Vinyl Lovers Only Read This....
Bill Poletti
[ Given the benefit of 20/20 hindsight, I'll likely regret ever
approving this post. At a minimum, it suffers from two major
problems. It is makes technically inaccurate assertions,
particularly in its characterization of dither. It also paints
those that disagree with the poster with a very broad, and
completely negative, paintbrush. But, Mr. Poletti has put his stake
in the ground, and it seems more appropriate to me that the readers
deal with the issues he raises than the moderators. So, have at it
folks. Do exercise a degree of common sense though. If this turns
into a shouting match, we'll shoot the thread in the back of the
head and be done with it entirely. -- jwd ]
VINYL LOVERS ONLY.
Fellow-vinyl lovers and music lovers,
We all know the real story about vinyl. True accuracy, sweetness,
correct tracking of the signal from the original recording, REAL HIGH
sampling rates, all that tech stuff. We know that there is superior
detail and that the queues are much more realistic. All we have to do
is turn to any review of a CD DAC or player and look at the undithered
1k sine wave plot. Not the shape of anything I've encountered in
nature. And NOT a sine wave.
Dithering the CD signal is the method of intentionally adding
distortion to the signal in order to remove the harshness. And the CD
lovers think they are getting the same thing as on the CD. And that
same dithering adds a noise floor. So much for a clean signal. Its
easy to hear that dithered distortion. Sorta sounds like sandpaper on
wood tracing the sound.
Well, I think we have a major opportunity....
I feel we are approaching the comparison between vinyl and CD all
wrong. The CD lovers are being closed minded and are not listening to
vinyl on systems capable of reproducing the wonders of the media.
They are arguing religion. Dogma, not real-world. They are listening
to their CDs without knowing what they are missing, or they have never
had a chance to make a real comparison between good vinyl on a well
set-up and tweaked-in table and a real top-of-the-line CD set-up
such as a Levinson DAC. We all know that even a Rega or circa 1980 AR
table will outperform the best of CD. They are the ones missing out.
If that's the way they are, then their credibility suffers.
I have a different tack and suggestion.
We tell em they're right. CDs are superior in every way. Us vinyl
lovers are just poor old tin-eared low-techers that won't let go of
the medium. We let em know that we think their superior technical
specifications always win out. After all, amps with the same specs
sound the same, don't they?
We convince them that out equipment really isn't good enough to hear
the superiority of CD.
So, that being the case, we're willing to trade them these superior
CD recordings for their records. Since records aren't really any
good, how about three records for each CD. Assuming that the records
are in good condition, of course.
Now, for those of us (vinyl lovers) that also like CD (myself
included), we can have both worlds. I just like vinyl better.
But vinyl isn't perfect. Not at all. Far from it! But it is
difficult to get reel-to-reel tape (I'm too young to have been around
as an audiophile during tape's peak) of second generation (from
mixdowns) of original masters. My experience is that there is at
least as big a gap between those recordings and vinyl as between vinyl
and CD.
My other suggestion is that we spend more time at live, un-amplified
musical performances. Always as eye-opener for me...
Though I make no apologies for the above statements, I am not trying
to be an elitist. Just being realistinc using my system and other for
whom I have a high amount of respect.
***************
OK, so I admit my deep cynacism in this post. But even Sony and
Philips have publicly stated their failings in making CDs even equal
the performance of better vinyl, much less surpass it. They have
clearly failed in their efforts to provide "perfect sound forever" to
those of us who still prefer to listen to music and not to read the
specifications provided by the techno-geeks,
Bill Poletti
Todd Jenkins
In article bi...@i1.net (Bill Poletti) writes:
>[ Given the benefit of 20/20 hindsight, I'll likely regret ever
I think I might agree... ;)
>VINYL LOVERS ONLY.
This post applies to everyone, not just vinyl lovers.
>We know that there is superior
>detail and that the queues are much more realistic.
The queues? Say again?
>All we have to do
>is turn to any review of a CD DAC or player and look at the undithered
>1k sine wave plot. Not the shape of anything I've encountered in
>nature. And NOT a sine wave.
Have you looked at the -80dB output of a 1khz sine wave from a
properly set up turntable? Considering your statement, I doubt it.
>Dithering the CD signal is the method of intentionally adding
>distortion to the signal in order to remove the harshness.
Errr....I don't think so...
>Its easy to hear that dithered distortion. Sorta sounds like sandpaper on
>wood tracing the sound.
If it is as you say, perhaps I should then comment, it's easy to hear
the nonlinearities in the vinyl playback system in phase, amplitude,
high-frequency droop, the requirements of heavy equalization to
provide the correct final amplitude, spl limitations in the low
frequencies, pops and crackles from errant bits of dust that couldn't
quite be removed, the necessity of a high-pass filter and it's
associated phase anomolies to prevent rumble and low-frequency
feedback out of the system, the inherent low-level feedback in the
playback system (i.e. sound waves feeding back into the pickup), etc.
Take your pick...I choose CD.
>The CD lovers are being closed minded and are not listening to
>vinyl on systems capable of reproducing the wonders of the media.
>They are arguing religion. Dogma, not real-world.
I might say the very same about vinyl lovers. Claiming absolute
superiority of one format over another is quite arrogant, don't you
think? Borders on religion, don't you think? A bit dogma-ist, don't
you think?
>We all know that even a Rega or circa 1980 AR
>table will outperform the best of CD. They are the ones missing out.
Actually, I don't think that I am missing out on much at all. I have
auditioned some of the best of both sources, and found both lacking.
However, I prefer CD by an overwhelming margin for many reasons.
Clarity, linearity, dynamic range and most of all
repeatability/reliability/ accessability. I have found that pressings
transferred to CD definitely sound different, and that rereleases are
quite often horrible (such as the '93 20-bit remastering of
Aerosmith's Greatest Hits CD) compared to the old tape I have laying
around. And I don't count adding euphonic distortions as
"outperforming" in general.
>If that's the way they are, then their credibility suffers.
Ah, but that isn't "the way they are."
>We tell them they're right. CDs are superior in every way. Us vinyl
>lovers are just poor old tin-eared low-techers that won't let go of
>the medium.
Actually, I suspect that you have heard the CDs available in the genre
you listen to (which I suspect is classical) and found them lacking.
I agree in general.
>I just like vinyl better.
And I just like CD better.
>My other suggestion is that we spend more time at live, un-amplified
>musical performances. Always as eye-opener for me...
I wholeheartedly agree. However, in the genre I typically listen,
this isn't an option. In fact, pretty much the only genre where this
is an option is classical and related areas. In most other areas,
such as blues, jazz, rock, r&b, alternative, new age, etc. this is not
generally an option (in some cases it is).
>Though I make no apologies for the above statements, I am not trying
>to be an elitist.
Well, it comes across very elitist and somewhat arrogant. Claiming
absolute superiority of analog over digital or vice versa is as
ridiculous as claiming that white paint is obviously a better color
for walls than tan, or that whites are obviously smarter than blacks
(or insert your ethnic group here). ***Please, no flames on race
here, the objective of bringing this up is to show how ridiculous such
a statement can be, and how offensive such a statement could be to
most people.***
>...But even Sony and
>Philips have publicly stated their failings in making CDs even equal
>the performance of better vinyl, much less surpass it. They have
>clearly failed in their efforts to provide "perfect sound forever" to
>those of us who still prefer to listen to music and not to read the
>specifications provided by the techno-geeks.
I would be very interested in seeing actual references to quotes by
Sony and Philips saying that CD is inferior to vinyl. And again, the
"perfect sound forever" was a pitch by marketing, and in no way
reflected the opinions of engineers involved in the project, who well
knew the limitations of the system. This red herring will never
die...
Perhaps this post is a waste of time, considering your obvious
overriding preference for vinyl and refusal to see any of it's
limitations. For the real world, CD is the source of choice. If we
could all have $5000 turntables and all listened to music at one place
and one place only (that being the perfectly set up listening room),
and all had the time and effort to spend to properly tune and retune
and retune the system, and that vinyl was transportable and impervious
to changes in atmosphere, then we might all listen to vinyl. However,
this is not the case, and it seems that you discount the many
advantages of CD for the few (but still significant) advantages of
vinyl. We all have our preferences, but to claim absolute superiority
is preposterous. Audio playback is simply a huge compromise...
Cheers!
Todd Jenkins
Project Engineer - EV/Altec Lansing
The opinions expressed herein do not represent those of EV/Altec Lansing
William Brent
On Jul 30, 1996 11:10:07 in article <Vinyl Lovers Only Read This....>,
'bi...@i1.net (Bill Poletti)' wrote:
>
>But vinyl isn't perfect. Not at all. Far from it! But it is
>difficult to get reel-to-reel tape (I'm too young to have been around
>as an audiophile during tape's peak) of second generation (from
>mixdowns) of original masters. My experience is that there is at
>least as big a gap between those recordings and vinyl as between vinyl
>and CD.
be without them. but pre-recorded reels are not the end all that
newcommers might think. generally; theye were dubbed onto inferior
stock at high speed, they have hiss and are subject to drop out (if
not maintained), a magnitized head will do as much damage as a worn
stylus. and reels were often re-issued at slower speed to save a few
pennies on stock and be able to reporduce them in half the time.
also columbia house would sell reels to club members, but these were
3.75ips clones of the non-club 7/5ips releases.
the exception I've found is the mid 1970 mag-tape releases, but these
were four track 'quad' releases. they compare verry favorable to the
vinyl (stereo or quad) of the same titles and I'd stack the Allman
brother's Filmore East reels against the CDs anyday (I have both as
well as the vinyl).
what regular issue reels did give people was ... no pops, ticks,
scartches, and superior seperation of the stereo tracks. but like all
media; its hit or miss. there is no panacea for me, so I use all
formats (though I did give up on 8tracks :-). I use DAT, but alway
make a backup on Beta-HiFi and 15ips reel. I buy CDs, and if the
sound is at least as good as the viny, I'll retire the vinyl (I can't
replace that, but can always get another CD).
but not everything I like will ever be available on CD, so even if I
did not like vinyl, I'd need to keep it around. and of course, as year
chases year... its a lot easier for me to read the liner notes on an
LP
bb
--
.,-*'`^`'*-,.__.,-*'`^`'*-,.__.,-*'`^`'*-,.__.,-*'`^`'*-,._
--==[> All spelling errors due to line noise <]==--
--==[> E-Mail me at bbr...@pipeline.com <]==--
_.,-*'`^`'*-,.__.,-*'`^`'*-,.__.,-*'`^`'*-,.__.,-*'`^`'*-,._
Tim Takahashi
Bill Poletti <bi...@i1.net> wrote:
> We all know the real story about vinyl. True accuracy, sweetness,
> correct tracking of the signal from the original recording, REAL HIGH
> sampling rates, all that tech stuff. We know that there is superior
> detail and that the queues are much more realistic. All we have to do
> is turn to any review of a CD DAC or player and look at the undithered
> 1k sine wave plot. Not the shape of anything I've encountered in
> nature. And NOT a sine wave.
I do find it amusing, make that ironic, that the published
"oscilloscope" tests are of analogue systems reproducing 1kHz square
waves (I'm thinking of the 1970's High Fidelity cartridge tests) at
high modulation levels where today we look at 1kHz sine waves at very
low levels from digital systems.
Somehow, with dramatic differences in sound possible with the same
recording equipment, only different microphone placement, all of this
argument over the post-production media seems to be getting a bit
stale.
-tim
SDuraybito
In article <4tmo58$n...@agate.berkeley.edu>, tjen...@michiana.org
(Todd Jenkins) writes:
> Perhaps this post is a waste of time, considering your obvious
> overriding preference for vinyl and refusal to see any of it's
> limitations. For the real world, CD is the source of choice.
[ quoted text deleted -- jwd ]
> However, this is not the case, and it seems that you discount the
> many advantages of CD for the few (but still significant) advantages
> of vinyl. We all have our preferences, but to claim absolute
> superiority is preposterous. Audio playback is simply a huge
> compromise...
What would you consider the advantages of vinyl?
[ If this question turns you to be more than it appears, a loaded
question if you will, this thread (or at least this sub-thread) is
history. Have a nice evening ... -- jwd ]
John Busenitz
On 30 Jul 1996, Bill Poletti wrote:
> deal with the issues he raises than the moderators. So, have at it
> folks. Do exercise a degree of common sense though. If this turns
> into a shouting match, we'll shoot the thread in the back of the
> head and be done with it entirely. -- jwd ]
> VINYL LOVERS ONLY.
Are vinyl lovers those inflatable dummies for those who have problems
with flesh & blood lovers?
> We all know the real story about vinyl. True accuracy, sweetness,
> correct tracking of the signal from the original recording, REAL HIGH
> sampling rates, all that tech stuff. We know that there is superior
I thought all that "tech stuff" didn't matter? Is this your opinion or
are you asserting it is fact?
> detail and that the queues are much more realistic. All we have to do
Care to back up your claims? These are just your opinions.
> is turn to any review of a CD DAC or player and look at the undithered
> 1k sine wave plot. Not the shape of anything I've encountered in
> nature. And NOT a sine wave.
So? What's your point? And exactly how many things have you
"encountered in nature"?
> Dithering the CD signal is the method of intentionally adding
> distortion to the signal in order to remove the harshness. And the CD
No, dither is something quite different than that.
> lovers think they are getting the same thing as on the CD. And that
> same dithering adds a noise floor. So much for a clean signal. Its
Surprise! The noise floor was always there, though it is lower than
the noise floor on analog media. And if the definition of dither was
"intentionally adding distortion to the signal", (which it definitely
is not), where does noise come in?
> easy to hear that dithered distortion. Sorta sounds like sandpaper on
> wood tracing the sound.
Is this just your opinion, or can you prove it?
> I feel we are approaching the comparison between vinyl and CD all
> wrong. The CD lovers are being closed minded and are not listening to
> vinyl on systems capable of reproducing the wonders of the media.
> They are arguing religion. Dogma, not real-world. They are listening
Again, where's your evidence? Just because someone likes a particular
medium better than another, and that medium is more accurate, that
person is closed-minded and dogmatic? Kids, this is a good example of
why one should stay in school.
> to their CDs without knowing what they are missing, or they have never
> had a chance to make a real comparison between good vinyl on a well
> set-up and tweaked-in table and a real top-of-the-line CD set-up
> such as a Levinson DAC. We all know that even a Rega or circa 1980 AR
> table will outperform the best of CD. They are the ones missing out.
No, we don't "all know that". This is your opinion, not a fact.
Please learn the difference.
> But vinyl isn't perfect. Not at all. Far from it! But it is
> difficult to get reel-to-reel tape (I'm too young to have been around
> as an audiophile during tape's peak) of second generation (from
> mixdowns) of original masters. My experience is that there is at
> least as big a gap between those recordings and vinyl as between vinyl
> and CD.
Wow! Finally something with a little sense. Yes, a well-mastered CD is
so close to the original tape that the sound quality gap between the
original and the vinyl is about the same as the gap between the CD and
the vinyl. Are you seeing the light?
> Though I make no apologies for the above statements, I am not trying
> to be an elitist. Just being realistinc using my system and other for
> whom I have a high amount of respect.
Then WHY do you assert your opinion as fact? Why do you insult those
with a different opinion than you? Why do you make so many bogus
technical claims? It seems as though you have no clue as to how sound
is reproduced and the concepts involved. By any chance might you have
skipped high-school physics class to be with your vinyl lover?
> OK, so I admit my deep cynacism in this post. But even Sony and
> Philips have publicly stated their failings in making CDs even equal
> the performance of better vinyl, much less surpass it. They have
Prove it! I've never heard of such. And even if so, that doesn't mean
it's the truth.
> clearly failed in their efforts to provide "perfect sound forever" to
There never were "efforts to provide 'perfect sound forever'". That is
merely propoganda.
> those of us who still prefer to listen to music and not to read the
> specifications provided by the techno-geeks,
Who's the techno-geek"? You are the one talking about "undithered
sinewaves" and making the (unsubstantiated) technical assertions.
You're a wanna-be techno-geek. Once you become the real thing, then
we'll listen to your "ditther is distortion" assertion.
Enough talk of lovers, vinyl or otherwise. Discussing audio
reproduction and trading opinions of various audio gear is what this
newsgroup is about, not asserting an opinion as fact and insulting
those who don't agree or are technically versed.
_____________________________________________________________
John Busenitz buse...@ecn.purdue.edu
P.U. ECE http://cernan.ecn.purdue.edu/~busenitz
Disclaimer: My statements do not represent Purdue University.
James Ker
Tod Jenkins wrote:
>I would be very interested in seeing actual references to quotes by
>Sony and Philips saying that CD is inferior to vinyl.
Well, interestingly enough, the followinng is a press release that the
consumer marketing support division of Philips Canada put out at the
beginning of 1987, in relation to availability of AKD cartridges. I
should also add that it was rather hastily withdrawn, and doubtless
they would now say that CD has moved on a bit!
Anyway:
"Product: AKD fouth generation cartridges
Yes, we still make phono cartridges!
Research and Development have never stopped, because we firmly believe
a strong future for our phonocartridges exist.
Regardless of Compact Disc or Digital Audio, or Audio and Video
Recorders, the old black vinyl disc is here to stay. Without getting
into discussion what's better or worse, over 800,000,000 black records
will still need good phono cartridges to be enjoyed by their owners.
The introduction of Compact Disc gave us a great challenge, and we are
proud to report that with out new range of phonocartridges, we can
achieve a more accurate and musical sound reproduction than with any
CD system. And isn't it music that it's all about?"
Interesting, isn't it?
James
David Bath
SDuraybito wrote:
> What would you consider the advantages of vinyl?
I would say it makes a much better frisbee than CDs do :)
--
David Bath mailto:ba...@aud.alcatel.com
Lightwave Systems Engineer Alcatel Telecom, Richardson, TX
Standard Disclaimers
Todd Jenkins
In article sdura...@aol.com (SDuraybito) writes:
>What would you consider the advantages of vinyl?
Without going overboard, there are a few factors that do indeed weigh
in vinyl's favor. The first and most obvious is the degree of
"euphonic distortion" in terms of harmonics. The effect is similar to
the comparison between an average tube and ss amp to me. This
manefests itself in a bit of "warmness" and "sweetness," etc. The
next, which ties into the same effects, is the loss of very-high
frequency content (above 12khz) which with many current tweeters is
aggravating (even the acclaimed Focal Ti series has irregular response
in the very-hf range with swept-sine wave testing and a breakup mode
around 18khz). The serious advantage is that with the tweeter, you
*could* simply place a very small inductor on the line of the woofer
which would give a very slight rolloff in the very hf range, but it
won't have the same audible effects as the initial signal drooping (at
least not in my experience). I have yet to hear a tweeter that can
accurately reproduce the 10-20khz band without irregularities...in
many cases it's better to slightly reduce the hash and harmonics than
not...
Ben Cannon
tjen...@michiana.org (Todd Jenkins) wrote:
>In article sdura...@aol.com (SDuraybito) writes:
>
>>What would you consider the advantages of vinyl?
