Micro Focus, creators of visual programming tools for software development, is pleased
to provide major funding for the Computer Chronicles, the story of this continuing evolution.
Welcome to the Computer Chronicles, I'm Stuart Shafay and sitting in this week for Gary Kildall
is Herb Lechner of SRI. Herb, what we have up on our Apple computer here today is the
Koala Micro Illustrator and my handy Koala Pad here and we see a menu of some artists'
choices and if I can, I'm going to try to communicate to you here in some computer graphics.
And so far it's working. If I get your name right. There we go. And I can even make it
a little more sexy than that. We'll give you a little brighter background color so it's
a kind of more upbeat message. Herb, computer graphics is the subject of our show today
and computer graphics in fact involves more than just this little thing we're playing
with or even video games which some people associate with computer graphics. There are
some pretty serious applications, aren't there? Very much so, Stuart. Computer fine art hangs
in galleries right alongside of traditional art. And of course in the commercial area,
computers aid animators and play a very important role in TV advertising today. And of course
the basic building blocks of computer art, that is the graphics subsystems, play a role
not only in art but in related activities such as computer-aided design and other graphic
applications. Okay, on today's program we'll be looking at one of the newest computer graphics
systems called Iris. We'll look at a computer art system called the PaintBox and we'll meet
two computer artists, real people that is. First, let's take a look at the kind of computer
art being done now on mainframe computers. Since the first light pen and graphics displays
of the 1950s, computer-assisted artwork has been associated with beautiful, if mathematical,
images of symmetry and abstraction. But the applications today extend far beyond repetitive
patterns. With enough computing power, computer graphics can extend an artist's reach to cover
the height and length of a wall or compress the time it takes to animate something from
a month to an hour. Using a digitizing tablet, MIT artist Ron McNeil chooses from a palette
of colors and pre-programmed images to create a collage that exists only in the computer's
memory. Before ever seeing a canvas, it undergoes some dramatic transformations. Starting with
materials from different sources like photographs or three-dimensional objects, the composite
can be manipulated to change in color, size, or geometric aspect. The screen-sized image
shown here is only an intermediate step toward the final hard copy, a painting that is 14
feet high and 48 feet long. Like a colossal paintbrush with a memory, the giant XY plotter
magnifies the completed artwork into a wall-sized mural, strip by strip and color by color.
As an ingenious collaboration between digital imagery and robotics, the giant plotter is
unique, but the special talents of computer graphics are increasingly used in another
area of the visual arts, animation. To an art form that was once the near-exclusive
province of film, computers offer a much faster method of animating images. Again, through
a digitizing tablet, an artist can create a figure by drawing in just enough points
to determine its size and rough shape. For a symmetrical object, all the computer needs
is the outline of one side, which is then shaped into a fluid form, regenerated, and
filled in with color and three-dimensional detail. The software calculates and reproduces
the characteristics of light and shadow that the object would possess in 3D. Finally, to
animate the drawing, the artist specifies some key camera positions from which the object
is viewed in space. The program then fills in the missing frames that simulate movement
between those points. Making a drawing come to life is the most tedious part of an animator's
job. Fortunately, it is also readily adaptable to digital processing. The systems used to
produce these images are not designed to replace artists, but to assist them. And with enough
human talent, they can mimic reality in detail or expand it to the fringes of our imagination.
Our guests around the table now are Michael Arendt. Michael is a computer artist and design
director with Aaron Marcus and Associates in Berkeley. And next to Michael is Don McKinney,
vice president with Silicon Graphics here in Mountain View in the Silicon Valley. Herb?
Michael, in your consultant role, you use all kinds of computer graphics equipment.
You have a setup here today for us. Do you want to tell us a little bit about this equipment
and what you're going to do? Well, this is an 8-bit microcomputer, which is very similar
to the type of computers that most of the viewers would use in their home, for instance,
to play games. And we use these type of systems, for instance, to create imagery for digital
directories and education software and such like that. And what exactly do you...show
us how you use this system here. Okay, well, the traditional artist tool is a brush. This
is an electronic artist tool, and it's using a graphics tablet, which you see here, and
an electronic pen. And with this, I can paint images and put graphics, combined graphics,
photographic images, and text on the screen. I can very quickly load in an image here,
for instance, to show you a raw photographic image that was input using a video digitizer,
which is simply a video camera. And once I get the image into the computer...
That's an unfortunate example, by the way. Michael, Bob, we'll go with it.
Once I get the image into the computer, then I can use this graphics tablet to take away
parts of the image that I don't want. As you can see here, it's loading in this microcomputer.
What do you mean by parts of the images you don't want, Michael?
Well, for instance, maybe I want to take out a background and put in another background,
or I want to take out areas where I would like to put in other types of things like
text and things like that, which I'll demonstrate here. I'll show you, for instance, I'm showing
you the build of an image. Don, while Michael is working here a second,
there are great memory demands, aren't there, in doing graphics? How do you accomplish that
with just a small microcomputer? Well, really, it depends on the exact amount
of memory. The memory requirements really are the requirements of the amount of color
that you have, and also what we call the resolution of the screen. So the higher and smoother
the image appears, the more memory is required. In the case of what you'll see later today,
we have approximately 3 million bytes of memory just used to store an image in real time.
