ELGAR, KOUSSEVITZKY, STOKOWSKI Accidental Stereo Recordings (1929-33) - PASC422

Amazing "accidental" stereo 78rpm recordings made by Elgar, Koussevitzky and Stokowski

New technologies and techniques produce astonishing stereo results from these unexpected early electric recordings

Shortly after the introduction of electrical recording, it became standard practice to make backups for wax matrices by simultaneously recording on a second cutting table, with the takes from the first being numbered 1, 2, etc. and the ones from the second 1A, 2A, etc. In the vast majority of cases, the two cutters were fed from the same microphone. Occasionally, however, each was run from its own microphone; and where those matching takes survive and have been released, it is possible to combine them to produce what has come to be called “accidental stereo”.

Since this phenomenon was first brought to light some thirty years ago, a number of matching “plain” and “A” takes have been proposed as possible stereo pairs; but until recently, the technology did not exist to prove whether they were or not. The problem lay in perfectly synchronizing the two sides: there could be differences in speed between the original cutting turntables, within each recorded side, or in the playback turntables. There was also the problem of “wow” caused by warping of the original discs or imperfect centering on the turntable. Any one of these factors could lead to a false positive – identifying two recordings made from the same setup as stereo based on subtle variations which caused them to be out of sync with each other.

The recent development of Celemony’s Capstan pitch-stabilizing program, used in conjunction with phase alignment software, finally enables these problems to be solved with a degree of accuracy hitherto unattainable. In successfully synchronizing some proposed stereo sides, others have proven to be only in mono. For example, Side 4 of Koussevitzky’s 1930 Tchaikovsky Pathétique came out in takes labeled 1 and 1A; however, analysis has proven that at least one of them was mislabeled, and that they are in fact identical. Similarly, two 1941 Stokowski recordings previously released in what was believed to be accidental stereo were shown to be no more than slightly out-of-sync mono.

The present release brings together nearly all the sides that have been identified publicly so far as Classical accidental stereo recordings. As the name implies, accidental stereo was never intended to be heard in synchronized form, and was not miked the way a stereo recording would have been. Most likely, one microphone was centered on the orchestra, while the other was pointed slightly off to one side. However, the joined-up results do suggest some directionality that accords with what is known of the orchestras’ seating patterns under these conductors. For example, a 1947 picture of Koussevitzky and the Boston Symphony on stage at Symphony Hall displays the basses lined up to the left and back with the timpani stationed on the right – exactly as we hear them in the Tchaikovsky.

One can best appreciate the added dimension accidental stereo brings at those points where the music blossoms from mono into stereo. Suddenly, what has been the auditioning of a historical artifact takes on a presence and a reality that puts us in the concert hall 85 years ago – a time machine like no other.

Mark Obert-Thorn