Three speed automatic turntable replacement

[William Sommerwerck]

None of these are forms of pre-distortion, because the effects
they compensate for are linear -- they are not distortion. (I'm
going to insist on this, because "distortion" has a clear, specific
meaning.)

I would argue that any difference between the input to a system
and its output would be "distortion" by definition.

That's not correct usage. Any linear transformation of a system's I/O
characteristics is, by definition, //not// distortion. Distortion is a
non-linear transformation.

One reason some early CDs sounded so bad is that they were made using
the cutting masters intended for driving cutting lathes. Everything that
was done to the master tapes to prepare them for making vinyls (and that
is a hell of a lot) is distortion, in my definition.

As far as I know, all the changes were linear changes, and were not
distortion,

stylus shape pre-distortion because the cutter is flat and the
playback stylus is a cone or an ellipse (and they need different
treatment during recording) (this is NOT Dynagroove)

It was a component of Dynagroove

It was also used long before (and after) Dynagroove came out,
by many labels.

Name one.

-- the only one that actually worked.

Which would explain why is was abandoned so soon after its introduction.

I don't know how long it took RCA to abandon Dynagroove. My understanding is
that RCA continued to use pre-distortion for some years. It was the only
part of Dynagroove that was a technically and aesthetically legitimate
improvement, and I doubt RCA wanted to immediately abandon the hardware for
it.

corrections because the cutter traverses the disk radially and the
playback stylus does not

This is a form of distortion, but it's not compensated for during
recording, because there is no standardized playback-arm geometry.

For information on this, and some of your other statements, I strongly
recommend that you get, and read, an earlier edition of John Eargle's
"Handbook Of Recording Engineering" -- earlier (2nd. ed. for instance),
because later ones make no mention of vinyl techniques, but they are
well covered in early ones.

If John Eargle's views differ from mine on these points, he is incorrect.

If the waveform you want out of the phono preamp is a square wave (or
near enough), what path must the groove cause the stylus to trace to
produce that? (Hint: it's *not* a square wave). Once you figure out what
the groove looks like for a square wave output, you'll realize that the
groove *never* really "looks like" the waveform of the "sound", and
that the conversion from one waveshape to the other is yet another
compromise, though not so much as the recording process. It is,
in fact, quite difficult to analyze the shape of the groove and
determine
"analytically" what the resulting sound waves should look like.

Actually, it's quite simple, because we know how cutting systems
were designed. In the electrical era, cutter heads were velocity
devices, with EQ applied to produce what the record label considered
a practical approximation of constant-amplitude recording.

OK. It's simple. So what *does* the groove look like?

Read the my preceding paragraph.

A mechanical playback method takes care of all that "automatically"
because the recording process is arranged so as to minimize the
*system* distortion, *assuming the reproduction is going to be by
means of a stylus in the groove*.

No, it isn't. If distortion in one part of the system cancels out
distortion
in another part, it's either dumb luck, or because somebody "fiddled"
with
things. This was particularly true in the days of acoustic recording,
when
recordings were intended to be played back on the reproducing equipment
made by the record manufacturer.

Please read what Western Electric said about its system of electrical
disk recording -- it was intended that each element of it be as perfect
as
technology then allowed, without errors in one part of the system being
compensated for in another part. This allowed for continual improvement
without having to periodically re-design everything.

I've read it. Eargle's book is a lot more recent, and embodies a whole
lot of knowledge and *practice* that was simply unknown at the time
the Western Electric stuff was written.

What does that have to do with the first paragraph of this section?

Any other playback process must perform all those things
explicitly in order to get good results -- i.e. it must simulate
the behavior of a stylus in the groove -- and that's not so easy.

Sure it is. The math to do this existed long before any of us was born.

And that is why vinyl recordings sound sooo good ...

I don't know whether you're being ironic, but on the assumption you're
not... About a year ago I sat down with a pile of audiophile LPs and
sonically browsed them. Many had really nice sound -- highly listenable --
but little of it sounded as much like //live// sound as the best SACDs (and
the rare CD). LPs can be highly euphonic -- but when push-pull comes to
shove, they simply aren't that //accurate//. As only a few had performances
I wanted to hang on to, I sold almost all my audiophile disks to Silver
Platters, and got $600 in store credit.

In general, mechanical analog recording stinks, and mechanical analog
playback is even worse. There's nothing inherently "correct" about it.
It's
amazing that phonograph records were as good as they (sometimes) were.

On that, we can absolutely agree.

 

[William Sommerwerck]

You may not be aware but digital sound also has such
schemes. Understand this (listen up y'all I ain't doing
this twice).......

Yes, we don't have all day, Vince.

OK you got wonderful 16 bit sound, it is great right?
Well what happens when you are at a piece of the
music that is at -66dB ? It ain't 16 bit no more, is it
even 8 bit ? Just that very soft part of it, understand ?

Well the people who designed the system DID figure
this out and as such there was a pre-emph signal on
CDs from day one. When it got below a certin level the
preemph would kick in and the deemph on the playback
side. This gave it more bits during the soft passages of
the music.

The pre-emphasis was optional, and was abandoned years ago. I believe some
CD players have no way of compensating for it.

CD quality has been quite surpassed by certain DVD audio
formats. In fact the Sony PCM-1 or PCM-F1 could take a
DVD-R or any VCR and record better than a CD. Almost thirty
years ago!

I don't think you know what you're talking about.

The PCM-F1 (and its Nakamichi equivalent, the DMP-100) could record at
14-bit or 16-bit resolution. That was its limit.

