historical classical recording techniques

[[email protected]]

Can anyone suggest classical recordings from the 60's, 70's, and 80's
which feature specific stereo techniques? I know there was plenty of
Blumlein at EMI and Decca Tree at Decca, but these things move in
trends (i.e. Deutsche Grammphon's transition from stereo to multitrack
orchestral recording in the 80's), and I'd really appreciate any
specific examples you might give.

Better yet, if there's a book in which someone's already brought this
information together, please clue me in.

Thanks a googol,
Matt

[William Sommerwerck]

Well, to start with...

The original RCA "Living Stereos" were spaced omnis, I believe.

[Harry Lavo]

"William Sommerwerck" > wrote in message
...
> Well, to start with...
>
> The original RCA "Living Stereos" were spaced omnis, I believe.
>

Bert Whyte's recordings on Everest were all made via ORTF placment, I
believe.

[PStamler]

On Jul 30, 1:46*pm, "William Sommerwerck" >
wrote:
> Well, to start with...
>
> The original RCA "Living Stereos" were spaced omnis, I believe.

As were the original Mercury Living Presence recordings -- both three
spaced U 47 mics in omni mode. The theory of three-mic stereo was
developed by Maxfield, who also led the Western Electric team that
developed electrical recording.

Peace,
Paul

[Scott Dorsey]

Harry Lavo > wrote:
>"William Sommerwerck" > wrote in message
...
>> Well, to start with...
>>
>> The original RCA "Living Stereos" were spaced omnis, I believe.

Spaced triads, yes.

>Bert Whyte's recordings on Everest were all made via ORTF placment, I
>believe.

The Bob Fine stuff on Everest was also done with spaced triads. It was
very popular because at the time there weren't many cardioids with decent
off-axis response.

There have been a bunch of sampler recordings out there for evaluating
miking methods. The hall makes such a huge difference you can't really
talk about one without the other.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[nebulax]

On Jul 30, 2:44*pm, " > wrote:
> Can anyone suggest classical recordings from the 60's, 70's, and 80's
> which feature specific stereo techniques? I know there was plenty of
> Blumlein at EMI and Decca Tree at Decca, but these things move in
> trends (i.e. Deutsche Grammphon's transition from stereo to multitrack
> orchestral recording in the 80's), and I'd really appreciate any
> specific examples you might give.
>
> Better yet, if there's a book in which someone's already brought this
> information together, please clue me in.
>
> Thanks a googol,
> Matt


Here's a site that talks about various classical music recording
techniques that were used at Mercury Records during the 'Living
Presence' era - http://www.soundfountain.com/amb/mercury.html

-Neb

[William Sommerwerck]

> Here's a site that talks about various classical music recording
> techniques that were used at Mercury Records during the 'Living
> Presence' era - http://www.soundfountain.com/amb/mercury.html

This is a fascinating article, which I've bookmarked for later, careful
reading. However, the author knows next to nothing about technology, and
there are technical and aesthetic errors. Don't take everything in this
article literally, or as absolute truth.

[nebulax]

On Jul 31, 8:36*am, "William Sommerwerck" >
wrote:
> > Here's a site that talks about various classical music recording
> > techniques that were used at Mercury Records during the 'Living
> > Presence' era -http://www.soundfountain.com/amb/mercury.html
>
> This is a fascinating article, which I've bookmarked for later, careful
> reading. However, the author knows next to nothing about technology, and
> there are technical and aesthetic errors. Don't take everything in this
> article literally, or as absolute truth.


Yeah, I was noticing some of that myself. It's also too bad that Wilma
Cozart Fine mastered the 'Living Presence' cd's with a DAT machine,
but I guess that was one of the few choices for a digital format at
the time.

-Neb

[William Sommerwerck]

> It's too bad that Wilma Cozart Fine mastered the 'Living Presence'
< CDs with a DAT, but I guess that was one of the few choices of
> digital format at the time.

That was just a few years ago. She could have gone with 88.2/24, if she'd
wanted.

