Mravinsky in Prague ... or is he?

The sober world of classical music recording is perhaps the last place in which one expects to find subterfuge or intrigue. Thus, readers of the review section in our last issue were probably as shocked as I to discover that one of Le Chant du Monde's (LCdM) reissues in their Shostakovich 25th Anniversary Edition was a cleverly disguised impostor.

For anyone who missed our January edition, the CD in question presents a recording of Shostakovich's Fourth Symphony from Gennadi Rozhdestvensky and the USSR State Symphony Orchestra. This first appeared in 1995 on the label Praga, a subsidiary of LCdM. Praga proclaimed the recording to be a Czech Radio broadcast from 1985. While reviewing the reissue in December, I was struck by suspicious similarities to the same performers' 1985 Moscow studio recording, currently available on a BMG/Melodiya twofer (74321 63462 2). Careful side-by-side comparison revealed that Praga's version is in fact the very same studio recording … but, as hard as this is to believe, with an audience's coughing and applause superimposed throughout!

It turns out that this is only the tip of a massive iceberg of misattribution, the full dimensions of which remain to be mapped. Indeed, I am chagrined to have ignored the alarm bell rung by Kenzo Amoh and Frank Forman in their painstakingly researched Mravinsky discography[1]. There, Amoh and Forman report that Praga's releases of Yevgeny Mravinsky conducting Shostakovich's Fifth, Sixth, Eleventh and Twelfth Symphonies[2] are all identical to previously issued recordings, none of which are Prague concert relays as Praga claim.

Regrettably, I discounted Amoh and Forman's findings in my reviews in DSCH No. 13 of Praga's Eleventh and Twelfth Symphony pressings, reissued in their Evgeny Mravinsky in Prague boxed set. Misinterpretation was the culprit. The discography argues that Praga's ostensibly live versions of these two symphonies are actually studio recordings because in each case, "the timings of the individual movements are the same, and, furthermore, [each] sounds like a studio recording"[1]. At the time, neither of those clues struck me as valid. First, the track timings of Praga's Eleventh differed from the timings on the 1959 Moscow studio recording that Amoh and Forman identified as its source (I did not then have the studio Twelfth identified in the discography to make the corresponding comparison). Second, how could one claim that the two Praga recordings sounded like studio productions with their conspicuous applause and other audience noises? On the strength of these data, I asserted, "I am confident that Praga are indeed supplying what they advertise."

Discovery of the misattributed Rozhdestvensky Fourth made me reassess Amoh and Forman's charges. My primary method for determining if two recordings were identical was to play both simultaneously in separate players, monitoring each recording through one headphone earpiece. When the recordings are the same, they remain in synchrony, whereas different performances always drift apart rapidly, even in the hands of the same musicians. Even when the same recording has been transferred to different CDs at slightly different speeds, it is still relatively easy to identify that the resulting gradual drift is due to transfer speed variance and not performance differences. I also used matching placement of coughs, stage noises and performance errors in two recordings to render a diagnosis of common origin.

First, I compared Praga's Eleventh with Revelation's reissue of Mravinsky's 1959 studio recording. This revealed that the two are indeed the same, although transfer speeds differ sufficiently to yield different track timings. Kenzo Amoh has since explained to me that when comparing movement timings for two recordings he calculates the ratio of each movement's duration to the total duration, and if the ratios deviate less than 1% between recordings, he concludes that they are identical. This method corrects for transfer speed variance between recordings, and should be reliable so long as transfer speed remains consistent within each recording.

As for Praga's Eleventh sounding like a studio recording: yes, it does, aside from the applause at the end and other brief episodes of audience noise. Amoh and Forman had not suggested that such sounds might be grafts, and it never occurred to me previously that these might have been added to camouflage the studio recording as a distinct concert performance.

