Dr. Dobb's Journal November, 2005
There were plenty of miracles, too. The problem is that we've gotten used to miracles. The impossible has become very nearly routine. Moore's Law has worked inexorably. A few years ago, a doubling of graphics capability was dramatically noticeable. Now, performance increases of a factor of 10 are just an incremental improvement. Of course, that's not true for games players. Video that looks great for office work is anemic when faced with Doom 3, and such games are crucial for selling video boards.
Graphics generation, rendering, and display demand more computer resources than any other task except big scientific projects. Some say graphics tasks actually require more resources than the 100,000 lines of Fortran simulations weapons designers run on their supercomputers.
Until recently, 3D graphics required a dedicated system, and those machines were expensive enough that you didn't do a lot else with them. For most of the years I went to SIGGRAPH, you seldom saw a Windows machine. Silicon Graphics and Sun machines were dominant, with Apple the only popular system visible. Then, slowly, Windows systems crept in. Apple never entirely vanished, but Apple and Sun system numbers diminished as Windows' presence grew. A few years ago it was news when Apple, Windows, and Silicon Graphics were about equal in numbers on the show floor. Then both Apple and SGI faded, and Intel boxes running Windows became dominant.
Exactly when it happened can be debated, but it's clear that as of now, you do not need a dedicated system to do 3D graphics, and many people have systems good enough to do 3D graphics creations. Of course "good enough" isn't enough for most graphics artists. They want the best; and while this year there are plenty of Apple systems, more than you'd see at any other show, the show was pretty well dominated by 64-bit Windows—which is to say, by AMD systems. Two years ago, everyone wondered whether AMD would survive. Now AMD 64-bit systems running 64-bit Windows are everywhere, and a 64-bit computer running a 64-bit OS is the desired system for everyone from Hollywood artists to high-end games players. This is definitely the year of the 64-bit system.
The most spectacular system we saw was the BOXX (http://www.boxxtech.com/) Quad 8-Core system, using four dual-core AMD 64-bit chips. Windows 64-bit XP Professional can't support more than two chips, so this beast was running Windows 64-bit Server 2003 but was being used as a workstation. See http://desktops.engadget.com/ entry/1234000670053142/ for more details, but think of this as a solid glimpse into the future.
When small computers first came out, I noted that they were the great equalizers for artists. Their first impact was on writing. I bought my first computer (Old Zeke, a Z-80 who is on display in the Smithsonian) to improve my writing productivity. Prior to small computers, writing was really hard work, unless you were an extraordinarily skilled typist. Worse, the mechanical work of writing interfered with editing and rewriting. Once you had a clean manuscript, it was very tempting to leave it be. Taking another pass would mean more retyping...
Small computers changed all that. Now, it's true enough that some writers are tempted to endless editing and revisions resulting in overwritten work that looks like the output of a committee, but that's not the computer's fault. The computer enables: When I am done with a manuscript, it is at least in potential precisely what I want it to be, and it will look no better or worse than anyone else's work. Moreover, the computer examines grammar and spelling, and while no one has to pay attention to its suggestions, they're often quite good, and you ignore them at your own risk. If the computer suggests your sentences are far too long, you might want to pay attention.
Small computers let every writer, beginner or long-established pro, have the best in production equipment, and moreover, choose the system that works best with the writer's talents and style. There's little advantage to be had from wealth. Moreover, the Internet is reducing the bargaining advantage of publishers over writers, and that revolution is nowhere near over.
The next equalization was in sound engineering. The typewriter, carbon paper, and Snopake had already gone a long way toward leveling the productivity playing field for writers long before the computer revolution, and any serious writer could afford the equipment to get into the game. It was different for musicians and vocal performers. Turning their talent into a salable product required expensive equipment, and in essence put them in thrall to recording studios. Writers were not entirely at the mercy of publishers. There was enough competition among publishers that our contracts, if not entirely fair, didn't reduce us to slavery. Not so with performing artists, whose contracts were incredibly unfair, and I don't use the word "incredible" very often. For a long time a vocal artist might earn almost nothing from a major hit record—the studio got just about all the revenue from record sales. The artist was expected to do live performances and appearances and live on that revenue. They never owned their own copyrights to either their songs or their performances, and their art was ruthlessly exploited by their blood-sucking publishers.
