comparison with the digital
> formats, one must add 12 db for the mentioned crest factor and 6 dB
> for the two polarities.
> The final result: 95+12+6=113dB
Again, RMS vs peak has nothing to do with it: peaks are musically
useful. Also, I don't know where you're getting polarities, you're going
to get seriously beat about the head on that one (by people capable of
pouncing on a '20 cps' typo). Nope and nope: in my book it's still 68 db
absolute max.
So, using your own reasoning as much as I felt was applicible, that
would leave us with:
CD: between 48 and 90 db usable dynamic range, absolute max
LP: around 68 db usable dynamic range, absolute max.
In either case, hitting the extreme values would take some doing: for
CD, it'd have to be very high resolution through the entire recording
process, reduced to 16 bit at only the very last stage, and we're also
assuming a very aggressive dither that loads lots of error energy in the
extreme highs to hit that kind of performance in the critical midrange:
some people find the sonic signature of this objectionable. For LP, you'd
have to make a direct-to-disc or something, and be playing it on a
seriously high-end turntable.
In practice, using junkier equipment, I suspect the real numbers are
more like:
CD: 48 db if you're lucky before it starts to sound compromised- with
some badly recorded CDs, this can be as little as 20 or 30 'clean' db.
Repeated processing and truncation to a 16 bit buss will do that.
LP: 40 db if you're lucky! Thankfully a lot of this is rumble and types
of distortion that don't damage the musical content as severely.
So, you wind up with roughly equal formats, but with _wildly_ different
'styles' of sound degradation, subject to personal preference of what type
of degradation annoys you most.
Chris Johnson
Richard D Pierce
Joseph Oberlander <josephob...@earthlink.net> wrote:
>> >Then you need to spend some time with a real Oscilloscope.
>>
>> I do, all the time, every day, and they DO NOT show jaggedness
>> unless that system is BROKEN.
>
>This is largely due to the fact that every poor d/a converter in the chain
>adds a little bit of noise/errors.
Absolutely completely totally false. I'm talkgin about an
example where one takes the signal, runs it through a single
A/D-D/A process at 16 bit/44.1 kHz, and ABSOLUTELY NO jaggedness
appears on the output AT ALL. None.
That jaggedness represents information that is outside the
baseband spectrum, and can ONLY exist if the system DID NOT
implement the REQUIRED anti-imaging filter properly. With that
prerequeiste filter in place, the original signal will be fully
reconstrcuted within the quantixation limits of the system with
NO jaggedness whatsoever.
Sorry, sir, I have seen been doing this sort of measurement and
work THOUSANDS of times and the ONLY time I have seen it is in a
sampler that is either hooribly and incompetently implemented
(and that MIGHT be the case with cheap music samplers, but not
in anything having to do with CDs) or the system is decidedly
broken
(Actually the only other time I have seen it is in sample rate
converters implemented in software on PCs and the like, but this
is simply another exampe of incompetent implementation).
Stewart Pinkerton
>I understand that my Otari MTR 12II 30ips machine is down -3db at
>28-30khz at 0db reference level, which is a hell of a lot better than
>CD. And it is only a "B" level professional recorder.
>
>I understand that my Teac 7030SL 15ips machine is down -2db at 24khz
>at 0db reference level on most tape formulations. And it is only a
>semi-pro (albeit top quality) machine.
>
>And I understand that my Schoeps small diameter mics in omni mode
>extended to 26khz.
>
>And that the new Sennheiser 800 microphone extends to 50khz.
>
>What is your point?
My point is that I entirely agree with your numbers, and they are less
than half an octave above the capability of 'plain Jane' 16/44 CD.
They are also an octave above the hearing limit of most adults.....
Steven Sullivan
: In article <9r737...@enews3.newsguy.com>, Steven Sullivan
: <sull...@gwis2.circ.gwu.edu> wrote:
:> Rob Gold <rgvi...@aol.com> wrote:
:> : 1) The Rube Goldberg contraptions known as vinyl LPs have no right
:> : sounding good at all, yet at their best they do. 2) Theoretically,
:> : CDs *should* sound magnificent, but too often in my experience they
:> : are flat, two-dimensional, uninvolving and lifeless-sounding. Go
:> : figure. 3) For the time being I'll keep building my LP collection on
:> : the cheap, and buy CD's (or whatever) to fulfill my new-music jones.
