Professional Amiga Graphics

PART THREE
Paul Hamilton airs some thoughts on improving your artwork - if you're sure you can be bothered...

When the Amiga was first launched, with the blessings of Andy Warhol, it was hailed as the first affordable artist's computer. The first graphics software written for it, Deluxe Paint, fulfilled that promise with tools and capabilites rivalled at that time only by high-end systems like the Quantel Paintbox.

Until quite recently it was still the only computer offering truly cutting-edge graphics tools at reasonable prices. Five years ago there were public domain raytracing packages running on 512k Amiga 500s with capabilities matched only by university minicomputers. At last year's Computer Graphics Expo I discovered that some of the 'professional' Silicon Graphics 3D packages selling for around £6,000 didn't offer 'advanced' features such as raytracing - features that we take for granted.

In the last few years the PC and Mac have largely taken over, and many core Amiga developers have dropped or are phasing out Amiga graphics development; it is a regrettable fact that our beloved platform is ailing, and probably has little hope of recovery. But despite current problems, we have at least a three-year headstart over computer artists using any other machine. What I find depressing is that we have almost nothing to show for it.

If you turn to the Readers' Gallery pages of (a magazine that rhymes with 'Anteater Doormat' - Ed) you will see some of the worst pictures ever put on public display. Any class of five-year-olds would produce more interesting work. That, at least, is my opinion, and in this article I will explain what's so bad about such pictures, and how you can do better.

Talent and Technique

There are two main reasons for bad computer graphics. On one hand, some tools make it very easy to create fairly polished looking work; this leads to pictures utterly devoid of any merit except technical competence, the same problem that occured in print when DTP software became accessible to people with no graphic design training. I remember my excitement when I first used Sculpt 3D and proudly rendered a picture of a mirrored ball on a chessboard. I had made a picture with perfectly realistic reflections and shadows, when on paper I could barely draw cartoon characters. Unfortunately, my critical faculties had gone on hold - it was so magically easy to do that I didn't stop to consider why on earth anyone would want to look at a picture of a ball and chessboard. After all, the world's art galleries are notably free of such pictures, no matter how well executed.

On the other hand, some software has such a steep learning curve that you can't make even simple pictures without immersing yourself in the technicalities of their production. This puts off most creative people, and means that these graphics tools are often being used by more scientific-minded computer enthusiasts. John Lasseter (director of Toy Story) has compared the results to paintings made by the chemists who mix the paint. Again, it is easy to get so caught up in the 'cleverness' of the techniques that you forget to give any thought to the content - the point of the picture.

All pictures, in any medium (in fact, all creative or artistic pursuits, from novels to computer games), can be regarded as having two components; the idea and the execution, or the creative and the technical. It is quite possible to have a good picture that is not very well drawn, and this is almost always better than a bad picture that is beautifully drawn. Of course, what we should be aiming for is a good idea that is also well executed.

Bad Pictures

The first type is the direct copy of someone else's picture. These are usually called things like 'Manga Babe' or 'Garfield' and accompanied by the subtitle, 'drawn freehand in 256 colours'.

Why?

Quite apart from being a breach of someone's copyright, they are pointless because they have already been drawn, and a device called a 'photocopier' can reproduce them quite competently. There is one reason why you might make such copies yourself - as an exercise to help you to study a particular artist's technique, or to improve your technical skills - but on no account should you ever show them to anyone else.

The second regular gallery offender is a subtler form of plagiarism. This is the 3D picture which uses models created by someone else, most commonly either spaceships or a cartoon gangster called Al (the latter was widely distributed some time ago as an advert for the modelling services of Viewpoint Datalabs). Such pictures stand out immediately because of inconsistencies of competence; the central model is beautifully made, but everything else about the picture is clearly the work of a different, less talented, hand - shoddy or non-existant scenery, amateurish lighting and composition.