[Robert Morein[_2_]]

"nebulax" > wrote in message
...
On Jul 30, 2:44 pm, " > wrote:
> Can anyone suggest classical recordings from the 60's, 70's, and 80's
> which feature specific stereo techniques? I know there was plenty of
> Blumlein at EMI and Decca Tree at Decca, but these things move in
> trends (i.e. Deutsche Grammphon's transition from stereo to multitrack
> orchestral recording in the 80's), and I'd really appreciate any
> specific examples you might give.
>
> Better yet, if there's a book in which someone's already brought this
> information together, please clue me in.
>
> Thanks a googol,
> Matt


Here's a site that talks about various classical music recording
techniques that were used at Mercury Records during the 'Living
Presence' era - http://www.soundfountain.com/amb/mercury.html

-Neb
----------------------------------------------------------------------------
---------
There is a guy who hangs out on aus.hi-fi, Iain Churches, who is a former
recording engineer for Decca. Has a tremendous depth of experience and
historical knowledge. If you hunt him down, I'd like to hear what he has to
say. The stupid asshole won't talk to me, he's too "high and mighty" and
believes all the **** about my worthless life, no job, etc, etc. Dickhead.

Bob Morein
(310) 237-6511

[Scott Dorsey]

nebulax > wrote:
>Yeah, I was noticing some of that myself. It's also too bad that Wilma
>Cozart Fine mastered the 'Living Presence' cd's with a DAT machine,
>but I guess that was one of the few choices for a digital format at
>the time.

What's wrong with DAT? It's not the most reliable thing around, but when
it works, it works fine.
--scott

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

[jazzman31[_2_]]

On Aug 1, 9:57*am, (Scott Dorsey) wrote:
> nebulax > wrote:
> >Yeah, I was noticing some of that myself. It's also too bad that Wilma
> >Cozart Fine mastered the 'Living Presence' cd's with a DAT machine,
> >but I guess that was one of the few choices for a digital format at
> >the time.
>
> What's wrong with DAT? *It's not the most reliable thing around, but when
> it works, it works fine.
> --scott
>
> --
> "C'est un Nagra. *C'est suisse, et tres, tres precis."

I was assuming they meant the limited freq response with dat. You're
stuck with those brick wall filters and comprimised top end when it
comes to re mastering on a better format. I could be wrong.

Rick

[Scott Dorsey]

jazzman31 > wrote:
>
>I was assuming they meant the limited freq response with dat. You're
>stuck with those brick wall filters and comprimised top end when it
>comes to re mastering on a better format. I could be wrong.

Bits is bits. You put a good converter on the front, it's as good
as any other 16-bit format. Like the CD it'll finally be issued on,
for instance.

Sure, the SV-3700 had crappy-sounding converters, but that is no reason
to avoid the format.

If you note, you'll see Ms. Fine was using Sony outboard converters.
They wouldn't be my first choice, but they have no brickwall filters
or compromised top end.
--scott

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

[William Sommerwerck]

>>> Yeah, I was noticing some of that myself. It's also too bad that
>>> Wilma Cozart Fine mastered the 'Living Presence' CDs with a
>>> DAT machine, but I guess that was one of the few choices for
>>> a digital format at the time.

>> What's wrong with DAT? It's not the most reliable thing around,
>> but when it works, it works fine.

> I was assuming they meant the limited freq response with DAT.
> You're stuck with those brick wall filters and comprimised top
> end when it comes to re mastering on a better format. I could
> be wrong.

Suppose you sample at 88.2kHz. You then have to downsample to 44.1kHz to
master the CD, right?

The trouble is that 44.1kHz sampling supports a bandwidth of no more than
22.05kHz. You therefore have to filter the 88.2kHz samples to a 22.05kHz
bandwidth before downsampling.

Right?

[nebulax]

On Aug 1, 4:05*pm, (Scott Dorsey) wrote:
> jazzman31 > wrote:
>
> >I was assuming they meant the limited freq response with dat. You're
> >stuck with those brick wall filters and comprimised top end when it
> >comes to re mastering on a better format. I could be wrong.
>
> Bits is bits. *You put a good converter on the front, it's as good
> as any other 16-bit format. *Like the CD it'll finally be issued on,
> for instance.
>
> Sure, the SV-3700 had crappy-sounding converters, but that is no reason
> to avoid the format. *
>
> If you note, you'll see Ms. Fine was using Sony outboard converters.
> They wouldn't be my first choice, but they have no brickwall filters
> or compromised top end.
> --scott


Actually, I missed the part about the outboard converters. The DAT
format itself was ok, but the internal A/D conversion circuits of
those machines usually left a lot to be desired. However, DAT really
sucked as a location film sound recording medium, due to moisture and
heat issues, tape sticking to the heads, etc. I had tapes eaten by a
not-fully-warmed-up portable DAT machine one more than one occasion!

-Neb

[Mike Rivers]

nebulax wrote:

> Actually, I missed the part about the outboard converters.

No, I don't think you did. That came later, but it certainly makes
sense, and DAT recorders have had digital inputs for most of
their life, though some that implemented SCMS wouldn't record
through the digital input unless the data stream carried the
proper code in the channel status block.

