Advertised as the complete Caruso, using the NoNoise technique, the Bayer edition is neither complete nor noiseless, and one is tempted to add that it is not Caruso either. Missing from the first G&T session are ''Dai campi'' and ''Celeste Aida''; among the 1914 recordings the Carmen duet with Alda is omitted though it can hardly be described now as unpublished as it has appeared in various issues, including the recent one on Club 99 ((CD) CL99–60, 7/90). It is possible that the compilers here were misled by the fact that the second (November 1902) versions of the Aida and Mefistofele arias were given the same catalogue numbers as the first; the matrix numbers, of course, are different but some confusion seems to prevail about this too, since what are listed as ''Matritzen-Nr'' are in fact catalogue numbers and of matrix numbers proper there is no trace.
The 'no-noise' claim (of which a good deal is made) gradually acquires a swelling accompaniment of hollow laughter. Volume 1 begins with the so-called Pathes (wrongly dated) and the Zonophones, with which the producers have made an effort and gained some success; but the first G&T (''Questa o quella'') is as noisy in respect of surface and what sounds like wear as any scruffy transfer from the early days of LP. Compare the EMI ((CD) CDH7 61046–2, 5/89) and the whole thing is shown up as nonsense: the latter (pre-Cedar) is much clearer and far less noisy. Similarly with the Adriana Lecouvreur excerpt, where EMI's transfer is preferable from the start (Bayer with their rougher copy cutting the first chord, which establishes the key). If these are relatively excusable on the grounds that originals are among the great rarities, there is no justification for the use of inferior copies in later issues. ''Magiche note'' (Die Konigin von Saba), the Leoncavallo La boheme arias, the 1910 ''Siciliana'', the Bizet Agnus Dei, the Destinn duet, from Il Guarany, Godard's Chanson de Juin, the second ''M'appari'': that is a short-list, merely a few of the well-known recordings of which it should certainly have been possible to find better originals. There is no need to invoke comparisons with EMI or RCA and their special resources; think of the immaculate copies used by Nimbus, or the infinite pains Ronald Philips used to take to find the best possible copies for his Olympus series. The first volume of a projected complete Caruso has recently been published on Pearl ((CD) EVC1—to be reviewed next month), I can vouch that there is a great deal more consistency, far greater evenness of surface, distinctly less noise.
Before coming to graver faults, it would be a pleasure to give credit where it is due. First, this is the first: a complete edition of Caruso on CD has been overdue, and Bayer have been the first to produce a virtually complete one. They have issued it in a neat box, each disc accompanied by a leaflet which, in addition to explaining the No Noise system, has one of Caruso's caricatures (the subjects unidentified, however) and a photograph of the tenor (but somewhat random in choice so that the Eleazar photograph graces recordings from 1912 instead of the last volume where it would be appropriate). Some of the transfers come out well: Faure's Crucifixus with Journet, La danza, the Farrar ''O soave fanciulla'', the Don Carlos duet, for example. There is also the sample disc, the 1902 ''Vesti la giubba'', played five times to illustrate stages in the transfer-process ('de-noising', 'de-clicking', 'enfiltering' and so forth). So, we listen intently and are probably rather impressed by all the critical listening and application of advanced technology that must have taken place in the studio. Meanwhile, the magnificent recording plays again and again, each time (presumably with nobody noticing) nearly a semitone high.
Pitches and speeds have always bedevilled the labours of transfer-makers: getting them exactly right has never been easy, and there is always somebody who disagrees. So we won't quibble here about quarter-tones. Up by a semitone are ''Amor ti vieta'', ''Mi par d'udir ancora'', ''Solemne in quest'ora'', ''O Mimi'', the Melba La boheme duet, ''Bianca al par'', ''Mal reggendo'', ''Di tu se fedele'', the ''Celeste Aida'' with recitative, the 1917 ''M'appari'', the Lombardi trio. Down by a semitone (these lists, by the way, have no pretensions to completeness) are about a dozen songs, including the first two with Mischa Elman. Then there are the anomalies: the first La partida is in F sharp minor, the second in G; the demonstration Pagliacci is above score-pitch, the others are more or less on it; the Aida duets with Homer have the first record in pitch, the second a semitone high; the Faust Garden scene duet gets ''Eternelle'' right, ''Il se fait tard'' wrong, and because the catalogue numbers put them in the wrong order (second-half first) that is how they are transferred here.
We are still not at the heart of what is wrong. As everybody knows, speed-and-pitch decisions concerning these old records affect the perceived character of a voice. In one way Caruso is remarkably well able to withstand these changes, his voice being so finely spun that a deviation upwards rarely produces flutter-vibrato and one downwards never suggests wobble. But in another he is peculiarly vulnerable, as he is to the other and wider uncertainties in the business of transfer. The brilliance of his tone (so exciting on the early gramophones) can become wearing and strident when the 'top' is over-emphasized, but equally the baritone weight and thickness are made oppressive if the lower frequencies are exaggerated. Many of these transfers are notably lacking in brightness and sheer freedom of sound if compared with others, let alone with the 78s themselves. In conjunction with low pitching the distortion is serious, and that is why at the start I extended queries about the proposed terms of reference even to the name of the artist himself.
It is a strange situation when one comes to the end of such a review and finds that everything written concerns the presentation, and nothing at all is said about the performances. There will be opportunities to remedy this quite soon, however, as the first volume in the Pearl series mentioned above is already out, and yet another complete run is promised by RCA in the near future.'