Small computers have pretty thoroughly leveled the playing field for vocalists and recording artists in general. Any small group can now own the equipment to do professional-level recordings. For under $10,000 they can buy a good Wintel computer with Avid Pro Tools or Sound Forge Professional (recently acquired by Sony; http://www.sonymediasoftware.com/products/soundforgefamily.asp), or a G5 Mac with a big screen and software. Their equipment will be capable of just about everything that used to be done by a million dollars' worth of amplifiers and mixers. Microphones and recording hardware can run another $20,000 or more, but those can be rented for a lot less, and there's the question of "good enough." A chamber orchestra will require higher end microphones than a rock group. Meanwhile, a few hundred dollars' worth of foam sound wall material, a good book, and a week's work can turn just about any room into a proper recording chamber. Finally, the Internet has done a lot to equalize marketing capabilities. Enough so, in fact, that some artist groups now think of live performances more for their publicity value as aids to recording sales, than as a primary revenue source—just as writers long thought of paid speeches as advertising with negative costs.
Better yet, most professionals use Avid's Pro Tools (http://www.avid.com/products/ xpressStudio/proToolsLE/), which David Em calls the gold standard of digital audio, and others use Sony Sound Forge (http://www.sonymediasoftware.com/Products/ShowProduct.asp?PID=961), but you don't need to spend that much money on software to have professional-level capabilities. GoldWave (http://www.goldwave.com/) offers GoldWave Digital Audio Editor and Multiquence Multitrack Mixer that between them cost under $100. And while Sound Forge and Pro Tools are far more popular, some professionals prefer the less costly GoldWave for its ease of use. Once again, it's a question of what's good enough. GoldWave can produce some really amazing effects.
Sony also offers an "entry level" program called Sound Forge Audio Studio at $70. I am not really qualified to choose between GoldWave and Audio Studio, but I will note that both can be learned with moderate difficulty, and while each does things a bit differently, once you learn one, it's a lot easier to learn the other. GoldWave is a fully professional-level program, and as I have said, a few prefer it to Sound Forge Professional. On the other hand, experience with the more limited Sound Forge Audio Studio is directly applicable to Sound Forge Professional. Either program and a good PC will let you do things that only studios with the most expensive equipment could accomplish just a few years ago.
The same leveling process is happening in the visual arts. It is now possible for nearly any artist to own all the hardware and software required to produce movies of the quality of, say, Shrek I or the first Jurassic Park film. Now true, making a full-length movie is sufficient work that almost no one is going to do that alone and unaided, but it won't be the equipment that prevents them. Moreover, the computers are raising individual productivity by a lot. Once you have good 3D models of your characters—say Shrek, Donkey, and the lovesick dragon—the software will do most of the scut work of moving them around. You don't have to draw motion frame by frame. Computers do that. You can then go in and tweak things, add or change expressions, put in extra motions and gestures, and so forth, again in the full expectation that the computer will help you seamlessly blend in your additions and corrections.
Don't misunderstand, it takes talent and lots of it to make even a short animated film; but the truly hard work is now mostly done by the computer. The computer levels the playing field and relieves the talented and untalented alike of junk work. The result is that artists can spend more time creating art and less on its implementation.
There's more. The production costs for movies with live characters is falling dramatically. Sky Captain and the World of Tomorrow was made by having all the actors work in front of a green screen, then all sets and scenery were added by computer editing. That was news when it came out. Now there are lots of movies made that way. A number of booths at SIGGRAPH featured equipment or software used in the production of Sin City, another "green screen" movie. Many more are planned. The green screen movie may not be the wave of the future, but it is no longer news when movies are made using this technique.
That's the present state of the art. Moore's Law predicts a doubling of computer power every 18 months or so. Given where we are on the Moore's Law curve, a doubling means a truly astounding jump in capability. Graphics arts hardware, though, doesn't follow Moore's Law. To the best I can tell, the doubling times are under one year. Given that, I think it's safe to predict that well before 2012, we will have at least one professional-quality feature-length animated film essentially created, directed, and produced by one artist.
The game of the month remains Everquest II, which says something about that game: It has held my interest for several months now, despite a number of other games that have piled up. I can also recommend Sid Meier's Pirates!. I still think the ships sail too fast, but I find I can get used to the speed. The game is the same wacky combination of spoof and action that made the original Pirates! game so much fun. Alas, it takes time from getting my Everquest II Paladin to level 50.
The computer books of the month are both from O'Reilly & Associates, and both about the Mac. The first, David Pogue's OSX Tiger, The Missing Manual, is another of the excellent O'Reilly "missing manual" series. I can say categorically that I have never seen a book in that series that you shouldn't buy if you use the products the books cover. I'm not sure I've ever said that about a series of books before. The other book of the month is Dave Taylor's UNIX for Mac OS X Tiger. This is a valuable introduction into the command-line processes hiding under the hood in your Mac OS X machine. If you use a Mac and you're not already a UNIX guru, this book will make your life easier. As we've observed often, with the Mac, nearly everything is very easy or impossible; but with OS X, some "impossible" things now become, not easy, but at least feasible, provided you know how to get at them through UNIX commands.
DDJ