Drawing is a system of design. Neither the appropriate choice of viewpoint nor beauty of technique is sufficient without a concern for composition. In composing a drawing, we manipulate the fundamental graphic elements of line, shape, and tone into figure-ground patterns that are coherent and which convey visual information. Through the organization and relationship of these elements, we define both the content and the context of a drawing. Planning this composition is therefore critical to the message it communicates.
The first step in composing a drawing is to determine the shape, size, and proportions of its field relative to the dimensions of the drawing sheet or board. This field should be large enough to incorporate a portion of the design context as well as space for the drawing title, graphic scale, and associated symbols.
The field of a drawing may be square, rectangular, circular, elliptical, or irregular. Rectangular fields are the most common and may be oriented either vertically or horizontally. Regardless of the shape of a drawing’s field, certain fundamental principles apply to the organization of elements within it.
Design drawings are reduced versions of full-size objects or constructions. In selecting an appropriate scale for a drawing, there are several factors to consider.
First, there is an obvious relationship between the scale of a drawing and the size of the drawing surface. The larger a design, the smaller its representation on the sheet or board; the smaller the design, the larger its scale can be. Also influencing drawing scale is the manner in which drawings are laid out in a presentation. For example, when plans, sections, and elevations comprise a set of cross-referenced information, their scale must enable the entire set to fit on a single sheet or board.
Second, the scale of a drawing regulates the perceived distance between the mind’s eye of the viewer and the representation of a design. Close-up views provide a detailed look at the features of the subject. Small-scale drawings increase this perceptual distance but enable the entirety of an idea to be grasped quickly. At the same time, these distant views minimize the amount of detail that can be depicted.
Large-scale drawings, on the other hand, are close-up views that allow a greater degree of detail and complexity to be revealed, as well as a greater range of tonal values to be rendered. As the scale of a drawing increases, the amount of detail required for legibility and credibility also becomes greater. Not including enough detail at larger scales can make a drawing look sparse and diagrammatic.
Finally, the scale of a drawing influences the type of drawing tool and technique we use. Fine-tipped instruments, such as pens and thin-lead pencils, encourage drawing at a small scale and enable us to focus on fineness of detail. Broad-tipped tools, such as color markers and charcoal, promote large-scale drawing and discourage the study of small-scale features.
Resolution refers to the ability of our visual system to resolve or distinguish two objects—from the scale of an on-screen pixel to a truck driving down a highway—even though they are very close together in our field of view. In drawing, our ability to resolve distinctions in a composition of lines, shapes and tonal contrasts is important to our reading of the image, which ultimately depends not only on how the image was created but also its size and the distance from which we view it.
The interaction between drawing medium and surface determines the relative smoothness or coarseness of a hand-drawn image; the results are immediately discernible to the eye and can be evaluated for proper levels of contrast and detail. Knowing the nature of the subject matter, the size of our drawing, and the distance at which it will be viewed, we can determine how smooth or coarse the graphics can or should be. For example, our eyes can resolve the textural qualities of the charcoal dust collected on the rough surface up to a certain distance, beyond which the distinctive pattern of darks and lights begin to blur and form, to the eye, smoother tonal gradations. On the other hand, the details of a small drawing done with a fine-tipped ink pen must be scrutinized from a relatively close distance to be appreciated.
While the size and resolution of an original hand drawing are self-evident, a digital image can vary in both size and resolution, depending on how the image is captured and the method by which it is output for viewing. This relationship between size, resolution, and visual texture is an important issue to understand when using digital images for a presentation. Depending on whether we are scanning, displaying or printing, we measure and express digital resolution in terms of samples, pixels, or dots per inch.
The following section refers specifically to raster images, which are made up of a rectangular grid of pixels and are resolution-dependent. Vector graphics, on the other hand, use mathematically based geometric primitives such as point, lines, curves and shapes to construct digital images. Vector images are resolution-independent and can be scaled more easily up to the quality of the output device, whether a monitor, projector, or printer.
