Bad metadata in 1782

[Rainer_Muller (Rainer)]

Here’s an example that Rovi and Musicbainz cannot always be blamed for bad metadata in Roon.

I have a Cappella Coloniensis recording of a symphony by JC Bach with the opus number op. 18/5 given in the booklet. Roon attributes the following:

Johann Christian Bach
Sonata (Duetto) for keyboard, 4 hands in A major, Op.18/5, CW A19 (T. 327/3)

The piece is clearly a symphony, not a keyboard duet. So Roon is wrong?

Not really. It seem that JC himself (or his publishers) are to be blamed for the confusion. In the New Grove, there are two entries for JC Bach’s op. 18, apparently unrelated:

Six Grand Overtures (E, B, D, D, E, D), 2 fl, 2 ob, 2 cl, 2 bn, 2 hn, 2 tpt, timp, str, op.18 (c1782); nos.1, 3, 5 for double orch

and

4 sonatas (C, D, E, G), hpd/pf, vn/fl, nos.1–4 in Four Sonatas and Two Duetts op.18 (c1781); CW xxxviii

Which shows again that we need a means in Roon to manually select the right composition. Usually, the automatic identification algorithm works quite well, but there are examples where it simply fails.

[Ludwig (Ludwig van Marbach)]

I’d be curious to see some screenshots, and to have more information on the album. Do you have an Allmusic.com link?

((Beethoven also had multiple publishers publishing op. 106 (for example) simultaneously, but in his case it was the same work. Still, he told them all that it was a new work, and had them all publish on the same day, so he got his fees from all the publishers, which he wouldn’t have if one had published a day later…))