Re: GG: Shuffling

[frank@clark.net]

On Thu, 31 Oct 1996, Daniel LaRusso wrote:
> On Tue, 29 Oct 1996, Scott Henderson wrote:
> > >>My CD copy of the 81 Goldbergs has only one track, containing all
> > >>the variations and both arias.  Where did you find your copy with 34
> > >>tracks?
> > >
> > >I would ineed like to know what kind of CD copy YOU have of the Gouldbergs!
> > >One track!!! I have the issue that comes with the Glenn Gould Edition on
> > >Sony. The CD number is Sony SMK 52619, and it has 32 tracks. Are there
> > >really CD's with no indexes for such an indexed work as the Goldberg
> > >Variations? Please, details!
> > 
> >         I have the same one-track version of the '81 Goldberg Variations.
> > When CBS Masterworks first released it in 1982 (MK37779) the aria and
> > variations were lumped together in a single track.  It's a nuisance, when
> > you want to review the performance of a single variation, to have to
> > fast-forward through several minutes of music.  When Sony reissued the
> > recording as part of the Glenn Gould Edition (SMK 52619), they wisely
> > partitioned it out onto 32 tracks.
> > 
> *I* have the '81 Goldbergs on a "Masterworks" CD (ie not the "edition"),
> and it has all 32 tracks. Hmmm.

I have the one track CD. The thing to look for is what in the 78 rpm days
was called the matrix number. This was a number assigned at the
recording studion to each *side* of a 78 rpm disc. Usually, the numbers
are in sequence with the music, but not always. (If the disc occupied an
odd number of sides, the last side was called a "filler." Most of the
time, the fillers were performed by the same artists recording the major
composition, but the can and have been anything.)

In those day before tape recording, an entire side was recorded at a time
(4 1/2 minutes max. usually for a 12" disc, 3 min. for a 10"). If the
artists were unsatisfied with their playing, they would play that side
over again. These different recordings are called "takes." Thus WAX 2321-3
is a Columbia matrix: W for Western Electric, whose system Columbia used
to make electrical recordings, A for England, X for 12" (nothing
for 10"). 2321 is the number and 3 is the third take. (This turns out to
be the 11th side of the first recording of the Tchaikovsky piano trio
(WIlliam Murdoch, piano, Arthur Catterall, violin, W.H. Squire, cello, a
great performance that has so far resisted reissue on CD. Actually, I
don't know what the take number was.)

The matrix and take numbers are located in the runoff groves of the disc,
sometimes in separate places. This informatio is invaluable to
discographers. Be warned, though, that there can be *mistakes* even so!

This method was retained in the LP era, but with a difference. Each side
still has a matrix number, but the other number does not represent a take
(a different performance) but a different pressing (called the "stamper"
number) of the same performance. But the sound can be re-enginered by
adding reverberation or whatever or taking previous manipulations out.
Many collectors feel that the first stampers of early stereo issues sound
the best and, as we all know, are willing to pay fantastic prices for
certain records on certain labels, esp. Victor and Mercury. Columbia and
DGG, however, are rarely accorded this "audiophile" status.

When a performance is reissued on a budget label, the original matrix
number is usually not retained.

Now, *MY* 1981 Gouldberg CD has DIDC 50035 on what would be a run-*in*
groove, since CDs play inside-out and also 11A10. Don't take the disc out
of the jewel box. The DIDC number is also below the catalog number on the
label, and usually is for LPs.

So, please go look at your 32-track issue and report these numbers back to
us.

To add to this confusion, there are at least 200 cases that *I*
know about of
a 78 rpm re-recording by the same, or even different, performers NOT to be
given a new catalog number. The only LP example that comes to mind is that
Vox issued the piano music of Mendelssohn, which included the concerti, on
four Vox boxes. Later it reissued the solo music alone with the catalog
numbers of the first three. Since box no. 1 was the same in both cases,
this means that three boxes were consolidated into two. 

I forgot to mention that the artists would not always make up their minds
on the spot but would listen and decide later. Quite often they would go
back to the studios to re-record certain sides. The studio logs are where
to go for the details, and *they* have mistakes, too. Company *catalogs*
also have errors, though this sounds like a contradiction in terms, since
the catalog number ought to be the number in the catalog!

Experience is the only way to become a craftsman discographer; there are
no general rules to resolve difficulties. By the way, the great pianist
Alfred Cortot recorded Liszt's La Campanella 21 times and was satisfied
with none of them. What's also interesting is that the same arabic
numeral would be retained into the electric era. Thus Chopin's Berceuse
has these takes and dates 

C 22502-1  1919.1.7
       -2  1919.1.8
       -3    "
       -4  1919.1.11
       -5  1920.1.27. Issued on single-sided Victor 74623 and
                      one side of double-sided Victor 6063 (the other 
                      side being S-S: Etude en forme de Valse)
       -6  1923.2.27
       -7  1923.2.27  Issued under the same catalog numbers, but with
                      the same issue of the S-S preserved. Cortot
                      made three re-recordings that year but was
                      satisfied by none.

CVE 22502-8  1926.10.26 These are electrics now.
         -9     "
         -10    "
         -11 1926.12.27
         -12    "
         -13 1926.12.28
         -14    "       Issued on double-sided 6752. Coupling is
                        an electric of Handel's H.B.Smith Variations
         -15

My thanks to Jim Cartwright, Immortal Performances, Box 8316, Austin, TX
78712, who put this tremendous amount of research inside one of his 78 rpm
catalogs. Alas, I caught five typos in it, and there may be others. NO
discography (this one is of Cortot's recordings in America, now reissued
on a two Biddulph CD set. Gould fans should buy it, for Cortot was very
distinctive, too.) is without errors, including my own.

I'm cross-posting this to 78-L, to see if I have portrayed these matters
of matrices and takes absolutely correctly. 

To members of 78-L: please keep the two Gould lists in any replies on this
thread.

Frank