>
>Without going overboard, there are a few factors that do indeed weigh
>in vinyl's favor. The first and most obvious is the degree of
>"euphonic distortion" in terms of harmonics. The effect is similar to
>the comparison between an average tube and ss amp to me. This
>manefests itself in a bit of "warmness" and "sweetness," etc. The
>next, which ties into the same effects, is the loss of very-high
>frequency content (above 12khz) which with many current tweeters is
>aggravating
I think I should add dynamic compression manifesting itself as
"greater detail" as well as trasnparency! I forst noticed this on a
very good FM station that I live next to, a song I recorded to DAT
sounded BETTER than the CD I have (exact same edit, etc) It had to be
compression.
>(even the acclaimed Focal Ti series has irregular response
>in the very-hf range with swept-sine wave testing and a breakup mode
>around 18khz).<snip> I have yet to hear a tweeter that can
>accurately reproduce the 10-20khz band without irregularities...in
>many cases it's better to slightly reduce the hash and harmonics than
>not...
You should give the newest Focal TC120TDX an audition! It uses a
small phasing cone to correct the earlier Ti series's very few
shortfalls. It's now *easily* the best tweter in the world!
The speakers I build with a pair of TDXs sound better (to me) than the
Wilson WATTs!!
--
Ben Cannon. art...@a.crl.com
"If you think it more, you will know it." -Joly
I have moved! To http://168.75.121.10/ben.htm
Not much, however the old adress no-longer works.
Matt Wenham
Todd Jenkins (tjen...@michiana.org) wrote:
> If it is as you say, perhaps I should then comment, it's easy to hear
> the nonlinearities in the vinyl playback system in phase, amplitude,
> high-frequency droop, the requirements of heavy equalization too
> provide the correct final amplitude ... ^^^^^^^^^^^^^^^^^^
I'm quite sick of CD users using this as an argument against vinyl
playback. Yes, there are many technical superiorities in CD playback,
but this is a red herring. You do not expect to be able to plug a CD
laser head straight into your pre-amp do you? You can't do it with a
tape head, FM aerial, or any LP cartridge which is remotely 'hi-fi'.
The signal coming off a CD requires orders of magnitude more
pre-processing to produce a line-level signal than that coming out of
a MM or MC cartridge. The fact that turntable manufacturers have
almost unanimously chosen to leave the details of the circuitry down
to amplifier designers is their loss. It is possible to produce a
state of the art MM/MC head amp which will easily fit inside most
turntables using only two or three op-amps per channel (external power
supply [best place for it] not withstanding). All that is then needed
is a passive pre-amp (volume control) and power amplification before
feeding to the loudspeakers. If ever CD playback gets this simple,
I'll be very surprised.
> In fact, pretty much the only genre where this
> is an option is classical and related areas. In most other areas,
> such as blues, jazz, rock, r&b, alternative, new age, etc. this is not
> generally an option (in some cases it is).
It certainly is an option for jazz listeners.
Matt Wenham.
---
The Analogue Addicts Archive, on line at:
http://www.york.ac.uk/~mjgw100/aaa.htm
Bob Myers
Matt Wenham (mjg...@mailer.york.ac.uk) wrote:
> Todd Jenkins (tjen...@michiana.org) wrote:
> > If it is as you say, perhaps I should then comment, it's easy to hear
> > the nonlinearities in the vinyl playback system in phase, amplitude,
> > high-frequency droop, the requirements of heavy equalization too
> > provide the correct final amplitude ... ^^^^^^^^^^^^^^^^^^
> I'm quite sick of CD users using this as an argument against vinyl
> playback. Yes, there are many technical superiorities in CD playback,
> but this is a red herring. You do not expect to be able to plug a CD
> laser head straight into your pre-amp do you? You can't do it with a
> tape head, FM aerial, or any LP cartridge which is remotely 'hi-fi'.
That's not the point, Matt. The fact is that LP playback REQUIRES
significant equalization (more properly, compensation for some nasty
things that have to be done to the signal during the process of making
the LP) which is done in the analog domain. (I put that qualifier on,
because it would certainly be possible to do the RIAA fix a lot
cleaner in the digital domain - but THAT would be WRONG, right? :-))
The CD doesn't have any such need - outside of the reconstruction
filter, which doesn't have to have a significant impact over the
majority of the audio frequency range, digital audio places no
intentional screw-ups like this in the analog signal path. The goal
of the D/A and reconstruction filter is to get the signal into analog
form as cleanly and LINEARLY as possible; the equalization required in
the LP playback chain is a necessary and significant NON-linearity in
amplitude vs. frequency response, and unless done VERY much more
carefully than is typical, throws in all sorts of interesting phase
effects as well.
> The signal coming off a CD requires orders of magnitude more
> pre-processing to produce a line-level signal than that coming out of
> a MM or MC cartridge. The fact that turntable manufacturers have
Really? Let's see - you read the CD with a laser, send the resulting
data to a DAC, and filter the output so you don't see anything above
22 kHz. THIS is preprocessing? I suppose you're going to throw in
the data decoding and error correction steps in there as
"pre-processing", but there's no justification for doing so: these
steps DO NOT have an effect on the data finally delivered to the DAC,
other than the rare occurence of the error correction stepping in and
returning the data to what it was SUPPOSED to be in the first place.
There is NO processing prior to the DAC in terms of anything which
intentionally modifies the audio signal.
On the other hand, the LP process is roughly -
1. "Read the data" by dragging the stylus through a groove. In this
process,
add the surface noise of the vinyl to the signal, plus add
distortions
caused by flexing of the groove walls, turntable speed variations,
and
so forth.
2. Convert the mechnical motion of the stylus to an electrical signal,
most commonly by wiggling a magnet in a coil or vice-versa. Add
distortion
due to the inherent non-linearities of these components, the
mechanical
resonances of the various structures, etc., and add noise. In an
analog system, you ALWAYS add noise at each stage - it is
unavoidable.
3. Amplify the signal, through a circuit which has a seriously
non-flat
response in terms of amplitude vs. frequency (intentional) and in
terms
of phase response (not intentional, but generally unavoidable).
Add
distortion (for obvious reasons, see above), and add noise (again).
You have now reached a point where the signal is equivalent to what the
DAC delivers; all processing after this is the same for either source, and
so isn't to be considered in evaluating differences in the systems.
Frankly, I can't see the case for LP playback, if you REALLY analyze
what's going on, as being either "simpler" OR somehow "more pure".
Bob Myers KC0EW Hewlett-Packard Co. |Opinions expressed here are not
Workstations Systems Div.|those of my employer or any other
my...@fc.hp.com Fort Collins, Colorado |sentient life-form on this planet.
Lon Stowell
In article <4u02n9$k...@agate.berkeley.edu> Ben Cannon
<art...@a.crl.com> writes:
>I think I should add dynamic compression manifesting itself as
>"greater detail" as well as trasnparency!
You left off that the compressed dynamic range will have a
higher average loudness and will sound like a wider dynamic
range to the untrained ear. Also the compressed dynamic range,
along with the highly distorted and monophonic upper treble will
make transients much more obvious, giving the illusion to the
untrained ear that the transients have more snap and sparkle.
JJMcF
It seems to me that what these discussions typically lack is
perspective. I have recordings on virtually all media from acoustic
78 to CD. For somebody like me whose musical preferences are
classical or jazz (or other primarily nonelectronified forms) the
difference between an acoustic 78 and a CD is less than that between
any recording and a live performance, by a long way. The differences
between CD and LP are minuscule, virtually undetectable, on this
scale. Having said this, however, I generally prefer to listen to
analogue recordings--the experience is artificial, of course, but
somewhat closer to real music because of some difficult-to-identify
more-mechanical quality of most CDs.
Richard D Pierce
In article <4u5sv3$m...@agate.berkeley.edu>,
Bob Myers <my...@hpfcla.fc.hp.com> wrote:
>Matt Wenham (mjg...@mailer.york.ac.uk) wrote:
>> Todd Jenkins (tjen...@michiana.org) wrote:
>> > If it is as you say, perhaps I should then comment, it's easy to hear
>> > the nonlinearities in the vinyl playback system in phase, amplitude,
>> > high-frequency droop, the requirements of heavy equalization too
>> > provide the correct final amplitude ... ^^^^^^^^^^^^^^^^^^
>> I'm quite sick of CD users using this as an argument against vinyl
>> playback. Yes, there are many technical superiorities in CD playback,
>> but this is a red herring. You do not expect to be able to plug a CD
>> laser head straight into your pre-amp do you? You can't do it with a
>> tape head, FM aerial, or any LP cartridge which is remotely 'hi-fi'.
>That's not the point, Matt. The fact is that LP playback REQUIRES
>significant equalization (more properly, compensation for some nasty
>things that have to be done to the signal during the process of making
>the LP) which is done in the analog domain. (I put that qualifier on,
>because it would certainly be possible to do the RIAA fix a lot
>cleaner in the digital domain - but THAT would be WRONG, right? :-))
Well, yes, LP's DO REQUIRE RIAA deemphasis, in addition to the
substantial gain needed. But, as I have pointed out here before, RIAA
equalization is not hard to get right (having said that, there are
MANY existance proofs out there that it's not hard to get wrong,
either). The RIAA equalization, being a minimum-phase function, is
completely and 100% correctable: it's a matter of carefule
implementation, that's all. You can get get it within 0.1 dB without a
lot of effort, and with similarily vanishingly small phase errors as
well. No big problem.
The problem comes from the fact that you are at the mercy of
electrical-mechanical (at the cutter stage) and mechanical-electrical
transducers, with all the attendant non-linearity and resonance issues
associated with such, and, despite the whinings of a variety of
ill-informed people to the contrary, it is at these transduction
interfaces where the greatest errors occur and that are the MOST
difficult to get even approximately right.
>intentional screw-ups like this in the analog signal path. The goal
>of the D/A and reconstruction filter is to get the signal into analog
>form as cleanly and LINEARLY as possible; the equalization required in
>the LP playback chain is a necessary and significant NON-linearity in
>amplitude vs. frequency response, and unless done VERY much more
>carefully than is typical, throws in all sorts of interesting phase
>effects as well.
Well, no, unless the designer REALLY screws it up (and, again, I point
to a variety of instances, both cheap and pretentiously expensive,
where the designer REALLY screwed it up), the actual phase an
amplitude errors in the entire RIAA emphasis-deemphasis chain can be
made VERY small. So it really isn't an a priori problem.
Again, it comes down to the fact that it's a LOT easier to get the
RIAA EQ correct than it is to get a phono cartridge even remotely
close. The phase and amplitude response of real cartridges is
embarrisingly bad, despite the rather limited, information-starved
"tests" and manufacturers claims to the contrary. The non-linearities
in the cutter heads and cartridges FAR exceed those found in the
emphasis and deemphasis.
>3. Amplify the signal, through a circuit which has a seriously non-flat
> response in terms of amplitude vs. frequency (intentional) and in terms
> of phase response (not intentional, but generally unavoidable). Add
> distortion (for obvious reasons, see above), and add noise (again).
Again, Bob, I have to disagree, and invite you to look at the work of
Baxandall and the like regarding how well one can get VERY cose to
completely eliminating these effects. As I mentioned above, the RIAA
preemphasis is a minimum-phase function, as is the deemphasis. If I
build a mesh that has the right time constants, the net response can
be flat with constant group delay: no big deal.
Fine, so you can build a phono preamp that accurately amplifies all
the faults of the phono cartridge. Whoopee.
The rest of your points about the faults of the media, the
non-linearities in the transduction process, etc., are certainly right
on, though.
--
| Dick Pierce |
| Loudspeaker and Software Consulting |
| 17 Sartelle Street Pepperell, MA 01463 |
| (508) 433-9183 (Voice and FAX) |
CAdams1471
In article <4u5sv3$m...@agate.berkeley.edu>, my...@hpfcla.fc.hp.com (Bob
Myers) writes:
>1. "Read the data" by dragging the stylus through a groove. In this
>process, add the surface noise of the vinyl to the signal, plus add
>distortions caused by flexing of the groove walls, turntable speed
variations,
>and so forth.
I have often wondered if it were not likely that "digital sound" was
simply a lack of analog distortion. The interaction of the stylus with
the record surface, for example.
Folks who have been use to hearing LP playback for many years may have
grown accustomed to the anomalies and simply find their absence from
digital playback to be disturbing.
To me, there is an elegance to digital playback due to its inherent
simplicity as so well described my Bob. Far more simple than vinyl
playback can ever be. (discounting acoustic 78 playback).
Christopher G. Adams
Bob Myers
Richard D Pierce (DPi...@world.std.com) wrote:
> Well, yes, LP's DO REQUIRE RIAA deemphasis, in addition to the
> substantial gain needed. But, as I have pointed out here before, RIAA
> MANY existance proofs out there that it's not hard to get wrong,
> either). The RIAA equalization, being a minimum-phase function, is
> completely and 100% correctable: it's a matter of carefule
> implementation, that's all. You can get get it within 0.1 dB without a
> lot of effort, and with similarily vanishingly small phase errors as
> well. No big problem.
Dick, this seems to be the only point on which we disagree, and I'm
not sure we really disagree at all.
I agree that in theory, the RIAA deemphasis can be done properly. And
I think we agree that, in actual practice, we can point to a large
number of examples where it ISN'T. That was really my point.
(Of course, a person feeling particularly annoying would at this point
note that there's probably not much sense in getting the deemphasis
right in phase response in the first place, given the horrors that the
LP system overall is going to inflict on the sound anyway...:-))
Hell, none of the stuff in audio, with the notable exception of your
particular black art, is all THAT hard to get right. Despite popular
opinions to the contrary, properly handling and amplifying signals
over such a limited range of amplitudes and frequency as is seen in
audio is NOT the electronic equivalent of brain surgery. In fact, it
is relatively EASY to do an acceptable job, as evidenced by the
plethora of designs which are "acceptable". I doubt that there's an
electron-pusher alive, degreed or not, who hasn't at some time penned
an audio amp design.
> Well, no, unless the designer REALLY screws it up (and, again, I point
> to a variety of instances, both cheap and pretentiously expensive,
> where the designer REALLY screwed it up), the actual phase an
> amplitude errors in the entire RIAA emphasis-deemphasis chain can be
> made VERY small. So it really isn't an a priori problem.
See, we agree. I didn't say it was an a priori problem. But if we
walk into the average audio outlet, and start grabbing boxes at
random, I wonder how many we'll find that don't get it right?
And we definitely agree on the myriad faults of the transducers that
have to do with the vinyl at either end of the
LP-making-and-playing-back process.
M...@magnequest.com
John Busenitz wrote:
>> deal with the issues he raises than the moderators. So, have at it
>> folks. Do exercise a degree of common sense though. If this turns
>> into a shouting match, we'll shoot the thread in the back of the
>> head and be done with it entirely. -- jwd ]
Ouch! Kevorkian would have been kind enough long ago to help put this
patient out of intractable miseries. But what do you pack JWD...or is
it the pipe that you offer? <G>
[ Sorry, if I told you, then I would have to kill you ... -- jwd ]
>> I feel we are approaching the comparison between vinyl and CD all
>> wrong. The CD lovers are being closed minded and are not listening to
>> vinyl on systems capable of reproducing the wonders of the media.
>> They are arguing religion. Dogma, not real-world.
Dogma pervades the whole issue on every side and every front. Each
camp refuses to recognize the preferences of the other and instead
gets into pissing matches about the techie stuff.
If you prefer analog (or digital) just simply state your preferences
and be done with it. Doug Purl (on a private mail list) proffered
that he was sick of tweakoids (and let me throw in techies) and the
often humorous explanations offered as foundations to preferences. I
have to agree...at best the arguments (from a purely technical basis)
never appear to be persausive...the only utility the appear to have in
these great debates of our times is to assauge the need for validation
of individual preferences of the folks arguing the matters at hand.
As a counterexample to the above...might I suggest...that there are a
host of practical or common sense issues that are for many consumers
(I merely suspect...can't prove) more relevant than which has greater
technical superiority.
For instance: cost of software, availability of titles, cost of
maintaining the system, planned obsolesence (which really pisses me
off), and etc.
After helping a friend move several thousand (say about 6K) records on
Saturday I was sort of wishing that he had been a cd addict. When I
am ready to move...I'll just have people pay me to haul off my
records...ha! And this does not mean that I am
anti-analogue...nope...but my back is still aching.
The whole notion of dead balls accuracy in playback is a joke at best
or (if we are being less forgiving) or worse is downright criminal.
And it is so for both analog and digital. No system that I have ever
heard comes remotely close...the colorations and deviations from a
"live" experience are so pervasive that I sense we must allow great
leeway in what any one person would care to adopt as their own studied
preference in playback equipment or software mediums. Everyone always
ignores the single most important factor in the enjoyment of
music...the "receiver" (the subjective auditor) whom brings with him
or her manifold prejuidices and differing listening and evaluative
skills and the like.
Now...go ahead shoot me my dear friend JWD...or just give me the pipe.
Mike
SDuraybito
In another thread <4u02n9$k...@agate.berkeley.edu>, Ben Cannon
<art...@a.crl.com> writes:
>I think I should add dynamic compression manifesting itself as
>"greater detail" as well as trasnparency! I forst noticed this on a
>very good FM station that I live next to, a song I recorded to DAT
>sounded BETTER than the CD I have (exact same edit, etc) It had to be
>compression.
I've been pondering this issue for a while. The attraction of SE
triodes may also fall under this category.
Question: Why?
First, is the dynamic range of the real world properly reproduced by
current audio playback technology? If not, what are the limiting
factors?
Second, is it possible that there exists an idealized dynamic range
that is most euphonic? It would seem this should be the same dynamic
range as the real world. Does loss of euphony occur linearly as
dynamic range decreases? Or are there some "magic" points at which
the ear/brain actually hears greater euphony?
Pete Goudreau
On 5 Aug 1996, mjg...@mailer.york.ac.uk (Matt Wenham) wrote,
> Todd Jenkins (tjen...@michiana.org) wrote:
>> If it is as you say, perhaps I should then comment, it's easy to hear
>> the nonlinearities in the vinyl playback system in phase, amplitude,
>> high-frequency droop, the requirements of heavy equalization too
>> provide the correct final amplitude ... ^^^^^^^^^^^^^^^^^^
> I'm quite sick of CD users using this as an argument against vinyl
> playback. Yes, there are many technical superiorities in CD playback,
> but this is a red herring. You do not expect to be able to plug a CD
> laser head straight into your pre-amp do you? You can't do it with a
> tape head, FM aerial, or any LP cartridge which is remotely 'hi-fi'.
Good point <g>...exactly what is the point, BTW...
A power amp offers significantly more processing than a preamp or
prepreamp, etc. just 'cause it's gotta provide tons'-'o-current into
nasty reactive loads...does that mean amps that exhibit technical
superiority (any way ya' slice it...) are inherently inferior, or do
you just not like the sound of 'em?
> The signal coming off a CD requires orders of magnitude more
> pre-processing to produce a line-level signal than that coming out of
> a MM or MC cartridge. The fact that turntable manufacturers have
You're kidding aren't you...
