The answer, as expected, is quite complicated.  Many
Hi Roger,

The answer, as expected, is quite complicated.  Many RCA pre-war metal parts survive, far more than HMV parts do, for example.  In the case of Stokowski and Toscanini, it depends on the specific recording.  Sometimes, the parts are worn out and highly distorted, as in the case of the 1947 Toscanini Tchaikovsky Pathetique Symphony.  My understanding is that at Columbia, classical parts were preserved while many pop parts were not.  Dennis Rooney would know more.

In the case of the Toscanini tapes, some tapes were stolen many years ago and some were scrapped many years ago.  Sometimes (especially with Toscanini), it's simply not clear what the edited master tape is.  
RCA's nomenclature changed every few years in the tape era, which means figuring out their naming conventions and numbering systems can be appallingly difficult.  Reverb was often added to records in the 1950s (and later), especially on subsequent re-releases of earlier material.  I'm not talking about the Toscanini electronic stereo Victrolas, which were an abomination (although a few of the mono only releases were decent.)  

I think the reason paperwork was lost was primarily due to frequent moves.  That, coupled with little to no institutional memory, plus limited space considerations meant that each new generation trashed their predecessors' materials.  In terms of audio preservation and storage, by most companies standards, RCA was actually quite good.  Still, the original tape boxes were often trashed in favor or unlabeled (except for the tape number) metal cans.  The difficulties go on and on...

Jon Samuels

On Thursday, March 27, 2014 10:57 AM, Roger Kulp <[log in to unmask]> wrote:
A few questions for Jon
Can one assume most of the metal parts for the Toscanini and 78/Stokowski era Philadelphia Orchestra were casualties of wartime scrap drives?
By added reverb do you mean when RCA made tapes for the electronic stereo reissues of the 60s and early 70s? Were they so short sighted that they did not keep earlier generations of the tapes?
Why were the original  session notes,marked scores,etc not saved?
The only thing I can think of was that RCA,like so many labels,did not believe the recordings made before three track were important enough to save,either from a cultural or a historical standpoint,even for an artist of the stature of a Toscanini or a Horowitz.

> Date: Thu, 27 Mar 2014 06:45:12 -0700
> From: [log in to unmask]
> Subject: Re: [ARSCLIST] Reissue sets.
> To: [log in to unmask]
> Dave, thanks for the kudos. 
> I can answer most
> Hi Dave and Don,
> Dave, thanks for the kudos. 
> I can answer most of your Toscanini questions.  At Sony, most of these large sets are copied (and re-ordered) from the previous large set.  However, if the compilers are aware of a later transfer, that's what they used.  So most of the recent set is copied from the early 1990s set, which is extremely variable in
 quality.  FYI, Dave, the 1941 Tchaikovsky Concerto with Horowitz was redone by me about 2004.  One reason it previously sounded so poorly is that some of the 78 sides were dubbed.  I found the undubbed parts, and used them.  I'm not certain that the Toscanini set includes my later transfer, but the 70 CD Horowitz box from 2009 (which also has extremely variable sound) definitely does.  I believe you'll find it a substantial improvement over what you remember.
> The Philadelphia recordings on the early 1990s CD are just a copy of the LP tape run through CEDAR.  The one in the recent set should be the 2006 transfers, which are substantially better.  Some people have been unhappy with the newer transfers, because of the use of different takes from the previous incarnation, and some problematic side-joins, but I think there is universal agreement that they sound better.
> It's very complicated why Toscanini recordings sound so poorly in general.  For one thing, the surviving tapes are often multiple generations down, frequently with added reverb.  For another, Toscanini, often requested changes upon changes, and the original engineers were often required to edit dubs in order to affect those changes.  FYI, at RCA, restoring older recordings is not always as simple as finding the edited session tapes and using them.  It isn't always clear which tapes those are, and even if you find them, it's often not clear if it's the same performance as the final edited master (they might have made changes to a later generation tape).  And at RCA, there are NO surviving edit plans or marked scores, and often the unedited session tapes (RCA ran two machines for each session) no longer survive so you couldn't re-edit it even if you wanted to.  By the three-track era,
 the unedited session tapes often survive, but not before.
>   Another problem is that on the edited workparts, the spicing tape often "bleeds" through, which sometimes leaves a residue that needs to be cleaned off, and even sometimes is corrosive and eats away at the oxide layer, leaving gaps in the tape.  Sometimes, edited workparts exist for one side (or even part of one side), but not the other.  In the case of the Munch Beethoven 9th for example, a three-track workpart exists only for the first LP side, while the rest only exists in a two-track mixdown which is off-pitch and heavily compressed.  Trying to match the sound between the two is a nightmare.  The older tapes were often stored improperly, and are no longer flat, etc. - you get the idea.  The point is there is sometimes a reason for variable sound.  One of the best sounding transfers I ever did, all I had to do was choose a playback
 curve put the two-track tape on the ATR, and then play and record it.  That's basically it.  I've also had the
>  reverse situation, where I've slaved for weeks over a recording, that sounds like absolute crap when I'm done.  (But improved absolute crap, I must admit.)  The point is, I guess, don't (always) judge a book by it's cover.
> Jon Samuels
> On Thursday, March 27, 2014 7:38 AM, Don Cox <[log in to unmask]> wrote:
> On 27/03/2014, DAVID BURNHAM wrote:
> > As I've mentioned here before, I've recently purchased the Westminster
> > box set, the Mercury Vol. 1 & 2, (but I haven't gotten 2 yet), and
> >
 most interesting, the latest incarnation of the complete Toscanini.
> > I'm amazed by the range of qualities of remastering in the Toscanini
> > set. Some are dreadful, like the late '40s early '50s versions of the
> > Beethoven Symphonies. These are painfully over-modulated with the
> > tympanies regularly going into distortion, (I didn't recognize the
> > names of the remastering engineers on these ones). One advantage of
> > this set is that it includes the recordings by the Philadelphia
> > Orchestra and the BBC Orchestra which weren't included in the '90s
> > set. The Philadelphia recordings are a marvel of recorded sound, high
> > fidelity even by today's standards, considering that these recordings
> > were made at roughly the
 same time as the wretched sounding Horowitz
> > Tchaikovsky concerto which was such a popular album; I've never heard
> > a
> > definitive version of why these were rejected for issue when they
> > were recorded. The first story I heard was that the recordings were
> > defective and unusable. This seems absurd since RCA had been recording
> > this orchestra in this hall probably more than they had recorded any
> > other orchestra - they should have been able to record them in their
> > sleep! The Mendelssohn "Midsummer Night's Dream" would have been a
> > particularly useful set to have on the market since nobody had
> > recorded this much of the score before. Also it's strange that, if
> > these records weren't to
 be issued, that Toscanini didn't re-record
> > the material with the NBC until several years later. 
> I have the complete 1990s edition (in its special case), and it does
> include all the Philadelphia recordings, on four CDs.
> My impression is that the problem with these recordings was that the 78
> rpm discs had very large numbers of clicks, because substandard wartime
> materials were used. The 1990 transfers do have very clicky backgrounds.
> I have the 2006 transfers in a 3-CD set. These are DSD transfers. The
> clicks are all gone and the string sound is much improved.
> Presumably the 2006 transfers are the ones used for the current complete
> edition.
> Regards
> -- 
> Don Cox
> [log in to unmask]