:>
:> 4) By deduction from (2) , if even *one* CD does *not* sound flat,
:> two-dimensional , uninvolving and lifeless, then the fault cannot be
:> in the compact disc medium or digital audio per se.
: Worth commenting on- the same is true for analog, of course. If
: even one vinyl record (a direct-to-disc extravaganza?) substantially
: outperforms the usually dismal technical performance cited, then...
The question reduces to: What, if any, commonly cited limitations of vinyl are
inherent to the medium -- as in, unavoidable even in the best case.
A good queston, I'd like to know the answer to.
No is denying that LPs sound good to some people, btw.
--
-S.
The 80's was a really hard thing to put up with. -- Steve Howe
Steven Sullivan
: Dick,
: unlike you, most of us can't talk to the late Mr. Wiener...
But like Mr. Pierce, you probably *do* have access to usenet archives
via Google.com, where Mr. Weiner's psots abotu audio are available.
: --
: You Know Who~
: "No matter how cynical you get,
: it is impossible to keep up." -- Lily Tomlin
: -----------------------------------------------------
: Cat stuff, Win Help & Purgatory
: @ http://You_Know_Who.home.att.net/
: `````````````````````````````````````````````
: "Richard D Pierce" <DPi...@world.std.com> wrote in message
: news:9r7di2$2ri$1...@bourbaki.localdomain...
: | In article <9r78le$uts$1...@bourbaki.localdomain>,
: | Chris Johnson <jinx...@sover.net> wrote:
: | >In article <9r6se...@enews2.newsguy.com>,
: | >ste...@pinkertons.fsnet.co.uk wrote:
: | >
: | >> Seious compromises? CD captures *everything* that is on any analogue
: | >> master tape up to more than 20kHz, which is above the limit of hearing
: | >> of the vast majority of adults. How is this a 'serious compromise'?
: | >
: | > This is an absurd claim. Talk to some mastering engineers before
: | >mouthing off in this manner... you've got no idea how foolish that sounds
: | >to someone who actually has the job of doing the capturing.
: |
: | Well, I, for one, HAVE and DO talk to a number of mastering
: | engineers all the time. And quite a number of them flatly
: | diagree with your mischaracterizations of their opinions and
: | flatly state the contrary of what you are claiming. I would only
: | refer you, for example, to the late Mr. Wiener, who mastered
: | some of the finest classical recordings in ANY medium, who
: | clearly contradicts what you claim.
: |
: | >
: | > It does not. It's hard work capture _most_ of the openness, ease and
: | >tonal richness of even 15 ips tape, and the professional standard is 30
: | >ips half-inch two-track tape. I question if you've ever heard _or_
: | >measured this.
: |
: | You may question it all you want, but I and quite a few others
: | have seen instances where self-proclaimed astute audiophiles
: | have been unable to distiguish the direct feed from such a tape
: | recorder from a 44.1/16 bit encode/decode process, using
: | equipment of their choice to listen.
: |
: | I thus, in light of these result, question whether YOU'VE heard
: | to many mastering engineers or heard and measured the results
: | you claim.
Arny Krueger
news:9r9hl...@enews4.newsguy.com...
> Howard Ferstler <hfer...@mailer.fsu.edu> wrote in message
news:<9r738...@enews3.newsguy.com>...
> > Denis wrote:
> > > The real undistorted dynamic range of RMS values(which really
> > > reflects the difference between piano and forte) of current CD
> > > format(44 kHz/16bits) is about 20 dB.
> > Where on earth did you get this information?
> This is simple:
As later comments will show, its also almost entirely false.
> 1. The well-known figure of 96 dB reflects the theoretical limit for
> the range of INSTANT signal values(the more precise result is even 98
> dB). Of course, each bit of quantization adds 6dB of this range if
> there are no limitations imposed by real A/D and D/A converters.
> The real-world equipment gets 93 dB average.
93 dB sounds about right, if you include the effects of adding
dither, etc.
> This is, again, the range of instant signal values between extreme
> positive and negative peaks. This fact simply reflects that there are
> 2^16=65536 possible discrete values between these extremes.