Let's see now, full scale at about 200uV with a dynamic range of 60dB
(being real, *real* generous here...) means that the smallest
resolvable (eletronically...let's not muddy the waters, shall we?)
signal is about 200nV meaning that the cartridge output at this point
is equal to the noise level. This means that the noise level out of
the cartridge is about equal to that of a room temperature 120 ohm
resistor. The circuitry necessary to bring this level up to line
level (about 1V rms) requires a gain of about 94dB (don't forget the
RIAA equalization). To keep the input noise from getting worse, the
first stage is going to have a noise level about equal to a 30 ohm
resistor and most of the available gain so that the signal to noise
ratio is only very slightly increased. This and the succeeding stages
must not produce distortions higher than this level or again the SINAD
will be degraded. This is not an easy task, in fact it is a complete
nightmare full of conflicting requirements and sonically degrading
compromises...try it sometime...
Now on the other hand, the output of a DAC IC is essentially already
at line level...unfortunately it is supplied in the presence of very
high levels of high frequency radiated and conducted noise emissions
that require incredibly high levels of cicuit design understanding to
simply get the signal out of the box without high levels of noise...a
completely different problem...and just as much of a nightmare...try
it sometime...
You were saying?
> almost unanimously chosen to leave the details of the circuitry down
> to amplifier designers is their loss. It is possible to produce a
No, it's simply mechanical guys leaving the electrical stuff to the
spark-Es...
> state of the art MM/MC head amp which will easily fit inside most
> turntables using only two or three op-amps per channel (external power
Yikes! Op-amps...arghhhh...see above.
> supply [best place for it] not withstanding). All that is then needed
> is a passive pre-amp (volume control) and power amplification before
> feeding to the loudspeakers. If ever CD playback gets this simple,
> I'll be very surprised.
<snip>
You might want to look at single box CD players ...it was in all the
papers...;-)
Cheers,
Pete
SDuraybito
[Detailed explanation of differences between digital and LP playbck
processing snipped]
>You have now reached a point where the signal is equivalent to what the
>DAC delivers; all processing after this is the same for either source,
and
>so isn't to be considered in evaluating differences in the systems.
>Frankly, I can't see the case for LP playback, if you REALLY analyze
>what's going on, as being either "simpler" OR somehow "more pure".
Bob, I think you're mixing things up a bit here. My view is that the
process of reading the CD pits up to the DAC is analogous to
stylus/groove contact to the first stage of RIAA equalization. Then
you can equate the DAC's work with RIAA stages.
It IS true that the DAC is a more efficient (fewer parts, etc) process
than RIAA equalization. HOWEVER, wouldn't you agree that tracking a
recording and converting it to a signal is more efficient than the
myriad processing functions required to get the admittedly more
accurate digital signal? Sure, you have to handicap LP playback due
to its greater errors (and lower error correction capability), but
when you objectively compare what's possible from LP playback and CD
playback, the LP process comes out the winner in terms of efficiency.
Whatever happened to the ideal of pursuing the most elegant solution
possible? Have we lost that art in the process of "being digital?"
Stewart Pinkerton
cadam...@aol.com (CAdams1471) writes:
>I have often wondered if it were not likely that "digital sound" was
>simply a lack of analog distortion. The interaction of the stylus with
>the record surface, for example.
>Folks who have been use to hearing LP playback for many years may have
>grown accustomed to the anomalies and simply find their absence from
>digital playback to be disturbing.
It may also be the case that many audiophiles have spent decades
building and tweaking their system around an LP source, balancing the
woolly bass and rolled-off treble with careful selection of amplifier
and speakers. Put an essentially ruler-flat CD source on the front of
such a system and of course the sound will be shrill and lacking in
bass. Rebalance the system with speakers having truly extended bass
and sweet treble and amplifiers sharing the same qualities and the
result will be a whole lot better. Ever notice how class A amplifiers,
silk-dome tweeters and really deep bass alignments are getting more
popular?
Steven Abrams
Richard D Pierce (DPi...@world.std.com) wrote:
> Well, yes, LP's DO REQUIRE RIAA deemphasis, in addition to the
> substantial gain needed. But, as I have pointed out here before, RIAA
> equalization is not hard to get right (having said that, there aree
> MANY existance proofs out there that it's not hard to get wrong,
> either). The RIAA equalization, being a minimum-phase function, is
> completely and 100% correctable: it's a matter of carefule
> implementation, that's all. You can get get it within 0.1 dB without a
> lot of effort, and with similarily vanishingly small phase errors as
> well. No big problem.
Unfortunately, getting RIAA deemphasis perfect may not have the effect
of correcting the emphasis put in during the recording process. How
many LPs are cut with perfect RIAA emphasis?!? Especially if you have
an old record collection.
~~~Steve
--
Steven Abrams abr...@cs.columbia.edu
Living is easy with eyes closed, misunderstanding all you see.
-Lennon/McCartney
Steven Abrams
In article <4u87a4$2...@eyrie.graphics.cornell.edu> sdura...@aol.com
(SDuraybito) writes:
> First, is the dynamic range of the real world properly reproduced by
> current audio playback technology? If not, what are the limiting
> factors?
I'll leave this to the experts to give the numbers. However, I'd
imagine that, the effective dynamic range of the average system under
average home listening conditions is NOT limited by the recording
media, if that media is a properly implemented 16-bit digital format.
The background noise and the setting of the volume knob on your system
probably conspire to give you a greatly reduced dynamic range. In my
office, for example, the hum of the computers and the A/C raise the
bacground noise level quite a bit.
> Second, is it possible that there exists an idealized dynamic range
> that is most euphonic? It would seem this should be the same dynamic
> range as the real world. Does loss of euphony occur linearly as
> dynamic range decreases? Or are there some "magic" points at which
> the ear/brain actually hears greater euphony?
In "average" listening environments, I wouldn't be surprised to learn
that people prefer dynamic compression. Some people even prefer
drastic compression, for example, in the car. I'm constantly turning
up the soft parts and turning down the loud parts on my car stereo.
Some CD players have a compression switch just for such occasions.
But none of this is probably germnaine to the conversation at hand.
More of issue is, I'd bet, what JJ has refered to on and off, here,
regarding how certain euphonic distortion gives the illusion of
*greater* dynamic range, hence the oft-spouted claim that LPs have
more dynamic range than their CD counterparts.
Getting back to the real question, why are you searching for a magic
number? Everyone is different and will have different preferences,
some for wider dynamic range, some for compressed range, and some for
the euphonic distortions that make it seem (to them) like greater
dynamic range.
Tim Takahashi
Bob Myers <my...@hpfcla.fc.hp.com> wrote:
>That's not the point, Matt. The fact is that LP playback REQUIRES
>significant equalization (more properly, compensation for some nasty
>things that have to be done to the signal during the process of making
>the LP) which is done in the analog domain.
I dont buy that argument. Analogue disk pre-emphasis curves have
varied widely over the years (beginning in 1925, with the first
Orthophonic 78s)... the RIAA standard was designed to reduce surface
noise. EMI 78's typically had NO pre-emphasis, btw.
As for being a problem in the analog domain, the RIAA curve consists
of a pair of staggered first-order low-pass filters. Not a very
complicated or contrived design.
>digital audio places no
>intentional screw-ups like this in the analog signal path.
I recall a "preemphasis" bit in the data stream. What possibly could
that be for?
-tim
Bob Myers
SDuraybito (sdura...@aol.com) wrote:
> Bob, I think you're mixing things up a bit here. My view is that the
> process of reading the CD pits up to the DAC is analogous to
> stylus/groove contact to the first stage of RIAA equalization. Then
> you can equate the DAC's work with RIAA stages.
> It IS true that the DAC is a more efficient (fewer parts, etc) process
> than RIAA equalization. HOWEVER, wouldn't you agree that tracking a
> recording and converting it to a signal is more efficient than the
> myriad processing functions required to get the admittedly more
> accurate digital signal? Sure, you have to handicap LP playback due
> to its greater errors (and lower error correction capability), but
> when you objectively compare what's possible from LP playback and CD
> playback, the LP process comes out the winner in terms of efficiency.
Ye gods, no. I'm not even sure what "efficiency" might mean in this
context, but the LP process is by any measure remarkably crude (esp.
the mechanically-based bits) and has MANY more potential sources of
error and distortion than the relatively simple digital process. The
digital process is non-intuitive and may appear to be on the level of
a "black box" to those not schooled in the art, but it IS actually
quite straightforward and simple in its fundamentals. There are
actually only TWO "processing functions" in digital which can impact
the final result, and which do not exist in the analog chain:
- Filter out all components above 22.05 kHz.
- Sample the resulting signal at 44.1 kHz and convert it to
digital data.
The subsequent steps in the digital domain - adding in the
error-correction bits, interleaving and encoding the data into its
final serial format for the CD - may appear complicated, but (a) do
not affect the actual data in terms of the resulting sound, and (b)
either work perfectly (at the encoding/recording end, or they don't
work at all. Digital circuits and math chips don't just "add a little
noise" to the signal, or cause a minor glitch in the phase response -
the data either gets through correctly, or it all goes to hell at
once.
> Whatever happened to the ideal of pursuing the most elegant solution
> possible? Have we lost that art in the process of "being digital?"
No; on the contrary, we have moved closer to that ideal. I consider
the digital medium "purer" (but I wish I could find a better word) in
the sense that it is simply a medium for the transfer of INFORMATION,
nothing more and nothing less, regardless of the "kind" of information
being conveyed. CDs carry audio information, but they don't really
have to be designed "knowing" that these bits represent sound; they
could just as easily be pictures, stock quotes, or the recipe for
Granny's apple pie. Neither the producer or the consumer is at the
mercy of the medium in this system, except to the extent that digital
audio is ruthless in exposing less-than-ideal practices at the front
end.
Mark Brindle
SDuraybito (sdura...@aol.com) wrote:
: It IS true that the DAC is a more efficient (fewer parts, etc) process
: than RIAA equalization. HOWEVER, wouldn't you agree that tracking a
: recording and converting it to a signal is more efficient than the
: myriad processing functions required to get the admittedly more
: accurate digital signal? Sure, you have to handicap LP playback due
: to its greater errors (and lower error correction capability), but
: when you objectively compare what's possible from LP playback and CD
: playback, the LP process comes out the winner in terms of efficiency.
:
: Whatever happened to the ideal of pursuing the most elegant solution
: possible? Have we lost that art in the process of "being digital?"
Kinda depends on your definitions of "elegant" and "efficient". On
one hand, an LP-system (or better, a wind-up mechanical Victrola) is
*conceptually* much simpler than any CD-player; so, in that sense, I
agree -- it's more "elegant".
OTOH, a state-of-the-art *implementation* of an LP playback system
requires fanatical attention to detail: ultra-precision bearings,
massivly heavy turntable platters, exquisitly fragile cartridges,
etc., etc. (And, the same is pretty much true of state-of-the-art
analog tape recorders.) Simple concepts -- which, unfortunately,
demand "virtually perfect" (i.e., EXPENSIVE) mechanical components.
CD players are the exact opposite -- conceptually complex (almost
beyond comprehension); but, they require *NO* twitchy mechanicals.
Methinks that should be worth a few points -- elegant and efficient to
*manufacture* -- even if the concept hurts your brain-nerve.
So, elegancewise, which is more efficienter and most betterest?
...who cares!
Mark
Bob Myers
SDuraybito (sdura...@aol.com) wrote:
> First, is the dynamic range of the real world properly reproduced by
> current audio playback technology? If not, what are the limiting
> factors?
> Second, is it possible that there exists an idealized dynamic range
> that is most euphonic? It would seem this should be the same dynamic
> range as the real world. Does loss of euphony occur linearly as
> dynamic range decreases? Or are there some "magic" points at which
> the ear/brain actually hears greater euphony?
I can't wait to see JJ's response to this, but in the meantime I'll
take a stab at a theory about dynamic range.
Again, I'm basing this on what *I* know of human perception, which in
my case means the limitations of visual perception rather than aural.
The human eye has an enormous "dynamic range" in terms of CAPABILITY;
it adapts to luminance levels over a range of a million to one or
better. But the *instanteous" range is much more limited - perhaps a
thousand to one - and it takes some time for the eye to become adapted
to higher or lower average light levels (as anyone who has gone from
bright sunlight into a dark room knows).
I would suspect that something similar may occur in our perception of
audio; the total dynamic range of real-world sound is certainly
enormous, well in excess of 100 dB. But can our ears adapt quickly to
perceive sounds over this entire range? If you are exposed to a
brief, extremely loud sound, will you be able to clearly hear a much
softer sound which IMMEDIATELY follows it? The above, plus personal
experience, would lead me to believe that you won't, and so
compressing the dynamic range of a recording could lead to the
perception of greater detail in that recording, even if such detail
were above the noise floor in the original "real world" situation.
But again, I would love to hear an expert opinion on this.
Stewart Pinkerton
sdura...@aol.com (SDuraybito) writes:
> It IS true that the DAC is a more efficient (fewer parts, etc)
> process than RIAA equalization. HOWEVER, wouldn't you agree that
> tracking a recording and converting it to a signal is more efficient
> than the myriad processing functions required to get the admittedly
> more accurate digital signal? Sure, you have to handicap LP
> playback due to its greater errors (and lower error correction
> capability), but when you objectively compare what's possible from
> LP playback and CD playback, the LP process comes out the winner in
> terms of efficiency.
To be efficient, you must first provide a baseline of performance and
then divide that by the work or cost necessary to achieve that
performance. When you OBJECTIVELY compare CD to LP, the performance
baseline is so much higher that 'efficiency' comparisons become very
difficult. Even if they had the same performance however, LP would
still lose out in any reasonable measure of efficiency. This is
especially true when you consider how close a $300 CD player such as
the Marantz CD-67 is to the best achievable from the medium when
compared with a $1,000 LP system, which must include the RIAA
headamp. Similarly, a state of the art CD player such as the Meridian
508-20 is around one fifth of the cost of a state of the art LP replay
system. The fact that 'loony tunes' transports and DACs are on sale at
ludicrous prices does not imply better performance in the way that say
the Forsell or SME decks and arms and the Lyra or van den Hul
cartridges are demonstrably superior to their less expensive
brethren. In terms of 'efficiency', it is worth noting that a top
phono cartridge is significantly more expensive than the Meridian
508-20 and is EXTREMELY fragile.
> Whatever happened to the ideal of pursuing the most elegant solution
> possible? Have we lost that art in the process of "being digital?"
As an engineer, I find the CD principle of music storage on an
archival material which may be PERFECTLY copied over many generations
(even domestically via CD-R) and which may be very nearly perfectly
reproduced by relatively low-cost and robust replay equipment, to be
an extremely elegant solution. While a top-quality LP replay system is
a beautiful piece of ultra-high precision engineering, I would hardly
call it an elegant device for reproducing a music signal and it is
without doubt extremely fragile (as is an LP) and not in the least
cost-effective.
Bill Poletti
In article <4uavf2$3...@eyrie.graphics.cornell.edu>,
pat...@popmail.dircon.co.uk (Stewart Pinkerton) says:
> Even if they had the same performance however, LP would still lose
> out in any reasonable measure of efficiency. This is especially true
> when you consider how close a $300 CD player such as the Marantz
> CD-67 is to the best achievable from the medium when compared with a
> $1,000 LP system, which must include the RIAA headamp. Similarly, a
> state of the art CD player such as the Meridian 508-20 is around one
> fifth of the cost of a state of the art LP replay system. The fact
> that 'loony tunes' transports and DACs are on sale at ludicrous
> prices does not imply better performance in the way that say the
> Forsell or SME decks and arms and the Lyra or van den Hul cartridges
> are demonstrably superior to their less expensive brethren. In terms
> of 'efficiency', it is worth noting that a top phono cartridge is
> significantly more expensive than the Meridian 508-20 and is
> EXTREMELY fragile.
[ quoted text deleted -- jwd ]
Gentlemen,
I'm not real sure I mentioned this, but I was not suggesting a
comparison between media at a given price point. However, there is
the point that a $300 CD player probably will (under most conditions)
outperform a similarly priced LP rig. Unless you have the possibility
of purchasing just the right used set-up. Then its iffy....
My comments are generally aimed at the much higher-end equipment.
If we compare systems at the $1,000 price point, many would argue
equality or LP superiority. Above that, it is my opinion, and the
opinion of many others, that CDs trail LP in the re-creation of music.
Yep, therer are a hellova lot of variables and table set-up is
considered by many a black art. But when it works....
Bill
Jeff Bernhard
CAdams1471 (cadam...@aol.com) wrote:
: In article <4u5sv3$m...@agate.berkeley.edu>, my...@hpfcla.fc.hp.com (Bob
: Myers) writes:
:
: >1. "Read the data" by dragging the stylus through a groove. In this
: >process, add the surface noise of the vinyl to the signal, plus add
: >distortions caused by flexing of the groove walls, turntable speed
: variations,
: >and so forth.
:
: I have often wondered if it were not likely that "digital sound" was
: simply a lack of analog distortion. The interaction of the stylus with
: the record surface, for example.
:
: Folks who have been use to hearing LP playback for many years may have
: grown accustomed to the anomalies and simply find their absence from
: digital playback to be disturbing.
I think there's obviously some truth here, but is it a simplification?
As improvements to turntables, tonearms and cartridges occur, the
sound they (re)produce has become less euphonic or distorted, and as
others have commented, the result is closer to the sound of good cd
playback. The anomalies to which we have become accustomed vary in
magnitude and in the best cases I wonder if these remnants are alone
sufficient to explain perceived differences. I am not arguing that
lp's will ever have as much dynamic range or don't have other
limitations, but rather that lp playback might, at its best, have
little of the problems that are often cited here to explain
preference. I'm speculating (excuse me). Perhaps even vanishingly
small amounts of these lp distortions (if they are ever that small)
are sufficient for the keen ears with which some of us claim to be
blessed (cursed).
--
Jeffrey Bernhard Concurrent Computer Corp.
Jeff.B...@mail.hcsc.com Voice: (954) 973-5496 Fax: (954) 977-5580
*** The opinions expressed herein are mine, not those of my employer! ***
David Fuller
I still have to think about the fellow at Masterdisk who spent his
life listening to laquer masters and making CD masters.
He was confused that while the digital master sounded "perfect"
compared to the analog, the analog playback was more engaging.
So he mixed the two worlds. He created a CD master with the pure
digital sound mixed with some vinyl whoosh (the sound you get on blank
grooves) and voila, a sound that was both perfect and engaging. Ended
up fooling a lot of audiophile friends with that one. I suspect if
you built a machine that injected typical vinyl sound artifacts into a
CD system you'd end up with something a large number of golden-ears
would rave about.
It just goes to show you - human perception is pretty versatile but it
also tunes itself in ways that can't be accurately calibrated or
measured.
--
Dave Fuller All opinions expressed are my own and not
Sequent Computer Systems those of Sequent Computer Systems, Inc.
dafu...@sequent.com
SDuraybito
In article <4u81oc$i...@agate.berkeley.edu>, cadam...@aol.com
(CAdams1471) writes:
> I have often wondered if it were not likely that "digital sound" was
> simply a lack of analog distortion. The interaction of the stylus with
> the record surface, for example.
>
> Folks who have been use to hearing LP playback for many years may have
> grown accustomed to the anomalies and simply find their absence from
> digital playback to be disturbing.
>
> To me, there is an elegance to digital playback due to its inherent
> simplicity as so well described my Bob. Far more simple than vinyl
> playback can ever be. (discounting acoustic 78 playback).
I've pondered this issue for over ten years now. I think there ARE
elements to LP playback that sound more euphonic due to various
distortion components, including not only surface noise but the
mechanical action of the sylus itself.