>The DAT
> format itself was ok, but the internal A/D conversion circuits of
> those machines usually left a lot to be desired.

Practically all A/D conversion left a lot to be desired in those days,
but better converters came along pretty quickly. I still have, and
occasionally
use, a Symetrix AD620 A/D converter in front of my DAT and CD recorder. It's
a 20-bit converter with dithering (if you want it, which you do when
going into
a DAT) that gave a fairly honest 16-bit stream.

> However, DAT really
> sucked as a location film sound recording medium, due to moisture and
> heat issues, tape sticking to the heads, etc. I had tapes eaten by a
> not-fully-warmed-up portable DAT machine one more than one occasion!

That was a characteristic of rotary head machines - ADATs and VCRs
included. There were two standard drum sizes for DAT. The small one used
in the popular Walkman DAT recorders was particularly fussy about
moisture. I suppose that with film sound work, there are times when you
don't
have time to give the gear a proper warm-up. For those times, we had
Nagras. And still do.



--
If you e-mail me and it bounces, use your secret decoder ring and reach
me here:
double-m-eleven-double-zero at yahoo -- I'm really Mike Rivers
)

[Laurence Payne[_2_]]

On Sun, 02 Aug 2009 14:51:41 GMT, Mike Rivers >
wrote:

>That was a characteristic of rotary head machines - ADATs and VCRs
>included. There were two standard drum sizes for DAT. The small one used
>in the popular Walkman DAT recorders was particularly fussy about
>moisture. I suppose that with film sound work, there are times when you
>don't
>have time to give the gear a proper warm-up. For those times, we had
>Nagras. And still do.

Isn't it all done to solid-state memory now?

[Scott Dorsey]

nebulax > wrote:
>Actually, I missed the part about the outboard converters. The DAT
>format itself was ok, but the internal A/D conversion circuits of
>those machines usually left a lot to be desired. However, DAT really
>sucked as a location film sound recording medium, due to moisture and
>heat issues, tape sticking to the heads, etc. I had tapes eaten by a
>not-fully-warmed-up portable DAT machine one more than one occasion!

Some of the DAT machines actually had pretty good converters; the HHB
portable certainly was better than most of what you see around today.

There were some awful ones too, especially on earlier machines. But
back then, using outboard converters was considered standard operating
procedure in a lot of studios so manufacturers didn't have a huge
motivation to improve the internal conversion. (Sadly, they also often
had lousy digital I/O too, in the case of Panasonic machines.)
--scott

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

[Mike Rivers]

Laurence Payne wrote:

> Isn't it all done to solid-state memory now?

I haven't figured out how to install solid state memory in my DAT. ;)


--
If you e-mail me and it bounces, use your secret decoder ring and reach
me here:
double-m-eleven-double-zero at yahoo -- I'm really Mike Rivers
)

[nebulax]

On Aug 2, 11:53*am, Laurence Payne > wrote:
> On Sun, 02 Aug 2009 14:51:41 GMT, Mike Rivers >
> wrote:
>
> >That was a characteristic of rotary head machines - ADATs and VCRs
> >included. There were two standard drum sizes for DAT. The small one used
> >in the popular Walkman DAT recorders was particularly fussy about
> >moisture. I suppose that with film sound work, there are times when you
> >don't
> >have time to give the gear a proper warm-up. For those times, we had
> >Nagras. And still do.
>
> Isn't it all done to solid-state memory now?
>


Yeah, pretty much all the new location film sound machines are either
hard-drive or solid-state memory. The last DAT machine (portable or
otherwise) went out of production a few years ago. I think you can
also still buy a new Nagra reel-to-reel machine, but the old ones are
so durable that I doubt many new ones are sold. And, I still prefer
the sound of analog Nagra machines myself, but I don't hear of any
producers requesting them these days.

-Neb

[Scott Dorsey]

nebulax > wrote:
>
>Yeah, pretty much all the new location film sound machines are either
>hard-drive or solid-state memory. The last DAT machine (portable or
>otherwise) went out of production a few years ago. I think you can
>also still buy a new Nagra reel-to-reel machine, but the old ones are
>so durable that I doubt many new ones are sold. And, I still prefer
>the sound of analog Nagra machines myself, but I don't hear of any
>producers requesting them these days.