I own a DMP-100. 25 years ago I made parallel live recordings with the dbx
700, a processor using a system similar to Sony's Direct Stream Digital. To
my ears, the dbx 700 was superior.

 

[josephkk]

On Tue, 14 Aug 2012 23:11:19 -0700 (PDT), jurb6006@gmail.com wrote:

Without what you erroneously call distorion everything you hear would sound hissy, to the point maybe of unlistenability. This also applies to tapes and therefore to the master tapes made to record music. Hiss upon hiss. Think AM was bad ?

If you are really interested i have a lucky recording. The surface noise
of the vinyl disc, the residual FM broadcast noise, and the residual NAB
tape hiss of a nice Ampex AX-300 semipro open reel tape deck are all about
the same level. If you know how to listen to you can hear the three
different noises clearly.

In an RIAA compliant phono preamp a calibrated EQ curve is applied. Actully if you see the "curve" it is actually a tilt. Two resistors and two capacitors per channel achieve this. It matches because they match the "time constants" of the opposite network applied to the signal when it was recorded.

 

[josephkk]

On Wed, 15 Aug 2012 06:25:17 -0700, "William Sommerwerck"
<grizzledgeezer@comcast.net> wrote:

You may not be aware but digital sound also has such
schemes. Understand this (listen up y'all I ain't doing
this twice).......

Yes, we don't have all day, Vince.

OK you got wonderful 16 bit sound, it is great right?
Well what happens when you are at a piece of the
music that is at -66dB ? It ain't 16 bit no more, is it
even 8 bit ? Just that very soft part of it, understand ?

Well the people who designed the system DID figure
this out and as such there was a pre-emph signal on
CDs from day one. When it got below a certin level the
preemph would kick in and the deemph on the playback
side. This gave it more bits during the soft passages of
the music.

The pre-emphasis was optional, and was abandoned years ago. I believe some
CD players have no way of compensating for it.


CD quality has been quite surpassed by certain DVD audio
formats. In fact the Sony PCM-1 or PCM-F1 could take a
DVD-R or any VCR and record better than a CD. Almost thirty
years ago!

I don't think you know what you're talking about.

The PCM-F1 (and its Nakamichi equivalent, the DMP-100) could record at
14-bit or 16-bit resolution. That was its limit.

I own a DMP-100. 25 years ago I made parallel live recordings with the dbx
700, a processor using a system similar to Sony's Direct Stream Digital. To
my ears, the dbx 700 was superior.

Fine. Dbx systems included dynamic compressors, like A-law and u-Law
telephony systems. Provided better dynamic range and other nice
properties at the cost of more processing, see also SACD.

?-)

 

I beg to differ. Everything you have heard in the last twenty years is somehow compressed. Not the data but the music. the texchnique was developed for the 8mm Sony digital audio format which ran on the same tapes as the camcorder. It would record IIRC 12 tracks, like an old eight track. You could switch tracks but to get back to the beginning unless you rewind the tape.

That process split the fields on the diagonal sweep of the heads into 12 sectors. The 8mm system had integral flying erase heads, just as PAL had an integral COMB filter. You COULD change tracks with ease, if the tape position was near.

But that is not the point. If you look at the specs for that system you find out a few things. First of all it's LIMITATIONS are that of an FM broadcast. Well two, in tandem to give you two channels. But you are only getting 15Khz.

Now the PCM-1 had it's own format, and really all I knowe is that is worked.. But I have in my posession the full manual on the PCM-F1 which explains it's operation in more detail than anyone could ever want. It has the normal mode which is 44,100 siteen bit. Your CDs are not that. The 16 bit audio is cropped down by EFM which separates the positive and negative, then it is further compressed down to eight bit words. Additionally there is a Dolby like noise suppression scheme. That is not quite high fidelity, yathink ?

The PCM-F1 doesn't use this, making it better than a CD out the box. But if tha isn't good enough, it has a 48Khz mode, which is what the big movie theaters etc. use. And there is no compression at all. Well they do use it but they have REAL sisxteen bit, CDs do not.

Yes, it was better even back then. What's more in case you didn't know, the ubiquitous CD is capable of four channel discrete sound. (reference a book called "Principles Of Digital Audio" circa 1990 or so for that)

You don't think I know what I am talking about ? I think you should have stopped before the "you know what you are talking about" part.

J

 

[isw]

In article <k0g7hg$5c2$1@dont-email.me>,
"William Sommerwerck" <grizzledgeezer@comcast.net> wrote:

--snip --

For information on this, and some of your other statements, I strongly
recommend that you get, and read, an earlier edition of John Eargle's
"Handbook Of Recording Engineering" -- earlier (2nd. ed. for instance),
because later ones make no mention of vinyl techniques, but they are
well covered in early ones.

If John Eargle's views differ from mine on these points, he is incorrect.

He was chief engineer for several recording companies. What are your
qualifications?

If the waveform you want out of the phono preamp is a square wave (or
near enough), what path must the groove cause the stylus to trace to
produce that? (Hint: it's *not* a square wave). Once you figure out what
the groove looks like for a square wave output, you'll realize that the
groove *never* really "looks like" the waveform of the "sound", and
that the conversion from one waveshape to the other is yet another
compromise, though not so much as the recording process. It is,
in fact, quite difficult to analyze the shape of the groove and
determine
"analytically" what the resulting sound waves should look like.

Actually, it's quite simple, because we know how cutting systems
were designed. In the electrical era, cutter heads were velocity
devices, with EQ applied to produce what the record label considered
a practical approximation of constant-amplitude recording.

OK. It's simple. So what *does* the groove look like?