Okay, Michael, I'm sorry, you're ready. Okay, now I've taken away the background.
Now I've put in a cartoon bubble here, and then I can load in another image here to show
you other things that I've... So you're loading images you've pre-prepared?
Yes, I pre-prepared these images. How did you go about getting the old background
out? Did you trace around it or something? Yes, you can block it out with a function
on the tablet using a black background blockout. And then so, as you can see, the image is
slowly building here, and then I'll show you the final image that I've created, and then
I can show you how you can manipulate that image other ways also.
What is the tablet you're using? Tell me how that works.
Well, it works by addressing various areas in the computer memory, and essentially it's
giving it a command either to put color onto the screen or putting geometric figures on
the screen like circles or grids. And you can also put text on the screen, as you see
here. Here's the final image I've built, which has a combination of a photographic image.
It has graphics in the background, and it has text also combined with those graphics.
And I've been able to do that with this tablet. I can also, for instance, do things like turn
the image upside down, or I can also flip-flop the image this way here, and then I can bring
it back up. I can also invert the colors, for instance. So I can make it negative or
make them in new colors. And then there are other functions on the tablet, too, that you
can do. For instance, you can create textures with what's called electronic brushes and
things like that.
Okay. Well, we're moving our way up from the simple Koala pad to something more sophisticated.
And, Don, that gets us into your area. And I think you have your iris system set up over
in the other part of the studio here, so let's take a walk over and see what iris does.
Don, how does your iris system here differ from what we've seen up until now?
Up until now, it's really, we've really been looking at microcomputers, really personal
computers. And this is really a different class of machine. The biggest difference,
though, that separates us out is the idea of us using custom chips developed by Silicon
Graphics for our own use to do these three-dimensional calculations. This example here is a Rubik's
cube. And what's going on here is we're calculating the location of all the endpoints of all the
faces so that we can rotate those in 3D and do all the hidden surface removals all in
a very real-time mode. Why don't we go on to the next demonstration here to give an
idea of how our machine can be used.
What kind of machine, what's the hardware we're using here, Don?
This is a terminal configuration that has a little floppy disk on top. But the basic
computer that we're using here is a Motorola 68000, often used in some of the other reasonably
high-performance microcomputers. The other chips that we actually have are called, what
we call our geometry engine, and it's a high-speed, three-dimensional floating-point calculation
unit that runs at 6.5 million floating-point operations per second. This speed, we are
planning to enhance to over 10 million floating-point operations just in a few minutes.
Okay, what's the demo here?
This one is a series of pictures of a series of buildings where we're doing a calculation
in real-time. So I'm going to angle up away from the buildings, and I'm going to zoom
back away. And as I'm zooming away, you can see this building going further and further
back. You can see that the light sourcing is done so that the certain surfaces that
are closer to the rays of sun are visible. Now what I'll do is I'll add buildings in
3D. So you can see as I move around the space, I can actually get the full panoramic effect
of this block, or I can even zoom in. So I can go through areas. Let's skip on to the
next. The next is...
Before we get to that, Don, what would the application be, say, of something like this
architectural thing you just showed me?
We have several customers that are in the A, E, and C market, the architectural, engineering,
and construction business. And we have several local companies that are doing piping diagrams
and piping calculations for intersections of pipes in, say, nuclear plants. Another
case might be for an architectural firm that would add software to our product and sell
it to the end-user architectural people for doing building design, for doing construction
of faces, landscape architectural design, even an interior decorating, use it to place
desk and furniture within a given room.
Okay. Tell me about this demo now.
Okay. This one is, again, it's a three-dimensional object. It's a rendering of a robot. So the
first thing I'm going to do here is get it a little bit larger. And then what I'll do
is I'll tip it over so you can actually look at this robot arm from the top. And I'll spin
the arm around. And now as I'm closer, let me tilt it back up so you can see it from
the front view. And in this case, what I'm doing is I have a one button. By touching
it, I can cause the shoulder to move the entire arm up and down. Or I can cause another button
can cause the forearm to move up and down. And in the last series, I can cause the pinchers
to actually open and close or to even grip something.
An application for this might be an auto-assembly plant where you're grabbing a metal part to
be welded into another area.
Okay. We have time maybe for one more, and I'll load your disk for you.
Okay. This is another sophisticated 3D application where we're actually computing the location
of an aircraft in flight. So the first thing that will happen is we'll be looking out the
front of the airplane at a building or a hangar. And then we're going to taxi up the runway,
make a right turn, taxi down the runway, make another right turn, and take off.
Okay. And you talked about the speed of about six and a half million. What was the unit's...
Floating point computations per second. Most computers are measured in MIPS, and some are
measured in what's called FLOPs. And the difference being that it's either an integer mathematics
or it's a floating point mathematics.
Okay. Let's see what this is.
