In 1979, the English label Rubini Records issued a four-LP set containing all of Fernando De Lucia’s G&T/Gramophone Company recordings. It was a fine set, and like many other music lovers I spent hours listening, enthralled by this fascinating tenor, who sounded as no one I had ever heard, either on records or in the flesh. I was astonished at the number of transposed arias, wondering if these records had been transferred at the proper speeds. I began collecting the original discs to determine for myself whether or not the Rubini LPs were accurate representations of De Lucia’s voice. After years of intense listening, I concluded that the pitches on the Rubini LPs are essentially correct. With the advantage of digital restoration techniques and the help of collectors who have generously loaned their precious discs in unworn condition, we now release these important recordings on CD in new transfers, as well as a selection of arias that De Lucia recorded for the Phonotype label after his retirement from the operatic stage.

De Lucia’s first recordings probably sold rather well since none of them is extremely rare, but unworn copies are not commonly found. Not as well-recorded as other G&T discs from the period, they tend to be noisy. Careful choice of stylus size and the judicious use of digital de-noising attenuates the noise to some extent, but too much noise reduction can adversely affect De Lucia’s vocal nuances.

The task of determining the correct speeds for the G&T recordings was relatively straightforward and involved a logical process for our production team. Coincidentally, the matrix numbers of De Lucia’s first recording session immediately precede those of Caruso’s second G&T session of 30 November 1902. Since the speed of the Caruso recordings is undisputed, it seemed logical to begin by playing the De Lucia discs at the same speed as Caruso’s, 68 r.p.m. This speed yielded pitches that were slightly flat, but increasing the speed by only 1.3 revolutions brought the De Lucia recordings solidly into key with his voice sounding absolutely natural and the transpositions quite plausible. Listening attentively to each of De Lucia’s subsequent recordings through 1908, we found that they all play at predictable speeds consistent with G&T recordings made by other singers. The session on 21 May 1909, during which ten songs were recorded, is problematic because there is an obvious change in De Lucia’s voice after the fifth song. Played at 70 r.p.m., an unusually slow speed for that time, he sounds correct in the first five songs, with plausible keys of B-flat, C minor, C major, F, and G. Raising the speed to 74.5 starting with the sixth song gives his voice the same timbre and yields keys of G, A Minor, and E, which are easily played by mandolin, violin, and guitar. Why the turntable speed would have changed during the session is unknown.

Establishing pitches that we feel are correct for De Lucia’s Phonotype recordings is far more daunting than for any of his other records. By the time he was making these records, the timbre of his voice had darkened somewhat, but he by no means sounded old and he could still create those magic pianissimi for which he was renowned. Michael Henstock’s thorough discographic research gives us a complete chronological list of these recordings, session by session, enabling us to establish patterns of speed consistency within each recording date.

The Phonotype recordings speeds vary widely from 69 r.p.m. to at least 83 r.p.m. and perhaps as high as 88! By this time, De Lucia was transposing almost everything down to one degree or another, so determining correct pitches for these recordings comes down to choosing a playback speed that produces the most convincing vocal timbre. Among all of De Lucia’s Phonotypes, there are three recordings that sound absolutely convincing at score pitch – “Giunto sul passo estremo” from Mefistofele (70.6 r.p.m.), “Mia madre” from Fedora (73 r.p.m.), and the first side of the Cherry Duet from L’amico Fritz (80 r.p.m.) Playing any of these discs below score pitch produces a ridiculous sound that no singer could possibly emit. The voice quality on these records can, therefore, be used as a kind of Rosetta Stone to help assess the correct pitches for other De Lucia Phonotypes.

Jeffrey Miller, Michael Aspinall, and I have spent countless hours listening intently to the Phonotype recordings we have included in this reissue. Sharing our thoughts and opinions as to the correct pitching of these controversial records, we discovered that in the end, we were in complete accord. As life-long musicians, our ears have been our primary guide, while our knowledge of old recordings has also informed our decisions. We know that everyone will not agree with our choices of repertoire nor our decisions on pitch, and that disagreements about the correct pitches for De Lucia’s recordings will continue, but we hope that our efforts will help keep De Lucia’s art alive and kindle new interest in this inimitable tenor.