This question is a little misleading and as a result earlier answers,
while correct, don't quite tell the whole story!  There are two quite
different questions to ask:


Question 1. Are Duo-Art rolls perforated in the UK better?

Here the answer is, on the whole, "Yes"; the rolls are better, for
the same reasons as already given: better paper, non-rusting spool
ends (plastic or wood) and good robust boxes.  I suspect they were
always better, and have not survived longer thanks solely to some
benign quality of the atmosphere.

It wasn't all good.  The dynamic coding of American master rolls was
fiddled with, possibly because of dynamic range, but also dynamic
system response time: the British machines were somewhat slower to
change dynamic level -- or, at least, the UK roll editor's piano was
slower!  Gaps between pedal slots were increased, presumably due to
a slower-acting pedal mechanism, which can make the music rather
choppy.  These issues were discussed during the 1920s in surviving
correspondence between US and UK roll editors (W. Creary Woods and
Reginald Reynolds, respectively).

Also, for those who like to sing along with their pianos, UK Duo-Art
dance rolls don't have words printed on them.  I think this was due to
copyright restrictions that increased royalty payments if words were
used, and then didn't allow word rolls to be used in roll libraries --
a great shame, in hindsight.


Question 2. Are UK Duo-Art recordings better?

The answer here is generally considered to be a qualified "No".  This
is because, after the shaky initial efforts, Aeolian had two distinct
approaches to recording dynamic coding:

(1) dynamic coding perforated in real time by the recording perforator,
as set by the recording engineer using hand-operated dials to note
[register] the performer's dynamics.

(2) terraced coding levels derived by an editor from the real-time
recording.

In the UK, Reginald Reynolds retained the real-time coding, including
every twitch of the engineer's hands even when no playing note is
affected.  Some pianos react badly to this, probably because the
ever-changing coding means that response times become critical.  The
earlier assumption about the slower response of the UK editing piano
may be the crucial factor, although it may simply be the zero-level
setting.  The letters from Creary Woods to Reginald Reynolds tell us
that the London rolls never played on American pianos even when new!

In the US, Creary Woods developed a method that only coded the levels
used by playing notes, eliminating the unnecessary stuff in between.
This lets the expression level stabilise before the note plays so is
less sensitive to variation between pianos.  It also naturally removed
quite a lot of the dynamic subtlety unless the editor worked hard to
replace it.

Only about 300 rolls were issued solely in the UK, some 285 in the
series starting from 01 and ending at 0365, and these retain their
edgy coding (as do Audiographic versions of them).  The vast majority
of higher-numbered UK issues are American dance rolls, normally issued
later in the UK when their show or film transferred from the USA,
although there are some UK classical issues in the list.

It seems reasonable to assume that Aeolian established the London
recording studio to record European artists for issue in the main
Duo-Art series, not simply for the UK market.  It seems rather a lot to
invest for 300 rolls, and in the 10 or 11 years it existed the London
studio must have recorded far more than 30 titles per year!

How many masters were sent to America to be issued in the main series,
being re-edited by Creary Woods to his standards?  There's no way of
telling.

Julian Dyer