However, there is a synthetic, some even would call mechanical,
quality to CDs that is absent from LP. Now, if anything, this sort of
quality should be more prevalent on LP, yet it is the CD that seems to
suffer from this effect.
My question to CD aficionados: Is the best CD reproduction you've
heard entirely free of any sense of synthetic sameness to the sound?
Mark Brindle
Bill Poletti (bi...@i1.net) wrote:
: If we compare systems at the $1,000 price point, many would argue
: equality or LP superiority. Above that, it is my opinion, and the
: opinion of many others, that CDs trail LP in the re-creation of music.
Even if that were true, you've failed to mention the fact that, for a
given level of performance, CD-player prices are falling rapidly while
LP system prices are rising. Thus, the "crossover point" is quickly
moving in a direction favorable to CDs.
Also note that even if you have an LP system that cost more than my
house, your LPs are getting somewhat worse-for-wear with every play;
whereas, my CDs are getting better every day -- the bits stay the
same, but affordable CD-players are constantly improving.
The economics seem clear. Instead of paying $5000++ for today's
not-sneered-at-on-RAHE LP rig, anyone (except maybe Bill Gates) would
be wise to spend ~$500 for a mid-end CD-player and invest whatever's
left in (audibly) appreciating assets. Buy more CDs!
...but, keep a penny to put on the tone-arm,
Mark
Mark Brindle
SDuraybito (sdura...@aol.com) wrote:
: My question to CD aficionados: Is the best CD reproduction you've
: heard entirely free of any sense of synthetic sameness to the sound?
My answer to digital antagonists: Record a few of your favorite LPs to
DAT using *YOUR OWN* vinyl playback system -- TT, arm, cartridge, head
amp, interconnects, spikes, Mpingo Disks ...whatever. Then, do a
direct A/B (X optional) comparison of the LP and the DAT.
I won't predict whether *you* can hear a difference -- or which one
*you* prefer; but, I suspect that you might be surprised by how well a
digital recording can approximate that mystical "vinyl sound". (Extra
credit: If you have access to a CD-R, cut a CD from the DAT.)
Any other test of LP vs digital is an apples-to-bullfrogs comparison.
Unless the digital signal is exposed to the unique sonic colorations
(euphonic or otherwise) of *your* LP playback system, the two media
might just as well have been recorded from different masters.
Yep, I've run this test many times -- not as an academic experiment,
but simply to transfer my LP collection to a more convenient medium.
Personally, I can't hear ANY difference (except that the clicks/pops
no longer "move around" from play to play). OTOH, I much *prefer* the
DAT copy -- 'cause my turntable can't flip platters...
...and it doesn't have remote control,
Mark
Curtis Leeds
Bob Myers wrote:
...>the LP process is by any measure remarkably crude (esp.
> the mechanically-based bits) and has MANY more potential sources of
> error and distortion than the relatively simple digital process. The
> digital process is non-intuitive and may appear to be on the level of
> a "black box" to those not schooled in the art, but it IS actually
> quite straightforward and simple in its fundamentals.
This is the classic error of confusing simple with crude. Part of the
beauty (and success) of the LP process rests in its sheer
simplicity. It is much like the internal combustion automobile
engine. Perhaps the Wankel rotary is the more elegant design, but the
IC engine is one of the most well understood devices made by man; it's
design and manufacture is much more precise and elegant than may first
meet the eye. The same is true of the LP: it may appear so simple as
to seem quaint. But that hardly makes it a crude technology. Listening
to LPs on a VPI TNT with an SME V will only reveal the crudeness in
the recording itself. Oftentimes, there is no substitute for a mature
technology.
********************************************************
Curtis Leeds cle...@mail.idt.net
"A man hears what he wants to hear, and disregards
the rest."
********************************************************
Stewart Pinkerton
bi...@i1.net (Bill Poletti) writes:
>If we compare systems at the $1,000 price point, many would argue
>equality or LP superiority. Above that, it is my opinion, and the
>opinion of many others, that CDs trail LP in the re-creation of music.
And many wouldn't. Similarly, in 1996 it is the opinion of the vast
majority of listeners that LP has had its day and that CD has not only
caught up but has surpassed vinyl, and that's just at 44.1/16!
None of this polemic of course has any kind of weight behind it, it is
just posturing and opinion. Where is there any shred of evidence that
LP at ANY cost is now in ANY way superior to CD?
>Yep, therer are a hellova lot of variables and table set-up is
>considered by many a black art. But when it works....
It's nearly as good as a decent $500 CD player. That IS what you meant
to say, isn't it? :-)
SDuraybito
In article <4u8hs6$o...@agate.berkeley.edu>, abr...@cs.columbia.edu
(Steven Abrams) writes:
> Getting back to the real question, why are you searching for a magic
> number? Everyone is different and will have different preferences,
> some for wider dynamic range, some for compressed range, and some for
> the euphonic distortions that make it seem (to them) like greater
> dynamic range.
Well, I'm intrigued by the idea. It relates to Peter Walker's
observation many years ago that every recording has an ideal volume
for playback.
Evidence of some sort of curve that suggests an ideal compression
spectrum would go a long ways towards designing audio equipment that
serves listeners and not technicians. I suppose Fletcher-Munson is a
good place to start - perhaps the error is made that Fletcher-Munson
compensation must be an exactly inverted signal.
Mark Brindle
SDuraybito (sdura...@aol.com) wrote:
: Evidence of some sort of curve that suggests an ideal compression
: spectrum would go a long ways towards designing audio equipment that
: serves listeners and not technicians.
Is it truly possible that you don't recognize the absurdity of this
statement? You're asking for an "ideal compression spectrum" -- which
of course, presupposes an "ideal listener preference" -- apparently,
to be SHARED by ALL "listeners"...
...but to be DESIGNED by "technicians"? Right?
Mark
Bob Myers
Curtis Leeds (cle...@mail.idt.net) wrote:
> Bob Myers wrote:
>> the LP process is by any measure remarkably crude (esp. the
>> mechanically-based bits) and has MANY more potential sources of
>> error and distortion than the relatively simple digital process.
>> The digital process is non-intuitive and may appear to be on the
>> level of a "black box" to those not schooled in the art, but it IS
>> actually quite straightforward and simple in its fundamentals.
> This is the classic error of confusing simple with crude. Part of
> the beauty (and success) of the LP process rests in its sheer
> simplicity. It is much like the internal combustion automobile
No, I am not "confusing simple with crude"; the overall LP process IS
crude, and it is FAR from simple, esp. in comparison with a digital
system where the sole job of the media is to deliver the bits to the
consumer. Cutting grooves in a disc in an attempt to approximate the
waveforms of sound IS crude, and the equipment needed to do this (and
to play back the result) is hardly simple. I agree that the LP system
is CONCEPTUALLY simple, but that's not a good measure of the
complexity of the actual process.
Have you ever seen what it takes to put those little grooves into your
beloved pieces of black vinyl? Simple, my eye....
Bob Myers | "One man's theology is another man's belly laugh."
my...@fc.hp.com | - Lazarus Long/Robert A. Heinlein
|
Stewart Pinkerton
Curtis Leeds <cle...@mail.idt.net> writes:
>Bob Myers wrote:
>> the LP process is by any measure remarkably crude (esp. the
>> mechanically-based bits) and has MANY more potential sources of
>> error and distortion than the relatively simple digital process.
>> The digital process is non-intuitive and may appear to be on the
>> level of a "black box" to those not schooled in the art, but it IS
>> actually quite straightforward and simple in its fundamentals.
>This is the classic error of confusing simple with crude. Part of the
>beauty (and success) of the LP process rests in its sheer
>simplicity. It is much like the internal combustion automobile
>engine. Perhaps the Wankel rotary is the more elegant design, but the
>IC engine is one of the most well understood devices made by man; it's
>design and manufacture is much more precise and elegant than may first
>meet the eye.
The elegant and simple designs are the two-stroke reciprocating engine
and the gas turbine. Incidentally, you make the classic error of
confusing specific with generic. All loudspeakers are not 'Tannoys' and
all internal combustion engines are not of the reciprocating piston
type. The industry standard 4-stroke reciprocating piston engine with 4
poppet valves per cylinder and twin overhead camshafts is fundamentally
crude but is most certainly neither simple nor elegant.
>The same is true of the LP: it may appear so simple as to seem
>quaint. But that hardly makes it a crude technology. Listening to LPs
>on a VPI TNT with an SME V will only reveal the crudeness in the
>recording itself. Oftentimes, there is no substitute for a mature
>technology.
Glad you mention this, now that the horse-drawn LP waggon has been
overtaken by CD, which is in its second decade and maturing rapidly.
For proof of maturity, take three CD players, say the Marantz CD-63 KI
Signature, the Sony XA3ES and the Denon DCD-1015. They use entirely
different transport mechanisms and DAC technology but sound remarkably
similar, VASTLY more so than would happen if you chose three LP
deck/arm/cart combinations. This convergence in performance is a sure
sign of any mature technology - LP can't do it because the process is
fundamentally too crude to be accurately reproducible!
LP is like the dancing bear, the real amazement is not that it dances
well, but that it can dance at all.
>********************************************************
> Curtis Leeds cle...@mail.idt.net
> "A man hears what he wants to hear, and disregards
>the rest."
>********************************************************
Your choice of signature quote is a REAL giveaway...............
[ And unfortunately, you're both well down the path of an argument over
personal preference that will never be settled, either here or in
the great beyond. Maybe it's time to move on to other fish. -- jwd ]
Matt Wenham
Bob Myers (my...@hpfcla.fc.hp.com) wrote:
> Ye gods, no. I'm not even sure what "efficiency" might mean in this
> context, but the LP process is by any measure remarkably crude ...
> There are actually only TWO "processing functions" in digital which can
> impact the final result, and which do not exist in the analog chain:
> - Filter out all components above 22.05 kHz.
> - Sample the resulting signal at 44.1 kHz and convert it to
> digital data.
Which if done 100% correctly, should have *no effect* on the signal below
22.05 kHz.
> The subsequent steps in the digital domain - adding in the
> error-correction bits, interleaving and encoding the data into its
> final serial format for the CD - may appear complicated, but (a) do
> not affect the actual data in terms of the resulting sound ...
But the circuitry needed to reverse such processing can affect the
overall sound by interacting with the DAC and analogue circuitry. This
can occur both through signal and power supply lines.
> and (b) either work perfectly (at the encoding/recording end, or
> they don't work at all. Digital circuits and math chips don't just
> "add a little noise" to the signal, or cause a minor glitch in the
> phase response - the data either gets through correctly, or it all
> goes to hell at once.
Then why don't all CD players sound the same, as Quad used to claim
for well engineered amplifiers? Is it all the fault of the analogue
circuitry? If so, why do different CD transports 'sound' different
through the same DAC? Digital circuits most certainly can just "add a
little jitter" (or even a lot) to the digital signal, which can have a
marked and measurable effect on the final analogue output.
Admittedly it's not the fault of the preceeding digital circuitry if
the DAC isn't designed to handle jitter (or indeed other forms of
noise in the digital domain) but as far as I am aware, it is
exceedingly difficult to build a DAC which is totally immune to
degradation of the signal fed to it. Therefore in terms of the
reproduction of the original signal, digital circuitry cannot be said
to either work or not work, there are shades of grey. Although two
digital signals might contain exactly the same data stream in terms of
1s and 0s, the DAC may produce different distortions of the original
signal when reproducing the two data streams.
Matt Wenham...
Bob Myers
jj, curmudgeon and all-around grouch (j...@research.att.com) wrote:
> Now, what range is euphonic? I THINK that's a mis-question, the answer
> is "the same as the original" to sound LIKE the original, but too MUCH
> a dynamic range can be disturbing, original or not.
I think this is the question I was heading for. "Sounding like the
original" doesn't seem to be the issue here - Mr. D. was pretty
specific in his original observation (unless I misread it, in which
case my apologies) that compression appeared to result in a sense of
greater DETAIL. My hypothesis is that this is due to something
similar to what I KNOW goes on visually - a reduction in contrast
ratio (which is the visual "dynamic range" of an image) can enhance
the perception of detail. In simple terms, you'll pick out details in
the dark part of an image better if the bright portions aren't causing
the eye to adapt its limited instanteous range upwards. My guess was
that a similar thing happens in sound, even when the sounds are not
simultaneous in time. If the range is too large, the ear would adapt
to a loud passage, and miss details in a quieter passage which
immediately follows. Compressing the range to match the
near-instantaneous range of hearing would then give the impression of
more detail. My problem is, I have no idea whether or not the ears
behave in this manner - I just know that the eyes do. (There's an
additional question of how quickly the adaptation, if it does occur,
happens.)
Pete Goudreau
On 8 Aug 1996, sdura...@aol.com (SDuraybito) wrote,
[quote deletion - wsr]
> However, there is a synthetic, some even would call mechanical,
> quality to CDs that is absent from LP. Now, if anything, this sort of
> quality should be more prevalent on LP, yet it is the CD that seems to
> suffer from this effect.
Yes there is a synthetic quality to nearly all CD playback *systems*
just as there is often an annoying funhouse mirror effect audible in
nearly all LP playback *systems*. Is it the medium or the system?
That's the real question and to my mind, and apparently many others'
as well, it hasn't been adequately answered.
I don't see how it is possible to state that this quality should be
present on LP more than on CD as the methods of storing and
reconstructing the original signal couldn't be more different.
> My question to CD aficionados: Is the best CD reproduction you've
> heard entirely free of any sense of synthetic sameness to the sound?
No, not even remotely. Not being an aficionado of CDs maybe this
answer isn't what you are looking for but there it is. The *method*
by which CDs are made and played has many technological pitfalls and
bloody few pieces of equipment exist that allow its theoretical
limitations to be met. However, IMHO, it is possible to get awfully
close to the original signal with a properly designed digital playback
system, although it unfortunately also allows for the accurate
reproduction of the producer's mucking about in the studio, the
soundman's upper mid hearing notches, cheap mics and processors, etc.
Cheers,
Pete
jj, curmudgeon and all-around grouch
In article <4ub07n$3...@eyrie.graphics.cornell.edu>
my...@hpfcla.fc.hp.com (Bob Myers) writes:
>SDuraybito (sdura...@aol.com) wrote:
>> Second, is it possible that there exists an idealized dynamic range
>> that is most euphonic? It would seem this should be the same dynamic
>> range as the real world. Does loss of euphony occur linearly as
>> dynamic range decreases? Or are there some "magic" points at which
>> the ear/brain actually hears greater euphony?
>I would suspect that something similar may occur in our perception of
>audio; the total dynamic range of real-world sound is certainly
>enormous, well in excess of 100 dB. But can our ears adapt quickly to
>perceive sounds over this entire range?
A well put question, Bob. First some discussion on masking ability vs.
absolute thresholds.
The ear can handle, without damage, about a 100dB dynamic range of
softest detectable to loudest w/o damage. It can go beyond the upper
limit while (not necessarly instantly) taking damage.
BUT, the most dynamic range the ear can handle between two different
frequencies is on the order of 70dB, best case, because of the
mechanics of the ear.
The best local (in frequency) SNR that the ear can handle is about
30dB for simultaneous masking, with up to 40dB under some very
specific conditions.
Hmm. The new McGraw-Hill handbook of digital stuff will have a chapter
dedicated to this in it, but it's not out yet.
Now, what range is euphonic? I THINK that's a mis-question, the answer
is "the same as the original" to sound LIKE the original, but too MUCH
a dynamic range can be disturbing, original or not.
--
Copyright alice!jj 1996, 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.
Tim Takahashi
Bob Myers <my...@hpfcla.fc.hp.com> wrote:
>jj, curmudgeon and all-around grouch (j...@research.att.com) wrote:
>> Now, what range is euphonic? I THINK that's a mis-question, the answer
>> is "the same as the original"
>I think this is the question I was heading for. "Sounding like the
>original" doesn't seem to be the issue here
But what original are we talking about? Streisand does not sound as
she appears on either CD, LP, Casette or 8-track. Or, is what we hear
what she's "supposed" to sound like?
Realistically... dynamic range for the sake of dynamic range is
exaggeration, a distortion. I doubt that I've experienced more than
60db dynamic range even at a live (unamplified) orchestral concert.
Special close-microphone techniques designed to exploit 90+ db dynamic
range are a distortion of the concert going experience.
>My hypothesis is that this is due to something
>similar to what I KNOW goes on visually - a reduction in contrast
>ratio (which is the visual "dynamic range" of an image) can enhance
>the perception of detail.
Better yet, why a black-and-white silver-gelatin print has a different
quality from a color chromagenic print, or a projected slide or an
image on a CRT.
In my perception, a silver-gelatin print has a artfully pleasant blend
of dynamic range (about 128:1) and resolution (5 l/mm if you're really
careful). It does lack information on color - but despite this, I
consider it the superior medium to record reality. Color prints, IMHO,
have a miserable dynamic range and disappointing color/contrast
interactions.
As a point of reference, my boss and I have a running disagreement
over whether it is better to scan prints or negatives for digital
editing (while I prefer to avoid digititis in my photography, I scan
negatives - he scans prints)
Back to audio....
-tim
Bob Olhsson
In article <4u8a9h$l...@agate.berkeley.edu>, abr...@cs.columbia.edu
(Steven Abrams) wrote:
>How
>many LPs are cut with perfect RIAA emphasis?!?
They're pretty close.
Especially if you have
>an old record collection.
The RIAA curve is actually RCA's New Orthophonic preemphasis that was
introduced as part of their 45 RPM single format in the early '50s.
Most cutting systems are alligned using a combination of test records
and the actual measurement of the relationship between frequencies as
they reflect a light pattern. Very little has changed. There have only
been a few commercial systems and a number of people have built their
own.
Typically cutting systems are trimmed to within 1 dB. The response
actually varies with the high-end declining as you move towards the
center so there is little point in becoming anal about RIAA
accuracy. Matching the cutting engineer's playback system is probably
the most accurate approach. Of course many people have the same
general complaints about the really flat cartridges such as the Shure
V-15s and Stanton 881s as they do about digital playback.
--
Bob Olhsson Audio | O tongue, thou art a treasure without end.
Box 555,Novato CA 94948 | And, O tongue, thou art also a disease
415.457.2620 | without remedy. == Jelal'uddin Rumi ==
415.456.1496 FAX |
SDuraybito
In article <4ufan0$i...@biosun.harvard.edu>, o...@hyperback.com (Bob
Olhsson) writes:
> Typically cutting systems are trimmed to within 1 dB. The response
> actually varies with the high-end declining as you move towards the
> center so there is little point in becoming anal about RIAA
> accuracy. Matching the cutting engineer's playback system is
> probably the most accurate approach. Of course many people have the
> same general complaints about the really flat cartridges such as the
> Shure V-15s and Stanton 881s as they do about digital playback.
Interesting assertion. I use a Shure Ultra 500 because I find that it
does provide a sense of very flat frequency response. But at the same
time, it seems to recreate leading transients quicker than CDs and
overall sounds far more "real" than CDs. There's none of the
synthetic or mechanical quality that I've come to hate from CDs.
The problems lie elsewhere...