Read the my preceding paragraph.

Yeah ... I didn't think you knew.

Isaac

 

[isw]

In article <73359959-a9ce-4bdc-bb74-23b6effbe289@googlegroups.com>,
jurb6006@gmail.com wrote:

"I would argue that any difference between the input to a system and its
output would be "distortion" by definition. "

Well I would argue something else, in fact I will.

Of the signal "modifications" applied prior to cutting -- to prevent
slope, displacement, and curvature overload, stylus tilt correction,
distortion of the groove so the output of a spherical (or elliptical)
stylus will better approximate the original signal, LF vertical limiting
to guarantee there always will be a groove to track, ..., a number are
not reversible during playback even in theory. I don't see how those can
be called anything except "distortions".

And, FWIW, John Eargle refers to the things done to the audio to make it
suitable for making vinyl disks as "distortions" ...

You may have made a blanket statement unknowing of certain aspects of hifi,
or you may be one of those purists that require the plastic for the faceplate
to be grown in th...... nevermind.

Nope. Just a retired EE and a long-time student of the audio recording
and reproduction process, with some experience in a recording studio and
at running a cutting lathe. And a serious pragmatist, when it comes to
what is necessary and sufficient for good *domestic* audio reproduction.

That statement would mean then that you object to RIAA equalization as well
as NARTB.

Of course I don't object to them; they are necessary to get even the
not-very-good reproduction that the analog media you mention are capable
of.

The reason RIAA equalization exists is because of the properties of vinyl.

The bass is cut to minimize stylus excursion at low frequencies; that
would be a problem no matter what the disk was made of.

Black vinyl is the toughest, and was produced by adding carbon to the
material, just like many plastics and similar things. Before that the laquer
or whatever records (78s) that were CUT also had alot of grit in them. This
made HISS. Since the high frequency waves ARE actually smaller, they decided
to crank them up on the recording side and do the EXACT opposite on the
playback side. It's like having to steer right when you have a flat tire on
the left.

Without what you erroneously call distorion everything you hear would sound
hissy, to the point maybe of unlistenability. This also applies to tapes and
therefore to the master tapes made to record music. Hiss upon hiss. Think AM
was bad ?

In an RIAA compliant phono preamp a calibrated EQ curve is applied. Actully
if you see the "curve" it is actually a tilt. Two resistors and two
capacitors per channel achieve this.

I have designed low noise/low distortion RIAA preamps; the better ones
have three breakpoints (3180 ľs, 318 ľs, and 75 ľs), plus a fourth HF
rolloff to keep the loop gain from dropping too low in the ultrasonic
range (improves stability), and sometimes a fifth to keep the sub-20 Hz
stuff from messing with the speakers.

It matches because they match the "time
constants" of the opposite network applied to the signal when it was
recorded.

Well, except for the unilaterally applied radial EQ (the cutting amp
boosts the highs towards the inner grooves -- the corresponding playback
cut is neither specified nor controlled; it just happens because the
groove "wiggles" get closer together there.

On tape the NARTB curve is THE standard, but for better results they do not
approach it separately. Very long ago they did, you could get an outboard
preamp and it would amplify your signal right from the tape heads, apply the
requisite NARTB equalization and give you flat response at "line level",
which is all pretty much what a tuner put(s) out. However in a tape deck
there is so much equalization to do with the heads that the NARTB, as I said,
is not considered sepatrately, and neither is it in playback. HOWEVER for
compatibility between different decks, it must be adhered to, that is the
only thing.

You may not be aware but digital sound also has such schemes. Understand this
(listen up y'all I ain't doing this twice).......

OK you got wonderful 16 bit sound, it is great right ? Well what happens when
you are at a piece of the music that is at -66dB ? It ain't 16 bit no more,
is it even 8 bit ? Just that very soft part of it, understand ?

Well the people who designed the system DID figure this out and as such there
was a pre-emph signal on CDs from day one. When it got below a certin level
the preemph would kick in and the deemph on the playback side. This gave it
more bits during the soft passages of the music.

Preemphasis was included in the CD spec, and was used at first, but it
is rarely used today; hasn't been for a long while.

This lack of bits situation is also abated by something called dither. Dither
is white noise added to the program material to prevent these digital
artifacts from cropping up during the soft passages of you favorite jam.

For CDs, the dither signal is rarely white noise; other spectra are more
useful.

All this info is in the Eargle book I mentioned earlier.

Isaac

 

[William Sommerwerck]

Everything you have heard in the last twenty years
is somehow compressed. Not the data but the music.
the texchnique was developed for the 8mm Sony digital
audio format which ran on the same tapes as the
camcorder. It would record IIRC 12 tracks, like an old
eight track. You could switch tracks but to get back
to the beginning unless you rewind the tape.

The only compression used was dynamic-range compression, which was (badly)
needed for 8-bit (!!!) digital recording. As you said, you could make
audio-only recordings. Pioneer produced a machine that could record 24 hours
on a single two-hour video tape, using six passes. (Pioneer's presenter at
the SCES remarked sarcasticly that some people could put their entire LP
collection on a single tape!)

If you look at the specs for that system you find ... its
LIMITATIONS are that of an FM broadcast. Well two,
in tandem to give you two channels. But you are only
getting 15Khz.

A good live broadcast on FM has far better sound that most CDs. 15kHz is
hardly a meaningful limitation.

The PCM-1 had it's own format, and really all I know is that
it worked. But I have the full [technical] manual, which explains
its operation in more detail than anyone could ever want.

Agreed! (I have it, too.)]

It has the normal mode, which is 44.1 kHz, 16-bit.

