Buxton, W. (1997). Artists and the Art of the Luthier. Computer Graphics:
The SIGGRAPH Quarterly, 31(1), 10-11.
Bill Buxton
Head, User interface Research, Alias | Wavefront, Inc.
Chief Scientist, Silicon Graphics, Inc.
The old adage is, "frustration is the mother of invention." For
me, it certainly is true. As a professional musician in the early1970's,
I was active trying to do live performance using electroacoustic instruments.
In one sense, the challenge was to extend the nature of "performance"
beyond the ultimate creative act of pushing the "Play" button
on a tape recorder in the concert hall. Such was the status quo of
the day.
I had some novel notion that computer-based "instruments", could
be worthy of the name (and hence capable of improvisation, worthy of practice,
and sensitive to subtle nuances of the performer's skill). Which brings
me back to frustration and invention. For me it was more efficient to learn
digital electronics and programming to design and implement my own instrument,
rather than learn to struggle (much less play) on what was commercially
available. (Of course this equation was facilitatedd by the not inconsiderable
help of then students, now giants of CG, such as Bill Reeves, Tom Duff and
Rob Pike).
This exercise turned out to be one of those life-altering decisions. Not
only did I end up with one of the greatest electronic performance systems
ever, and a bunch of new friends, I ended up learning a lot about design
for the artist, since the system that we developed (Buxton, Fogels, Fedorkow,
Sasaki & Smith, 1978; Buxton, Sniderman, Reeves, Patel & Baecker,1979;
Buxton, Reeves, Fedorkow, Smith & Baecker, 1980; Buxton, Patel, Reeves,
& Baecker, 1981; Buxton, Patel, Reeves & Baecker, 1982) was used
by, and influenced by, musicians from around the world. I also discovered
that in the grand scheme of things, there are three levels of design: standard
spec., military spec., and artist spec. Most significantly, I learned that
the third was the hardest (and most important), but if you could nail it,
then everything else was easy. After my work with artists, my research career
at the University of Toronto and Xerox PARC was relatively simple.
Well, sort of. Every once in a while, artists resurfaced and interrupted
my otherwise tranquil life as a researcher and academic. One prototypical
example occurred in about 1987 when Alain Fournier and I were co-directors
of the UofT graphics lab. Having implemented what we thought was a significantly
new airbrush into a paint program, we invited an artist from the CBC, Peter
Softley, in to tell us how great we were. Tell us he did. In a few choice
words (not suitable for reproduction is such an august ACM publication)
he made it clear that it was obvious that neither we - nor anyone else who
had implemented an airbrush - had ever seen one, much less used one. A key
point, among others, was that nobody ever uses an airbrush without a frisket,
or stencil in the other hand. The one-handed wonders that we and everyone
else were producing, were completely incapable of capturing the essence
of the art.
The result of Peter's visit was that he took our whole lab to his studio
for an airbrush lesson. What we really got was a lesson (a useful euphemism
for "kick in the ass") to reaffirm the importance of involving
the artist (a.k.a. user, customer - but all too often "victim")
in the design process. But the reality is that despite Peter's efforts,
it has taken me ten years to be able to finally bring an airbrush to market
that even begins to be worthy of the monicker.
What I knew for myself in my music system, and Peter so delicately reminded
me in the domain of paint programs, was that the ten years that this has
taken are nothing special - at least when contrasted to the years that the
artists themselves have invested in developing their unique skills. While
the essence of the artist is reflected in their work, it is rooted in skill
- skill which is hard earned, and therefore worthy of respect by the instrument
builder, or "luthier." But it is precisely these same skills which
are so poorly captured by most computer-based tools. I maintain that the
skills (and therefore needs) of the artist are different from those of,
say, an accountant. Yet, based on the tools used, when I walk through Disney
Feature Animation, I can hardly tell if I am in the accounting or character
animation department.
For me, where the proverbial rubber meets the road in all of this is at
the level of the input/output transducers that provide the physical interface
that captures the artist's gesture in a form that can be understood by the
technology. This is where I feel that the greatest potential and missed
opportunity lies.
The status quo in the CG world is the "good old" graphical user
interface, characterized by its windows, icons, mouse, pointing paradigm.
But a mouse is like a bar of soap, and the only time it's appropriate to
draw with a bar of soap is Halloween. So where does this leave the artist
the rest of the time? The issue here is one of priorities and relative economics.
Let's look at a couple more conventional "technologies." In contrast
to a mouse, if I were to ask you how much does the bow of the first violinist
of the New York Philharmonic cost, what would you answer? (Remember, I'm
speaking about the bow, not the violin.) The answer, equally true for the
first chair of almost any good symphony, is, about the cost of an entire
SGI workstation. Likewise, if you ask what does a full set of top of the
line sable water colour brushes cost, the answer is, about the same as the
cost of a top of the line Macintosh computer.
Now for me, I couldn't tell the difference between the bow of the top professional
compared to a beginner's, any more than I could tell the difference between
a good sable brush and the brush that came with the $2.95 watercolor set
that I had in grade school. But then, I'm neither a violinist nor a watercolor
artist. But that is not the point. Because of the huge investment in skill
that these artists have made, and the potential that lies behind this skill,
these artists deserve tools worthy of their investment. My claim is that
- for the most part - they have been largely short-changed by the computer
industry.
In my view, it is time for this to change. The one-size-fits-all general
purpose GUI that has dominated the industry is simply not worthy of the
latent talent that might otherwise be manifest through the tools that we
create. The tools must begin to reflect both the diversity and attention
to quality, that we see in more conventional media, such as the symphony
orchestra or the tools found at an art college. My frustration is that it
has taken so long for this to come about. But this is matched by my optimism
that things are about to change. If we do it right, and pay attention to
the Peter Softleys of the world, we might even get it right.