SDuraybito
In article <4udaeu$k...@agate.berkeley.edu>, dafu...@sequent.com
(David Fuller) writes:
> So he mixed the two worlds. He created a CD master with the pure
> digital sound mixed with some vinyl whoosh (the sound you get on
> blank grooves) and voila, a sound that was both perfect and
> engaging. Ended up fooling a lot of audiophile friends with that
> one. I suspect if you built a machine that injected typical vinyl
> sound artifacts into a CD system you'd end up with something a large
> number of golden-ears would rave about.
Personally, I'd love to try any machine that would make CDs as
enjoyable as LPs. I've been waiting for fifteen years now. Is there
any hope for me?
[ Obviously a rhetorical question ... :-) jwd ]
Richard D Pierce
In article <4ug5j3$1...@eyrie.graphics.cornell.edu>,
Matt Wenham <mjg...@mailer.york.ac.uk> wrote:
>Bob Myers (my...@hpfcla.fc.hp.com) wrote:
>> The subsequent steps in the digital domain - adding in the
>> error-correction bits, interleaving and encoding the data into its
>> final serial format for the CD - may appear complicated, but (a) do
>> not affect the actual data in terms of the resulting sound ...
>But the circuitry needed to reverse such processing can affect the
>overall sound by interacting with the DAC and analogue circuitry. This
>can occur both through signal and power supply lines.
First, there's no evidence that changes in the number of corrected
errors has any effect on power draw. If that's the assertion, it's a
trivial job to measure the changes. Where's the data?
But assuming it DOES have such an effect on the power supply, if that
effect causes problems in the DAC an analogue circuitry, it's a sure
mark of rather gross incompetence or gross negligence on the part of
the person who designed it.
While it is relatively simple to design analog circuitry with high
power supply rejection capability, AND it's also relatively easy to
design stable, well isoltaed supplies that DO NOT suffer from these
problems, that's absolutely no assurance that there is not a
collection of ifots out there who do just that.
However, because there ARE such idots is no excuse to whitewash the
entire principle.
>> and (b) either work perfectly (at the encoding/recording end, or
>> they don't work at all. Digital circuits and math chips don't just
>> "add a little noise" to the signal, or cause a minor glitch in the
>> phase response - the data either gets through correctly, or it all
>> goes to hell at once.
>Then why don't all CD players sound the same, as Quad used to claim
>for well engineered amplifiers? Is it all the fault of the analogue
>circuitry?
If they don't sound the same, that's certainly one VERY good place to
point the finger.
>If so, why do different CD transports 'sound' different
>through the same DAC?
Because the guy who designed the DAC is an idiot, that's why.
>Digital circuits most certainly can just "add a
>little jitter" (or even a lot) to the digital signal, which can have a
>marked and measurable effect on the final analogue output.
If and only if the output timing on the DAC is closely correlated to
the incoming timing errors. And if that is the case, it's a sign of
bad design and implementation.
>Admittedly it's not the fault of the preceeding digital circuitry if
>the DAC isn't designed to handle jitter (or indeed other forms of
>noise in the digital domain) but as far as I am aware, it is
>exceedingly difficult to build a DAC which is totally immune to
>degradation of the signal fed to it.
No, it is NOT exceedingly difficult. It is, in fact, relatively
straightforward. It simply requires that you DO NOT assume that the
incoming bitstream timing is essential to determining outgoing D/A
clocking. The incoming digital stream has ALL the information about
the outgoing clock rate you need sitting right there in bits 6-7 of
the first byte of the channel status word (or in bits 24-27 of in
consumer mode). It WILL tell you, straight away, unambiguously that
you had better clock at 32 kHz, 44.1 kHz or 48 kHz.
>Therefore in terms of the
>reproduction of the original signal, digital circuitry cannot be said
>to either work or not work, there are shades of grey. Although two
>digital signals might contain exactly the same data stream in terms of
>1s and 0s, the DAC may produce different distortions of the original
>signal when reproducing the two data streams.
Only if the implementation is bad or broken.
Once again, everyone, THERE IS NO GREY AREA IN THE TIMING. That is,
unless, you choose to ignore what's being handed to you and
incorrectly use the incoming rate to control your clock. And even if
you do it that that way, you can still, without an extraodinary amount
of difficulty, do a really bang up job of doing it well.
But that STILL does not prevent a LOT of manufacturers from doing it
EXTREMELY badly. And, as a service to the broadest possible spectrum
of consumer budgets, they are known to fuck it up royally at all price
points.
--
| Dick Pierce |
| Loudspeaker and Software Consulting |
| 17 Sartelle Street Pepperell, MA 01463 |
| (508) 433-9183 (Voice and FAX) |
Bob Myers
Tim Takahashi (t...@me.rochester.edu) wrote:
> Realistically... dynamic range for the sake of dynamic range is
> exaggeration, a distortion. I doubt that I've experienced more than
> 60db dynamic range even at a live (unamplified) orchestral concert.
> Special close-microphone techniques designed to exploit 90+ db
> dynamic range are a distortion of the concert going experience.
In many cases, I'd have to agree; still, this may not be relevant to
the question of whether not intentionally compressing the dynamic
range could contribute to a heightened PERCEPTION of detail (at the
expense, of course, of accuracy in the absolute sense). I'm
completely in the dark as to how the ear responds to extreme changes
in sound level over the short term, even when "extreme" is taken to
mean something far less than 90+ dB.
>> My hypothesis is that this is due to something similar to what I
>> KNOW goes on visually - a reduction in contrast ratio (which is the
>> visual "dynamic range" of an image) can enhance the perception of
>> detail.
> Better yet, why a black-and-white silver-gelatin print has a different
> quality from a color chromagenic print, or a projected slide or an
> image on a CRT.
Well, actually, we understand these differences pretty well. I'm not
sure that this is the appropriate forum to get into this, though.
> As a point of reference, my boss and I have a running disagreement
> over whether it is better to scan prints or negatives for digital
> editing (while I prefer to avoid digititis in my photography, I scan
> negatives - he scans prints)
Personal preference, adding in heaping quantities of dependency on the
scanner, display, and color models and color correction in use.
Bob Myers | my...@fc.hp.com
Senior Engineer, Displays | Note: The opinions presented
Workstation Systems Division | here are not those of my employer
Hewlett-Packard Co., Ft. Collins, CO | or of any rational person.
Stewart Pinkerton
mjg...@mailer.york.ac.uk (Matt Wenham) writes:
>Then why don't all CD players sound the same, as Quad used to claim
>for well engineered amplifiers? Is it all the fault of the analogue
>circuitry?
Having eliminated the real cheap rubbish (as Quad would), don't they?
Are you SURE they don't, to an insignificantly tiny degree?
>If so, why do different CD transports 'sound' different
>through the same DAC?
Do they? How do you KNOW they do? Not, I hope, because you've been
reading reviews by journalists whose jobs would disappear if they found
no differences between a Teac T-1 and a Mark Levinson 31..........
>Digital circuits most certainly can just "add a
>little jitter" (or even a lot) to the digital signal, which can have a
>marked and measurable effect on the final analogue output.
True, but that should have NOTHING to do with the transport.
>Admittedly it's not the fault of the preceeding digital circuitry if
>the DAC isn't designed to handle jitter (or indeed other forms of
>noise in the digital domain) but as far as I am aware, it is
>exceedingly difficult to build a DAC which is totally immune to
>degradation of the signal fed to it.
Not really, reclocking (that's REAL reclocking, not messing about with
VCXOs) from FIFO buffers has been around for donkeys years in most
data transmission fields, indeed it's an ESSENTIAL feature of all CD
transport electronics, to vary the rotational speed of the disc
thereby maintaining constant groove velocity and hence constant
average data rate. The data are clocked out of the buffer at an exact
44.1 kHz rate and the disc speed is adjusted to keep the buffer half
full. It's after this point that everything is allowed to go nasty in
many systems.
In the absence of the common usage of true reclocking for DACs, it is
basically PATHETIC that there is no industry standard for master clock
linkage from DAC to transport in ALL outboard DACs. It's been fifteen
years, for heaven's sake! Some makers such as Linn and Arcam do it but
in a proprietary way, although of course there is the 'professional'
SDIF-2 standard which uses three data lines for left channel, right
channel and word clock. As far as I am aware, this is not used by ANY
domestic equipment.
Pete Goudreau
On 9 Aug 1996, mjg...@mailer.york.ac.uk (Matt Wenham) wrote,
> Bob Myers (my...@hpfcla.fc.hp.com) wrote:
>> The subsequent steps in the digital domain - adding in the
>> error-correction bits, interleaving and encoding the data into its
>> final serial format for the CD - may appear complicated, but (a) do
>> not affect the actual data in terms of the resulting sound ...
> But the circuitry needed to reverse such processing can affect the
> overall sound by interacting with the DAC and analogue circuitry. This
> can occur both through signal and power supply lines.
Only through poor engineering of the product. If the digital section
interferes with the analog sections' operation then the designer
hasn't an earthly clue of standard engineering practices in this realm
of design. Case in point is the board layout of the Assemblage
DAC-1...we got quite a laugh over this one here as there are so many
profoundly fundamental errors present in this design that it is
amazing that it even works...it's amazing what copper foil and
sub-mini coax can do to fix these things...;-) Analog circuitry's
noise susceptibility is a very common problem in other industries and
the techniques for their solution are well known...they're even in
(shudder, gasp...) books. These may be found in your local university
engineering library or technical bookstores, if a designer were to be
so inclined...
>> and (b) either work perfectly (at the encoding/recording end, or
>> they don't work at all. Digital circuits and math chips don't just
>> "add a little noise" to the signal, or cause a minor glitch in the
>> phase response - the data either gets through correctly, or it all
>> goes to hell at once.
> Then why don't all CD players sound the same, as Quad used to claim
> for well engineered amplifiers? Is it all the fault of the analogue
> circuitry? If so, why do different CD transports 'sound' different
> through the same DAC? Digital circuits most certainly can just "add a
> little jitter" (or even a lot) to the digital signal, which can have a
> marked and measurable effect on the final analogue output.
Once again...appallingly poor engineering (or lack thereof) of the
circuitry involved in different players.
No, it isn't the fault of the analog circuitry, it is the fault of the
designer who laid down the schematic and the board layout.
Again, stupid designers who cannot make something as simple as an
outboard DAC independent of the well know and characterised errors
present in the inbound data stream.
Yes, they do, but that is a completely irrelevant statement...the
digital circuitry induced additive jitter (correlated or not) can be
removed by a number of methods downstream at the DAC ICs where jitter
actually has any meaning at all (beyond bit and frame loss, of
course). Any measureable difference in the output signal can be laid
at the feet of incorrect or substandard implementation of the digital
to analog conversion circuitry, ignoring differences due to purely
analog design issues, of course.
There are however different digital circuits that are aimed at
achieving the same result and they do indeed sound different but this
is not because they create differing amounts of additive jitter, it is
because they perform the same function slightly different e.g., higher
accumulator word width and more FIR taps yield a higher stopband
attenuation in the oversampling output as in the difference between
the NPC '5813 and the '5824.
> Admittedly it's not the fault of the preceeding digital circuitry if
> the DAC isn't designed to handle jitter (or indeed other forms of
> noise in the digital domain) but as far as I am aware, it is
> exceedingly difficult to build a DAC which is totally immune to
> degradation of the signal fed to it. Therefore in terms of the
> reproduction of the original signal, digital circuitry cannot be said
> to either work or not work, there are shades of grey. Although two
> digital signals might contain exactly the same data stream in terms of
> 1s and 0s, the DAC may produce different distortions of the original
> signal when reproducing the two data streams.
That's true, the preceeding digital circuitry should have nothing at
all to do with the reconstruction of the analog version of the signal.
It's been done for decades in other industries, shame those designing
high-end equipment haven't actually worked at a real job doing
this...explains a lot.
There are no shades of grey, given a particular digital implementation
of the requisite signal processing, the responsibility of generating
the analog signal lies solely at the feet of the D/A circuit and its
ability to suppress jitter induced distortion. Analog domain errors
are another story altogether...
If the two DACs use the same processing algorithms and computational
resolution then the differences will be due to the inability to
suppress jitter at the DAC ICs, assuming of course that the two units
have the same analog stages. If both units fully suppress jitter then
all that is left is the differences in analog stage performance which
can be quite a bit given the state of the art of design...
All in all, the DAC cannot produce distortions from the digital data,
given no errors, if jitter is suppressed at the conversion clock. Any
distortions produced by the DAC in the absence of jitter are solely
analog errors and as such are mutually exclusive of the digital plant.
Cheers,
Pete
Tim Takahashi
Bob Myers <my...@hpfcla.fc.hp.com> wrote:
>Tim Takahashi (t...@me.rochester.edu) wrote:
>> Realistically... dynamic range for the sake of dynamic range is
>> exaggeration, a distortion.
> In many cases, I'd have to agree; still, this may not be relevant to
> the question of whether not intentionally compressing the dynamic
> range could contribute to a heightened PERCEPTION of detail (at the
> expense, of course, of accuracy in the absolute sense).
I think that SD's grasping for straws has lead us to debate a moot
point. Kind-of-like talking about "micro-harmonics."
I've seen no evidence of intentional (or unintentional) dynamic range
compression in any of my tube gear, when used in typical operation. It
does not explain the classic "tube" sound....
Now, as I've written before - the problem is more with the "digital"
remastering than with the digital media - many remastering jobs seem
to have been swayed by some perverse need to improve "dynamic range."
Noise gating, companding, NONOISE artifacts are noticable - and seem
to be common techniques in the remastering of older source material.
I give as example the RCA Caruso series "Stockham/Soundstream"
remasterings, or the NONOISED Fletecher Henderson series. The
resulting sound is totally unlike a acoustical or early electrical 78,
not "musical" in my book, and has a peculiar dynamic range (surface
noise removed, but lacking "dynamics")
Actually, as a demonstration disk, nothing comes close to the 1909
Verdi : Celeste Aida of Caruso. On my wind up Victrola, the dynamic
range is unbelievable..... it must peak at 110, 120db?
(ear-shattering loud)
-tim
Robert Orban
In article <4u02n9$k...@agate.berkeley.edu>, art...@a.crl.com says...
>I think I should add dynamic compression manifesting itself as
>"greater detail" as well as trasnparency! I forst noticed this on a
>very good FM station that I live next to, a song I recorded to DAT
>sounded BETTER than the CD I have (exact same edit, etc) It had to be
>compression
I design broadcast compressors, and I think your conjecture is sound
if you understand that most all FM stations use multiband compression
these days. A 'competitive' pop music station will typically compress
in four to six frequency bands. This means that these compressors are
always performing an 'automatic re-equalization' operation on their
inputs. If the inputs don't have as much high frequency spectral
content as the audio processor expects, the highs will be dynamically
boosted. This can give the impression of 'greater detail'.
Matt Wenham
Richard D Pierce (DPi...@world.std.com) wrote:
> >... as far as I am aware, it is
> >exceedingly difficult to build a DAC which is totally immune to
> >degradation of the signal fed to it.
> No, it is NOT exceedingly difficult.
I stand corrected.
> The incoming digital stream has ALL the information about
> the outgoing clock rate you need ...
So, let's say we build a DAC unit which takes the output from a
digital source, FIFO buffers it with a nice pile of RAM, and then
clocks the data out to the DAC according to the relevant bits of the
incoming data stream. This unit should be free from all effects of
jitter etc. from the digital source. If it's this simple, why on earth
do DAC builders still use the incoming data stream to control the
clock? It appears that this is a much simpler and elegant method than
trying to 'lock on' to what is potentially a very broken signal
entering the DAC unit.
Yes, I have changed my tune before anyone asks,
Matt Wenham.
jj, curmudgeon and all-around grouch
In article <4ug5lk$1...@eyrie.graphics.cornell.edu>
my...@hpfcla.fc.hp.com (Bob Myers) writes:
>Compressing the range to match the near-instantaneous range of
>hearing would then give the impression of more detail. My problem
>is, I have no idea whether or not the ears behave in this manner - I
>just know that the eyes do. (There's an additional question of how
>quickly the adaptation, if it does occur, happens.)
To some extent they do, but since such gain riding also introduces
distortion, the amount you can include varies with listener.
A better way is sometimes to use multi-band compression, of course,
like most recording labs. :-)
John Shriver
I'm not sure that the sample rate subcode bits are correctly coded on
all CD's, or come out correctly on all SPDIF outputs. (I haven't
checked on this at all, but my recollection is that the subcode bits
were very sloppy in the early years. There is documentation in this
issue...)
Of course, this would only require the DAC to correctly default to
44.1 when the subcode bits are bogus. Not a toughie. (There's maybe
0.01% of home users who have anything running at 32 or 48 -- a DAT
deck.)
Also, there are times when you can't use an independent clock.
Consider the Meridian powered loudspeakers with Digital input. If
they used independent clocks and buffer memories, how would the two
channels stay in sync? You would have to make one a clock master, and
put a clock cable between them. They would also have to be very
consistent on startup conditions on the FIFO RAM.
Also, a DAC with a big FIFO could be rather distressing to a DJ. It's
going to have to have some delay between first bit in and first sample
out, to let the FIFO fill halfway.
(Does anyone have the Red Book or IEC clock tolerance specs for SPDIF?
That's what we need to compute how many seconds of buffer are needed.)
Richard D Pierce
In article <4un7ks$e...@biosun.harvard.edu>,
Matt Wenham <mjg...@mailer.york.ac.uk> wrote:
>If it's this simple, why on earth do DAC builders still use the incoming
>data stream to control the clock?
Got me on that one, Batman. I certainly don't pretend to speak for those
people. Why not ask them?
>It appears that this is a much simpler and elegant method than
>trying to 'lock on' to what is potentially a very broken signal
>entering the DAC unit.
Ah, but it's not a broken signal. That's the point. As long as the bit
data is intact, it, by the very definition of the protocol, is not
broken.
Unless, of course, you choose to ignore the very definition of the
protocol.
>Yes, I have changed my tune before anyone asks,
Ah, but is your tune the C=440 that Seigfried has pontificated as the most
musical of all pitches? :-)
Matt Wenham
On Mon, 12 Aug 1996, John Shriver wrote:
> I'm not sure that the sample rate subcode bits are correctly coded on
> all CD's, or come out correctly on all SPDIF outputs ...
> Of course, this would only require the DAC to correctly default to
> 44.1 when the subcode bits are bogus. Not a toughie. (There's maybe
> 0.01% of home users who have anything running at 32 or 48 -- a DAT
> deck.)
I'm sure it would be perfectly possible to establish the rate of the
incoming signal as it went into the buffer, and then switch the
internal clock accordingly.
> Also, there are times when you can't use an independent clock.
> Consider the Meridian powered loudspeakers with Digital input ...
Yes, I can't think of any consistent way of doing it without a clock
line, unless the designer is happy that the FIFOed signal passing to
the slave speaker is going to be clean enough to drive the DAC
directly after passing down a few feet of cable.
> Also, a DAC with a big FIFO could be rather distressing to a DJ. It's
> going to have to have some delay between first bit in and first sample
> out, to let the FIFO fill halfway.
Hmmm... this is rec.audio.high-end though isn't it? I don't think club
DJs tend to be that bothered by fidelity - not in the clubs I've ever
been to anyway. SPLs seem far more important. For radio DJs who have
things cued up some time before they hit the 'play' button, surely it
would be possible to read the first N frames of data off the disc in
advance. The play button would then start the output from the buffer,
and the input to the buffer from the disc would start N frames ahead
of the start of the actual cueing point.