In hindsight, however, the doctoring is ham-fistedly transparent. Not only is a loud ovation tacked onto the end of the symphony, but also seven seconds of ambient concert-hall noise are inserted between the second and third movements. Note that the movements of this symphony are supposed to be played without pause, as presented in the studio original.

As if this were not enough, the gremlins add extra "evidence" that this is a live concert by superimposing more audience noise from 0:20 to 0:32 in track 3, resulting in abruptly different acoustics than in the surrounding passages.

In my review of Praga's original release of this Eleventh in DSCH No. 9, I noted the then-unexplained break between second and third movements. Unfortunately, the differences in transfer speed and acoustics between Praga's version and Revelation's tricked me into perceiving all manner of interpretive and performance differences between the two. I apologise to the readership for this embarrassing mistake.

I have since confirmed all of Amoh and Forman's reports of Praga misattributions save the case of the Mravinsky Fifth, for which I have not yet acquired a pressing of the 1978 Vienna concert that they list as its origin. I do not doubt that Amoh and Forman are correct here too, especially since Mr. Amoh's identification of homogeny is based not only on movement timings but also on identical placement of coughs[1].

It would be difficult to argue that these cases are other than intentional efforts to mislead the buying public. Most damning is the addition of applause and other audience noises to the studio recordings of Shostakovich's Fourth, Eleventh and Twelfth Symphonies. More subtle, but no less devious, is the listing of recording information for the Fifth and Sixth Symphonies that coincides perfectly with actual Mravinsky concert programmes in Prague on those dates[3].

Casting my net beyond symphonies, I hauled in additional forgeries. Rozhdestvensky's reading of Shostakovich's Scarlatti transcriptions, opus 17, on the same Praga CD as the Fourth Symphony, hails from Moscow, not Czech Radio. This is a straight copy, with no concertgoers added, of the original Russian recording currently available on the BMG/Melodiya twofer entitled Shostakovich Orchestral Works.

I have also confirmed suspicions I aired in our last issue that the 25th Anniversary Edition's recordings of the Taneyev Quartet sounded as if they were transferred from LP. Simultaneous-playback tests show that the Taneyevs' Fourth and Fifth Quartets masquerading on Praga as Czech Radio broadcasts are actually Melodiya releases. By the time this edition went to press, I had not yet secured the original Melodiya version of the Taneyevs' Seventh Quartet to compare with Praga's; however, given the latter's LP-surface noises, I would wager a large sum that this Seventh also comes from the same Melodiya set.

Further tests reveal that the 25th Anniversary Edition's performances of Shostakovich's Sixth and Fifteenth Quartets by the venerable Beethoven String Quartet were originally Russian recordings of the Ostankino State TV and Radio Company.

At present count, three of the four Evgeny Mravinsky in Prague volumes present performances in which Mravinsky was elsewhere. I have so far identified seven of the ten Praga discs in LCdM's 25th Anniversary Edition as containing bootleg recordings (the remaining five CDs in this 15-disc edition offer modern recordings from LCdM's subsidiary Saison Russe, to which no suspicions adhere). I predict that my continuing investigation will uncover further examples of mislabelling on Praga.

There do seem to be genuine articles among the fakes, however. I have tried but failed to identify alternate origins for Praga's versions of Shostakovich's Tenth Symphony (PR 7250053 in the 25th Anniversary Edition), Prokofiev's Sixth Symphony, and Bartok's Music for strings, percussion and celesta (PR 256019 and 256016, respectively, in Evgeny Mravinsky in Prague). All of these feature Mravinsky as conductor. Kenzo Amoh's Mravinsky concert listing[3] confirms that Mravinsky programmed each of these works in Prague concerts on the dates given by Praga (albeit dating the Prokofiev to 26 May rather than 25 May 1967 as stated on Praga's original release). To date, all simultaneous-playback tests I have conducted to compare these recordings with others of Mravinsky's have shown them to be distinct.

