Composition in PhotoPaint 10

Draw will be serving PhotoPaint throughout so let's set up Draw to match the specs we need for our PhotoPaint document.

In PhotoPaint, open a new document, 700w x 700h x 72ppi, black background.

figure 1 : Page Options in Draw

Draw is a little more involved :

• Open the Options ('Ctrl J'), navigate to the 'Page | Size' options, and emulate the settings of figure 1. Click on the 'Save Custom Page' option and call it 'Shapes' (this will allow you to recall it easily from the Workspace property bar).
• Open the default RGB Palette (Window | Color Palettes). We'll be using it exclusively.
• Most importantly, use the same Color Management settings in Draw as you have them set up in PhotoPaint (Tools | Color Management). This will ensure a minimum color shift between copy/pasting across applications.

We'll be dancing back and forth between the two programs so - if your system will allow - it's in your best interest to leave both open throughout.

Now, take a good look at the finished image at top (and, of course, refer to it as often as you wish); We'll be building this image from the bottom up (from the Object Docker's viewpoint), introducing each shape as they arrive. For now, notice especially that most of the background objects are weaker (less chroma/saturation) whereas the foreground objects (the 'snake', 3 small shapes, and text objects) are bright and bold (more chroma/saturation). Note also there is much more 'weaker' information making up this document than there is 'bolder'. All of this is on purpose to get the viewer's attention according to what we want them to focus attention on! More specifically it actually allows the viewer to focus his attention on the impactual elements without becoming overly distracted with their surrounding. All are very important aspects for creating simple composition : show the viewer what you want him to see then communicate it with color and form (and words, if necessary). Alright, enough. Let's get on with it...

Starting in PhotoPaint, add a new object (Object | Create), rename it 'backdrop' then set the Fill well to 90% black and fill it with the Paint Bucket ('F').

We'll focus on the first three objects : the torus (donut!), the large cube, and the array of lines. The torus was created in a 3D program and is available as part of the download files. Get to the Open dialog and open 'torus.psd'. Copy/Paste the torus into your document then center it (if necessary). Save your work and let's move into Draw.

figure 2 : Creating the Cube in Draw

The Cube. Figure 2 has the shape and it is simple. But I assume nothing in the exercise so here's how it's done.

In Draw, place a vertical guide at 350 pixels. Select the Rectangle tool and create a perfect square, 190 x 190 px; convert it to curves ('Ctrl Q'). Enable 'Snap to Guides' (View | Snap to Guidelines) and snap the right side of the rectangle to the guide.

Press the Space bar (for the Pick tool) then click on the square to get the skew & rotation handles. Ctrl-drag the left skew handle up -30 degrees (look at the Status bar for input).

Click on the square again and press NUMPAD+ to duplicate. Ctrl-drag the duplicate to the right until it snaps to the guide.

Make a duplicate of this rectangle and drag it straight up until the bottom of this shape meets with the top of the former (figure 2a).

Enable 'Snap to Objects' now. Using the nodes, drag the upper-left node to the upper-left corner of the original rectangle, then drag the upper-right node to replace the position of the previous node (figure 2b & c).

Remove the guideline and disable 'Snap to Guide'. Click on the Interactive Fill Tool ('G'). Utilizing the 'Snap to Objects', drag out a linear fill across the top rectangle (figure 3). In the property bar set the two colors to : From R105 G70 B35, To R204 G136 B68. Fill each of the remaining boxes using the same fill but according to the direction implied by figure 3. Group all three sides and save your work.

Throughout this exercise, when not mixing grays we will be mixing Reds & Browns (the Browns will become various once we get back in PhotoPaint).

The Array of Lines. Add a page to this document then select the Freehand tool. Create a horizontal line, 200 px in length, set the Outline Width to 2.0. Input the position coordinates to x 150, y 550.

Press the Space bar and click the line for the skew-rotate handles. Ctrl-drag the left skew handle up 45 degrees (figure 4a). Duplicate then mirror the line horizontally; set the y-coordinate to 50 (figure 4b). Select the Interactive Blend tool, drag from one line to the other then change the number of steps to 31 (figure 4c). Save your work.

Now, copying/pasting one object at a time, place the cube and line array into your PhotoPaint document and take a look at figure 5. I've set up two visual clues for setting the coordinates of each object (look at the final piece, also). The 'torus' object is centered and stays there. Arrow #1 in the figure points to the angle of the line matching the angle of the cube. Arrow #2 points to where the line, cube, and torus all meet. Hopefully, this will guide you nicely to the approximate position.

Figure 5 : Arranging the Objects

When the dust has settled, rename the objects : 'cube_lrg' and 'line_array' accordingly. The 'line_array' object should be on top when viewing the Object Docker. Let's move on to blending these.

Blending - stated simply - is to have an object affect the object(s) below it, or behind it. Two distinct areas for blending objects are : 'Merge Mode' and 'Opacity'. We will focus on these two properties but work mainly with opacity. In themselves, they are very powerful for composition. Both of these are - of course - in the Objects Docker.

Select the 'torus' then reduce its Opacity to 9%. Because its color and texture is so intense (high chroma/saturation) we can go very low and still the object is clearly visible. Select the 'line_array' and reduce its Opacity to 80%.

For the 'cube_lrg' we want to modify it first by intensifying its range of color. Go to Image | Adjust | Contrast Enhancement. Set the Input Value Clipping to 70 / 150 (figure 6). This will spread the tonal range of our browns and add some wonderful saturate value to them (pushing them toward the reds).

