GB2247385A - Video graphics effects - Google Patents

Abstract

A soft pen effect is generated in a computer graphics system by superimposing a plurality of translucent pen lines of varying width. The softness of the line may be varied by adding or subtracting layers or by varying the decrement in pen width from layer to layer. <IMAGE>

Description

VIDEO GRAPHICS EFFECTS FIELD OF THE INVENTION This invention relates to video graphic effects and more particularly to the production of soft pen effects.

BACKGROUND TO THE INVENTION A soft pen is an effect achieved in a 2D painting system in which the translucency of the pen line varies across the width of the line. Typically, a line will have a translucency approaching 1 at the centre (opaque) degrading towards 0 at the line edges.

2D pixel based painting systems achieve soft brush effects via local algorithms while the pen is tracing a path on the screen. 2D object based drawing systems, which are capable of supporting re-editing and animation of a drawing or picture, often ignore soft brush effects.

The present invention aims to provide a technique for obtaining soft pen effects in a 2D object-based graphics system.

SUMMARY OF THE INVENTION This invention is defined by the independent claims, claims 1 and 18.

In its broadest form the invention produces a soft pen effect by superimposing at least two pen layers of varying width.

The innermost or narrowest layer may be fully opaque but the remaining layers are translucent. The translucency of the layers increases towards the edge of the pen line. The opacity profile of the soft pen is linear in a preferred embodiment but may be non-linear. Non linearity may be achieved by applying a selected function to calculate the opacity of each layer or by applying a non-linear function to determine the change in the layer width from layer to layer.

The number of layers is chosen preferably so that there are two layers per pixel.

The invention is not limited to lines drawn by pens but is equally applicable to any line produced by a graphics system, for example, a shape stored in an object library.

An embodiment of the invention will now be described, by way of example, with reference to the accompanying drawings, in which: BRIEF DESCRIPTION OF DRAWINGS Figure 1 is an example of a pen line obtained using a square nib; Figure 2 shows how a soft line effect may be generated; Figure 3 shows a cross-section on the line A-A in figure 2; Figure 4 shows how pen lines with varying opacity profiles may be produced; Figure 5 shows how the pen line of figure 4 may be adjusted to maintain softness with zoom; and Figure 6 shows, in block diagram form, hardware for implementing the invention.

DESCRIPTION OF PREFERRED EMBODIMENTS Figure 1 shows a typical pen line obtained using a square nib. The pen line may also be referred to as a brush stroke.

the pen line illustrated is generated by a method disclosed in our common dated copending application no. claiming priority from Australian application no PK0617 filed on 13th June 1990, the contents of which are incorporated herein by reference.

The present invention is not concerned with the manner in which the brush stroke or pen line is constructed and further detail will not be given here. The pen line may be constructed with a nit of other than square shape, for example of round nib, as described in our common dated copending application number claiming priority from Australian application PK 0619 filed 13th June 1990, the content of which is incorporated herein by reference. The present invention is not concerned with the shape of the pen nib and further explanation will not be given.

The pen line is defined as the region swept out by the pen nib following a path of arbitrary complexity. The pen line in figure 1 produced by a square nib following a parabolic path but the invention is applicable to any pen nib shape following any path.

The soft pen effect is achieved by constructing the pen line from a number of layers of lines of varying thickness. In the figure 2 illustration, the pen line is made up of 4 such lines, the relative line widths are illustrated in figure 3. The profile shown in figure 3 is a step profile which approximates to a trapezoidal profile, the shape of the trapezoid being governed by the softness parameter of the pen.

The softness parameter defines how much of a pen is soft.

A softness of zero means that the pen is hard: it has a square opacity profile. Softness values between zero and one define linearly the proportion of the pen which is soft. This in effect is a set of pens with trapezoidal opacity profiles.

It should be noted that a pen may be hard even if the pen line is semi-translucent. The pen softness is a measure of change in opacity rather than actual opacity.

Two parameters define the layering of the pen to make it soft, the number of layers and the decrement in pen radius (figure 3) for each layer.

The maximum pixel width of the soft part of the pen, derived from the softness parameter defines how many layers are required. It is preferred to have two layers per pixel although the invention is not limited to this configuration and the number may be reduced when rendering at a lower overall image quality.

The decrement in pen radius is the width w of the soft part of the brush divided by the number of layers. In the illustration of figure 3 the decrement is w/4.

The layered pen is generated using a standard scan line rendering technique.

The layers of a soft pen are generated as separate objects during the preparation stage of rendering. During this stage pen line outlines are generated, open curves are closed, the layered pen line is generated and the layer opacities precalculated.

