Hi John:

I'm not sure what era Capitol you are referring to.

I think the very early Capitol stereo, the first round of session recorded in Pittsburgh, may have 
been done on a 2-track machine with 2 mics. Soon afterward, Irv Joel and another engineer (did a lot 
of jazz recording at Capitol NYC but I can't remember his name right now) travelled around in a 
little "micro-van" with a "portable" Ampex 300 3-track. I don't think they used more than 3 mics, at 
least early on. The results of those sessions would have probably been mixed down to a 2-track 
disk-cutting master.

I don't know how Capitol's Hollywood recordings were done, but I assume 3-track with multiple mics, 
mixed to a 2-track cutting master. I would guess the same about Houston.

I do know some facts about the Capitol/EMI/Angel recordings made by Carson Taylor in the late 60s 
and early 70s, in Chicago and Cleveland. Taylor used a 4-channel (8-track) 3M Dynatrack recorder at 
the sessions, and about a dozen microphones (counting his coincident stereo mics as 2 mics). Back at 
the Capitol Tower, he mixed his session tapes to 2 channels and dubbed them to 2-track NAB, and then 
edited the 2-tracks into the master. Original USA LPs were cut from these tapes. A former Capitol 
mastering engineer has told me that Taylor sometimes or always edited two master tapes, and one was 
sent to England for the original-issue EMI LPs. Somewhere along the line, dubs were made, and it's 
apparent that the EMI reissue CDs and SACDs were made from these dubs. John Marks, a writer for 
Stereophile, has discovered that at least two of the Cleveland recordings were dubbed so as to end 
up with fast/sharp pitch on the EMI CDs and SACDs. It's unclear how or why this happened. John did 
spectrum analysis on the CDs/SACDs and on dubs of original LPs that he and I had. The original LPs 
are perfect A=440, whereas all EMI CD and SACD resissues we could find are A=445 to 446. These are 
the Oistrakh recordings with Cleveland/Szell. We tested the Gilels-Szell recordings, made by 
Columbia engineers, and they are dead-on A=440. We also tested the CD and HDTracks downloads of 
later Szell/Cleveland recordings made by Carson Taylor and they are A=440. And we tested 
Chicago/Giulini recordings made by Taylor, and they are A=440. So something went wrong in dubbing 
the master for those particular Cleveland-Szell recordings, and it's not clear yet where or why the 
error occured. John Marks is still digging.

-- Tom Fine

----- Original Message ----- 
From: "John Haley" <[log in to unmask]>
To: <[log in to unmask]>
Sent: Wednesday, March 26, 2014 6:28 AM
Subject: Re: [ARSCLIST] Fwd: [ARSCLIST] "Why Vinyl Is the Only Worthwhile Way to Own Music"

> RE Jon Samuel's eye-opening post about multiple generations tape dubs at
> RCA, "aging" the product every time reissues were done, even into the CD
> era (pretty horrible news, actually), does anyone one know if this was also
> the practice with EMI/Capitol?  I have recently been comparing some Capitol
> classical LP's reissued over time, and I would swear that the earlier ones
> have more "life" in them.
> Back to RCA, I guess that means that the Victrola classical LP's that so
> many collectors seem to have prized (at least in the past), were made from
> later generation tape copies too.
> Best,
> John Haley
> On Wed, Mar 26, 2014 at 6:17 AM, Tom Fine <[log in to unmask]>wrote:
>> As far as I know, the only RCA Living Stereo issues that are all from
>> first-generation tapes are the BMG SACD/CD discs. Jon Samuels and Mark
>> Donahue will correct me if I'm wrong, but my understanding is that all of
>> the 3-track session tapes were mixed to 2-tracks, which were then edited
>> into "master" tapes. And, a third generation dub was often used to cut LPs.
>> For the earliest stereo recordings, the ones made on the RCA 2-track
>> machine at 30IPS, at least the earliest stereo LPs were cut from
>> first-generation tapes. Later LPs may have been cut from dubs (they must
>> have been, because the first generation tapes were still in good playing
>> condition 50 years later).
>> RCA began using 3-track session recorders circa 1956. So the 1954 and 1955
>> stereo recordings were all or almost all 2-track and the first-generation
>> LPs were almost all cut from first-generation tapes.
>> It is also my understanding that Columbia was late to stereo, but jumped
>> right in with 3-track, so all of their stereo LPs were cut from
>> second-generation or later 2-track tapes.
>> Back in the day, people either couldn't hear the sound degradation from
>> each generation of tape dubbing, or refused to acknowledge it, or felt it
>> was minimal and harmless compared to disk-to-disk or disk-to-tape dubbing
>> they had probably done earlier in their careers.
>> -- Tom Fine
>> ----- Original Message ----- From: "Carl Pultz" <[log in to unmask]>
>> To: <[log in to unmask]>
>> Sent: Wednesday, March 26, 2014 5:48 AM
>> Subject: Re: [ARSCLIST] Fwd: [ARSCLIST] "Why Vinyl Is the Only Worthwhile
>> Way to Own Music"
>>  OMG. The horror is finally revealed. Did this also apply to the .5
>>> "audiophile" LP series? They sounded lousy to me.
>>> -----Original Message-----
>>> From: Association for Recorded Sound Discussion List
>>> [mailto:[log in to unmask]] On Behalf Of Jon Samuels
>>> Sent: Tuesday, March 25, 2014 11:41 AM
>>> To: [log in to unmask]
>>> Subject: Re: [ARSCLIST] Fwd: [ARSCLIST] "Why Vinyl Is the Only Worthwhile
>>> Way to Own Music"
>>> Early CD transfers had another problem. At RCA (and, from what I
>>> understand, to differeing degrees at other record companies), edited reels
>>> from the recording sessions were never used in the early CD days. The main
>>> reason had to do with bookkeeping (and Jack Pfeiffer's belief that no one
>>> could hear the difference, and therefore was not worth the trouble to
>>> track
>>> down, find and physically restore the edited masters). From it's earliest
>>> LP days, RCA maintained a system where every LP side (and later every CD)
>>> had to have its' own tape. That meant that if an LP was re-issued with a
>>> different number, the later LP master would at best be a first-generation
>>> dub of the previous LP. Unfortunately, they took this a couple of steps
>>> further. Three-track (and higher) masters were always mixed down to
>>> 2-track
>>> for the LP. (One of the reasons this was done was to deliberately reduce
>>> the dynamic range in the LP master before the cutting stage.) Each new
>>> re-issue's tape master was a dub of the most recently released LP tape
>>> master, not the original one. They continued this practice with early CDs.
>>> Also, in later LP years, they often dubbed early 30 ips tape masters to 15
>>> ips, and used those for later LPs with the same issue numbers. The
>>> consequences of these factors was that early CD masters were sometimes as
>>> much as seven or eight generations down from the original session tapes.
>>> (It also explains why collectors often prefer earlier LP issues.)
>>> The first RCA CDs that used the edited session tapes (called workparts in
>>> RCA parlance) rather than dubs as their source material were the Artur
>>> (now
>>> Arthur) Rubinstein CD series released in 1984, which was produced on CD by
>>> Max Wilcox. That didn't become pretty much standard practice around
>>> until around 1988/9 (and even then, not in every instance).
>>> This doesn't even allow for the improvement in the quality of digital gear
>>> over the past thirty years (a subject written about here many times).
>>> Best,
>>> Jon Samuels