> Now we will subtract.
> First, 6 dB out for single polarity(from zero to peak) - one gets
> 93-6=87 dB.
This was already reflected in item (1). Therefore this entire item
is falsified. We're still at 93 dB.
> Next, The most significant distortion of analog signal in digital
> domain is well-known quantisation noise. The level of this distortion
> is defined by least significant bit(16th in our case). The lower the
> signal value, the smaller the number of quantization steps between the
> negative and positive peaks - the higher this distortion. Moreover,
> these high-order distortions are quite noticeable by our ears, and if
> a complex signal like music is dealt with, there will appear a broad
> spectrum of intermodulation components, which will fill all the
> audible range! The reasonable lower limit for the signal level (peak
> value!) is therefore at least -40 dB. At this level the THD already
> reaches 1% in real-world A/D - D/A chain.(Look at the test results for
> any CD player or DAC.)
> Here we get 87-40=47dB.
All digital systems are properly dithered. Properly done dithering
which is not uncommon, completely eliminates quantization distortion.
Therefore this entire item is falsified. We're still at 93 dB.
> Next, The crest factor, i.e., the ratio between peak and RMS values
> for a pure sine wave is 1.41 or 3dB, that of a square wave is 1 or 0
> dB, and this ratio for the music signal is at least 3 or 10 dB, even
> this signal is heavily processed. The sounds of some acoustical
> instruments, for instance, cymbals, may have this crest factor
> exceeding 20dB. So, let it be 12 dB.
> Here we get 47-12=35dB.
This was already reflected in item (1). We're still at 93 dB.
> Next, let me again notice the complex nature of the musical signal.
> This signal contains a lot of spectral components including those of
> very low level. These low-level components determine the fine
> structure of signal and are to be preserved. Our ears together with
> the hearing analyzer in our brain can distinguish the coherent
> components which are some 15 dB below the wideband background. The
> trained ear of a radio operator can trace the sound of Morse code even
> at lower levels. It is the proven fact. Therefore to be sure these
> least audible components do not fall below the mentioned -40 dB, we
> must get the allowance.
> Finally, 35-15=20dB.
All digital systems are properly dithered. Properly done dithering,
which is not uncommon, allows people to hear signals that are many dB
(at least 15dB) below the quantization level. Therefore, this entire
item is falsified. We're still at 93 dB.
> Of course, these 20dB is not a hard figure. Depending on the signal
> complexity and the use of pre-emphasis, one can add some 6 or
> sometimes 10 dB. But no more!
> The dither makes just a little help, it reduces the THD figure from
> 30% at the levels below -90 dB to mere 10%. The distortion figures at
> higher levels change a little. Indeed, even this help is necessary to
> make the digital distortions less noticeable.
All digital systems are properly dithered. Properly done dithering,
which is not uncommon, completely eliminates low-level distortion.
Therefore, this entire item is falsified. We're still at 93 dB.
Joseph Oberlander
> What, you have a problem with CD sounding good without having to spend
> $10k on the replay gear? Interesting 'high end' approach to life.
No, just that they both are old technology that needs upgrading.
I didn't like my records much more than I like my CDs.
> >Con:
> >Cd's suffer greatly from poor engineering and mastering and are
> >often "slapped onto the disc" as if proper attention isn't
> >required(the "digital solves everything" view)
>
> Also a problem with LP, note.
Absolutely. I rate both as having flaws and problems.
> I think you need to revise your numbers - or are you simply referring
> to MP3 at 512 kb/sec bitrate? Still not CD quality, note.
Okay - I did some checking. it would actually be at 88 or 96K
sampling - whatever that works out to in MP3-speak.
> >5-channel audio would also be a reality - not the compressed
> >junk on movies, but real 5-channel recordings of symphonies and
> >such at 512 bit or better.
>
> 6 channels at 24/96 is part of the DVD-A spec, no need for nasty
> compression techniques - with apologies to JJ! :-)
Well, then, I'm happy. I guess they've been doing their homework :)
So why aren't they releasing plain music on them yet? I seems like it
would solve all of the problems and give you surround in the process.
(all the room reflections and ambience - so it feels like a concert
hall or jazzz club or...)