To reproduce an image, a scanner uses a charge-coupled device (CCD) or other sensor to sample portions of the original image. The higher the samples per inch (SPI), the higher the resolution of the scanned image and the more faithful the scan is to the original. Many manufacturers use dots per inch (DPI) in place of SPI in specifying the resolution capability of their scanners, but technically there are no dots in the scanned image until it is printed.
When scanning a hand drawing or photograph, we should know the final output method to insure that we scan at the proper resolution. For example, a scanning resolution that may be acceptable for posting on the web could result in poor quality printed output.
Scanners produce raster images whose scanned resolution can be resized and resampled using image editing software. Because most scanned images require some type of image editing, scanning at a slightly higher resolution is often advantageous. It is easier to delete unneeded resolution after scanning than it is to restore lost resolution during editing.
Digital cameras, like scanners, use an electronic sensor to capture images. Camera resolutions are usually expressed in terms of megapixels or how many millions of pixels it can record in a single image. For example, a camera that captures 1600 × 1200 pixels produces an image with a resolution of 1.92 million pixels, which is rounded up to 2 megapixels for marketing purposes.
Higher camera resolutions provide more pixels to work with when creating large prints or cropping images.
When creating images for on-screen display or posting on the web, we should be thinking in pixels per inch (PPI). Computer monitors typically display images at 72 or 96 PPI but high-resolution monitors may display more pixels per inch. Creating and scanning images beyond the screen resolution of a monitor is a waste of image data if the image will never be printed and unnecessarily increases file sizes and download times. If an image is to be printed at the same size as the original or larger then increasing the scanning resolution of the image will provide us with the additional image data needed to maintain good quality output resolution. Note, too, that the same picture on a low-resolution monitor looks larger than it does on a higher resolution monitor because the same number of pixels are spread out over a larger area.
Print resolution, measured in dots per inch (DPI), refers to the dots of ink or toner an imagesetter, laser printer, or other printing device can place within an inch to print text and graphics. Most printers print the same number of dots horizontally and vertically. For example, a 600 DPI printer will place 600 tiny little dots across one inch horizontally and 600 dots in a vertical inch.
In general, the more dots per inch a printer is capable of outputting, the sharper and clearer the printed image. Correspondingly, the lower the DPI of a printer, the less fine detail it can print and the fewer shades of gray it can simulate. Because screen resolution is typically lower than print resolution, low-resolution images that look fine on-screen will almost always print poorly.
The quality of printed output depends not only on the resolution of the printer but also the type of paper used. Some types of paper absorb ink more readily than others, resulting in the ink dots spreading out (dot gain) and effectively reducing the DPI of the image. For example, because ink spreads more on newsprint, the ink dots must be further apart than on high-quality coated paper, which can accept more closely spaced ink dots.
In practice, SPI and PPI are often used interchangeably, and DPI is frequently used in place of one or both terms. However, each sample, pixel or dot in a digital image behaves differently, depending on whether it is scanned, viewed on screen, or printed. When working with digital images, a particular challenge is reconciling the difference between the size and resolution of a scanned image, how the image appears on-screen, and how it will print.
To illustrate this:
When printed at 300 DPI, the first two digital images produce the same 3" × 5" prints but the 300 PPI image will look better than the 96 PPI image because it squeezes in more dots per inch. A 6" × 10" 300 DPI print of the third image spreads the same number of dots as the first image over a larger area. This is a method for producing a larger print size of an original image while maintaining an overall sense of quality.
When printed at 600 DPI, the three digital images produce prints that vary greatly in size.
When printing presentation sheets or boards, a range of resolutions from 150 to 300 DPI will produce good- to high-quality output. Resolutions higher than 300 DPI would increase the print quality but the degree of enhancement may not warrant the larger files sizes. On the other hand, resolutions less than 150 DPI may result in coarse or blurry images that lack detail and subtle variations in color and tone. The range of 150–300 DPI, therefore, is a general guideline and can be adjusted depending on print size and method of printing.
Presentations that are to be viewed on-screen or posted to a website can have a lower resolution than for print because the majority of displays have a pixel density of between 72 and 150 PPI. These monitors would not be able to display any additional pixel information beyond their native screen resolution. Although advancements in technology are increasing the pixel density of monitors, image resolutions between 100 and 150 PPI are generally enough for good image quality. For presentations designed to be projected as a slide show or animation, the resolution should match the resolution of the digital projector.