Matt Wenham
---
The Analogue Addicts Archive, on line at:
http://www.york.ac.uk/~mjgw100/aaa.htm
Richard D Pierce
In article <4uo8dv$9...@agate.berkeley.edu>, John Shriver
<j...@shiva.com> wrote:
> I'm not sure that the sample rate subcode bits are correctly coded on
> all CD's, or come out correctly on all SPDIF outputs. (I haven't
> checked on this at all, but my recollection is that the subcode bits
> were very sloppy in the early years. There is documentation in this
> issue...)
Actually, it's not the subcode bites, it's the channel status bits to
be precise.
In the AES/EBU implementation, bits 6-7 of channel status code byte 0
are implemented as follows (from IEC 958):
bits 6-7 Encoded sampling frequency
bit 6 7
state "0 0" Sampling frequency not indicated, Receiver
default to 48 kHz and manual override or
auto set enabled.
"0 1" 48 kHz sampling frequency. Manual override
or auto set disabled
"1 0" 44.1 kHz sampling frequency. Manual override
or auto set disabled
"1 1" 32 kHz sampling frequency. Manual override
or auto set disabled
Notice there is a default condition "0 0". Here, one can
"automatically" derive the sampling frequency from the incoming data
rate. Now, this does NOT imply one must be tightly locked in the
short-term to that clock. The differences are sufficiently gross
between the alternatives that one has to simpy make a coarse magnitude
judgement, and proceed with the answer as if you got it directly from
the explicitly set bits.
In consumer mode, it's encoded in bits 24-27 (again from IEC 958):
Fs (sampling frequency)
bit 24-27 = 0 0 0 0 44.1 kHz
0 1 0 0 48 kHz
1 1 0 0 32 kHz
All other combination are reserved for future
standardization.
There is no equivalent "default" condition.
> Of course, this would only require the DAC to correctly default to
> 44.1 when the subcode bits are bogus. Not a toughie. (There's
> maybe 0.01% of home users who have anything running at 32 or 48 -- a
> DAT deck.)
Correct. Again, it's not rocket science to determine the difference
between 32, 44.1 and 48 kHz very quickly and unambiguously, and choose
your local clock source accordingly.
> Also, there are times when you can't use an independent clock.
> Consider the Meridian powered loudspeakers with Digital input. If
> they used independent clocks and buffer memories, how would the two
> channels stay in sync? You would have to make one a clock master,
> and put a clock cable between them. They would also have to be very
> consistent on startup conditions on the FIFO RAM.
Well, consider what a hypothetical inconsistancy would get
you. Assume, purely for academics, that their FIFOs are 64 samples
deep (I have no idea if they are). Let's also assume that if the data
is sent out to the two destinations simultaneously, it will arrive
there simultaneously, except for gross (VERY gross) cable-induced
jitter, equal to the FIFO depth (this is so close to a physical
impossibility as to be ludicrous to assume, but let's do it
anyway). Let's look at the WORST case condition, where one is
completely full and the other completely empty.
64 samples corresponds to a time difference of about 1.4 ms, not
insubstantial at all. So if such a condition DID occur, it would
correspond to an 18 inch movement of one speaker relative to the
other. Hey, that's a LOT!.
But to assume that, you MUST assume that the cable is introducing a
1.4 millisecond difference in instantaneous delay. That, my friends,
corresponds to jitter on the order of 1,400,000 ps!!! Anyone care to
invoke what magic would cause such a miraculous feat of propogational
legerdomain? I can't, and I am sure no one else could.
In fact, we find the lack synchrony between two such independent DACS
is on the order of 1-2 sample in actual experiments
> Also, a DAC with a big FIFO could be rather distressing to a DJ.
> It's going to have to have some delay between first bit in and first
> sample out, to let the FIFO fill halfway.
No, the FIFO does NOT have to wait until it is half-full before
anything is clocked out. The goal is to attempt to maintain a LONG
TERM average half-full depth, not an INSTANTANEOUS condition of
half-full. That's what the fifo is for.
Again, assume a FIFO depth of, say, 1,000 samples. Do you REALLY think
the DJ is going to hit that CUE button with an accuracy 11 ms?
> (Does anyone have the Red Book or IEC clock tolerance specs for
> SPDIF? That's what we need to compute how many seconds of buffer
> are needed.)
Which tolerance? Short term or long term? Big difference. Maximum
sample rate error. Several levels of accuracy are allowed in the SPDIF
spec, and the channel status bits are set accordingly:
CLOCK ACCURACY
bit 28-29 = 0 0 Level II
= 0 1 Level III
= 1 0 Level I
= 1 1 Reserved
5.3.4.1 Level I: high accuracy mode
Tolerance of transmitting sampling frequency shall be
within += 50 x 10^-6
5.3.4.2 Level II: normal accuracy mode
All receivers should receive a signal of +- 100 x 10^-6 of
nominal sampling frequency.
5.3.4.1 Level III: variable pitch shifted clock mode
The signal in this mode may be received by specially designed
receivers.
Note. - the frequency is under consideration. A value of
+-12.5% is envisaged.
Now, again, this does not state whether the error is short-term or
long term (although by other descriptions, they seem to imply long
term). If we are talking short term errors, then we can calculate FIFO
depth. If long term, the only thing we have to do is pull the sampling
frequency.
+- 50 ppm implies a fifo depth of 4.4 samples worst case. This would
handle the equivalent of 113 nanoseconds of peak-to-peak jitter,
that's 113,000 picoseconds of jitter.
+- 1000 ppm implies a fifo depth of 90 samples worst case. This would
handle the equivalent of 2000 nanoseconds of peak-to-peak jitter,
equivalent to 2,000,000 picoseconds.
(There were some simplifying assumption to the above
back-of-the-envelope calculations. For example, we assume that the
sample rate variation are the maximum imaginable in all cases and tax
the system to its worst. Exact figures based on realistic jitter
spectrum are left as an excercise to the student.)
Like we've said: it ain't rocket science to get it right.
Frank Gales
Matt Wenham wrote:
> So, let's say we build a DAC unit which takes the output from a
> digital source, FIFO buffers it with a nice pile of RAM, and then
> clocks the data out to the DAC according to the relevant bits of the
> incoming data stream. This unit should be free from all effects of
> jitter etc. from the digital source. If it's this simple, why on earth
> do DAC builders still use the incoming data stream to control the
> clock? It appears that this is a much simpler and elegant method than
> trying to 'lock on' to what is potentially a very broken signal
> entering the DAC unit.
Hello, the following information is partly second hand from a
developer of Audio-DA-Converters, and partly my own ideas. Matt, your
idea with the RAM is the same I had too, when I asked the before
mentioned developer about the whole jitter thing. He said that with
this method you get rid of the jitter introduced by the transport,
that's okay. But through the RAM you get new jitter, which makes this
solution rather senseless. RAMs have a too much too slow rising time,
and the following "digital" (in quotation marks because signal form is
important vis-a-vis normal digital usage) components have a certain
amplitude range where these detect a positive or negative signal edge.
-----------------
/
/
- - - - - - - - - - - - -
/|
/ | signal edge detection
/ | voltage range
/ |
- - - - - - - - -|- - - -
/| |
/ | |
---------- | |
| |
signal edge detection
time variance
By this there exists a signal edge detection time variance which is a
cause for jitter. The more "digital" stages are following the more
time errors are introduced. I don't know if the detection time
variance changes in one component with temperature. Here the
engineers in this group with deeper knowledge are asked. So you will
say that this effects don't matter because the whole thing is clocked
and there are only the jitter effects of the clock signal to be
considered. But I think that it can't be this simple. When the jitter
of the digital audio signal is large enough signal edges of this
signal can fall into the detection intervall of following digital
component with the result that there is a certain propability p that a
high signal is detectet and 1-p that a low signal is detected. Then it
becomes fuzzy logic. This is something to really avoid even in normal
wholy digital gear. But I don't know if this really matters. Please
gentlemen feel free to correct me.
In my opinion the signal form of the DAC digital input can have an
influence on the analog ouput signal, which makes the whole think to
be considered analog digital. If it would be as simple as to build
computers, all CD players would be built in Taiwan and would be
included in every PC sold in the world.
One question I have in this area. Is the digital audio signal in a
CD-player in serial or in parallel form or in both. I know that the
signal on CD and on digital connection cables is serial.
It would be very informative to speak with an audio-DAC developer. I
don't know very much about the inner structure of DACs as electronic
components and the time restrictions to consider when using these.
The mentioned developer (Carlos Candeias of Candeias Audio who also
built the DAC part of the newest CEC CD-player) stated that he uses
only very very fast and very small (1 bit) buffer memories in his
gear, because larger buffers will introduce as much jitter as these
prevent.
Thank you for your attention (I want repeat that I don't know if all
the things I wrote are right, so consider these as suggestions which
can be basis of a discussion)
Have fun
Frank
Don't forget to listen to the music, when listening. The music is the
most important part of the whole think. Professional musicians and the
guys with ghettoblasters and cheap walkmen don't have our
problems. Maybe these are closer to the music as we are.
--
--------------------------------------------------------------
Frank Gales Erwin-Schroedinger-Strasse
Digital Systems Institute D-67663 Kaiserlautern / Germany
Electrical Engineering
University of Kaiserslautern E-mail: ga...@rhrk.uni-kl.de
Steven Abrams
In article <4ug8t1$1...@eyrie.graphics.cornell.edu> sdura...@aol.com
(SDuraybito) writes:
> In article <4ufan0$i...@biosun.harvard.edu>, o...@hyperback.com (Bob
> Olhsson) writes:
> > same general complaints about the really flat cartridges such as the
> > Shure V-15s and Stanton 881s as they do about digital playback.
> Interesting assertion. I use a Shure Ultra 500 because I find that it
> does provide a sense of very flat frequency response. But at the same
> time, it seems to recreate leading transients quicker than CDs and
> overall sounds far more "real" than CDs. There's none of the
> synthetic or mechanical quality that I've come to hate from CDs.
>
> The problems lie elsewhere...
Rephrasing: Bob says, "lots of people have similar complaints about
red delicious apples that they do about oranges."
SD says, "I eat golden delicious apples and I have none of those
complaints. The problem must lie elsewhere."
~~~Steve
--
Steven Abrams abr...@cs.columbia.edu
Living is easy with eyes closed, misunderstanding all you see.
-Lennon/McCartney
jj, curmudgeon and all-around grouch
In article <4u8kg8$a...@cnn.Princeton.EDU> t...@me.rochester.edu (Tim
Takahashi) writes:
>Bob Myers <my...@hpfcla.fc.hp.com> wrote:
>> That's not the point, Matt. The fact is that LP playback REQUIRES
>> significant equalization (more properly, compensation for some nasty
>> things that have to be done to the signal during the process of making
>> the LP) which is done in the analog domain.
> I dont buy that argument. Analogue disk pre-emphasis curves have
> varied widely over the years (beginning in 1925, with the first
> Orthophonic 78s)... the RIAA standard was designed to reduce surface
> noise. EMI 78's typically had NO pre-emphasis, btw.
Phonograph stylii, magnetic ones, are velocity sensors. Without the
RIAA curve or something like it, they wouldn't work.
Your claim of surface noise is just wrong. Yes, without an RIAA curve
there would be lots of surface noise, but more than that there would
be no bass or low midrange at any significant level, because of the
PHYSICS of the situation.
Please get your facts right.
> I recall a "preemphasis" bit in the data stream. What possibly could
> that be for?
It was put there as part of a political process in which some people
worried about noise at high frequencies. It's use is discouraged by
every recording person I've ever talked to, from the gentlemen from
Philips I argue with at standards meetings to the gentlemen from Japan
who I do likewise. (Yes, audio standards meetings.)
In practice the preemphasis is almost NEVER set, and much equipment
doesn't even pay attention to it.
Andre T. Yew
Frank Gales <ga...@rhrk.uni-kl.de> writes:
>But through the RAM you get new jitter, which makes this
>solution rather senseless. RAMs have a too much too slow rising time,
>and the following "digital" (in quotation marks because signal form is
>important vis-a-vis normal digital usage) components have a certain
>amplitude range where these detect a positive or negative signal edge.
That may be, but it sounds like a simple problem to solve with
a buffer or if you're really worried, a synchronizer circuit. The
real problem, I think, with using RAM with a realtime stream is
latency: do I want to go forward a track and have to wait a little bit
for the RAM fifo to fill up? Or when watching a movie, does the RAM
introduce enough of a latency to make them talk out of sync?
>The more "digital" stages are following the more
>time errors are introduced.
But what if you put a synchronizer circuit right before the
DAC? That will mop up all time errors.
>One question I have in this area. Is the digital audio signal in a
>CD-player in serial or in parallel form or in both. I know that the
>signal on CD and on digital connection cables is serial.
The vast majority of interconnect busses for consumer DACs are
serial.
--Andre
--
PGP public key available
Mark Brindle
Frank Gales (ga...@rhrk.uni-kl.de) wrote:
: One question I have in this area. Is the digital audio signal in a
: CD-player in serial or in parallel form or in both. I know that the
: signal on CD and on digital connection cables is serial.
In *every* CD player/transport, serial bits from the optical pickup
are first converted to (14-bit parallel) EFM "symbols". Each 14-bit
symbol is then used to index a "ROM lookup table" which returns the
appropriate (8-bit parallel) data byte. (This is 14-to-8 decoding.)
After 14/8 decoding, the data-bytes are stored (in a RAM buffer) for
de-interleaving, error correction, more de-interleaving, more error
correction, and finally, assembly into frames of (16-bit parallel)
audio words. (This frame contains 6 stereo samples == 24 bytes.)
Since the CIRC cross-interleaving process "scatters" every "instant of
music" over *dozens* of (7350 hz) CIRC frames, the frame assembly
process requires a buffer depth of at least several *milliseconds*,
(i.e., MILLIONS of nanoseconds!). After frame-assembly, the 16-bit
audio words are clocked out of the buffer, (still in parallel), and
sent to the local D/A chip and/or the S/PDIF transmitter chip.
Thus, *every* CD-player/transport includes the functional equivalent
of a "Digital Lens"; it's *ABSOLUTELY REQUIRED* for audio decoding!
Of course, it *is* possible for "new" jitter to be introduced while
copying data-words from the buffer to the S/PDIF transmitter or D/A;
but, there's *NO WAY* for bit-timing (or pit-spacing) irregularities
*in the optical pickup bitstream* to "sneak through" to the outputs.
Yes, load variations in CD-drive mechanism MIGHT feed-back through the
power supplies and cause jitter in the output; but, this jitter WILL
NOT be related to jitter in the "raw bits" at the opto-pickup because
nothing is servo'd to the "raw bits". The disk-speed servo takes its
signal from "synch pulses" that appear only once in every 7350 hz
frame (i.e., once in every 588 "raw bits"). And, the focus and
tracking servos don't care at all about 1's and 0's -- they just worry
about the shape and location of the reflected laser-beam.
regards,
Mark
Tim Takahashi
Bob Myers <my...@hpfcla.fc.hp.com> wrote:
> The fact is that LP playback REQUIRES significant equalization
Tim Takahashi <t...@me.rochester.edu> wrote:
> I dont buy that argument. Analogue disk pre-emphasis curves have
> varied widely over the years (beginning in 1925, with the first
> Orthophonic 78s)... the RIAA standard was designed to reduce surface
> noise. EMI 78's typically had NO pre-emphasis, btw.
JJ, famous net curmudgeon :
> Phonograph stylii, magnetic ones, are velocity sensors. Without the
> RIAA curve or something like it, they wouldn't work.
Now, while I havent engineered LP mastering equipment, I have designed
and built LP playback gear. I am missing something here.
Perhaps I should say :
"EMI 78's, typically, require no treble de-emphasis"
The specified playback equalization for a 1930's EMI has no treble
de-emphasis (of course, this doesnt imply that the recording was made
with no pre-emphasis, sorry).
The Decca FFRR scheme was odder yet. In the playback domain, a
second-order low-pass filter was imposed over about 5khz (to kill the
surface noise?)... but there was no other form of treble de-emphasis -
most other schemes use a first-order low-pass RC network.
Of course, 1930's 78rpm disks were engineered to play over mechanical
acoustical phongraphs. Surely the phonograph horn, driven by the
diapraghm, and, in turn, stylus would also be defined as a "velocity"
sensitive device.
Don't the "knees" in the RIAA "curve" imply a transition from
"constant velocity" to "constant amplitude" recording schemes?
-tim
David Bath
> But through the RAM you get new jitter, which makes this
> solution rather senseless.
Obviously you have not read the ongoing discussions in this newsgroup.
Any jitter in the data stream is ENTIRELY cancelled by using an
accurate crytal controlled clock at the DAC with a RAM FIFO feeding
it. Re-clocking digital data using FIFOs is a standard method in
digital communications systems. We've been doing it here for over 20
years. Too bad the CD designers aren't able to research prior art.
--
-------------------------------------------------------------------
David Bath mailto:ba...@aud.alcatel.com
Lightwave Systems Engineer Alcatel Telecom, Richardson, TX
Standard Disclaimers
John Busenitz
On 13 Aug 1996 j...@research.att.com wrote:
> It was put there as part of a political process in which some people
> worried about noise at high frequencies. It's use is discouraged by
> every recording person I've ever talked to, from the gentlemen from
> Philips I argue with at standards meetings to the gentlemen from Japan
> who I do likewise. (Yes, audio standards meetings.)
Does this have something to do with the so-called "aperture loss"
associated with sampled systems or systems that have some sort of
aperture involved in the recording/playback process (like the width of
the magnetic strip on a tape head). I recall Dick Small saying he
worked on a special filter to compensate for this when he was employed
at his former job (KEF?). It was for some type of digital audio
system, and he recalled that the magnitude respone of the system was
down almost 3 dB at 20 Hz, but this magnitude rolloff was "phaseless",
that is, the phase stayed at zero, so a conventional boost (1st order)
wouldn't work as ideally, so a high-Q LP filter was used, since a n
anti- aliasing/reconstruction filter was needed anyway.
Nice run-on sentance, eh?
_____________________________________________________________
John Busenitz buse...@ecn.purdue.edu
P.U. ECE http://cernan.ecn.purdue.edu/~busenitz
Disclaimer: My statements do not represent Purdue University.
jj, curmudgeon and all-around grouch
In article <4uvdjf$e...@tolstoy.lerc.nasa.gov> John Busenitz
<buse...@ecn.purdue.edu> writes:
> Does this have something to do with the so-called "aperture loss"
No.
Aperature loss is a very easily understood phenominon that can be
directly addressed and calculated by treating the aperature as a
window that is convolved with the original signal, or multiplied with
the transform. Reversing it involves creating a filter that inverts
the aperature function.
That is, of course as long as the aperature function has no in-band
near-zeros. Systems with singularities rarely work well.
jj, curmudgeon and all-around grouch
In article <4usps7$i...@eyrie.graphics.cornell.edu>
t...@me.rochester.edu (Tim Takahashi) writes:
> Of course, 1930's 78rpm disks were engineered to play over mechanical
> acoustical phongraphs. Surely the phonograph horn, driven by the
> diapraghm, and, in turn, stylus would also be defined as a "velocity"
> sensitive device.
