Klarinet Archive - Posting 000364.txt from 1999/02

From: Oliver Seely <oliver@-----.EDU>
Subj: Re: [kl] Notation Software
Date: Tue, 9 Feb 1999 17:36:08 -0500

At 01:11 AM 2/9/99 -0500, Don wrote:

>I know that the OCR (optical character
>recognition) programs for text are really finicky. You spend more time
>correcting the results of the scan than you would have typing in something
>from scratch.

Having owned and used MIDISCAN (2.5) for several years now I can second
what Don wrote, but with some rather specific qualifcations. I'm working on
the Beethoven Trio, opus 11 right now. The Adagio movement is only 4 pages
long, so
I scanned it but fixed a couple of general things before generating the
MIDI file
then importing it into Finale. BIG mistake. Lots of 32nd note runs which
from
my experience with MIDISCAN would come out O.K. in Finale if I had just
cleaned them up first. (Lots of them were outside the zone of
interpretation.)
Ah, well, it's only 4 pages. My normal procedure is
to scan at 400dpi, nothing less now. I don't know if the current version
of MIDISCAN
takes TIF files with higher resolution, but if it does, I'd use it. I
discovered a great
difference in quality between 300dpi and 400dpi. I'd go to 600dpi,
particularly where
there are tiny staves for the other instruments attached to the piano part.
I make sure that the pages I scan are as close to horizontally/vertically
perfect
as I can make them and the cleaner they are the better. When I did K.268 I
did
a xerox enlargment of 1.61 on a miniature score before scanning and it
worked out O.K.
but it would
have gone a lot faster had it been cleaner. Don't try to do a miniature
score directly
because it will give you terrible results. Beethoven Opus 103 octet (2nd
and 3rd
movements) is in my pile of things to do and the enlargement I had to do
there was
1.75, but it looks like it will be clean enough to give me good results.
MIDISCAN works
best with staves of 0.3 inches
or 7 mm height or larger on A4 or 8.5 x 11" paper. After scanning and
during program
interpretation it is always difficult to know without some experience
whether to use a
wide, normal or narrow interpretation zone. Different problems can occur
with each
depending upon the number of high and low notes connected with each staff and
the dynamics and articulation markings.

Still, when a score is clean, MIDISCAN saves me a lot of time.

Oliver

-------------------------------------------------------------------------
Unsubscribe from Klarinet, e-mail: klarinet-unsubscribe@-----.org
Subscribe to the Digest: klarinet-digest-subscribe@-----.org
Additional commands: klarinet-help@-----.org
Other problems: klarinet-owner@-----.org