Towards the end of his review of the earlier recording, under Bernhard Klee, JW remarked that the opera was ''unlikely to improve much upon its record of two productions in 150 years''. But of course one revival makes another twice as likely, and here we are with a second recording just a little more than two years later. If only the second were twice as good, or even half as good, that would make this newly created choice a blessing. Instead, it is better in one respect, worse in another, and altogether makes rather a nuisance of itself.
Jos van Immerseel puts the case for the new recording in his introductory note: a performance ''with the instruments as Mendelssohn would have known it [sic], shows the orchestration in an even more remarkable light''. And 'light' is the word, for not only is there a greater clarity, with a shifting of balance away from the strings, but the 'weight' is lightened too, and even more than in Bernhard Klee's recording we feel the affinity with Mozart. This is only 1827 after all, or earlier still, back to 1824, if we go not by the date of the opera's premiere but of its conception. The trouble is that this interest in the orchestration has subordinated the singers. In orchestral passages a normal volume level of playing is perfectly adequate, but the voices have one fidgeting to turn it up. They are (appropriately) lightweight voices and they needed good forward placing if they were to establish their dramatic characters with any vividness.
The Klee is certainly better in that respect, and in one instance, that of Camacho (he whose wedding this turns out not to be), Klee has the more effective casting. Otherwise Immerseel has singers whose voices are rather more firmly placed than Klee's, including a particularly good Sancho Panza in Urban Malmberg. In the matter of their Don Quixote they are fairly even—which is exactly what the voice-production of the two singers is not. Klee's William Murray and Immerseel's Ulrik Cold have it as an assumption in common that the character of the doleful knight can be established by the infirmity with which he sings a sustained note, as on the word ''Halt!''.
So choice is uncertain. The point to hold on to is that the opera itself is well worth hearing. That understates. Orchestrally, this is a magical score, alight with the joy in composition which we know in the young Mendelssohn of the Octet and A Midsummer Night's Dream Overture. Dramatically it is less sure, but there are some highly effective atmospheric touches, and the ensembles show remarkable skill in construction. Immerseel describes it as possibly ''the most brilliant opera ever written by a youthful composer''. Mendelssohn was 16, and it is true that we have to look to Mozart to rival that. As to the recordings, they have far more in common than they have of difference: preference will depend on the relative value the listener places upon period instruments and stage presence. In both, the accompanying booklet leaves much unnoted: for instance, neither mentions that the original had spoken dialogue. For that matter, neither of the productions makes any attempt to solve obvious problems, such as the start of the ensemble in Act 2 (No. 17) where everybody exclaims about a voice which nobody has heard.
'