Analog People in the Digital World
I hear this term more and more lately, and the people who use it throw it around as if they've stumbled upon something new and unique. After all, who isn't an analog person? The term is meant to describe someone who doesn't understand the digital world, but to me it highlights a profound misunderstanding of both worlds.
One of my first college term papers was on George Lucas and his contribution to the digital age with his "Edit Droid", the first non-linear digital editing system. A follow-up paper discussed computer animation in film, titled "Art in Motion in Motion Pictures - The Development of Computer Technology", and covered films like Oliver and The Abyss. That was almost 20 years ago when I wrote those papers. In my mind I've been an analog person in a digital world most of my 39 years of life, which is probably why I ended up in the field of graphics technology. I may not be working in film, but my contributions to the graphics industry gives me peace of mind that I am still contributing in a round about sort of way.
3D animation (actually, animation in general) has always fascinated me. I created my first 3D animation on an 80286 with a math co-processor using 3D Studio 2.0. For a 30 second, 320x200 animation, it took the computer over 4 days to render. And this was nothing more than a flyover of a static 3D model. A single 640x480 truecolor frame took 4 hours to generate. Now, we have consumer level 3D graphics accelerators that can generate over 1000 times the amount of detail in real time at over 60 frames per second at high-definition 1920x1080 resolutions. I am floored at how far we've come in such a short time. An XBox360 or Playstation 3 has more rendering and processing power than some of the most powerful computers on the planet at the time I wrote those papers. As far as we've come, though, there is still another barrier to break to get to real-time photorealism, and that's radiosity calculations. Some have gotten close, but the best photo-realistic CGI still takes a phenomenal amount of processing power, hence the multi-processor rendering farms at ILM. It's one thing to take a light source, hit an object and cast a shadow like we see in many computer games of today, but it's something entirely different to calculate the surface properties of the object and the way it bounces light onto other objects. In the real-world, we use flags and barn doors and bounce cards and scrims to tell the light where to go, and it's hard enough to control it. Imagine, then, having to mathematically model every single particle in a beam of light and track it through space and time, bouncing off of and through objects in your scene, and how some objects split it or change it to different wavelengths, changing the color of the beams, thus changing the color of the objects, which reflects or refracts to other objects, etc., etc. Now add another light source or two or three or twenty. Add a filter to a source to change the color temperature. The amount of information to calculate becomes staggering. And we've only lit the scene. It's a whole other set of calculations to model it and to make it move realistically. And then there's another set of calculations to simulate the lens on your virtual camera and how the light passes through the various elements.
I could go on forever, but I'd lose my audience. I probably already lost half of you after the first paragraph, but that's okay. I'm reminiscing.
Bottom line is we are all still just analog people in an analog world. Realistic digital people are the stuff of science fiction. At least for now. Our interface to the digital world as it exists today will remain analog and will always require analog to digital conversion or vice versa. For you to even see these words requires that your eyes pick up the light emanating from your computer screen, sent there through analog waves. If you're blind, you can hear these words through digital voice synthesis and then through analog waves of sound entering your ears, vibrating your ear drums. And to reply to this post will require input in an analog form.
I can hear it now. "That's just so... analog."
Welcome to my world.
One of my first college term papers was on George Lucas and his contribution to the digital age with his "Edit Droid", the first non-linear digital editing system. A follow-up paper discussed computer animation in film, titled "Art in Motion in Motion Pictures - The Development of Computer Technology", and covered films like Oliver and The Abyss. That was almost 20 years ago when I wrote those papers. In my mind I've been an analog person in a digital world most of my 39 years of life, which is probably why I ended up in the field of graphics technology. I may not be working in film, but my contributions to the graphics industry gives me peace of mind that I am still contributing in a round about sort of way.
3D animation (actually, animation in general) has always fascinated me. I created my first 3D animation on an 80286 with a math co-processor using 3D Studio 2.0. For a 30 second, 320x200 animation, it took the computer over 4 days to render. And this was nothing more than a flyover of a static 3D model. A single 640x480 truecolor frame took 4 hours to generate. Now, we have consumer level 3D graphics accelerators that can generate over 1000 times the amount of detail in real time at over 60 frames per second at high-definition 1920x1080 resolutions. I am floored at how far we've come in such a short time. An XBox360 or Playstation 3 has more rendering and processing power than some of the most powerful computers on the planet at the time I wrote those papers. As far as we've come, though, there is still another barrier to break to get to real-time photorealism, and that's radiosity calculations. Some have gotten close, but the best photo-realistic CGI still takes a phenomenal amount of processing power, hence the multi-processor rendering farms at ILM. It's one thing to take a light source, hit an object and cast a shadow like we see in many computer games of today, but it's something entirely different to calculate the surface properties of the object and the way it bounces light onto other objects. In the real-world, we use flags and barn doors and bounce cards and scrims to tell the light where to go, and it's hard enough to control it. Imagine, then, having to mathematically model every single particle in a beam of light and track it through space and time, bouncing off of and through objects in your scene, and how some objects split it or change it to different wavelengths, changing the color of the beams, thus changing the color of the objects, which reflects or refracts to other objects, etc., etc. Now add another light source or two or three or twenty. Add a filter to a source to change the color temperature. The amount of information to calculate becomes staggering. And we've only lit the scene. It's a whole other set of calculations to model it and to make it move realistically. And then there's another set of calculations to simulate the lens on your virtual camera and how the light passes through the various elements.
I could go on forever, but I'd lose my audience. I probably already lost half of you after the first paragraph, but that's okay. I'm reminiscing.
Bottom line is we are all still just analog people in an analog world. Realistic digital people are the stuff of science fiction. At least for now. Our interface to the digital world as it exists today will remain analog and will always require analog to digital conversion or vice versa. For you to even see these words requires that your eyes pick up the light emanating from your computer screen, sent there through analog waves. If you're blind, you can hear these words through digital voice synthesis and then through analog waves of sound entering your ears, vibrating your ear drums. And to reply to this post will require input in an analog form.
I can hear it now. "That's just so... analog."
Welcome to my world.
Total Comments 0