Normal mode is 44.056 kHz, 14-bit. 16-bit recording gains two bits by
weakening the error correction, making uncorrectable errors more likely. (I
never had problems with 16 bit.)

Your CDs are not that. The 16 bit audio is cropped down by EFM which
separates the positive and negative, then it is further compressed down to

eight bit words. Additionally there is a Dolby-like noise suppression
scheme. That is not quite high fidelity, yathink?

If you're talking about CDs, that's mostly wrong. The CD format is linear.
There is no "Dolby-like noise suppression scheme" of ANY sort. EFM means
"eight to fourteen modulation", which maps eight bits of each 16-bit data
word into a 14-bit "space". This reduces read and pressing errors, and makes
it easier to correct those that do occur.

The PCM-F1 doesn't use this, making it better than
a CD out of the box.

The PCM-F1 is essentially identical to CD. Both are 16-bit, and their sample
rates differ only by 0.1%.

But if that isn't good enough, it has a 48Khz mode, which is what
the big movie theaters etc. use. And there is no compression at all.
Well, they do use it but they have REAL sixteen bit, CDs do not.

You're confusing the PCM-F1, CD, and DAT. DAT has 48kHz sampling (plus 44.1
and 32). The others don't. NONE of these systems use lossy compression. And
they're all "real" 16-bit.

Most systems of theatrical digital sound use some form of lossy compression.
But I believe non-lossy compression is beginning to appear. (Someone fill me
in, please.)

Yes, it was better even back then. What's more, in case you
didn't know, the ubiquitous CD is capable of four channel
discrete sound. (Reference a book called "Principles Of Digital
Audio" circa 1990 or so for that).

The original Red Book standards defined a four-channel disk that ran at
double speed (halving the playback time). Unfortunately, the idiots at
Philips neglected any provision requiring two-channel-only CD players to be
compatible with four-channel disks.

As long as we're on it... The Philips idiots also missed the opportunity to
double the playback time for mono recordings.

You don't think I know what I am talking about ? I think you
should have stopped before the "you know what you are
talking about" part.

To quote Alexander Pope: "A >>little<< learning is a dangerous thing."

 

[William Sommerwerck]

For information on this, and some of your other statements, I strongly
recommend that you get, and read, an earlier edition of John Eargle's
"Handbook Of Recording Engineering" -- earlier (2nd. ed. for instance),
because later ones make no mention of vinyl techniques, but they are
well covered in early ones.

If John Eargle's views differ from mine on these points, he is incorrect.

He was chief engineer for several recording companies. What are your
qualifications?

I know what I'm talking about. Does Mr Eargle?

I didn't say John Eargle was wrong. I said I was right, and that if his
views (which I don't know in detail) were different, then he's wrong. I
stand by that.

If the waveform you want out of the phono preamp is a square wave (or
near enough), what path must the groove cause the stylus to trace to
produce that? (Hint: it's *not* a square wave.) Once you figure out
what
the groove looks like for a square wave output, you'll realize that
the
groove *never* really "looks like" the waveform of the "sound", and
that the conversion from one waveshape to the other is yet another
compromise, though not so much as the recording process. It is,
in fact, quite difficult to analyze the shape of the groove and
determine "analytically" what the resulting sound waves should look
like.

Actually, it's quite simple, because we know how cutting systems
were designed. In the electrical era, cutter heads were velocity
devices, with EQ applied to produce what the record label considered
a practical approximation of constant-amplitude recording.

OK. It's simple. So what *does* the groove look like?

Read my preceding paragraph.

Yeah ... I didn't think you knew.

The paragraph reads: "In the electrical era, cutter heads were velocity
devices, with EQ applied to produce what the record label considered a
practical approximation of constant-amplitude recording."

Do you know the difference between velocity and constant-amplitude
recording? If a recording approximates constant-amplitude, then the groove
waveform will approximate the original waveform. Right? RIGHT?

 

[isw]

In article <k0j0n5$ffh$1@dont-email.me>,
"William Sommerwerck" <grizzledgeezer@comcast.net> wrote:

For information on this, and some of your other statements, I strongly
recommend that you get, and read, an earlier edition of John Eargle's
"Handbook Of Recording Engineering" -- earlier (2nd. ed. for instance),
because later ones make no mention of vinyl techniques, but they are
well covered in early ones.

If John Eargle's views differ from mine on these points, he is incorrect.

He was chief engineer for several recording companies. What are your
qualifications?

I know what I'm talking about. Does Mr Eargle?

Why don't you just google on his name and decide for yourself?

I didn't say John Eargle was wrong. I said I was right, and that if his
views (which I don't know in detail) were different, then he's wrong. I
stand by that.


If the waveform you want out of the phono preamp is a square wave (or
near enough), what path must the groove cause the stylus to trace to
produce that? (Hint: it's *not* a square wave.) Once you figure out
what
the groove looks like for a square wave output, you'll realize that
the
groove *never* really "looks like" the waveform of the "sound", and
that the conversion from one waveshape to the other is yet another
compromise, though not so much as the recording process. It is,
in fact, quite difficult to analyze the shape of the groove and
determine "analytically" what the resulting sound waves should look
like.

Actually, it's quite simple, because we know how cutting systems
were designed. In the electrical era, cutter heads were velocity
devices, with EQ applied to produce what the record label considered
a practical approximation of constant-amplitude recording.

OK. It's simple. So what *does* the groove look like?

Read my preceding paragraph.

Yeah ... I didn't think you knew.