Again, there are legitimate reasons for using someone else's model. You might have been commissioned to produce a picture or animation featuring a very complex object, say a particular brand of car or a human skeleton, and for some reason you had to do this with computer graphics rather than photography; fair enough, rather than spend weeks trying to build an accurate model, it is easier to pay to use someone else's if it is available. But the gallery entries I'm talking about are supposed to be showcasing your own ability, and so using someone else's model is cheating - and it is doubly senseless when the picture has no other redeeming quality.

The third recurring type, and my current bete noire, is the 'spaceship, planet & lensflare' picture (SPL). This has replaced the 'chrome ball on chessboard' as the number one gallery cliche, and yet such pictures are still being created - and winning prizes. As a rule, these pictures display no artistic merit whatsoever, and are frequently offenders under both of the other categories above, using third-party models and/or being copies of stills from Star Trek or Babylon 5.

If you are ever tempted to create a spaceship, planet & lensflare picture, please just don't. Please.

The common factor linking the three types of pictures described above is that the viewer is supposed to admire the person who created them for being 'clever'. In other words, the creator is showing off something that they are proud of because it required some effort or skill. The problem is that the 'cleverness' was in the technical execution; the result is utterly worthless.

Many of you will join me in wondering why the three types of pictures above are still winning gallery prizes. I suppose it is because the gallery selections are made by computer journalists rather than the magazines' art directors. As to why an art competition is being judged by people with no artistic knowledge or ability, I really don't know. Perhaps we should write to the editors of such magazines and ask them to do something about it. After all, the Amiga press is the public face of the Amiga community; and no doubt some more cynical or stupid Amiga artists are still churning out these pictures deliberately because they know they will win prizes. Perhaps AUI's gallery competitions (which will definitely not be printing any of the pictures described above) can start a trend by rewarding talent and originality.

Think

Therefore, the following suggestions for improving your artwork are nothing to do with technique; they are about professionalism in the thinking behind your pictures.

Before you create your next picture, ask yourself one important question: why? What is the point of the picture? There are many possible reasons: to entertain, to inform, to communicate, to share an emotion or an idea, or merely to decorate. (And if you're being inspired by something that's already been done, you'd better be damn sure that your version is going to improve on it in some respect.)

Once you've decided what the aim of the picture is, you can think about the best way of achieving that aim. This should automatically start to improve your artwork.

While technical proficiency should not normally be the point of a picture, there are a few skilled (as distinct from creative) jobs within the field of computer graphics; for instance, 3D modelling is now a possible full-time occupation. If you are very technically minded, but have no aspiration to be an artist, building 3D models to other people's designs may be a good career; and so your portfolio will need to feature pictures of detailed, accurate models. In this case prospective employers won't necessarily care if the pictures are any good, only if the models are well made, and so it is acceptable to aim for no more than technical competence. However, your chances of getting a job will be far higher if you stand out from the crowd a little, so even if you are aiming for a non-creative job like this I would suggest you put some thought and creativity into your pictures.

Even if your spacecraft is beautifully modelled, is it really going to get the undivided attention of a prospective employer if it's sat between a planet and a lensflare like everybody else's?

Bearing in mind that in this case the point of your pictures is to show how beautifully built your models are, ask yourself how you can best display them to their advantage. One idea would be to look at how well-made, desirable objects are displayed in the real world; perhaps render your models in glass cases as if on show at a museum, or standing on a mirror lit by spotlights like a sports car at an auto show.

Assuming that you accept that the content of a picture is more important than its execution (which I am of course presenting as an opinion, not a fact), the problem then is chiefly one of creativity - the old question of where you get ideas from. Without wishing to sound glib, this is actually very easy. Oops, I just sounded glib. Bah.

All of us have incredible creative faculties that we use, unconsciously, every day. If you find it difficult to think of ideas I would strongly advise you to read a book on creativity - the best one I have read is Impro by Keith Johnstone, in the drama section of most libraries. Read it, read it, read it.

The essential concept behind all creativity-boosting methods is that your subconscious, or 'right-brain', can effortlessly create infinite ideas, but your conscious or 'left-brain' suppresses or stops you from accessing them - in other words, the very act of trying to think of an idea makes it almost impossible for you to do so (don't ask me, I just work here). The solution is to learn methods of tricking or distracting your conscious mind, so you can access the subconscious.