Note that digital resolution also depends on viewing distance. An image that looks pixelated when viewed up close might appear to be of high quality if large enough to be viewed from a greater distance.
In addition to adjusting the resolution of a digital image, we can also crop the image to alter its size, proportions, and figure-ground relationships. Cropping a digital image retains the desired portion of the image and cuts the remainder. Masking, on the other hand, involves creating a window through which we view a selected portion of the image. The size, shape and position of the opening controls what we see and do not see of the original image.
Raster images are usually cropped, while vector images are typically masked. Once cropped, a raster image cannot regain the material that was cut. A masked vector image is more flexible since we can manipulate and adjust the size, shape and position of the mask.
The size of a graphic image relative to the size of its field determines how we read the figure.
Situating the drawing in a large field emphasizes its individuality. The space between a drawing and the edge of a sheet typically should be similar to or larger than the dimensions of the drawing.
If we enlarge the drawing or reduce the size of its field, its figure begins to interact visually with its background. The field begins to have a recognizable shape or figural quality of its own.
Enlarging the drawing or reducing its field still further establishes an ambiguous figure-ground relationship in which the field elements can also be seen as figures.
When a paraline drawing, perspective drawing, or other graphic image does not have a rectangular shape, it tends to float in its field. We can stabilize the image with either a title block or a horizontal band of color or value.
When framing a drawing, avoid using a double or triple mat. Doing so can create the impression of a figure on a background that itself has a background. Therefore, attention would be diverted away from the figure, where it belongs, to the frame around it.
Drawing composition concerns the relationships among the parts of a graphic image rather than the rendering of any particular part. We can apply certain principles of visual design to regulate the organization of a drawing composition to promote a sense of order and unity.
The following ordering principles, in promoting unity, do not exclude the pursuit of variety and visual interest. Rather, the means for achieving order are intended to include in their patterns the presence of dissimilar elements and characteristics.
In scanning an image, the eye is attracted to certain graphic elements. The eye seeks out areas of:
We can also stress the importance of an element by isolating it within the drawing composition. We use these points or areas of interest to define the focus of a drawing. In each case, a discernible contrast must be established between the dominant element and the subordinate aspects of the composition. Without contrast, nothing can dominate.
There may exist not one but several focal points in a drawing. One may dominate while others serve as accents. We must be careful that multiple centers of interest do not cause confusion. When everything is emphasized, nothing dominates.
In any drawing, there will naturally be a mix of shapes and tonal values. How we organize these elements should result in a visual sense of balance. Balance refers to the pleasing, harmonious arrangement of elements or proportion of the parts in a design or composition. The principle of balance involves reaching equilibrium among visual forces of weight, compression, and tension in a drawing.
There are two principal types of balance: symmetrical and asymmetrical. Symmetry refers to the exact correspondence in size, shape, and arrangement of parts on opposite sides of a dividing line or axis. Bilateral or axial symmetry results from the arrangement of similar parts on opposite sides of a median axis. This type of symmetry leads the eye to the mediating axis in a quiet manner.
Radial symmetry results from the arrangement of similar radiating parts about a center point or central axis. This type of symmetry stresses the centerpoint or middleground of a composition.
We recognize asymmetry by the lack of correspondence in the size, shape, or tonal value of elements in a composition. In order to achieve visual or optical balance, an asymmetrical composition must take into account the visual weight or force of each of its elements and employ the principle of leverage in their arrangement. Elements that are visually forceful and attract our attention must be counterbalanced by less forceful elements, which are larger or placed farther away from the center of gravity of the composition.
Explore ways to compose this fragment of a Spanish town within a larger drawing field. How could you compose the view to emphasize the town’s position atop a hill? How could you alter the composition to stress instead its relationship to a distant mountain range?
Explore ways to compose this view of the Sydney Opera House, designed by Jørn Utzon in 1956, within a larger square or rectangular field. How could you place the structure to emphasize its soaring roof forms as well as their relationship to the harbor on which they front?
