KOUSSEVITZKY Beethoven: Symphony No. 9 (1947) - PASC300

Further notes on the RCA red vinyl 78rpm discs and their pros and cons
(from the Pristine Classical Newsletter - Editorial, 29 July 2011)

As I began writing this column last week I'd just started transferring the sixteen sides which make up Serge Koussevitzky's 1947 recording of Beethoven's Symphony No. 9. Recorded in Tanglewood in August of that year with the Boston Symphony Orchestra, four soloists and the Music Festival Chorus, the recording was one of a relative handful to be issued on the RCA Victor Red Seal De Luxe label.

It seems my mention of this label's clear red vinyl 78s brought back a lot of memories for readers of this column - and judging by what you wrote it sounds like they were something of a mixed blessing. Some people loved them, others were far less enthusiastic.

What I had, in this recording, was the rare opportunity to play sides which had never been played before. Even of those three or four sides which had, they had only received a single play using a modern, lightweight stylus, and would therefore have been just about as good as new.

The format was hailed by the RCA Victor marketing department in 1945 as the future of music reproduction and the biggest advance in 45 years in playback technology. It's an interesting claim, and one which I'd dispute: the almost overnight switch from acoustic to electric recording across the world in mid-1925 surely remains the biggest single leap forward in recording technology - everything else has (I would argue) been more about evolution than revolution.

Thus the new plastic discs of 1945 also coincided with the commercial development of much greater frequency range recordings than had previously been offered, something Decca in the UK made a big deal about with their ffrr recordings. Others chose not to repeat the experience of 1925 and say anything to suggest their back catalogue had been rendered obsolete overnight. So the RCA red vinyl discs were also able to offer a  greater frequency range than earlier offerings - but they weren't telling anyone this, and to be clear, the same was also true of their shellac discs.

However, shellac's days were numbered. Wartime use of vinyl-like discs for radio transmissions had probably done much to point the way forward, with the LP and 45rpm microgroove discs of 1948 onwards being the logical conclusion.

In 1945 the technology wasn't quite there yet, and the Koussevitzky 78s tell us as much about what was right about the new plastic records as was still yet to be mastered. The first few turns of all of the sides displayed something which LPs never really fully conquered: a heavy low-frequency rumble which gradually diminished as the side went on. But on these discs it was far more pronounced than anything you'd hear on an LP - or on a regular shellac 78 - and got worse as I compensated for the poor bass response of the recording in question.

Not to worry - it's not too difficult to filter out. I'm guessing the greater energy of the disc rotating at 78rpm only emphasises a problem common to other vinyl discs. On a number of sides it was accompanied by a regular thump, indicating the disc wasn't as flat as one would like - to the extent that the final side jumped ever so slightly out of the groove on each turn, something easily edited out, but which would probably not have happened at the more sedate speed of an LP!

So that's the bottom end, what about the top? Well, at its best it was clear, crisp and clean. In theory a record turning at this kind of speed should be much better able to handle higher frequencies than one running at 33rpm, and I would expect a modern 78rpm pressing from a high resolution source to deliver good sound at the top end. But here we're dealing with a 1947 recording and technology optimised for shellac. We get end of side distortion, a really noticeable treble roll-off (common to most 78s), and a lot of swish.

Swish is created at the time of cutting grooves in a record, and in a direct-to-disc system such as was used prior to magnetic tape, there's no chance to re-cut the disc in order to fix it. I'll be clear and say right now that I don't know precisely what causes it, but it bedevils 78s and early LPs and is exaggerated by digital noise reduction which normally misidentifies it as content and leaves it alone, only putting into greater sonic relief.

So it's no great surprise to find it here on a number of sides, leaving the slow, tedious job of manually selecting each swish and trying to bring it under control or eliminate it. with 68 minutes of music and a swish potentially 78 times per minute - well you do the maths.

Finally we come to sides 4 and 13. I'll quote now from notes in a Koussevitzky Discography published in the Spring 1990 issue of the ARSC (Association for Recorded Sound Collections) Journal:

"Due mainly to the extremely noisy surfaces typical of so many late postwar Victor pressings, the shellac 78s (DM-1190) make for hard listening. However, the red vinyl 78s (V.DV-12) came out fairly well. Sides 4 and 13 have a particularly distorted sound on both shellac and vinyl editions similar to other sides that are known to have been recut. Victor was capable of making fairly decent sounding recuts. Some of the bad ones, including the two sides at hand, sound as though the disc being copied was played with a stylus fouled with dirt."

What this suggests is that these two side, for some reason, had to be dubbed - a "recut" - from another disc master (or copy) before pressing masters could be made, usually as a result of some kind of problem in the mastering or manufacturing process. Both sides 4 and 13 were especially noisy and had a fuzzy distortion to them, well described here.

I was able to cut through the worst of this with various restoration tools, and the difference in sound quality is far less glaring in the finished transfer than it had been to begin with. What is confusing to me is that this problem seemed to have been avoided in the later microgroove issues of this recording, which makes one wonder how they got through quality control for the 78s.

One final thing to report on these discs, and one which would have been equally the case on the shellac editions. The Boston Symphony Orchestra, at least when replayed at the correct speed, appears to have been tuned to A=450Hz rather than the more common A=440. Were the Victor disc cutters really rotating at 78rpm, as this would indicate a speed of 76.25rpm for cutting if the orchestra was playing at concert pitch. Or did the BSO play sharp?

If the latter is the case then I'll have to apologise to pitch purists - I've made the roughly 2% correction required to repitch the recording to A=440Hz.(Wikipedia tells me that these days the BSO tunes to A=442...)

At this stage I feel like I've successfully trashed the recording to the extent that nobody will be interested in hearing it, so let me now redress the balance. Despite all the faults outlined above - most of which are common to 78s and LPs to one degree or another - the red vinyl discs did indeed deliver probably the best available sound quality at the time. It was a technology still in its infancy, and one which would be improved upon in the years to come, but they were very much on the right track - and the sound quality I've been able to extract from the records is remarkably refined and clear. Just about all the shortcomings have been within the capacity of modern remastering software to remedy, and XR remastering has dug deep into the recorded sound to bring out the previously recessed bass and make the whole sound more direct and immediate.

The ARSC notes comment:

"Victor obviously had a great deal of trouble recording anything sonically acceptable in the Music Shed. The sound has a distant quality, but it is possibly the result of putting the mikes back aways from the orchestra to capture ambience, of which a reasonable amount exists on this recording. A & R man R. Gilbert, would have probably achieved better results by putting his mikes closer to the orchestra, and forgetting about trying to make this recording sound as though it had been made in a fine concert hall."

I would suggest that the XR rebalancing of the tonal response of the recording actually achieves what the sound recording engineers were attempting to achieve - but have a listen to the first movement and judge for yourself. It doesn't sound "distant" to me - in fact I like it a lot!

I always like to find something special to mark numerical milestones in our catalogue if I can - and this record more than fits the bill for our 300th orchestral release. It's provided me with a fascinating technological diversion and with a rather nice recording to boot. Once again, a big thanks to Al Schlachtmeyer for the discs - another major contingent of which arrived here a couple of days ago and await careful unpacking over next few days. In the meantime I hope you'll enjoy the first fruit of this remarkable collection!