One trick I find useful is to make odd associations. Pairing any two unrelated ideas will often suggest an interesting, disturbing, or amusing image. If your mind goes blank even when you try to think of unrelated ideas, trying writing down a big list of objects and situations and picking two items at random.

Another method for coming up with ideas is to take the opposite approach - follow associations of connected ideas. For example, your subject might be a spaceship - which is a large vehicle - which is in many ways similar to a normal ship or boat - models of which are traditionally displayed in a bottle - so how about putting your spaceship in a bottle, on a futuristic pub mantelpiece? Instead of producing yet another SPL offender, you now have a picture that might raise a smile, and will certainly stand out from the crowd.

Of course, once you've got over the stumbling block of actually having ideas, you still need to cast a critical eye over them and weed out the bad ones. There must have been a time, probably back in the 18th century, when pairing a mirrored ball and a chessboard was daringly original.

Practical Advice

Look for ways of improving the staging and composition. A tilted camera, or a tight close-up, can often add interest to fairly mundane subjects. As a general rule, avoid perfectly symmetrical compositions; they are almost always a bad idea, and often make a picture look dull or lifeless.

If you are inexperienced at lighting shots, research the subject - look at books on interior design or photography for ideas. If you normally use just one light, try adding a second, 'fill' light (make it about a quarter as bright as the main lightsource and position it roughly opposite it, to add interest to shadowed areas).

Don't get stuck with the idea that 2D and 3D graphics are rigidly separate and shouldn't be mixed. Be particularly wary of thinking of raytraced pictures as 'perfect' - they can often be improved by painting over the top or retouching details. Remember that it is the finished picture that must be judged as being good, or clever - not the process of creating it.

As the saying goes, you can't fool Mother Nature. It is a bad idea to try to model realistic organic objects on a computer because we are all experts on their appearance. To date, almost no one has reproduced a living creature on a computer accurately enough to look convincing, because we are so familiar with reality that we will notice the slightest mistake. The effects in Jurassic Park would probably not have impressed anyone if we'd ever seen any real live dinosaurs. This doesn't mean you can't include animals or people in your 3D pictures - but I would suggest that you stylise or caricature them rather than aim for accuracy. Alternatively, if you have to include a convincing creature, scan a photo of a real one and use that; or buy a high-quality model from a library, such as Viewpoint.

Be aware of cliches, and avoid or subvert them. Try to use techniques in unexpected or unusual ways. Experiment, and question basic assumptions you've made about the use of your tools. For example, I noticed that the release of Lightwave led to a glut of pictures and animations of spaceships with gratuitous lensflares, to the extent that I vowed never to use them. Then one day I realised that the lensflares themselves weren't the cliche, it was the way they were automatically being used on rendered science fiction scenes, and a bit of experimentation led to my 'Christmas Angel', which whatever you think of it is at least not cliched (NB - this and other pictures are on my home site, (Nice try. - Ed.)).

If you have even the slightest ambition to earn money from your artwork, enter every AUI gallery competition from now on (see overleaf). You will probably find that some of the themes don't appeal to you; bear in mind that all professional artists have to learn to work to briefs and deadlines whether they like them or not. You will rarely have the luxury of being able to turn down work, and you may as well start getting used to it now.

The Bottom Line

If there is one thing the world has quite enough of, it is mediocrity.

Every picture you show to anyone else should have at least one other good point besides being well-executed. Follow this advice and you will have become more professional, and thus more likely to earn money from your work, than 95% of your competition.

At least, that's what I think; my judgement is probably clouded by my naive belief that art has the power to change people's lives, and you may know better. Please feel free to send in your comments, criticisms and questions (and remember that reasoned argument impresses more than incoherent abuse). I will now step down from this soapbox and quietly get on with the task of starving in my garret.

This article first appeared in Amiga User International.
(c)Paul Hamilton 1996