I get a call to drag the 4.2 out of the closet now and then, but mostly
from sound effects people or soundtrack music folks. No dialogue, really.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[William Sommerwerck]

> >>>> Yeah, I was noticing some of that myself. It's also too bad that
> >>>> Wilma Cozart Fine mastered the 'Living Presence' CDs with a
> >>>> DAT machine, but I guess that was one of the few choices for
> >>>> a digital format at the time.
> >
> >>> What's wrong with DAT? It's not the most reliable thing around,
> >>> but when it works, it works fine.
> >
> >> I was assuming they meant the limited freq response with DAT.
> >> You're stuck with those brick wall filters and comprimised top
> >> end when it comes to re mastering on a better format. I could
> >> be wrong.
> >
> > Suppose you sample at 88.2kHz. You then have to downsample to 44.1kHz to
> > master the CD, right?
> >
> > The trouble is that 44.1kHz sampling supports a bandwidth of no more
than
> > 22.05kHz. You therefore have to filter the 88.2kHz samples to a 22.05kHz
> > bandwidth before downsampling.
> >
> > Right?
> >
> As a delivery medium to the stamper, it's a 1 to 1 mapping, fine.

I don't think you caught the point.

[William Sommerwerck]

>> I don't think you caught the point.

> Maybe not. I probably agree with your point, but what is it? :)

The point is that higher sampling rates don't necessarily enable you to have
a less-abrupt anti-aliasing filter -- at least, if you're going to
down-convert to 44.1kHz.

You can't resample from 88.2s/s to 44.1s/s without first filtering the
88.2s/s samples to remove all the content above 22.05kHz.

[William Sommerwerck]

"Soundhaspriority" > wrote in message
...
>
> "William Sommerwerck" > wrote in message
> ...
> >>> I don't think you caught the point.
> >
> >> Maybe not. I probably agree with your point, but what is it? :)
> >
> > The point is that higher sampling rates don't necessarily enable you to
> > have
> > a less-abrupt anti-aliasing filter -- at least, if you're going to
> > down-convert to 44.1kHz.
> >
> > You can't resample from 88.2s/s to 44.1s/s without first filtering the
> > 88.2s/s samples to remove all the content above 22.05kHz.
> >
> Well, if you record at 88.2, and downconvert in software, you can choose
> your algorithm. Recording at 44.1, the choice is made at the time of the
> recording.
>
> Some time after 2000, Cirrus Logic introduced an IIR filter, with a
gradual
> slope, as opposed to the traditional choice of a FIR filter with a sharp
> slope. On playback, one tester of an Audiofire unit incorporating Cirrus
> converters noticed a low level ultrasonic artifact, but remarked that it
> actually sounded better than another unit with higher specced AKM
> converters, presumably due to less phase shift. I think AKM recently moved
> in this direction as well.
>
> By recording at 88.2, the filter cutoff is way above the audio band. So
you
> can play with it in post. Software algorithms have at least the potential
to
> be better than hardware, because they are easier to change, and silicon
> realestate is not a consideration.
>
> Bob Morein
> (310) 237-6511
>
>

[William Sommerwerck]

>> You can't resample from 88.2s/s to 44.1s/s without first filtering the
>> 88.2s/s samples to remove all the content above 22.05kHz.

> Well, if you record at 88.2, and downconvert in software, you can
> choose your algorithm. Recording at 44.1, the choice is made at
> the time of the recording.

> Some time after 2000, Cirrus Logic introduced an IIR filter, with a
gradual
> slope, as opposed to the traditional choice of a FIR filter with a sharp
> slope. On playback, one tester of an Audiofire unit incorporating Cirrus
> converters noticed a low level ultrasonic artifact, but remarked that it
> actually sounded better than another unit with higher specced AKM
> converters, presumably due to less phase shift. I think AKM recently
> moved in this direction as well.

> By recording at 88.2, the filter cutoff is way above the audio band. So
you
> can play with it in post. Software algorithms have at least the potential
to
> be better than hardware, because they are easier to change, and silicon
> real estate is not a consideration.

Sorry for the accidental pointless repost. Slip of the finger.

The original post suggested that oversampling /automatically/ gives a slower
rolloff, or eliminates the need for filtering altogether.That's what I was
jumping on.

[Arny Krueger]

"jazzman31" > wrote in message

> On Aug 1, 9:57 am, (Scott Dorsey) wrote:
>> nebulax > wrote:
>>> Yeah, I was noticing some of that myself. It's also too
>>> bad that Wilma Cozart Fine mastered the 'Living
>>> Presence' cd's with a DAT machine, but I guess that was
>>> one of the few choices for a digital format at the time.
>>
>> What's wrong with DAT? It's not the most reliable thing
>> around, but when it works, it works fine.
>> --scott

Agreed.