The same applies to Praga's submission of Shostakovich's First Violin Concerto (PR 256019 in Evgeny Mravinsky in Prague; PR 7250052 in the 25th Anniversary Edition), even though this struck me as especially fishy for three reasons. First, Praga credit this to David Oistrakh with Mravinsky leading the Czech Philharmonic Orchestra in a live recording from May 1957, but there is no applause afterwards, even though the track plays long enough after the final note for its reverb to die away fully. My experience is that this work often evokes an instant response from an audience. Second, an audience can indeed be heard throughout the rest of the work, but the volume of their input varies, increasing suddenly in the interval between the first and second movements. This doesn't sound as if the concertgoers are simply releasing en masse the coughs they were stifling during the first movement; the recording's balance changes, as if an audience soundtrack has been turned up. Third, the Music Library of the St. Petersburg (Leningrad) Philharmonic has no record of this concert[1].

Of course, the first two clues are circumstantial, and as for the third, many of Mravinsky's programmes are missing from the Music Library's collection[3], so the absence of one for the date in question does not prove that Praga's recording is inauthentic.

The Praga discs caught dealing illegitimate venue and date information have still credited the correct performers, so the obvious paternity test for Praga's Violin Concerto is the only other commercially available version of the work by Oistrakh with Mravinsky[1, 4], a 1956 studio recording recently reissued on RCA Red Seal's 2-CD set David Oistrakh: The Essential (74321 72914 2). There are, however, numerous differences in performance and engineering artefacts between the two that would have been difficult to manufacture in the editing booth. Oistrakh's recordings of the work with Mitropoulos and Rozhdestvensky do not mirror Praga's either.

LCdM appeared to be unaware of the counterfeits lurking in their Praga holdings when I informed them of these findings. Their director, Christian Girardin, assured me in March that they are investigating my allegations and that he will inform DSCH when the investigation is complete. The company had not supplied a statement for publication by the time we went to press.

As a result, I am still unable to specify when and how these misattributions entered the production queue. All of these recordings are released under exclusive licence from Czech Radio, so it is entirely possible that their mislabelling and plastic surgery occurred long before Praga acquired them. Regardless of the origins of this mess, it is now incumbent upon LCdM to mop up. Until then, buyers should tread on tiptoe.

Notes

Special thanks to Louis Blois for providing access to recordings in his collection for this investigation.

[1] Amoh, K. and F. Forman. 2000. Yevgeni Mravinsky Legacy: A Recording Listing 1938 - 1984. Japanese Mravinsky Society, Tokyo, Japan. To purchase, contact Kenzo Amoh, No. 107, 1-12-23, Den-enchufu, Ota-ku, Tokyo, 145-0071, Japan; Tel/Fax (03) 3721-3080; e-mail amohkenz@ce.mbn.or.jp. The 1999 edition, Legacy of Yevgeni Mravinsky: Discography, is accessible online at http://plaza19.mbn.or.jp/~yemravinsky/discography.htm.

[2] Amoh and Forman credit Daisuke Kisagami for first noting that Praga's version of the Fifth Symphony was identical to Mravinsky's 1978 Vienna concert, and Hiroshi Yasuda for first identifying the true origins of Praga's Sixth, Eleventh and Twelfth Symphonies. Kenzo Amoh confirmed their findings in all four cases.

[3] Amoh, K. 2000. Yevgeni Mravinsky: A Concert Listing 1930-1987. Japanese Mravinsky Society, Tokyo, Japan. Purchase details above[1].

[4] Geffen, P. 2000. David Fiodorovich Oistrakh: Discographie. Online at http://www.oistrakh.com/discographie_geffen.html. Last accessed 28 June 2001.

[5] Soviet-era recording data are notoriously unreliable. For Mravinsky performances, recording information comes from [1]. For all other recordings, information comes from the notes to the releases used for comparison, and should be regarded with caution.

[6] Releases identified with "[1]" were used by Kenzo Amoh. All other releases were those I used.

W. Mark Roberts
Top