I don’t think there’s
much I can usefully add to the comments
I made in the opening
volume of this series in which I
laid out the technical complexities
of the Welte-Mignon system and its inherent
limitations as a recording mechanism.
The second volume, to which nevertheless
I looked forward, brings us Grieg and
Strauss as well as pianistic titans
such as Pugno and Hofmann, a Liszt and
Beethoven specialist in Lamond and examples
of musicians, famed in their day, who
never made commercial disc recordings,
such as Fannie Bloomfield Zeisler. For
the purposes of this review I’d specifically
direct interested readers to a review
I wrote of her Piano Rolls, a double
CD set on Pierian, for extended comments
on her performances and the system utilised
to reproduce her rolls.
Again I should state
my own view, which is that these rolls
should certainly be utilised by inquisitive
listeners, most particularly in the
case of those pianists who never made
disc or cylinder recordings. Otherwise
exceptional caution should be maintained
in respect of rhythmic and tonal matters
and questions of touch and depth of
tone – some of the essences of pianism.
How else to account for Hofmann’s undated
Freiburg-recorded Mendelssohn, a piece
he recorded commercially. If one takes
the Bell Telephone Hour live performance
in 1944, admittedly much later than
this roll and well into his decline,
we can nevertheless hear tremendous
subtlety of tonal gradation and expressive
phrasing. At a similar tempo – almost
identical in fact and arguing at least
for consistency of interpretation by
Hofmann – the piano roll is rhythmically
stiff, lacking in colour and subtlety
and tonally and emotively dead. Was
Saint-Saëns so capricious in his
playing of Chopin’s F sharp minor Nocturne?
We have his G & Ts and I would counsel
anyone astonished by this performance
to take refuge in those instead. Lamond
made a number of recordings of Un Sospiro
and we know that he was an eventfully
changeable artist in this work, taking
between 3.56 to play it in 1921 to 4.35
in 1925 (I’ve not heard the 1941 Decca)
– with the 1925 and 1936 performances
clocking in slightly under that last
timing. It’s not just that the roll
lasts 5.24 – he had more time to play
it on a roll, after all - so much as
it sounds utterly mechanical and soulless.
In the case of Bloomfield Zeisler one
can make a comparison between this transfer
and that by Pierian. Naxos’s Steinway-Welte
Reproducing Piano is very clear and
recorded with immediacy and brightness,
with the drawback that the action is
noisy. Pierian use a Feurich Welte piano
in a domestic setting – more veiled
and cloaked, not unattractively, though
the piano is sometimes out of tune.
I should add that the rolls have been
transferred at a slightly different
speed as well – there’s twenty seconds
difference and this does beg a question
all of its own.
In the end the unrelieved
brightness and tonal homogeneity of
these performances becomes wearying.
The inherent rhythmic problems and tonal
limitations and post editing complexities
argue for, as I said, a high degree
of caution. For those performers who
never recorded these are still valuable
tools. For others these are adjuncts,
no more, to their recorded discography.
Jonathan Woolf