Shown is a plan diagram of the Capitol Complex at Islamabad, Pakistan, designed by Louis Kahn in 1965. Explore how you could achieve a balanced composition within a rectangular field. How could you maintain the same balance within a square field? How would rotating the plan 90° affect the compositional possibilities?
Harmony refers to consonance—the pleasing agreement of the parts in a design or composition. While balance achieves unity through the careful arrangement of both similar and dissimilar elements, the principle of harmony involves the careful selection of elements that share a common trait or characteristic:
Perhaps the most natural way to produce harmony in a drawing is the use of a common medium and technique throughout the composition. Employing the principle of harmony too rigorously can result in a unified but uninteresting composition. Drawings need diversity as an antidote to monotony. But variety, when carried to an extreme for the sake of interest, can result in visual chaos and a fragmented message. It is the careful and artistic tension between order and disorder—between unity and variety—that enlivens harmony. Stability and unity emerge from stimulating the action of contrasts as well as the union of similarities.
There are times when we may want to combine both a hand drawing and a digitally created representation into a single composite image. When doing so, we should be careful to control the style, range and contrasts in line weights and tonal values to ensure a harmonious relationship exists between analog and digital drawings.
A range of digital techniques exist for modeling and simulating the lighting of three-dimensional forms and spaces. The simplest approach is ray casting.
Ray casting is a technique that analyzes the three-dimensional geometry of forms and determines the illumination and shading of surfaces based on their orientation to an assumed light source. The primary advantage of ray casting is the speed with which an illuminated three-dimensional image or scene can be generated, often in real-time. This makes ray casting a useful tool in preliminary design to study the solar consequences of the massing and composition of building forms and the shadows they cast. See pages 166–167 for examples.
Ray casting, however, does not take into account the way light travels after intersecting a surface and therefore cannot accurately render reflections, refractions, or the natural fall off of shadows. For this, ray tracing is necessary.
As a ray of light travels from its source to a surface that interrupts its progress, it may be absorbed, reflected, or refracted in one or more directions, depending on the material, color, and texture of the surface. Ray tracing is a digital technique for tracing these paths to simulate the optical effects of illumination.
Local illumination is a basic level of ray tracing that is limited to direct illumination and the specular reflections of light rays. While local illumination does not take into account the diffuse inter-reflection of light among the surfaces in a three-dimensional space or scene, some ray tracing programs can approximate this ambient light in their lighting algorithms.
A better predictor of how a space would be illuminated by any number of light sources is global illumination. Global illumination techniques use sophisticated algorithms to more accurately simulate the illumination of a space or scene. These algorithms take into account not only the light rays that are emitted directly from one or more sources. They also track the light rays as they are reflected or refracted from one surface to another, especially the diffuse inter-reflections that occur among the surfaces in a space or scene. This enhanced level of simulation comes at a cost, however. The process requires time and is computationally intensive, and should therefore be used only when appropriate to the design task at hand.
When using color in design drawings, we should be careful to consider the range of hues, intensities and values and how they are distributed across an image. Among these properties of color, value is the most critical in how we perceive an image’s compositional elements and relationships. Areas of high contrast attract our attention more emphatically than areas of low contrast. High-key images having a predominance of light values appear delicate, airy and ethereal. Low-key images having a predominance of darker values impart a moody, earthy feel.
The intensity of hues should be in proportion to the scale of the image or model. Just as a full-size image is scaled down to fit the size sheet or board, the intensity of colors should also be toned down in a model.
When specifying colors in a digital environment, it is important to consider the type of output for which we are designing. For digital monitors and projectors, the array of colored lights is produced in an additive manner. For printed output, the color pigments produce the range of colors through a subtractive process.
RGB is an additive color model in which white is reproduced by the superimposition of the three primary colored lights—red, green and blue—and black is the absence of light. Red, green and blue lights can be added together in various ways to reproduce the spectrum of colors we see. The main purpose of the RGB color model is for the sensing, representation, and display of images in electronic display systems, such as digital cameras, scanners and projectors, computer monitors, and televisions.