> I was assuming they meant the limited freq response with
> dat. You're stuck with those brick wall filters and
> comprimised top end when it comes to re mastering on a
> better format. I could be wrong.

If some one were going to use classical recording techniques, why do they
need more than the 24 KHz bandpass of DAT?

Jazzman, if you were *really* going with 60's classic production techniques,
you'd use analog tape and be stuck with head bumps, wow, flutter, scrape
flutter, tape saturation, azimuth losses, print though, a high noise floor,
etc.

You do realize that compared to real world analog tape in the 60s, DAT
quality was a gigantic breath of fresh air?

However, if you want to recreate legacy micing techniques today you'd be
wise to get a Microtrak or a Zoom. Press the right buttons and you've got
24/96.

[Arny Krueger]

"William Sommerwerck" > wrote in
message

>>>> Yeah, I was noticing some of that myself. It's also
>>>> too bad that Wilma Cozart Fine mastered the 'Living
>>>> Presence' CDs with a
>>>> DAT machine, but I guess that was one of the few
>>>> choices for
>>>> a digital format at the time.
>
>>> What's wrong with DAT? It's not the most reliable thing
>>> around, but when it works, it works fine.
>
>> I was assuming they meant the limited freq response with
>> DAT. You're stuck with those brick wall filters and
>> comprimised top end when it comes to re mastering on a
>> better format. I could be wrong.

> Suppose you sample at 88.2kHz. You then have to
> downsample to 44.1kHz to master the CD, right?

"Supposing that you sample at 88.2" basically means you are into numbers for
the sake of numbers.

> The trouble is that 44.1kHz sampling supports a bandwidth
> of no more than 22.05kHz. You therefore have to filter the 88.2kHz
> samples to a 22.05kHz bandwidth before downsampling.

DAT supported 48 KHz sample rate recording and playback. IOW 24 KHz Nyquist.

If memory serves, the pro machines supported analog recotrding at both 44
and 48, but the consumer machines could only record analog at 48.

[Arny Krueger]

"William Sommerwerck" > wrote in
message
>>> I don't think you caught the point.
>
>> Maybe not. I probably agree with your point, but what is
>> it? :)

> The point is that higher sampling rates don't necessarily
> enable you to have a less-abrupt anti-aliasing filter --
> at least, if you're going to down-convert to 44.1kHz.

Agreed, that is how it is supposed to be done.

> You can't resample from 88.2s/s to 44.1s/s without first
> filtering the 88.2s/s samples to remove all the content above 22.05kHz.

There is evidence that some commercial resampling software breaks this rule.
They leave in a little aliasing, I guess as some kind of EFX.

[Scott Dorsey]

Arny Krueger > wrote:
>DAT supported 48 KHz sample rate recording and playback. IOW 24 KHz Nyquist.

Also 44.1, and also 32 ksamp/sec although there were two 32 ksamp/sec formats
which were incompatible.

>If memory serves, the pro machines supported analog recotrding at both 44
>and 48, but the consumer machines could only record analog at 48.

Generally true but not always.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[William Sommerwerck]

> If memory serves, the pro machines supported analog recording at
> both 44.1 and 48, but the consumer machines could only record
> analog at 48.

I have a Sony 87ES -- presumably a "prosumer" machine -- that records at 48,
44.1, and 32. I sometimes use 32 to record FM broadcasts, simply because you
get 4 hours of recording with no practical loss of frequency response.

If you were going to release recordings at 48kHz, you could probably get
away with no low-pass filtering _at all_, because signals would have to be
above 28kHz to "fold" back into the audio band (and I doubt any mics get/got
out that far). But, again, once you resample to 44.1kHz, you'd better be
sure there's no significant program material above 22kHz.

I have a sneaking suspicion that, for most acoustic music recorded on an
analog deck, the low-pass filter itself probably does more sonic harm than
any small amount of over-22kHz components that wind up in the audible
passband.

[William Sommerwerck]

"Arny Krueger" > wrote in message
...
> "William Sommerwerck" > wrote in
> message

>> You can't resample from 88.2s/s to 44.1s/s without first
>> filtering the 88.2s/s samples to remove all the content
>> above 22.05kHz.

> There is evidence that some commercial resampling software
> breaks this rule. They leave in a little aliasing, I guess as some
> kind of EFX.

As I suggest in the other post, there might not be enough energy in that
region to justify "hard" filtering.