The paragraph reads: "In the electrical era, cutter heads were velocity
devices, with EQ applied to produce what the record label considered a
practical approximation of constant-amplitude recording."

Do you know the difference between velocity and constant-amplitude
recording? If a recording approximates constant-amplitude, then the groove
waveform will approximate the original waveform. Right? RIGHT?

Nope. Not even close.

An electrical generator (a magnetic cartridge) provides a constant
voltage in response to a constant velocity. To get a more-or-less square
wave, the groove is a triangle wave. The slopes equate to the flat top
and bottom, and the inflection points to the rising and falling edges.

Isaac

 

[William Sommerwerck]

The paragraph reads: "In the electrical era, cutter heads were velocity
devices, with EQ applied to produce what the record label considered a
practical approximation of constant-amplitude recording."

Do you know the difference between velocity and constant-amplitude
recording? If a recording approximates constant-amplitude, then the
groove
waveform will approximate the original waveform. Right? RIGHT?

Nope. Not even close.

It's dead-on.

An electrical generator (a magnetic cartridge)...

There are non-magnetic electrical generators. You meant a /velocity/ device.

provides a constant voltage in response to a constant velocity.

Only if it's a velocity generator.

To get a more-or-less square wave, the groove is a triangle wave.

Yes -- out of an UNEQUALIZED magnetic pickup.

Do I have to explain it all over again? Seems so.

When you cut a phonograph record with an electromagnetic head, you get a
constant-velocity recording. That is, groove modulation is huge for low
frequencies, teensy for high frequencies.

There two basic problems with this. The extreme low-frequency excursions
require a very wide groove pitch (reducing playing time) and the high
frequencies are "down in the dirt" (making the sound noisier than we might
like).

The solution adopted by the recording industry was to cut the bass and boost
the treble to produce an approximation of constant-amplitude recording. I
say an "approximation", because a huge amount of boost and cut would be
needed to get constant amplitude from (say) 50Hz to 20kHz (about 50dB). In
practice, the bass cut stops at about 500Hz and the treble boost begins at
about 2kHz (which needs only about 30dB).

The result is a recording that's much closer to constant amplitude. If one
looks at a constant-amplitude recording under a microscope, it will resemble
the original acoustic waveform -- not the temporally integrated waveform
produced by an un-EQ'd magnetic cutter head..

When recording using RIAA (or other) EQ is played back with a
constant-amplitude pickup, the output is essentially flat, exactly what we
want in a cheap phonograph. When played with a magnetic (velocity) pickup,
the constant-amplitude waveform is differentiated, requiring integrating
equalization in the amplifier. Which is what phono preamps provide.

I realize this can be confusing, but the preceding is correct. You need to
think it through.

 

[spamtrap1888]

On Aug 16, 7:31 am, "William Sommerwerck" <grizzledgee...@comcast.net>
wrote:

Everything you have heard in the last twenty years
is somehow compressed. Not the data but the music.
the texchnique was developed for the 8mm Sony digital
audio format which ran on the same tapes as the
camcorder. It would record IIRC 12 tracks, like an old
eight track. You could switch tracks but to get back
to the beginning unless you rewind the tape.

The only compression used was dynamic-range compression, which was (badly)
needed for 8-bit (!!!) digital recording. As you said, you could make
audio-only recordings. Pioneer produced a machine that could record 24 hours
on a single two-hour video tape, using six passes. (Pioneer's presenter at
the SCES remarked sarcasticly that some people could put their entire LP
collection on a single tape!)

If you look at the specs for that system you find ... its
LIMITATIONS are that of an FM broadcast. Well two,
in tandem to give you two channels. But you are only
getting 15Khz.

A good live broadcast on FM has far better sound that most CDs. 15kHz is
hardly a meaningful limitation.

The PCM-1 had it's own format, and really all I know is that
it worked. But I have the full [technical] manual, which explains
its operation in more detail than anyone could ever want.

Agreed! (I have it, too.)]

It has the normal mode, which is 44.1 kHz, 16-bit.

Normal mode is 44.056 kHz, 14-bit. 16-bit recording gains two bits by
weakening the error correction, making uncorrectable errors more likely. (I
never had problems with 16 bit.)

Your CDs are not that. The 16 bit audio is cropped down by EFM which

separates the positive and negative, then it is further compressed down to
eight bit words. Additionally there is a Dolby-like noise suppression
scheme. That is not quite high fidelity, yathink?

If you're talking about CDs, that's mostly wrong. The CD format is linear..

If you don't count quantizing error and aliasing effects.

There is no "Dolby-like noise suppression scheme" of ANY sort. EFM means
"eight to fourteen modulation", which maps eight bits of each 16-bit data
word into a 14-bit "space". This reduces read and pressing errors, and makes
it easier to correct those that do occur.

 

[isw]

In article <k0lm6i$t5m$1@dont-email.me>,
"William Sommerwerck" <grizzledgeezer@comcast.net> wrote:

The paragraph reads: "In the electrical era, cutter heads were velocity
devices, with EQ applied to produce what the record label considered a
practical approximation of constant-amplitude recording."

Do you know the difference between velocity and constant-amplitude
recording? If a recording approximates constant-amplitude, then the
groove
waveform will approximate the original waveform. Right? RIGHT?

Nope. Not even close.

It's dead-on.

An electrical generator (a magnetic cartridge)...

There are non-magnetic electrical generators. You meant a /velocity/ device.

provides a constant voltage in response to a constant velocity.

Only if it's a velocity generator.

And kindly describe another kind of electromagnetic generator?

To get a more-or-less square wave, the groove is a triangle wave.