When we enlarge a digital image we can see that the image is in fact made up of a large number of pixels, each with their own color and value, which is determined by the intensity and optical combination of the three sub-pixel colors: red, green and blue. Varying the intensity of the red, green and blue optical colors will produce the full range of colors we use within the digital environment. Typically, the intensity of each color is divided into 256 levels of intensity—calibrated along a scale from 0 to 255. An intensity of 0 relates to no intensity of the color, while a level of 255 indicates full intensity of that color. As such, an RGB value of 0,0,0 would result in the color black (no light intensity for any of the colors) and an RGB value of 255, 255, 255 would result in the color white (full intensity for red, green and blue). Each color in the digital spectrum is assigned a specific RGB value, which indicates the intensity of each of the three optical primary colors: red, green and blue.
RGB is a device-dependent color space—different devices detect or reproduce a given RGB value differently, since the color elements (such as phosphors or dyes) and their response to the individual R, G, and B levels vary from manufacturer to manufacturer, or even in the same device over time. Thus an RGB value does not define the same color across devices without some kind of color management system.
CMYK is the acronym for the four colored inks used in the printing process—cyan, magenta, yellow and black. CMYK is a subtractive color model because the four colored inks used in color printing—cyan, magenta, yellow and black—subtract brightness from the typically white background of the paper, with black resulting from the full combination of colored inks. Each of these colors absorbs certain wavelengths of light, with the colors we see being the colors that are not absorbed. By using a halftone of dots for each color, the full spectrum of printed colors can be achieved.
In the digital environment, tonal values are displayed in either an additive manner using light on a monitor or in a subtractive manner using the pigment from a printer or plotter. On a display screen, the intensity of light shown in a pixel will determine the tonal value. There are 256 levels of light intensity which create a corresponding 256 distinct tonal values of gray—with an intensity level of zero corresponding to black and a level of 255 corresponding to a color of white (full light intensity).
Since we design and evaluate architecture in relation to its environment, it is important to incorporate the context in the drawing of a design proposal. In each of the major drawing systems, we do this by extending the ground line and plane to include adjacent structures and site features. In addition to the physical context, we should indicate the scale and intended use of spaces by including human figures and furnishings. We can also attempt to describe the ambience of a place by depicting the quality of light, the colors and textures of materials, the scale and proportion of the space, or the cumulative effect of details.
These elements are simply parts of a greater whole, and the amount of interest and attention we give them should be proportional to their importance in the overall composition. Therefore, the following guidelines apply to the drawing of contextual devices:
The viewer of a drawing relates to the human figures within it, becomes one of them, and thus is drawn into the scene. Therefore, in the drawing of architectural and urban spaces, we include people to:
The figures we use to populate a drawing should be in scale with the environment. Therefore, we need to draw human figures in proper size and proportion. We can divide the standing human figure into seven or eight equal parts, with the head being 1⁄7 or 1⁄8 of the total body height.
In orthographic multiview drawings, we can simply scale the five- or six-foot height. Remember that in orthographic projection, the height and width of elements remain constant regardless of their depth within the projected view. We can also scale the height of human figures in paraline views. Since the view is downward, the figures should have some degree of roundness to indicate their volume.
In perspective drawings, the placement of human figures can indicate not only spatial depth but also changes in level. It is generally easiest to begin by locating where each person is standing. Then extend this spot vertically and place the eyes of the head of each figure on the horizon line. Once the height of a figure is established, we can use the principles of linear perspective to shift the figure right or left, up or down, or into the depth of the perspective. Figures above or below the level of the spectator should first be sized as if on the same level, and then shifted up or down as required. When drawing people in a sitting position, it is usually best to first draw a person standing alongside the chair. Then use the established proportions to draw the same person sitting down.
The human figures we employ to indicate scale and use also become important elements in a composition and should not conceal or distract from the focus and essential features of a drawing. Utilize both groups and solitary figures and the principle of overlap to convey depth.
We indicate activity in a drawing by the number, disposition, posture, and dress of human figures. The figures should convey the nature of the activity and be appropriate to the place and setting. The manner in which we draw them should answer the fundamental question: What activity should go on in this room or space?