Just out of curiosity... Does anyone here know why the earliest RCA CD
transfers sometimes sounded Just Plain Awful? I remember comparing a
Horowitz CD with its "HP" remastering. The latter not only sounded like a
different recording, but a different performance!

[Scott Dorsey]

William Sommerwerck > wrote:
>
>I have a Sony 87ES -- presumably a "prosumer" machine -- that records at 48,
>44.1, and 32. I sometimes use 32 to record FM broadcasts, simply because you
>get 4 hours of recording with no practical loss of frequency response.
>
>If you were going to release recordings at 48kHz, you could probably get
>away with no low-pass filtering _at all_, because signals would have to be
>above 28kHz to "fold" back into the audio band (and I doubt any mics get/got
>out that far). But, again, once you resample to 44.1kHz, you'd better be
>sure there's no significant program material above 22kHz.
>
>I have a sneaking suspicion that, for most acoustic music recorded on an
>analog deck, the low-pass filter itself probably does more sonic harm than
>any small amount of over-22kHz components that wind up in the audible
>passband.

On the 87ES that might be the case. For a really weird example, though,
try one of the decks like the SV3700 which use the same filter constants
for both sample rates. You'll find that even with source material that
doesn't have a lot of high end (ie. no harpsichord) that there's a radical
difference in sound between the two rates. This is due to changes in aliasing
and it shows how severe a problem the aliasing really is.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[[email protected]]

Arny Krueger > wrote:
: If memory serves, the pro machines supported analog recotrding at both 44
: and 48, but the consumer machines could only record analog at 48.

The Tascam DA-P1 (which was basically a pro machine) and the DA-20
(which was basically a consumer machine) both supported 48.

[Scott Dorsey]

William Sommerwerck > wrote:
>
>Just out of curiosity... Does anyone here know why the earliest RCA CD
>transfers sometimes sounded Just Plain Awful? I remember comparing a
>Horowitz CD with its "HP" remastering. The latter not only sounded like a
>different recording, but a different performance!

Lots of different reasons. While I tend to blame the PCM-1610 and PCM-1630
for much of what was wrong with early CDs, a lot of it also had to do with
the fact that labels were in such a rush to get their back catalogue out onto
CD that they weren't really paying attention to what they are doing... hell,
there is a performance of Ravel's Bolero out there that was released from a
quarter-track 7.5ips reference dub, and it's clear they didn't even bother
putting it on a quarter-track machine. Much of these were done by kids who
had no experience, from tapes that were many generations down from the
originals, too.

There were some shops where they "always" used certain EQ constants in the
mastering chain and when they moved from LP to CD nobody thought it was
important to rethink any of that.

Consequently... when you got people who don't care what they are doing and
people who don't know what they are doing, you get bad sound. And the
PCM-1610 sure didn't help.
--scott

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

[William Sommerwerck]

>> I have a sneaking suspicion that, for most acoustic music recorded on an
>> analog deck, the low-pass filter itself probably does more sonic harm
than
>> any small amount of over-22kHz components that wind up in the audible
>> passband.

> On the 87ES that might be the case. For a really weird example, though,
> try one of the decks like the SV3700 which use the same filter constants
> for both sample rates. You'll find that even with source material that
> doesn't have a lot of high end (ie. no harpsichord) that there's a radical
> difference in sound between the two rates. This is due to changes in
> aliasing and it shows how severe a problem the aliasing really is.

How do you know what the source of the difference in sound is? One would
assume aliasing would be quite irritating sounding, as the aliased
components are anharmonic.

[Scott Dorsey]

William Sommerwerck > wrote:
>>> I have a sneaking suspicion that, for most acoustic music recorded on an
>>> analog deck, the low-pass filter itself probably does more sonic harm
>than
>>> any small amount of over-22kHz components that wind up in the audible
>>> passband.
>
>> On the 87ES that might be the case. For a really weird example, though,
>> try one of the decks like the SV3700 which use the same filter constants
>> for both sample rates. You'll find that even with source material that
>> doesn't have a lot of high end (ie. no harpsichord) that there's a radical
>> difference in sound between the two rates. This is due to changes in
>> aliasing and it shows how severe a problem the aliasing really is.
>
>How do you know what the source of the difference in sound is? One would
>assume aliasing would be quite irritating sounding, as the aliased
>components are anharmonic.

Because nothing changes other than the clock frequency, everything else
remains the same.

The aliasing products are at very low levels, but they are still very annoying.