Yes -- out of an UNEQUALIZED magnetic pickup.

Do I have to explain it all over again? Seems so.

When you cut a phonograph record with an electromagnetic head, you get a
constant-velocity recording. That is, groove modulation is huge for low
frequencies, teensy for high frequencies.

There two basic problems with this. The extreme low-frequency excursions
require a very wide groove pitch (reducing playing time) and the high
frequencies are "down in the dirt" (making the sound noisier than we might
like).

The solution adopted by the recording industry was to cut the bass and boost
the treble to produce an approximation of constant-amplitude recording. I
say an "approximation", because a huge amount of boost and cut would be
needed to get constant amplitude from (say) 50Hz to 20kHz (about 50dB). In
practice, the bass cut stops at about 500Hz and the treble boost begins at
about 2kHz (which needs only about 30dB).

The result is a recording that's much closer to constant amplitude. If one
looks at a constant-amplitude recording under a microscope, it will resemble
the original acoustic waveform -- not the temporally integrated waveform
produced by an un-EQ'd magnetic cutter head..

When recording using RIAA (or other) EQ is played back with a
constant-amplitude pickup, the output is essentially flat, exactly what we
want in a cheap phonograph. When played with a magnetic (velocity) pickup,
the constant-amplitude waveform is differentiated, requiring integrating
equalization in the amplifier. Which is what phono preamps provide.

I realize this can be confusing, but the preceding is correct. You need to
think it through.

In fact, I did not "think it through" until I happened to actually LOOK
AT THE GROOVE of the "square wave" track of a test record. Then I
thought "well, that's odd ..." and then I realized what must be going on.

Isaac

 

[isw]

In article
<8748f609-59dc-4aff-8ee6-f5f48ac5a217@f4g2000pbq.googlegroups.com>,
spamtrap1888 <spamtrap1888@gmail.com> wrote:

On Aug 16, 7:31 am, "William Sommerwerck" <grizzledgee...@comcast.net
wrote:
Everything you have heard in the last twenty years
is somehow compressed. Not the data but the music.
the texchnique was developed for the 8mm Sony digital
audio format which ran on the same tapes as the
camcorder. It would record IIRC 12 tracks, like an old
eight track. You could switch tracks but to get back
to the beginning unless you rewind the tape.

The only compression used was dynamic-range compression, which was (badly)
needed for 8-bit (!!!) digital recording. As you said, you could make
audio-only recordings. Pioneer produced a machine that could record 24 hours
on a single two-hour video tape, using six passes. (Pioneer's presenter at
the SCES remarked sarcasticly that some people could put their entire LP
collection on a single tape!)

If you look at the specs for that system you find ... its
LIMITATIONS are that of an FM broadcast. Well two,
in tandem to give you two channels. But you are only
getting 15Khz.

A good live broadcast on FM has far better sound that most CDs. 15kHz is
hardly a meaningful limitation.

The PCM-1 had it's own format, and really all I know is that
it worked. But I have the full [technical] manual, which explains
its operation in more detail than anyone could ever want.

Agreed! (I have it, too.)]

It has the normal mode, which is 44.1 kHz, 16-bit.

Normal mode is 44.056 kHz, 14-bit. 16-bit recording gains two bits by
weakening the error correction, making uncorrectable errors more likely. (I
never had problems with 16 bit.)

Your CDs are not that. The 16 bit audio is cropped down by EFM which

separates the positive and negative, then it is further compressed down to
eight bit words. Additionally there is a Dolby-like noise suppression
scheme. That is not quite high fidelity, yathink?

If you're talking about CDs, that's mostly wrong. The CD format is linear.

If you don't count quantizing error and aliasing effects.

Quantization is nicely taken care of by the dither, and aliasing by the
oddly named "anti-aliasing low-pass filter" which is an integral (and
mandatory) part of the process.

Isaac

 

[William Sommerwerck]

In fact, I did not "think it through" until I happened to actually
LOOK AT THE GROOVE of the "square wave" track of a test
record. Then I thought "well, that's odd ..." and then I realized
what must be going on.

I still have my CBS Labs test LPs, and don't even need to look at the
square-wave disk to know that the groove doesn't resemble a triangle
waveform. IT CAN'T, because an LP cut with RIAA equalization comes much
closer to being constant-amplitude than constant-velocity.

-----------------

After writing that, I decided to pull out the STR 112 test disk. It turns
out it //does// have triangle waveforms. Here's what the liner notes say...

"Two groups of four 1000 Hz square-wave [sic] test bands are provided on the
STR-112, Side A, at the outer and inner radii of the disc. ... The actual
modulation excursion, as viewed under a microscope, is a triangle wave.
Therefore, playback with an ideal velocity responsive [sic] pickup will
yield a square wave."

In other words... This is NOT a constant-amplitude recording, and it is to
be used //without// RIAA equalization.

"The square wave modulation allows rapid appraisal of stylus-tip mass,
damping, and tracking."

But the disk doesn't have "square wave modulation".

If the conclusion isn't clear, I'll beat it into the ground with a sledge
hammer. This disk DOES NOT test what it claims to test. You cannot test the
"square wave" response of a pickup by applying a grossly different stimulus!

What the disk actually tests is the response to a triangle wave. The fact
that a (nominal) square wave pops out of a velocity-responding pickup is
beside the point -- especially because an LP mastered with RIAA EQ //would//
have a groove displacement that looked a lot like a square wave -- the
pickup's response to which is what we're supposed to be testing.

QED. Case closed.

