Information Technology Reference
In-Depth Information
is not straightforward because the segmentation
of the symbols is a complex problem, ever more
with an eager interpretation process.
distinction between the input symbols and fast
writing speed . Actually, one of the goals of the
proposed system, called “ Presto ”, was to make
it possible to write music in a faster way than on
paper thanks to the use of pen interaction and
the definition of new faster gestures to replace
the classical way to draw musical symbols. As a
consequence, the gestures are also thought to be as
small as possible. Of course, in order to facilitate
their learning, the new gestures are as intuitive
for musicians as possible, but are still different
from the classical ones. For instance, the gesture
to draw a filled-note head is a dot. In order to add
an accidental to a note head, the user has to draw
a small vertical segment starting on this head,
uprising for a sharp, descending for a flat. In the
same manner, a durational dot is added to a note
head by drawing a small horizontal segment start-
ing on this head. The second column of Table 1
presents some of the gestures of this system.
Finally, in terms of input process speed, the
performances of Presto are positive, because ex-
perimentations highlight that the software enables
musicians to copy music in the average time of
72% of that required to copy on paper.
Forsberg, Dieterich, and Zeleznik presented in
1998 a pen-based musical score editor, called the
Music Notepad ”; a more recent version of this
composition system is available on the internet
(Forsberg, Holden, Miller, & Zeleznik, 2005);
the gestures we explicit in this paragraph and in
the third column of Table 1 corresponds to this
last version. Once again, the authors exploit a
new alphabet to write classical musical notations,
which is in fact quite close to the one proposed
by Anstice et al. (1996) and Ng et al. (1998). For
instance, quarter-notes can be drawn almost the
same way as on paper, but accidentals can not: a
gesture starting on a note head and uprising on
the right is interpreted as a sharp, descending on
the right as a flat. A durational dot is added by
tapping on the note head.
We proposed a pen-based system dedicated
to the editing of musical scores (Macé, Anquetil,
existing pen-Based musical score
editors
In order to deal with both of the problems presented
in previous section, the main solution consists in
changing the gestures , that is the way musical
symbols are drawn, and to constrain the user to
a new way of writing musical scores. Therefore,
the authors often define new alphabets, which
are designed to reduce the similarity between the
different symbols in order to make their discrimi-
nation easier. These new alphabets are generally
unistrokes, in order to facilitate the interpretation
process: as a symbol is always drawn with only
one stroke, the hand-drawn shape recognizers can
be called each time the user draws a stroke. If an
interpretation for this stroke is found, then it is
replaced by the corresponding musical symbol.
On the contrary, if no possible interpretation
is found, then the system has been unable to
recognize the stroke: it is rejected . It disappears
from the screen, and the user has to draw it again.
Naturally, the definition of new alphabets for
musical symbol composition tends to find the
best compromise between ease of recognition for
systems and ease of learning for users. Neverthe-
less, this solution has a major drawback: the user
has to get used to this new alphabet, and does not
write musical notations the same way as on paper.
We believe that this creates a gap between the
musician and the computer which often implies,
for instance, a rupture in the creation process in
comparison with the use of paper.
One of the first pen-based systems was
proposed in 1996 by Anstice et al. and then
enriched in 1998 by Ng, Bell, and Cockburn; it
was based on this idea. The authors exploited
the three primary criteria proposed by Goldberg
and Richardson (1993) for their pen-based text
input system, which are: ease of learning , high
Search WWH ::




Custom Search