We can create digital figures from photographs using image-processing software as well as retrieve them from online resources. The same principles that govern the scale, clothing, placement, and gesturing in hand drawing should apply to the use of digital images of people in architectural settings.
The ability to produce photorealistic images of people is seductive. Keep in mind that the graphic style with which we populate architectural drawings should not distract or detract from the architectural subject matter. The figures should have a similar level of abstraction and be compatible with the graphic style of the drawn setting.
Bring a pen, pencils, and a sketchbook to a public place where people gather. Practice drawing the people you see. Draw people standing as well as sitting; sketch small, distant figures as well as closer ones. Begin by working out the structure, proportions, and gesture of each individual, then build up a sense of volume, and finally add necessary details. Start slowly in the first session. In each succeeding session, gradually shorten the time you take to draw each figure and reduce the amount of detail accordingly.
In this linear perspective, use analytical lines and the principles of convergence to transfer the human figure to locations A, B, C, and D.
In addition to people, there are other elements we can use to suggest the context of a drawing. These typically include topography and entourage—the landscaping and other environmental features shown in the rendering of a building.
In addition to indicating scale, trees and other landscape features portray the geography and character of a site, whether hilly or flat, wooded or barren, urban or rural. This entourage should never compete with but rather act as a foil for the design that is being illustrated.
Drawing trees and shrubs follows in a similar manner to how we construct a scene. We begin with the branch structure, following the pattern of growth from the ground upward. To this framework, we can add the overall shape and massing of the foliage, paying close attention to texture, tonal value, and degree of opacity and transparency. Be economical. The amount of detail rendered should be consistent with the scale and style of the drawing.
Trees and other plant materials are important means of providing tonal values and textures in a drawing. How we portray these natural elements is therefore a consideration in planning the tonal range and pattern of a composition.
When drawing trees, pay attention to structure, shape, scale, and purpose.
Image-processing software provides the means to manipulate photographic images of an existing site and landscape and adapt them for use in describing the context for an architectural design.
As with digital images of people, the ability to produce photorealistic images of trees and other landscape elements can be seductive. Keep in mind that the graphic style of site and contextual elements should not distract or detract from the architectural subject matter. Their graphic description should have the same level of abstraction and be compatible with the graphic style of the drawn setting. This can be done by adjusting the transparency, brightness/contrast and color saturation of these images. Multiple filters can also be used on these components to mute the level of detail of the context in order to match the detail in the rest of the drawing.
Bring a pen, pencils, and a sketchbook to a public park. Practice drawing a variety of trees and other plant life you see. Draw small, distant trees as well as closer ones. Begin by blocking out the branch structure of the subject. Over this framework, build up the shape, texture, mass, and tonal value of the foliage.
Draw a series of timed sketches of a deciduous tree from direct observation. Begin with a 5-minute drawing, then do a 3-minute sketch, and end with a 1-minute sketch. Build up each drawing from structure to shape and tonal value. Repeat this exercise for a conifer.
Draw a series of sketches of a deciduous tree from direct observation. Begin by drawing the subject from a distance of 25 feet. Move to a distance of 50 feet and draw the same tree again. Draw the tree one more time from a distance of 100 feet. Each time you move farther away, pay attention to how the foliage shifts from texture gradient to a shape of tonal value. Repeat this exercise for a conifer.
The type and arrangement of furnishings are important indicators of use and activity in a space. Their placement reminds us that there should be places on which to sit, lean, rest our elbow or foot, or to simply touch.
Drawing furniture in conjunction with people helps establish their scale and maintain the proper proportion of their parts. Except when furniture is the subject of a design, use real, well-designed examples, and proceed from their geometric basis. Once the structural framework for the form is established, we can add indications of material, thickness, and details.
We include a variety of vehicles—cars, trucks, buses, and even bicycles—to indicate roadways and parking areas in exterior scenes. Be realistic with their placement and scale.
Drawing vehicles in conjunction with people helps establish their scale. Use real examples whenever possible, and as in the drawing of furniture, proceed from their geometric basis. If we overdraw these elements and include too much detail, they can easily distract from the focus of a drawing.