 

[josephkk]

On Wed, 15 Aug 2012 19:33:38 -0700 (PDT), jurb6006@gmail.com wrote:

I beg to differ. Everything you have heard in the last twenty years is somehow compressed. Not the data but the music. the texchnique was developed for the 8mm Sony digital audio format which ran on the same tapes as the camcorder. It would record IIRC 12 tracks, like an old eight track. You could switch tracks but to get back to the beginning unless you rewind the tape.

That process split the fields on the diagonal sweep of the heads into 12 sectors. The 8mm system had integral flying erase heads, just as PAL had an integral COMB filter. You COULD change tracks with ease, if the tape position was near.

But that is not the point. If you look at the specs for that system you find out a few things. First of all it's LIMITATIONS are that of an FM broadcast. Well two, in tandem to give you two channels. But you are only getting 15Khz.

Now the PCM-1 had it's own format, and really all I knowe is that is worked. But I have in my posession the full manual on the PCM-F1 which explains it's operation in more detail than anyone could ever want. It has the normal mode which is 44,100 siteen bit. Your CDs are not that. The 16 bit audio is cropped down by EFM which separates the positive and negative, then it is further compressed down to eight bit words. Additionally there is a Dolby like noise suppression scheme. That is not quite high fidelity, yathink ?

The PCM-F1 doesn't use this, making it better than a CD out the box. But if tha isn't good enough, it has a 48Khz mode, which is what the big movie theaters etc. use. And there is no compression at all. Well they do use it but they have REAL sisxteen bit, CDs do not.

Yes, it was better even back then. What's more in case you didn't know, the ubiquitous CD is capable of four channel discrete sound. (reference a book called "Principles Of Digital Audio" circa 1990 or so for that)

You don't think I know what I am talking about ? I think you should have stopped before the "you know what you are talking about" part.

J

You are incorrect on some points.

CDs are true 44100 samples per second, 16 bits per channel, stereo. Any
compression is applied before translation to disc; that does not mean that
there isn't any, it is just not in the process from audio stream (analog
or digital) to disc to what comes out of a standard player.

From what you say you have no idea what EFM is or does. Let alone how it
is used in making CDs.

?-)

 

[josephkk]

On Sat, 18 Aug 2012 04:15:30 -0700, "William Sommerwerck"
<grizzledgeezer@comcast.net> wrote:

In fact, I did not "think it through" until I happened to actually
LOOK AT THE GROOVE of the "square wave" track of a test
record. Then I thought "well, that's odd ..." and then I realized
what must be going on.

I still have my CBS Labs test LPs, and don't even need to look at the
square-wave disk to know that the groove doesn't resemble a triangle
waveform. IT CAN'T, because an LP cut with RIAA equalization comes much
closer to being constant-amplitude than constant-velocity.

-----------------

After writing that, I decided to pull out the STR 112 test disk. It turns
out it //does// have triangle waveforms. Here's what the liner notes say...

"Two groups of four 1000 Hz square-wave [sic] test bands are provided on the
STR-112, Side A, at the outer and inner radii of the disc. ... The actual
modulation excursion, as viewed under a microscope, is a triangle wave.
Therefore, playback with an ideal velocity responsive [sic] pickup will
yield a square wave."

In other words... This is NOT a constant-amplitude recording, and it is to
be used //without// RIAA equalization.

"The square wave modulation allows rapid appraisal of stylus-tip mass,
damping, and tracking."

But the disk doesn't have "square wave modulation".

If the conclusion isn't clear, I'll beat it into the ground with a sledge
hammer. This disk DOES NOT test what it claims to test. You cannot test the
"square wave" response of a pickup by applying a grossly different stimulus!

Wrong. Oh and BTW the velocity of the track IS (approximately) a square
wave. Think that through.

What the disk actually tests is the response to a triangle wave. The fact
that a (nominal) square wave pops out of a velocity-responding pickup is
beside the point -- especially because an LP mastered with RIAA EQ //would//
have a groove displacement that looked a lot like a square wave -- the
pickup's response to which is what we're supposed to be testing.

QED. Case closed.

And you really missed a lot. The output of the cartridge is a approximate

square wave for that track, if you run that through RIAA playback
equalization you get a very different signal.


?-)

 

[William Sommerwerck]

In other words... This is NOT a constant-amplitude recording,
and it is to be used //without// RIAA equalization.

"The square wave modulation allows rapid appraisal of
stylus-tip mass, damping, and tracking."

But the disk doesn't have "square wave modulation".

If the conclusion isn't clear, I'll beat it into the ground with a
sledge hammer. This disk DOES NOT test what it claims to
test. You cannot test the "square wave" response of a pickup
by applying a grossly different stimulus!

Wrong.

I just /knew/ someone was going to claim that. It's called "blindly
believing a manufacturer's claim despite all evidence to the contrary".

The LP doesn't "know" what sort of a pickup is playing it. The groove is
engraved with a triangle wave, and the behavior of the pickup's mechanical
system represents its response to a TRIANGLE WAVE driving function. What
part of that don't you (or CBS) understand? The fact that a velocity pickup
converts the triangle into a square wave is beside the point -- yea, even
meaningless. If I played the disk with a amplitude-responding pickup (such
as a Sonotone) its output would be a TRIANGLE WAVE. Would the test then be a
SQIARE WAVE test? According to CBS's perverted use of language and logic, it
would be.

Oh and BTW the velocity of the track IS (approximately) a square
wave. Think that through.

Of course. You're repeating what I said. But the fact that a velocity pickup
has an output that resembles a particular waveform DOES NOT mean you're
testing how the pickup responds to a groove modulation of that sort.