But, roughly speaking, the diaphram that the stylus connected to was
working above its resonance, rolling of at 12dB/octave, one
compensating for the other.
Frank Derks
Tim Takahashi wrote:
[ quoted text deleted -- jwd ]
> Don't the "knees" in the RIAA "curve" imply a transition from
> "constant velocity" to "constant amplitude" recording schemes?
No they do not, the first time constant is used to extend the bass
response from the LP. The acceleration of the needle in the groove
would be to high without this correction. 78rpm disks are spinning
faster so there is less acceleration ! But more playing time was
wanted so the speed was lowered. The second curve gives the higher
frequencies more cutting room on the disk wich lowers the distortion
and gives a better signal to surface noise ratio.
In digital the trick was also provided in the red book standerd. It
was used to lower the distortion and the noise for the early DA
converters with resolutions of 14 bit.
Frank Derks
Khafara
Gee, if "Dynamic Compression" is what makes analogue vinyl more
pleasing to the ear (and more accurate) than CD, then how does one
explain the fact that a halfway-well-done analogue recording of a
symphony provides much better stereo images and soundstaging than any
CD one can name?
I'd like to see a CD reproduce the three-dimensional soundstage of a
concert hall - or any other venue - as well as does the RCA Soria (or
the Classic Records) LP of The Royal Ballet. ANd you don't need a
$3000 cartridge to hear this, either: both the $189 AudioTechnica
ATML-440 and the $800 Clearaudio Gamma-S transmit this quite nicely,
thank you.
Tamara Baker
The Absolute Sound
rec.audio.high.end
Steven Abrams
Hi, Tamara. Welcome to Rec.Audio.High-End. You may have forgotten to
don your asbestos suit before entering this portion of Usenet.
In article <4v49dn$s...@biosun.harvard.edu> kha...@aol.com (Khafara)
writes:
> Gee, if "Dynamic Compression" is what makes analogue vinyl more
> pleasing to the ear (and more accurate) than CD, then how does one
> explain the fact that a halfway-well-done analogue recording of a
> symphony provides much better stereo images and soundstaging than any
> CD one can name?
One of the things you'll find annoys people here is when opinion is
stated as a fact. You see, you may find that a halfway well-done
analaogue recording provides better images and soundstaging than any
CD, but I may find the exact opposite is true.
Another of the things that is annoying is seeing positions
represented. "Dynamic Compression" is one of the a number of euphonic
distortions present in vinyl, all of which can make some people
perceive greater soundstages, increased dynamic range, or what have
you. JJ knows quite a bit more about these things than I do, but one
thing is for certain: They do not make LPs more accurate than CDs,
although they may make them seem that way. That is fine; preference
is a matter of personal taste, while accuracy is not.
> I'd like to see a CD reproduce the three-dimensional soundstage of a
> concert hall - or any other venue - as well as does the RCA Soria (or
> the Classic Records) LP of The Royal Ballet. ANd you don't need a
> $3000 cartridge to hear this, either: both the $189 AudioTechnica
> ATML-440 and the $800 Clearaudio Gamma-S transmit this quite nicely,
> thank you.
Since I'm not familiar with this particular recording, I can't
comment. However, it may well be that no other recording or recording
media in the universe will produce greater 3-D images for you than
this record.
As they say, to each his (or her) own.
Leon Vick
Khafara <kha...@aol.com> wrote in article
<4v49dn$s...@biosun.harvard.edu>...
> I'd like to see a CD reproduce the three-dimensional soundstage of a
> concert hall - or any other venue - as well as does the RCA Soria
(or
How long has it been now (sixteen years?) since HP and TAS rejected CD
out of hand and started spewing invective about it? All that time
which could have been spent exploring the problems perceived, testing
solutions, leading the industry and listeners into the future instead
of back to the past, all wasted. I stopped buying vinyl in 1980. I
stopped reading TAS in 1986. I've have never regretted either
decision.
Richard D Pierce
In article <4v49dn$s...@biosun.harvard.edu>, Khafara <kha...@aol.com>
wrote:
>Gee, if "Dynamic Compression" is what makes analogue vinyl more
>pleasing to the ear (and more accurate) than CD,
Who said it made it more accurate? You did, who else?
>then how does one
>explain the fact that a halfway-well-done analogue recording of a
>symphony provides much better stereo images and soundstaging than any
>CD one can name?
One explains it purely by noting that this is you personal opinion, it
is NOT a universal truth that exists because you simply declared it to
be so.
>I'd like to see a CD reproduce the three-dimensional soundstage of a
>concert hall - or any other venue -
Well, I wouldn't like to see it, I'd like to hear it. And I do. I hear
superb soundstage presentations on some LP's and some CD's. I here
completely awfule presentations on many LP's and many CD's (for
example, the original LP releases of Chapuis performing the Bach organ
works has the most gawd-awful imaging imaginable. The CD re-release
has the same gawd-awful imaging, but the bass is marginally better).
>Tamara Baker
>The Absolute Sound
Oh gawd, another thought policeperson from the Arbiters of Absolute
TRVTH(tm).
Stewart Pinkerton
kha...@aol.com (Khafara) writes:
>Gee, if "Dynamic Compression" is what makes analogue vinyl more
>pleasing to the ear (and more accurate) than CD, then how does one
>explain the fact that a halfway-well-done analogue recording of a
>symphony provides much better stereo images and soundstaging than any
>CD one can name?
It's NOT more accurate! Stop it, this has been done to death, it's not
more accurate, it's not, it's not, it's not (sorry).
This one might name the following -
Linn Records, The Winged Lion, The Palladian Ensemble
Water Lily, Trumpet & Organ, Ronald Thomson and Emma Lou Diemer
Proprius, Jazz at the Pawnshop, assorted Swedes :-)
Nimbus Records, Beethoven 9th Symphony, The Hanover Band
Telarc, Late Night Brubeck, Dave Brubeck et al
These are all live recordings with an excellent sense of acoustic,
unenhanced by soggy LP artifacts and tube reverberation :-)
>I'd like to see a CD reproduce the three-dimensional soundstage of a
>concert hall - or any other venue - as well as does the RCA Soria (or
>the Classic Records) LP of The Royal Ballet. ANd you don't need a
>$3000 cartridge to hear this, either: both the $189 AudioTechnica
>ATML-440 and the $800 Clearaudio Gamma-S transmit this quite nicely,
>thank you.
Don't have any of those, but I do have lots of other good LPs that do
nearly as good a job as the average CD. Did you have an actual point
to make or is this going to be a round of yes it is, no it isn't?
Khafara
At least LPs don't suffer from error-"correction" software or "the
vinegar syndrome".
It would be lovely if digital were everything the manufacturers say it
is. Alas, it is not. It was rushed onto the market a good five to
ten years early, in order to combat the decade-long slump of the
1970s.
It would be even better if the manufacturers (who, in the interests of
combatting "format fatigue", want us to buy into DAT and - ye gods! -
MiniDisc) were really interested in "perfect sound forever", but,
well... look at the record:
- The 44.1 kHz format was picked so that one could make the CD
transport as small as possible and still fit concerts like Beethoven's
Ninth onto a single disk.
- Formats superior to the 44.1 (such as the 3M format, on which was
recorded Rickie Lee Jones' Girl at Her Volcano) got squished.
- DAT (which is being touted as "almost as good as regular digital")
and MiniDisc (which is being touted as "almost as good as DAT") are
getting heavy commercial pushes, while HDCD and other formats with
more sonic promise than either DAT or MiniDisc are being relatively
ignored, except by - you guessed it - audiophile companies such as
Reference Recordings and Chesky.
To me, it's obvious: The big movers and shakers will try to get away
with the lowest quality they can. It was true in the pre-stereo years
(compare RCA Victor's releases in America to what the British enjoyed;
RCA had no real competition in the US, and so could be as lazy as they
wanted to be), and it is true today. The only way they will move
towards quality is if they forced to do so.
Tamara Baker
The Absolute Sound
rec.audio.high.end
Richard D Pierce
In article <4v49i3$s...@biosun.harvard.edu>, Khafara <kha...@aol.com> wrote:
>At least LPs don't suffer from error-"correction" software or "the
>vinegar syndrome".
A superb arguing technique worthy of any alley brawl!
Dear Tamara, CD's have error correction because the mechanisms exists
to correct whatever errors may opccur, and despite yours and other
luddites claims to the contrary, it is not "guesstimation", but exact
replacement, no more no less.
LP's despite their demonstrably gross and oerwhelming numbers over
errors, don't do error correction BECAUSE THEY CAN'T, not because they
don't need it.
>well... look at the record:
Yes, but this time let's temper our view (as unattractive as the
notion seems to you people at AbNon) with facts instead of hearsay,
rumors, deliberate smear campaigns and outright lies and distortions
(yes, folks, this is the same gorup who several years ago published a
"proof" that CD's can't possibly have 16 bits, bolster by the most
preposterous collection of absolute nonsense as to make utter fools of
themselves):
> - The 44.1 kHz format was picked so that one could make the CD
>transport as small as possible and still fit concerts like Beethoven's
>Ninth onto a single disk.
Wrong. 44.1khz was picked because the existing digital mastering
quipment of the day was based on storage on NTSC video storage. The
selection of 74 minutes as the minimum supported length determined the
diameter, given the bit density and trackl pitch specs proposed at the
time.
Whether or not 44.1 kHz is or is not suitable is one question, but
your revisionist nonsense still falls.
> - DAT (which is being touted as "almost as good as regular digital")
Touted by whom? More runor-mongering on your part or will you surprise
us with real facts?
>and MiniDisc (which is being touted as "almost as good as DAT")
By whom?
> are getting heavy commercial pushes,
Wrong, again, your revisionist bullshit is at odd with the facts. The
DAT consumer market is dead, the entire industry (with the possible
exception of yourself in isolation) considers it so. It is, however,
THE most prevelant two-channel R/W digital storage medium in the
professional and radio markets, FAR exceeding ALL other formats
combined, and rivalling and well possibly exceeding two-track analog
tape.
>while HDCD and other formats with more sonic promise
HDCD is based on,. among other things, based on PM's claims of some
vague "Army" study of gunshot hearing, a study that they have utterly
failed to produce, despite many requests. The latest claim is that it
is "classified." This is bullshit. If i is classified, then they have
violated various national sevurities regulations in using it. If it is
NOT classified, where is it?
HDCD has also stacked the deck against ANY fair comparisons by require
this ridiculuous 6 dB gain difference, thus effectively removing NAY
possibility of comparison listening, despite many luddites whining
claims to the contrary.
>To me, it's obvious:
Agreed, to ANYONE who has an agenda, to ANYONE who makes stuff up as
they go along, to ANYONE who engages in revisonist histoy, to ANYONE
who doesn't bother to understand the principles involvedm, such as
error correction and the like, ANYTHING you want will be obvious.
You may like LP's all you want, nobody will fault you for it (and few,
it may be true, will even care). There are many LP's that I like to.
But PUHLEASE spare us the grand tradition of uninformed smear tactics
your publication is famous for. Instrauct us with facts, enlighten us
with your opinions and your likes and dislikes, but please keep your
uninformed, technically and historically wrong diatribes to yourself.
Stewart Pinkerton
DPi...@world.std.com (Richard D Pierce) writes:
>>Tamara Baker
>>The Absolute Sound
>Oh gawd, another thought policeperson from the Arbiters of Absolute
>TRVTH(tm).
Hmm, I seem to recall another periodical called 'truth'. Oh yes of
course, I'm forgetting my basic Russian, it's 'Pravda'.
Tim Takahashi
In article <4v5d8h$1...@biosun.harvard.edu>, Leon Vick <lv...@nando.net> wrote:
>How long has it been now (sixteen years?) since HP and TAS rejected CD
>out of hand and started spewing invective about it?
HP and TAS had quite a bit to say about the sound of digitially
mastered LPs, long before CD came out. Looking back on those records,
I can see a preponderance of Telarc Cannons and not a lot of music.
For the most part, I have to agree with HP and clan. While the typical
classical LP release of the late 1970's wasnt so hot, the digitial
recordings weren't a step in the right direction.
>I stopped buying vinyl in 1980.
Since CD didnt hit until the summer of 1984, what did you buy in the
interim? Open-Reel? Cassette? 8-track?
-tim
Leon Vick
Tim Takahashi <t...@me.rochester.edu> wrote in article
<4v7isd$7...@agate.berkeley.edu>...
> HP and TAS had quite a bit to say about the sound of digitially
> mastered LPs, long before CD came out. Looking back on those
records,
> I can see a preponderance of Telarc Cannons and not a lot of music.
Apples and oranges here? Telarc certainly made sure nobody would say
that digital couldn't have bass, but that was a management decision.
The music was fine. The performances weren't always first class but I
can still find aural and musical virtue in the digital LP of Carmina
Burana (for example) compared to most contemporary RCA, Angel, DGG,
Columbia and similar vinyl mashings. I also have some Harmonia Mundi,
direct-to-disk Crystal, and other very pricey LPs that are good.
> ...While the typical classical LP release of the late 1970's wasnt
so
> hot, the digitial recordings weren't a step in the right direction.
Thank heaven for Telarc. If somebody hadn't been there with digital
experience and a commitment to the medium, we still might be listening
to the sound Nipper's claw in every groove. Telarc proved that digital
could sound a lot better than anybody thought, based on what others
were doing. While some good LPs were made in the '70s, the
improvements made since were surely done for self-preservation.
> Since CD didnt hit until the summer of 1984, what did you buy in
the
> interim? Open-Reel? Cassette? 8-track?
Heh, heh, heh. Not much Tim, and most of that was $15 digital as in
Fresh Aire. Let me add that I haven't bought an LP since, and never
looked back. After thirty years of bad pressings, scratches, warps,
inter-groove distortion, etc., the vinyl industry had all they would
ever get from me, regardless of what they made. Having since built a
system that works for CDs, I don't feel a bit deprived either.
jj, curmudgeon and all-around grouch
In article <4v49dn$s...@biosun.harvard.edu> kha...@aol.com (Khafara)
writes:
>Gee, if "Dynamic Compression" is what makes analogue vinyl more
>pleasing to the ear (and more accurate) than CD, then how does one
>explain the fact that a halfway-well-done analogue recording of a
>symphony provides much better stereo images and soundstaging than any
>CD one can name?
Well, I'm not agreeing with your particular preference on
soundstaging, that's not how it sounds to me, but "how does one
explain" could exactly be that the soundstage is empahsized BECAUSE of
the dynamic compression bringing up the lower-level reverberation
signals in level relative to the peak.
Not only are the two thinks you cite not contradictory, they are at
least somewhat likely to show correlation, at least with small levels
of peak compression and low-level boost.
jj, curmudgeon and all-around grouch
In article <4v49i3$s...@biosun.harvard.edu> kha...@aol.com (Khafara)
writes:
> - DAT (which is being touted as "almost as good as regular digital")
Not from where I live. DAT is touted as a pro platform, a better
platform than CD, at least at the AES and whereever I've been.
>and MiniDisc (which is being touted as "almost as good as DAT") are
>getting heavy commercial pushes, while HDCD and other formats with
>more sonic promise than either DAT or MiniDisc are being relatively
>ignored, except by - you guessed it - audiophile companies such as
>Reference Recordings and Chesky.
Well, you assert that "HDCD and other formats with more sonic
promise", implying that of HDCD, and some other formats. Care to
provide some evidence that HDCD has "more sonic promise" <Note, I'm
not saying it doesn't, but I'm waiting to see how you provide
evidence), or that there are other formats being ignored?
I'm aware of the 96kHz DAT players, what else is there that
you're alluding to? Enlighten us, please.
jj, curmudgeon and all-around grouch
In article <4v5d8h$1...@biosun.harvard.edu> "Leon Vick"
<lv...@nando.net> writes:
> All that time which could have been spent exploring the problems
> perceived, testing solutions, leading the industry and listeners into
> the future instead of back to the past, all wasted.
Yeah. I've invited more than one CD-hater to visit. I've managed to
convince exactly one to visit and sit down and listen.
Surely, some of the euphonic effects of LP really ought to be
connected to a switch, furthermore we ought to UNDERSTAND the "basic
set" that wouldn't bug anyone (might not be enough for some, but would
be the "basic set" at least) and provide some gain, but Noooo, all I
get is "blind tests are crap" and "what's with this blind test stuff,
do YOU wear a blindfold when you do it with your wife". (yes, really,
somebody said that to me. s/he can self-id if s/he cares to). Will
anyone help out, listen? Yeah. A few. And those people aren't the ones
complaining.
Stewart Pinkerton
kha...@aol.com (Khafara) writes:
>At least LPs don't suffer from error-"correction" software or "the vinegar
>syndrome".
Ever heard of the RIAA curve? Now there's an error compensation
nightmare for you! I don't know why you put "correction" in quotes,
since digital error correction works almost perfectly, i.e. there will
normally be less than two and probably zero uncorrected errors in an
entire CD - try getting through any LP track without half a dozen
clicks! The 'vinegar syndrome' is more likely to be the use of a CD
source in a system set up to balance the 'molasses syndrome' of the LP
source, a full CD system is just clean, flat and natural from organ
pedal to triangle.
>It would be lovely if digital were everything the manufacturers say
>it is. Alas, it is not. It was rushed onto the market a good five
>to ten years early, in order to combat the decade-long slump of the
>1970s.
Quite so, but in 1996 it is indeed very close to 'perfect sound
forever' and over the past three years has pulled ahead of vinyl in
absolute terms, the best feature being that the finest CD player money
can buy is not only vastly superior to any LP source, it also costs
less than a top of the line cartridge!
> - DAT (which is being touted as "almost as good as regular digital") and
>MiniDisc (which is being touted as "almost as good as DAT") are getting
>heavy commercial pushes, while HDCD and other formats with more sonic
>promise than either DAT or MiniDisc are being relatively ignored, except
>by - you guessed it - audiophile companies such as Reference Recordings
>and Chesky.
Wrong, DAT is generally regarded in the industry as superior to CD and
of course is now available at 96kHz or 20-bits, and soon
both. MiniDisc is of course a low-quality medium, but makes a heck of
a good joggers music system. The jury is definitely out on HDCD, which
of course has irreversible coding built into the CD data. The only
format which offers a genuine sonic advance is DVD, or 20-bit 96kHz
DAT, which is a superb recording medium but will never have the
software base that CD has and DVD will.
>To me, it's obvious: The big movers and shakers will try to get away with
>the lowest quality they can. It was true in the pre-stereo years (compare
>RCA Victor's releases in America to what the British enjoyed; RCA had no
>real competition in the US, and so could be as lazy as they wanted to be),
>and it is true today. The only way they will move towards quality is if
>they forced to do so.
No argument there, I still have a V2000 video recorder :-(, and I was
furious when Rupert Murdoch steamrollered analogue ASTRA TV into
commercial dominance for satellite broadcasting against the vastly
superior D-MAC system.
OTOH, look at the superb quality available to 'the masses' for <$1,000
nowadays - compare a Marantz CD-67, Yamaha AX-590 and B&W 601s against
what you could get for the same money ten years ago. I had one of the
original Marantz CD-63 top-loaders and it cost more than that whole
rig, with no comparison at all in the sound quality!