The next stage is the calculation of intersections of objects with scan lines. These intersections are sorted in order of increasing x; the technique is described in more detail in our copending application No claiming priority from Australian application PK 0619 referred to previously.

The final stage is colouring and at this point the separate layers are treated as part of the same pen to minimise opacity calculations which will be described. During colouring, scan line intersections are processed from left to right, and opacity/colour is calculated at each intersection using a reverse painter's technique, which processes objects which intersect in order of decreasing display priority until a fully opaque object or the background is found; Opacity summation is non-linear to ensure convergence on full opacity. At this stage the precalculated pen layer opacities are utilised to obviate opacity/colour calculations for all but the top pen layer.

Opacity is calculated as follows: during preparation the opacity of each pen layer is calculated and stored. The opacity of the top (frontmost and narrowest) layer is the opacity of the pen for an n-layer pen where the top layer is numbered 1 and the bottom n, the opacity of the ith layer is: ith layer opacity = i * (pen opacity/n) this equation hold for a linear softness profile.

During the colouring phase, when the top layer of a layered pen is encountered, the opacity is directly known and any layers below are ignored. Processing continues with the next object below the top layer which is not a layer of the same pen.

Although the technique described provides a linear opacity profile it is possible to construct a soft pen having a non-linear profile by applying the appropriate function when calculating the opacities of each brush layer. The number of layers, the off-set of each layer and the actual opacity of each layer can all be controlled to produce interesting and artistic profiles. Alternatively the increase in line width from layer to layer could vary according to a non-linear function. Figure 4 shows an example of a soft pen with a non-linear width profile.

It was mentioned previously that the number of layers should be chosen such that there are two layers per pixel.

Clearly the resolution and line quality will decrease during zoom. This problem may be overcome by either of the following methods: Firstly, more layers can be added to the pen line so that the layers have the same relative width and resolution quality.

Alternatively the number of layers could be kept constant and the ratio of the width of the top layer to the base layer is adjusted adaptively such that the soft region remains the same size on the screen and has the same resolution quality. Examples of the two techniques are shown in figures 5 a) and b).

Figure 6 shows in outline hardware suitable for deriving layered pen line as described herein. The lines are conveniently defined using a graphics tablet 10 and stylus which inputs coordinate data through serial part 12 to the central processor 14. The processor 14 and the processor accelerator 16 perform the majority of the mathematical calculations required to derive each of the layers and perform the rendering and opacity calculations.

Claims (18)

1. A method of generating a soft line in a computer graphics system comprises superimposing a plurality of lines in layers of different widths, at least all but the narrowest line being translucent.

2. A method according to Claim 1, wherein the soft line comprises a plurality of lines arranged in order of increasing line width.

3. A method according to Claim 2, wherein the increase in line width from layer to layer is constant.

4. A method according to Claims 1, 2 or 3, wherein the softness of the soft line is determined by a softness parameter indicative of the proportion of the line width which is soft.

5. A method according to Claim 4, wherein the line has a trapezoidal opacity profile.

6. A method according to any preceding claim, wherein the softness of the line is varied by varying the number of superimposed layers.

7. A method according to any preceding claim, wherein the softness of the line is varied by varying the change in line width between component layers of the soft line.

8. A method according to any preceding claim, wherein the opacity of a given layer is derived from the opacity of the top layer and, wherein the ith layer opacity is given by: ith layer opacity = i * (top opacity/n)

9. A method according to any of Claims 1 to 7, wherein the opacity of the layers varies non-linearly.

10. A method according to Claim 1 comprising: generating each of the layers of the soft line; calculating the opacity of each layer; calculating the points at which the layers intersect scan lines; sorting the intersection points; and calculating the opacity and colour of the soft line at each of the intersection points for the uppermost layers only.

11. A method according to any of Claims 1 to 9, wherein the narrowest layer is translucent.

12. A method according to any preceding claim, wherein the resolution of the soft line is maintained during zoom by adding further layers to maintain a constant relative line width.

13. A method according to any one of Claims 1 to 11, wherein the resolution of the soft line is maintained during zoom by adaptively adjusting the ratio of top layer width to bottom layer width to maintain the size of the soft portion of the line constant as a display.

14. A method according to any preceding claim, wherein each of the layers is hard having a constant opacity across its width.

15. A method according to any preceding claim, wherein the number of layers is twice the pixel width of the soft part of the line.

16. A method according to any preceding claim, wherein the soft line is generated by a pen.

17. A method according to any of Claims 1 to 15, wherein the soft line is generated from the outline of stored object.

18. Apparatus for generating a soft line in a computer graphics system comprising means for generating a plurality of lines of varying width and means for superimposing the generated lines to produce a soft line, wherein at least all but the narrowest line are translucent.