Okay. So what I'm doing now is I'm sitting at the runway. I'm going to accelerate the
airplane as it goes toward the building. I'm turning the rudder now using the mouse and
locking in on a direction. And then I'm going to sweep around and get back onto the main
runway. And now as I accelerate the plane still further, raise the flaps, just like
in a normal takeoff, I can switch between the viewing angle of the pilot looking out
the front of the window to the viewing angle of the...from the tower. And now I'm off the
ground.
That's very impressive. Okay. If you watch television, you are familiar with computer
graphics. Some of the effects you see on the Computer Chronicles are, in fact, generated
by computer. Well, we'll see a very sophisticated video computer graphics system in just a minute.
Joining us now is Anne Chase, who is a freelance computer artist, and Kevin Prince. Kevin is
engineering manager at MCI Quantel. Now, Anne, you played around with a picture of me at
the break just before we got to the segment. That was animation, I take it. And how do
you do that with this system?
Well, basically, I can find that animation stack in the library and show you how I set
it up just by tapping. It's cell by cell is how it's set up. And here you can see that
I just drew the tie and rolled it up on your neck.
So you captured the real picture and then drew on top of the real picture and then just
did the cell animation.
Right. Just cycled the animation in.
Does this machine do all that with just commands that you give it?
Yeah, basically, just your artistic ability and, you know, what you can do using the capabilities
of the machine.
Kevin, what are the capabilities of the machine? What kind of hardware or software is supporting
this system?
Well, hardware-wise, we're based around the 68,000 processor, as previous people were
talking about. But we've got a lot of dedicated hardware to do all the fast wipes, the actual
painting and drawing in the system and changing all the brush sizes, et cetera. It's taken
us some time to develop that. But associated with that is, probably more importantly, is
the actual software interaction with the user. Everything you see is obviously a very large
operating system that we've had to divide for the system.
Anne, could you show us, you mentioned drawing and brush strokes, some of the free hand capability
of the system?
Basically, what I'll do here is just wipe the canvas and you can see that I'm in a paint
mode. And this shows me my brush size and I'll take a white here and you can see that
I can just paint right on the screen. I can choose a different color just by tapping and
say a larger brush size. I can paint over the top of that. Then I can go back and choose
that mixed color right off the image and store it to use later.
Now, this is obviously much finer drawing, greater resolution than some of the earlier
things we saw at the beginning of the program. That's a function of what, Kevin?
That's basically the way we do our store manipulation, the way we actually store the picture. We
have various methods of mapping the image that we're actually trying to work with into
the store, very different to individual pixel operations which the other machines would
be working with.
And Paintbox is, in fact, in use right now in television graphics. How is it being used?
Most of the networks have actually got several of these systems and they very often use the
machine for their news production. And practically every night you'll see some form of over-the-shoulder
shot that's probably been generated on the Paintbox. But hopefully it's so good you
would never notice.
Now, this system gives you the capability of mixing that which is artistically drawn
with that which is real-life photograph sort of a thing. How is that done?
Well, actually I can just go to the live source and you can see that right now we're taking
in a live video source of what we're doing. And at any time I can, I don't even want to
look at myself, let's just tap something down and see what we have here. There we have me.
Okay, so you have just captured a still frame from this live coverage of what you're doing
right now and now you can paint with it.
Right. Or I could, let's say I'd like to create a stencil. And just very quickly I can go
in here with a stencil medium and just outline something and fill it.
And you're filling, why are you filling that now?
What I'm going to do is a cut and paste technique. This is a stencil medium that's laid over
the image. What I'll do is just...
This is the word processing of art. Cut and paste.
Absolutely, yes. And then I'll just paste it up.
At this point we're taking a portion of that image and we've now got, as you can see, two
images.
There we are.
I think that's a huge improvement.
That's right.
And you seem to have a lot of fun doing this. Is there resistance to an artist, or shall
we say a pure artist who's used to dealing in brushes and paints, getting comfortable
with using this technology?
Maybe at first it's a little difficult to use the menu, to learn to think and actually
read at the same time while you're drawing. But once you've worked with the system for
a short period of time, it's just like second nature. Very simple to use. Set up in artist's
terms.
It's clear how valuable this would be to a television media. What about the area of fine
art, Anne? Is computer art catching on there too, or growing?
Definitely, definitely. It's the wave of the future. That's what's going to be happening
in art in a couple of years. In fact, it already is.
Should brush manufacturers worry?
No, no, no.
They're not in trouble yet?
Not yet.
Well, then we have about a minute left. How far can you go with this technology? Will
we get to the point where we could be replaced by animated people and synced up with words?
I sincerely hope not. The way I see it at the moment is that we're providing the means
by which an artist can extend their capabilities. We can turn a reasonable artist into a good
one and an extremely good one into a superb artist. All we're doing is allowing them to
make their artwork in a much shorter space of time and in the medium in which it's required.
Well, that's fascinating. You don't exist anymore on that picture, Kevin, but we know
you're here. We're all out of time. Thanks so much for being with us. Thank you for joining
us on this edition of the Computer Chronicles.
Micro Focus, creators of visual programming tools for software development, is pleased
to provide major funding for the Computer Chronicles, the story of this continuing evolution.