>Tamara Baker
>The Absolute Sound
>rec.audio.high.end
Interesting sig., does this mean that TAS is now part of r.a.h-e and
you are an official spokesperson? :-)
Happy listening!
Steven Abrams
In article <4v49i3$s...@biosun.harvard.edu> kha...@aol.com (Khafara)
writes:
> At least LPs don't suffer from error-"correction" software or "the
> vinegar syndrome".
Tamara,
In what way is the ability to unambiguously correct for errors in the
media something to be suffered from? And why is this in quotes? Do
you understand how it works?
If I were in a crass mood, I'd make sexist suggestsions as to what you
might really mean by the vinegar syndrome, but that might get me in
trouble.
The rest of your post is so full of misinformation which others have
tackled that it doesn't warrant any of my time.
> To me, it's obvious: The big movers and shakers will try to get away
> with the lowest quality they can. It was true in the pre-stereo years
> (compare RCA Victor's releases in America to what the British enjoyed;
> RCA had no real competition in the US, and so could be as lazy as they
> wanted to be), and it is true today. The only way they will move
> towards quality is if they forced to do so.
Good morning to you. How long have you been in this business, and
when was the first time you realized this? This situation is not
unique to audio; it's the same with every single consumer commodity.
Why? Because to spend any time or effort to make it better than you
need to in order to sell it is very bad business.
Peter Needham
Stewart,
>It's NOT more accurate! Stop it, this has been done to death,
>it's notmore accurate, it's not, it's not, it's not (sorry).
Yes, but vinyl IS more musical. In the end, it depends if you deem
accuracy or musicality more important.
Peter
--
Peter Needham (415) 432-1234 24 hours
72662...@compuserve.com
Stewart Pinkerton
Peter Needham <72662...@CompuServe.COM> writes:
>Stewart,
>>It's NOT more accurate! Stop it, this has been done to death,
>>it's notmore accurate, it's not, it's not, it's not (sorry).
>Yes, but vinyl IS more musical. In the end, it depends if you deem
>accuracy or musicality more important.
>
It also depends what you mean by musical. It's well established that
if you make a DAT or CD-R recording of the output from a record deck
and replay that recording, it's indistinguishable from the
original. Bingo - musicality from digital!
If you want to carry on in a time warp with a fragile and inaccurate
storage medium requiring VERY expensive and fragile mechanical
engineering for replay, then feel free. I for one will continue to
work on my system in order to bring the original performance closer to
my listening room. Now that's what I call musical!
John Busenitz
On 26 Aug 1996, Peter Needham wrote:
> Stewart,
>
> >It's NOT more accurate! Stop it, this has been done to death,
> >it's notmore accurate, it's not, it's not, it's not (sorry).
>
> Yes, but vinyl IS more musical. In the end, it depends if you deem
Prove it! This is only your opinion. There are very many who think
that CD digital audio is more musical than vinyl. It seems that
musicality is somewhat subjective. So in the end, since all are equal
from a subjective (musicality) standpoint, it would seem that the
medium with the greater accuracy (CD) would be the better medium,
generally speaking. Of course, it is the opinion of the individual
that determines their favorite medium, so this is not to say that
everyone should use CDs.
> accuracy or musicality more important.
What if CD has both? Then it doesn't matter.
[We've been down this road before. Let's not turn this into the same
old CD vs. vinyl thread - wsr]
SDuraybito
In article <4vsea6$h...@tolstoy.lerc.nasa.gov>, pat...@popmail.dircon.co.uk
(Stewart Pinkerton) writes:
> It also depends what you mean by musical. It's well established that
> if you make a DAT or CD-R recording of the output from a record deck
> and replay that recording, it's indistinguishable from the
> original. Bingo - musicality from digital!
That's not the defintion of musical. That's the definition of
accurate. The two things are different. Please try to understand.
Tom Krehbiel
: Of course, it is the opinion of the individual
: that determines their favorite medium...
Unfortunately, that's far from the case. It tends to be the
individual's musical interests that determines that person's "favorite
medium," to the extent that anyone might have a favorite.
If you're a fan of the latest hits in any musical genre, CDs will be
your favorite medium. If you're a devotee of classic jazz, you're
going to be buying, holding, and playing a lot of LPs...particularly
since the reissue flood has slowed to a trickle. Same thing applies
to classical recordings from a certain length of time back...but to a
slightly lesser extent due to the bargain-priced CD lines.
In fact as a committed jazz listener, I regularly slap a stack of
shellac on the old Garrard in order to hear performances that never
made it onto LP, let alone CD.
One can only have a free choice of "favorite medium" if the music one
wishes to hear is available on all formats. And that has never
happened yet in the history of recorded sound. The trend was
established when the gramophone came out to challenge the phonograph.
Eventually the phonograph faded away...but not before a lot of ink was
shed over the relative advantages of discs vs. cylinders. (General
opinion was that cylinders sounded better, but discs were easier to
manufacture and easier to deal with at home...the first victory of
convenience and economics over sound quality.)
John Busenitz
On 28 Aug 1996, GlassAmps wrote:
> As a vinyl lover, I don't think the issue is purely vinyl vs CD. I
> believe that at the heart of the issue is digital VS analog. Simplu
> put, in analog there are an infinite number of intervals between as
> quite as it can be and as loud as it can be. In digital, slightly
Can you prove this?
> louder than 'her' is a stair-step up. That value, then, holds through
> a range until the next stair step. Sort-of like a stepped volume
> control, vs a continuously variable one. Digital Vinyl seeems to be
> less edgy than digital CD, but perhaps that is because the trajectory
> of flight of the groove and the stylus smooth out the stairstepping.
What?!? Low-pass filtering? Are you asserting that CD-playback systems
don't have reconstruction filters or anti-aliasing filters? Just what
are you saying? If one put a low-pass filter on digital audio it
sounds "less edgy"?
> Yes I know about 44kHz etc. but I am concerned about the number of
> steps between 0 and 100+ db.
It's called "Dynamic Range", and CD digital audio has more of it than
vinyl, by far. No contest. It has to do with the noise floor. Go read
about it and come back when you have enough of a clue not to
regurgitate that same tired old crap we hear every month.
(Moderators, can't we cut out these uneducated and blatantly wrong
assertions? It's like posting, "I like vinyl so it's better than CD.")
Bob Myers
GlassAmps (glas...@aol.com) wrote:
> As a vinyl lover, I don't think the issue is purely vinyl vs CD. I
> believe that at the heart of the issue is digital VS analog. Simplu
> put, in analog there are an infinite number of intervals between as
> quite as it can be and as loud as it can be. In digital, slightly
> louder than 'her' is a stair-step up. That value, then, holds
> through a range until the next stair step. Sort-of like a stepped
> volume control, vs a continuously variable one. Digital Vinyl
> seeems to be less edgy than digital CD, but perhaps that is because
> the trajectory of flight of the groove and the stylus smooth out the
> stairstepping.
This keeps cropping up, but is still incorrect.
"Analog" systems do not, and cannot, provide "an infinite number of
intervals" in the signal, nor do they in any way provide better
amplitude resolution than the 16-bit digital standard. Such arguments
fail to consider two factors.
First, in any practical system, the amplitude resolution is limited by
noise; you cannot unambiguously reproduce or determine a level beyond
the limit imposed by noise, since noise by definition is random
fluctuations in the signal amplitude. And the fact is that the
current digital standard (at 16 bits/sample) has an inherent noise
limit FAR below that which can possibly be achieved in any practical
analog standard. So the digital system can reproduce any amplitude
with AT LEAST as much accuracy as any current analog system, and in
fact far exceeds the analog systems' performance in this regard.
This can also be shown from a different perspective: an "infinitely
precise" system implies the capability to convey an infinite amount of
information. But this is clearly impossible in any band-limited
channel with ANY amount of noise present. It is impossible to create
a real system, using either analog or digital encoding, with
"infinite" accuracy, precision, or resolution (pick your favorite
word), for this very reason. So the original assumption - that
"analog" provides an "infinite" number of ANYTHING - is fundamentally
in error.
The original statement also hints at another erroneous notion, which
is that an output from a "digital" source is somehow "stairstepped" or
has some form of discontinuity simply from "being digital". This
argument fails to consider the action of reconstruction filtering,
which is a REQUIRED part of any digital system (as it prevent
high-frequency "aliasing", which is what such "stairsteps" REALLY
represent).
(And even if the "stairsteps" were not perfectly removed, how much of
the components that they represent - which are ALL above 22.05 kHz -
do you think you would hear?)
My guesses as to the reasons that you think vinyl sounds "less edgy"
than CD have to do with poor use of the CD medium, and the fact that
the LP really ISN'T a very good medium when it comes to accurately
delivering high-frequency information. Whether you prefer the result
or not is another matter entirely, but don't blame THAT on some
supposed distortion which is being ADDED by the CD. The truth is
quite the reverse.
Bob Myers KC0EW Hewlett-Packard Co. |Opinions expressed here are not
Workstations Systems Div.|those of my employer or any other
my...@fc.hp.com Fort Collins, Colorado |sentient life-form on this planet.
GlassAmps
As a vinyl lover, I don't think the issue is purely vinyl vs CD. I
believe that at the heart of the issue is digital VS analog. Simplu
put, in analog there are an infinite number of intervals between as
quite as it can be and as loud as it can be. In digital, slightly
louder than 'her' is a stair-step up. That value, then, holds through
a range until the next stair step. Sort-of like a stepped volume
control, vs a continuously variable one. Digital Vinyl seeems to be
less edgy than digital CD, but perhaps that is because the trajectory
of flight of the groove and the stylus smooth out the stairstepping.
Yes I know about 44kHz etc. but I am concerned about the number of
Stewart Pinkerton
sdura...@aol.com (SDuraybito) writes:
Please try to understand what IS being said, as opposed to what you'd
LIKE to have been said. The DAT or CD-R recording of the LP replay
system is indistinguishable from the sound of the original LP replay
system, not the original performance! Hence, any 'musicality' in the
LP system is retained by the digital recording.
It follows that so-called LP 'musicality' is a departure from truth -
i.e. it's a CREATED PERFORMANCE, not a reproduction and certainly not
a high-fidelity reproduction. Now, if you wish to create your own
performance with no regard to the original, that's just fine and
dandy. I'll take my reproduction straight up, no rocks and no soda.
The mere fact that you separate 'accurate' and 'musical' is a bit sad,
really. If a reproduction system is truly accurate, it cannot fail to
be musical. Vinyl and single-ended triodes are mere rose-tinted
escapism, alas nostalgia is not what it was (sigh).
Richard D Pierce
In article <501dlh$a...@biosun.harvard.edu>,
GlassAmps <glas...@aol.com> wrote:
>As a vinyl lover, I don't think the issue is purely vinyl vs CD. I
>believe that at the heart of the issue is digital VS analog. Simplu
>put, in analog there are an infinite number of intervals between as
>quite as it can be and as loud as it can be. In digital, slightly
>louder than 'her' is a stair-step up. That value, then, holds through
>a range until the next stair step. Sort-of like a stepped volume
>control, vs a continuously variable one.
This myth has been proffered SO many times as "proof" of something,
it's almost getting to be a mantra.
No, LPs CANNOT HAVE an infinite number of intervals, they simply
cannot. To do so requires that the LPs have an infinite dynamic range
(the relation between dynamic range and resolution was established a
half century ago. No amount of hand-waving on the part of anyone has
yet even slightly dislodged it). This is because inorder to encode on
ANY medium (be it LP, CD, or the telephone, for that matter)
differences between levels that are infinitesimally different requires
that the medium have absolutely NO NOISE whatesoever.
You would find it VERY difficult to support an assertion that LPs have
no noise.
The presence of the noise, be it the noise resulting from the thermal
random noise in the air itself, to the noise in the microphone and
preamps, to noise in the mixing and mastering process, to noise in the
LP, your phono preamp, and whatever, all contribute to causing the
entire reproduction chain to irretrievably loose resolution. The
presence of random, unpredictable noise hopelessly ambiguates ANY
information in the signal below a level determined by the noise. That
information is lost. The noise in the system places an ultimate lower
limit upon how small a change can be faithfully recorded.
Now, the problem with asserting that LPs MUST have better resolution
(and you have SPECIFICAALY defined resolution as the ability to
differentiate between very cmall changes in signal level, not some
esoteric redefinition of the term) flies completely in the face of the
inevitable fact that LP's compared to nearly any other consumer music
delivery media (with the exception of cassettes and 8-tracks) has MORE
noise. More noise means poorer resolution, the noise inevitably and
irretrivably ambiguates larger changes in signal. A medium with lower
noise floor, BY DEFINITION, is capable of preserving SMALLER
differences in signal changes than one with a higher noise floor.
Regrettably, LPs DO have a higher noise floor than CDs.
Further, your assertion that CDs possess some "stepiness" is an
assertion that is directly testable by observation. We should easily
measure this stepiness. Well, in some cases of faulty design, we can
see some, but in the vast majority of the cases where its obvservable,
it occurs at levels BELOW the point where the signal would have long
since been buried in the noise floor of an LP.
Further, in the medium itself, it is the process of dithering at the
recording and mastering process that ensures that the "stepiness" you
suiggest simply doesn't exist.
The problem here is that your assertions seem "intuitively correct."
Unfortunately, the much of the intuition and the "common sense" that
people use in this topic is simply dead wrong, and much of it has been
propogated by ill-informed editors and authors in the much of the
audiophile press (as one example, a recent "expose" on the "horrors"
of digital attempted to "prove" that CDs have far less than 16 bits of
resolution. Unfortunately, the author of this article had every single
technical point completely wrong: he had utterly failed to even do the
slightest bit of research into the topic, and made a complete fool of
himself and his magazine, though it was neither the first or last
time).
Now, none of this says you have to LIKE CDs of LPs or anything. It
DOES say that attempting to bolster your preference by using the
"stepiness" as "proof" simply fails because the premise is wrong. You
may like what you like.
>Digital Vinyl seeems to be
>less edgy than digital CD, but perhaps that is because the trajectory
>of flight of the groove and the stylus smooth out the stairstepping.
If your assertion is true, then it predicts three VERY important
things:
1. That there are, in "digital vinyl" LP's, EASILY observable steps.
Unfortunately for your assertion, when one inspects these LPs under
a microscope, there is not the slightest bit of evidence that they
exist.
2. If vinyl DID have infinite resolution (and it provably does not),
then these steps would be easily visible. If the steps existed in
the digital input (which they provably do not), then there is
something wrong in the LP process that makes them disappear. In
other words, if they are missing in the LP, then the LP hasn't
enough "resolution" to preserve them.
3. If these steps DID exist, and the "trajectory" skips over them, it
is a sign of SEVERE mistracking. Hardly an endorsement of the process.
>Yes I know about 44kHz etc. but I am concerned about the number of
>steps between 0 and 100+ db.
Well, we'll take it on face value that you do. Unfortunately, this
only covers one dimension of the problem. I suggest you explore the
tightly intertwined concepts of noise, dynamic range and resolution,
as they serve precisely the same role in the amplitude domain that
sample rate, the Nyquist frequency, bandwidth and resolution in time
do in the time domain.
Have at it, you'll find it a very fascinating topic if, for no other
reason, because it shows how one's intuition can be so at odds with
physical reality, The same is true for many other aspect of
high-fidelity reproduction. It's also true in relativity and quantum
mechanics as well, but that's for another news group or three.
Scott Johnson
In article <501dlh$a...@biosun.harvard.edu>,
GlassAmps <glas...@aol.com> wrote:
>As a vinyl lover, I don't think the issue is purely vinyl vs CD. I
>believe that at the heart of the issue is digital VS analog. Simplu
>put, in analog there are an infinite number of intervals between as
>quite as it can be and as loud as it can be. In digital, slightly
>louder than 'her' is a stair-step up. That value, then, holds through
>a range until the next stair step. Sort-of like a stepped volume
>less edgy than digital CD, but perhaps that is because the trajectory
>of flight of the groove and the stylus smooth out the stairstepping.
A little physics lesson for you, here.
The width of a record groove is about .1 mm; its height is about the
same (I don't know either figure exactly; they're close.)
Compact disks have a dynamic range of 96db; there are 2^16 = 65536
different possible values of amplitude for any sample. With the
digital error correction found in the CD format, each of these can be
stored EXACTLY.
Too duplicate that with an LP groove, the needle must be able to
detect 65536 different discrete positions (at least), all within a
space of .1mm. Doing some quick math, this means the needle must be
able to detect, without error of about 1.5 nm. That's NANOmeters.
Or, if you prefer, 15 Angstroms.
If we go to 20 bits of resolution, this figure becomes less than 1
Angstrom.
How can LPs do this? They can't. LPs (and all other analog media,
for that matter), do NOT have infinite resolution. At the very best,
their resolution is equivalent to the lattice structure of the
underlying medium; and they don't even get THAT close.
Rather than being quantized, though, sufficient noise is present in
the recording and playback processes that any quantization which MIGHT
be imposed by the lattice structure is dithered out. The result is it
kindasorta APPEARS to be continuous, with a lot of added noise.
Add proper dither to a A/D converter and what do you get? The same
effect. A properly dithered signal will resemble the original with
only a noise source added--no nasty quantization effects.
And when it comes to the level of the noise floor, CD blows LP away.
Completely.
Of course, this doesn't mean that if you like LPs, you're a bad guy or
anything. Far from it.
All it means is that the suggestion that LPs have infinite resolution
is hogwash. They have no such thing.
Scott
--
/--------------------------------------------------------------------------\
|Scott Johnson -- Professional (sometimes) SW Engineer and all-purpose Geek|
|I don't speak for nobody but myself, which everyone else is thankful for |
\--------------------------------------------------------------------------/
Stewart Pinkerton
bf...@acsu.buffalo.edu (Tom Krehbiel) writes:
>One can only have a free choice of "favorite medium" if the music one
>wishes to hear is available on all formats. And that has never
>happened yet in the history of recorded sound. The trend was
>established when the gramophone came out to challenge the phonograph.
>Eventually the phonograph faded away...but not before a lot of ink was
>shed over the relative advantages of discs vs. cylinders. (General
>opinion was that cylinders sounded better, but discs were easier to
>manufacture and easier to deal with at home...the first victory of
>convenience and economics over sound quality.)
Gee whiz, you don't suppose that was because cylinders had a constant
linear velocity, like CD? :-)
Actually, on second thoughts that's a serious question, could it be
that people reacted badly to the deterioration as 78s moved closer to
the inner grooves?
Steven Abrams
sdura...@aol.com (SDuraybito) writes:
> In article <4vsea6$h...@tolstoy.lerc.nasa.gov>, pat...@popmail.dircon.co.uk
> (Stewart Pinkerton) writes:
> > It also depends what you mean by musical. It's well established that
> > if you make a DAT or CD-R recording of the output from a record deck
> > and replay that recording, it's indistinguishable from the
> > original. Bingo - musicality from digital!
>
> That's not the defintion of musical. That's the definition of
> accurate. The two things are different. Please try to understand.
Record Deck == britishspeak for turntable.
Therefore, his post reads like this: If you take the output from a
TURNTABLE and make a DAT or CD-R recording of it, the output is
indistinguishable from the original. If the original was musical,
then the CD-R or DAT is, too.