In generally the SV3700 is a textbook example of how not to design digital
audio interfaces.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[PStamler]

On Aug 3, 9:41*am, "William Sommerwerck" >
wrote:

> If you were going to release recordings at 48kHz, you could probably get
> away with no low-pass filtering _at all_, because signals would have to be
> above 28kHz to "fold" back into the audio band (and I doubt any mics get/got
> out that far).

I'm afraid you're wrong about that. Hang a good condenser mic in front
of, say, a bluegrass banjo, or a mandolin, or a drumkit, and record at
96k, then look at the result with a spectrum analyzer (the one in
Audition will do fine). There's plenty of signal up to the Nyquist
frequency. The widest-band mics will probably be transformerless, but
I recently had the occasion to measure the electronics of a
(transformer-coupled) Neumann KM 84, and they're flat out to about
100kHz.

Peace,
Paul

[Arny Krueger]

"Soundhaspriority" > wrote in message

> > wrote in message
> ...
>> Arny Krueger > wrote:
>>> If memory serves, the pro machines supported analog
>>> recotrding at both 44 and 48, but the consumer machines
>>> could only record analog at 48.
>>
>> The Tascam DA-P1 (which was basically a pro machine) and
>> the DA-20 (which was basically a consumer machine) both
>> supported 48.
> They all supported recording at 48. As Arny remarks, many
> consumer machines omitted analog 44 support.

This resulted in some *interesting* tapes that flipped between 44 (recorded
via the digital input) and 48 (recorded via the analog input) on a
track-by-track basis.

[nebulax]

On Aug 3, 10:41*am, "William Sommerwerck" >
wrote:
> > If memory serves, the pro machines supported analog recording at
> > both 44.1 and 48, but the consumer machines could only record
> > analog at 48.
>
> I have a Sony 87ES -- presumably a "prosumer" machine -- that records at 48,
> 44.1, and 32. I sometimes use 32 to record FM broadcasts, simply because you
> get 4 hours of recording with no practical loss of frequency response.
>
> If you were going to release recordings at 48kHz, you could probably get
> away with no low-pass filtering _at all_, because signals would have to be
> above 28kHz to "fold" back into the audio band (and I doubt any mics get/got
> out that far). But, again, once you resample to 44.1kHz, you'd better be
> sure there's no significant program material above 22kHz.
>
> I have a sneaking suspicion that, for most acoustic music recorded on an
> analog deck, the low-pass filter itself probably does more sonic harm than
> any small amount of over-22kHz components that wind up in the audible
> passband.


I think you're right. The main advantage I've heard attributed to wide-
band electronics (i.e.- pre-amps that can pass signal out to 120k or
more) is the absence of filtering, and whatever effects those filters
have on the signals down the freq spectrum. Plus, if a pre-amp can
reproduce 120k, then 22k doesn't even make it breathe hard. I've used
an Avalon 737sp channel strip with a sweep eq that allows you to make
a boost up to to 32k, and even with it set at that extreme, it seems
to add more 'air' to a vocal track. So, though you might not
technically be able to hear above 22k, you can still tell the effects
those freqs are having on the rest of the signal.

-Neb

[jazzman31[_2_]]

On Aug 4, 3:16*pm, nebulax > wrote:
> On Aug 3, 10:41*am, "William Sommerwerck" >
> wrote:
>
>
>
> > > If memory serves, the pro machines supported analog recording at
> > > both 44.1 and 48, but the consumer machines could only record
> > > analog at 48.
>
> > I have a Sony 87ES -- presumably a "prosumer" machine -- that records at 48,
> > 44.1, and 32. I sometimes use 32 to record FM broadcasts, simply because you
> > get 4 hours of recording with no practical loss of frequency response.
>
> > If you were going to release recordings at 48kHz, you could probably get
> > away with no low-pass filtering _at all_, because signals would have to be
> > above 28kHz to "fold" back into the audio band (and I doubt any mics get/got
> > out that far). But, again, once you resample to 44.1kHz, you'd better be
> > sure there's no significant program material above 22kHz.
>
> > I have a sneaking suspicion that, for most acoustic music recorded on an
> > analog deck, the low-pass filter itself probably does more sonic harm than
> > any small amount of over-22kHz components that wind up in the audible
> > passband.
>
> I think you're right. The main advantage I've heard attributed to wide-
> band electronics (i.e.- pre-amps that can pass signal out to 120k or
> more) is the absence of filtering, and whatever effects those filters
> have on the signals down the freq spectrum. Plus, if a pre-amp can
> reproduce 120k, then 22k doesn't even make it breathe hard. I've used
> an Avalon 737sp channel strip with a sweep eq that allows you to make
> a boost up to to 32k, and even with it set at that extreme, it seems
> to add more 'air' to a vocal track. So, though you might not
> technically be able to hear above 22k, you can still tell the effects
> those freqs are having on the rest of the signal.
>
> -Neb

Makes a lot of sense and likely is the reason why vinyl (and tape) can
sound "better" to some i.e. a gradual rolloff of frequencies above 20k

Rick

[Robert Orban]

In article >, says...
>
>
>Harry Lavo > wrote:
>>"William Sommerwerck" > wrote in message
...
>>> Well, to start with...
>>>
>>> The original RCA "Living Stereos" were spaced omnis, I believe.
>
>Spaced triads, yes.

