An introduction to visualisation
Published 02 June 2009
|Written by Jamie Gwilliam|
Page 1 of 2
Jamie Gwilliam, Autodesk’s visualisation and 3D application specialist, discusses some of the basics of composition and what we need to consider when creating a competent computer generated image and animation. This article covers many of his dos and don’ts, learnt during his professional career in architectural and product visualisation.
Before discussing what should form the basis of a good Computer Generated Image (CGI) we need to remind ourselves of the most important part of the process and, an area which is often overlooked, the purpose of the visual. We need to ask ourselves: ‘What are the key points to communicate to the viewer?’ It’s all too easy when working on an image to lose track of the main reason for it. Instead we get carried away with the small, insignificant details. These small technical design details will often add little to the overall impact, mood and effect of the animation or visual. It’s always worthwhile to draw up a list of five to seven important features that need to be communicated. This can be helped by re-visiting the original mood-boards for the scheme or product.
By helping to re-define the important differentiators of the project (for example the strategic balcony view or building’s footprint), we can ensure the imagery enforces the important key points to the viewer and end client. Remember, your job when producing a 3D image is to portray the unique points of the project in the clearest way possible, and not just to make a pretty picture. With photorealism we aim to fool the audience into believing that what they see is real. With technical visuals, our sole aim is to educate the viewer in the clearest way possible.
3D visuals generally fall into one of two categories — to either sell an idea or to improve upon the design. Both areas often require two different styles of visualisation. For example, a marketing image will be quite different to one required for a planning submission or massing model. A good 3D image doesn’t depend on the level of photorealism but on how well it matches its purpose. Once we have outlined our brief and viewers, we start to create our test compositions.
If this is the first visualisation project upon which you are embarking then you should create as many test ‘shots’ as possible. Create quick block previews from all angles within the project. These do not need to be fully materialised renders, but instead can be simple hidden-line screen grabs or greyscale renders. By producing grey renders, we can also check our 3D model for potential defects, which may have been overlooked. At this stage, all we are concerned with is the form and mass. This part of the process should be all about speed and experimentation. Treat the project as though it were a real-world development in which you’re running around with a digital camera. Take as many shots as you feel are necessary, then throw away the ones which don’t work. This will help you learn which angles and compositions work well for your next project. Ensure, however, that you give no more than seven concept shots to the lead architect or developer to choose from. Showing too many concept shots will often just lead to confusion and the client asking for a mix of a few, as they’re unable to make a decision.
As a general rule, dramatic angles with large perspectives work well for tower blocks, whereas small dwellings benefit from a more refined approach where you would use a more natural lens. One concept which sits well with most visuals is to add a ‘Dutch Camera’ or ‘Dutch Angle’ effect. This is an early cinematic effect which often adds interest to an otherwise standard shot. Often, as seen in Figure 1, the horizon line needs to be tilted, and one of the perspective lines to run into the image’s corner (top right). It’s worth noting that this technique works extremely well if you wish to enforce the notion of speed and movement. The tilted angle will always add a level of drama to the visual too. Now look at the grey image (top left), and see how the composition’s impact is lacking. On a side note, this grey look is a great way to test angles, without being distracted by colour. It also has the added luxury of a faster render/production time.
This technique can be seen in many of the visualisation specialists’ work. Beware when creating animations, however, as it is easy to overdo this effect, and can result in a sea-sick end client. Also, experiment with the frame or image size to see what suits the effect. Don’t get stuck in producing a standard A4-proportioned visual. In the same way the Dutch angle will add to the mood of a 3D image, a change in image proportion will result in a different mood. For example letter box proportions add to the notion of speed as a result of the larger horizon line which is available to the viewer.
Try to understand the basics of how real world cameras work. Read up on principles of physical cameras and try to understand the basic terminology. Often the best 3D imagery and artists employ real-world principles. If anyone in your office has a SLR, then these are people who will become great 3D visualisers as they should already have an understanding of composition and common photographic terminology. This includes concepts such as shutter speed and film ISO. Many of these terms are used within the 3D visualisation process and as such, the two worlds are overlapping more than ever. Many visualisation beginners fall into the trap of forgetting they’re in control of the camera and instead do all the shots at a ‘safe’ eye level. For extra drama in an interior image, try placing the camera in the bottom corner of the room and focus the target upwards to the ceiling. This will have the added benefit of making the space seem wider than it actually is, and will look less like a snap-shot.
Much in the same way that we look to cinematography and photography for inspiration, we should also implement their standard photographic and painting techniques. One such method is the ‘rule of thirds’.
This practice of splitting the image into nine equal imaginary boxes requires the horizon line to be placed on or near to one of the box’s horizontal lines. This ensures the horizon is not distracting to the viewer by cutting the image in half. Implementing this simple technique the eye is held within the image for longer. The intersection of the box lines can also be used for the benefit of the 3D artist by adding strong focal points to this area of the visual. This can be illustrated by the two Spanish Antarctic base huts pointing towards the viewer forming a central area of curiosity, in turn retaining the audience’s interest (see Figure 2 overleaf). This remains one of the biggest challenges in visualisation. These meeting points are powerful places to add objects in an interior scene or focus on key exterior parts of the development. Next time that you’re using your compact camera, look out for this grid, on the camera’s viewfinder or screen.
The rule of thirds ensures we are able to hold the viewer’s attention in the image, but by adding subtle elements into the visuals and animations we can control or predict where the viewer’s eye will travel. There are many procedures we can implement, but it’s often the simplest that has the most effect. Now let us look back at the previous expedition imagery. We can see how the audience’s attention is drawn into the image by more of these imaginary visual lines. We can imagine these lines drawn along the two front facing huts, roofs towards the centre of the image, forming a ‘V’ – thus leading us into the important background detail. In this case, the effect is subtle, but can often be more obvious by literal lines created by a road and path or a power cable for a product. It is these simple techniques which distinguish a good, captivating image from an average one.
|< Prev||Next >|