If you want to call the STR 112 a test of transient behavior, go right
ahead, because it arguably is. But it DOES NOT test the pickup's response to
square-wave modulation. What would you call its behavior for a
amplitude-responding pickup?

And you really missed a lot. The output of the cartridge is an
approximate square wave for that track, if you run that through
RIAA playback equalization you get a very different signal.

No, I missed nothing. You are guilty of blindly believing what "experts",
without thinking about it.

You can easily find books that say that simply sampling a continuous
waveform converts it into digital data. But that doesn't make it true.

 

[josephkk]

On Wed, 22 Aug 2012 04:38:31 -0700, "William Sommerwerck"
<grizzledgeezer@comcast.net> wrote:

In other words... This is NOT a constant-amplitude recording,
and it is to be used //without// RIAA equalization.

"The square wave modulation allows rapid appraisal of
stylus-tip mass, damping, and tracking."

But the disk doesn't have "square wave modulation".

If the conclusion isn't clear, I'll beat it into the ground with a
sledge hammer. This disk DOES NOT test what it claims to
test. You cannot test the "square wave" response of a pickup
by applying a grossly different stimulus!

Wrong.

I just /knew/ someone was going to claim that. It's called "blindly
believing a manufacturer's claim despite all evidence to the contrary".

The LP doesn't "know" what sort of a pickup is playing it. The groove is
engraved with a triangle wave, and the behavior of the pickup's mechanical
system represents its response to a TRIANGLE WAVE driving function. What
part of that don't you (or CBS) understand? The fact that a velocity pickup
converts the triangle into a square wave is beside the point -- yea, even
meaningless. If I played the disk with a amplitude-responding pickup (such
as a Sonotone) its output would be a TRIANGLE WAVE. Would the test then be a
SQIARE WAVE test? According to CBS's perverted use of language and logic, it
would be.

First off you have to find/have/create an amplitude responding pickup. Do
that first then re-enter the discussion.
Second a triangle wave is the integral of a square wave, or the same
relation between amplitude and velocity response.

Oh and BTW the velocity of the track IS (approximately) a square
wave. Think that through.

Of course. You're repeating what I said. But the fact that a velocity pickup
has an output that resembles a particular waveform DOES NOT mean you're
testing how the pickup responds to a groove modulation of that sort.

If you want to call the STR 112 a test of transient behavior, go right
ahead, because it arguably is. But it DOES NOT test the pickup's response to
square-wave modulation. What would you call its behavior for a
amplitude-responding pickup?


And you really missed a lot. The output of the cartridge is an
approximate square wave for that track, if you run that through
RIAA playback equalization you get a very different signal.

No, I missed nothing. You are guilty of blindly believing what "experts",
without thinking about it.

You can easily find books that say that simply sampling a continuous
waveform converts it into digital data. But that doesn't make it true.

But they will claim that the z-transform make it so when challenged. Just

as you argue about velocity versus amplitude responding pickups. Find me
a phonograph pickup that is amplitude responding. Then we have something
to discuss. Hint, stylus connected to a diaphragm and horn is not one.

I constantly fight with other engineers that claim that just because it is
on some manufacturers' data sheet it is right/true. e.g. that the name of
the standard is RS-232 when the cover of the current 15+ year old (1997)
version of standard says TIA-232. Likewise network connectors which are
IEC 60603-7-x (See page 13 of TIA-568.2) or 8P8C modular instead of RJ-45
(which not even phone companies use any more). I will bet that you have
seen such issues but rarely dug into it to the finish.

?-)

 

[William Sommerwerck]

First off you have to find/have/create an amplitude-
responding pickup. Do that first then re-enter the discussion.

They're commonly found in cheap turntables, especially of the USB ilk.

Second, a triangle wave is the integral of a square wave,
or the same relation between amplitude and velocity response.

Yes, of course. That's simple calculus, which I took in high school almost
50 years ago.

But they will claim that the z-transform make it so when
challenged. Just as you argue about velocity- versus
amplitude-responding pickups. Find me a phonograph
pickup that is amplitude responding. Then we have
something to discuss.

First of all, the issue has nothing whatever to do with whether
amplitude-responding pickups exist, ever have existed, or ever could exist.
We are talking about a manufacturer claiming that a test record shows how a
pickup responds to (mechanical) square-wave modulation, when the disk
doesn't have such modulation.

Piezoelectric transducers are basically amplitude-sensitive. Crystal and
ceramic pickups were manufactured for decades, but gradually disappeared as
magnetic pickups grew less expensive and tracked at lower forces. There have
been "good" ceramic pickups (Sonotone and Weathers, for example), but they
were rare. However, you can still find ceramic pickups in cheap turntables.

http://www.knowzy.com/computers/audio/digitize_your_lps/usb_record_player_turntable_comparison.htm

(The author's claim that anti-skating is an absolute necessity is debatable,
to say the least.)

If you want to get super-ultra picky about it, many ceramic pickups are
mechanically equalized to compensate for the ~12dB shelf in response when
playing RIAA LPs. This doesn't change the fact that the pickup is,
fundamentally, an amplitude-responding device.

I constantly fight with other engineers that claim that just because it is
on some manufacturers' data sheet it is right/true. e.g. that the name of
the standard is RS-232 when the cover of the current 15+ year old (1997)
version of standard says TIA-232. Likewise network connectors which are
IEC 60603-7-x (See page 13 of TIA-568.2) or 8P8C modular instead of RJ-45
(which not even phone companies use any more). I will bet that you have
seen such issues but rarely dug into it to the finish.

It appears that we are equally annoyed by the spread of misinformation.