There was a recent discussion of this on the Ampex mailing list. I
had been under the mistaken impression that both Living Stereo and Living
Presence used three spaced omnis but was corrected. This was true for
Living Presence, but it turns out the Living Stereo often used more than
three mics. There was a link to a fascinating film made by RCA ca. 1957
about the entire record production process from original recording to
vinyl in listeners' homes. (Unofrtunately, I don't have the link handy.)
The film showed more than three mics in use to record the Boston Symphony
in what was at that time very early stereo. The master recorders appearing
in the film were Ampex 300s and it appeared that there were two machines,
both quarter-inch, so this was evidently before the era when recorded
their masters on three-track 1/2" tape.

[Scott Dorsey]

Robert Orban > wrote:
>There was a recent discussion of this on the Ampex mailing list. I
>had been under the mistaken impression that both Living Stereo and Living
>Presence used three spaced omnis but was corrected. This was true for
>Living Presence, but it turns out the Living Stereo often used more than
>three mics. There was a link to a fascinating film made by RCA ca. 1957
>about the entire record production process from original recording to
>vinyl in listeners' homes. (Unofrtunately, I don't have the link handy.)
>The film showed more than three mics in use to record the Boston Symphony
>in what was at that time very early stereo. The master recorders appearing
>in the film were Ampex 300s and it appeared that there were two machines,
>both quarter-inch, so this was evidently before the era when recorded
>their masters on three-track 1/2" tape.

A lot of those recordings were spaced triads with spot mikes added here
and there, this being the real world and so forth. I'd still call those
spaced triad recordings.

Or are you talking about a longer array of omnis, like a line of five?
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Robert Orban[_2_]]

In article >, says...
>
>
>Robert Orban > wrote:
>>There was a recent discussion of this on the Ampex mailing list. I
>>had been under the mistaken impression that both Living Stereo and Living
>>Presence used three spaced omnis but was corrected. This was true for
>>Living Presence, but it turns out the Living Stereo often used more than
>>three mics. There was a link to a fascinating film made by RCA ca. 1957
>>about the entire record production process from original recording to
>>vinyl in listeners' homes. (Unofrtunately, I don't have the link handy.)
>>The film showed more than three mics in use to record the Boston Symphony
>>in what was at that time very early stereo. The master recorders appearing
>>in the film were Ampex 300s and it appeared that there were two machines,
>>both quarter-inch, so this was evidently before the era when recorded
>>their masters on three-track 1/2" tape.
>
>A lot of those recordings were spaced triads with spot mikes added here
>and there, this being the real world and so forth. I'd still call those
>spaced triad recordings.
>
>Or are you talking about a longer array of omnis, like a line of five?

As far as I know, you are correct about "spaced triads plus spot mics."
However, my understanding of the essential difference between Living Presence
and Living Stereo is that Living Presence used *only* three mics. Both RCA
and Mercury produced some excellent sounding recordings in that era, so
it seems that the lesson to be learned is that the number of mics is less
important than the artistry of the balance engineers who placed the mics and
(for Living Stereo)who determined the balance when mixing the spot mics into
the mains.

The film implies that RCA had sufficient budget to rehearse with the
full orchestra while settling on final mic placement and balance...and this
with union musicians in Chicago, Boston, and elsewhere.

It's actually not surprising. In the decade before Living Stereo, RCA created
a bona-fide mass culture superstar with the NBC Symphony and Toscanini (who
unfortunately died just a few years before the Living Stereo era) and
classical music had a mass appeal that seems unbelievable by today's
standards. (Think the Bell Telephone Hour, Amahl and the Night Visitors,
Leonard Bernstein's Young Peoples' Concerts, etc.) The Bernstein series of
course was on CBS, and Columbia (not Mercury) was really RCA's arch-rival
during the Living Stereo era, despite Columbia's reputation for heavy-handed
multi-micing and overcooked treble.