Once you're finished, reduce its Opacity to 12%. Witness how, even with the low opacity, the hue is still retained throughout the object. But because of the low opacity settings, it interacts seamlessly with the other objects. The result? A nice overall blending of all our objects (all visible but none overt) as well as a splash of low-value color (the cube). This approach allows color to be introduced to the viewer while not taking away the overall input of the bold elements that will be introduced later in the exercise. Save your work.

Let's get back into Draw to create the sequence of dots that are located on the right-hand side of the image. The grid of circles will be made using the blend tool.

Create a new page in your Draw document and start by creating a perfect circle, 26 x 26 px. Fill it with black and remove the outline. Finally, set the position coordinates to x 0 and y 700.

Use the NUMPAD+ keystroke to create a duplicate and set the x-coordinate for this object to 700.

Select the Interactive Blend Tool and drag from dot to dot; set the number of steps to 23 for a total of 25. Press the Space bar, go to Arrange | Break Blend Group Apart; ungroup all and the group the objects (figure 7a).

When a blend is broken apart we end up with 3 objects : the middle object being the grouped array created by the blend. To do a blend on all the objects (forthcoming) - as a group - we must ungroup all first, then re-group them.

Make a duplicate and set the y-coordinate to 0 (figure 7b).

With the Blend tool you will now be able to blend both rows of dots, creating a large grid (figure 7c). Set the number of steps to 23, as before.

Just as we had done earlier, break apart the blend then ungroup; but leave them all ungrouped this go round.

Press Esc to deselect the grid and consider : we're going to create our own style of gradient effect. We won't be using the Gradient tool as it won't provide what we need. Rather, we are going to be selecting square 'blocks' of this grid and gradually apply color value to them. It's a bit of a tedious task but the effect is fun and (more) interesting.

Figure 8 shows what we will look like once finished. The idea is to begin a selection from between the dots (figure 9). There's enough gap between the dots to do so. You'll select an entire 'square' of the grid then add the color. Continuing this procedure will be to select smaller 'squares' of dots, changing the color each time until you come to the point of deleting a final 'square' of dots at the center of the grid. There will be a bit of a trick toward the final stages so work with me as we go.

What follows is the layout for the selection process in rows/columns, attributing the colors accordingly :

Row/Column	Color
1	Black (already there)
2	80% black
3	70% black
4-5	60% black
6	40% black
7	30% black
8	Red
9	White

Once you've done the white 'square' you'll no longer have a visual. But because we're going to need it anyway, we'll create a black rectangle the size of the document to bring the white into focus.

Double-click on the Rectangle tool. This automatically supplies our rectangle to spec. Fill it with black and remove the outline, then send it to back (Arrange | Order); finally, lock the object (Arrange | Lock Object).

Now that we can see our white, select the final 'square' to delete it (figure 10 is a partial close-up).

Group all of the dots; select and unlock the rectangle. We'll use it as a container for the grid.

Select the grid of dots, go to Effects | Power Clip | Place Inside Container.... Click once on the black area at the center with the large arrow. And that's it! Save your work.

To get the 'globe' look we'll use a Fish-eye lens attached to a circle. Select the Ellipse tool, create a perfect circle, 600 x 600 px. Set its xy-coordinates both to 350 and remove the Outline.

Open the Lens Docker (Window | Dockers) and choose 'Fish Eye' with 75% Rate; click 'Apply' if necessary. Save your work and double-click on the Pick tool to select everything.

Make a copy of the grid of fish-eyed dots and migrate over to PhotoPaint. You'll want to paste these into a new document (Edit | Paste). We'll need to manipulate these guys a bit before introducing them into our original document. Note the Docker pastes it as 'Object 1'.

In order for the dots' background to be transparent, we need to knock out the black info. We'll do this with the Color Mask dialog.

Go to Mask | Color Mask... and set the properties as you see them in figure 11.

If you've not dealt with this dialog before and wonder how to get the Black into the color well...double-click on the well and the Color Mixer will appear. Select your Black and you're done.

You should have a bunch of marching ants doin' the wave. Cut the selection ('Ctrl Sh UpArr') then delete 'Object 2' in the Docker. You'll be left with the dots only with just a trim of black surrounding them. No worries. That black will blend just fine (figure 12).

As you can see, our original row/column of black dots are missing and that's what we want. Because of this, though, we want to trim away the excess around the dots; likewise we will be using only half of these guys so do some more trimming besides.

Select the Rectangle Mask tool ('R'), set the Mask Style to 'Fixed Size', 670 x 670; click in the upper-left corner then center it (Mask | Align...). Go to Image | Crop | Crop to Mask to trim it up; remove the mask.

We now need the greater half of the grid on the right-hand side. What I mean by 'greater half' is we want to include the middle columns in its entirety. We can accomplish this with the Freehand Mask tool easily. Select it ('K') and mask out the dots accordingly (figure 13, partial close-up).

Once you're satisfied with your mask, make a copy ('Ctrl C'), migrate over to your original document and paste; rename the object, 'dots'. Open the Align and Distribute dialog (Object | Arrange) and check off 'Center - Horizontal' and 'Right - Vertical'. This gets us quickly over to the right but we don't want to be all the way over. What we do want is for the shape of the sphere to match the edge of the 'torus' object. So use the nudge arrows to move the dots left until it meets up with it. It's important here to get it a close as possible as the next step will require it.

We're going to enhance the curve of the spherized dots by stroking a mask with the Art Brush. Ensure that 'grid_dots' is still selected then enable 'Lock Object Transparency' (at the foot of the Objects Docker).

Select the Circle Mask tool ('J), set it 'Fixed Size' at 600 x 600; click one in the upper-left hand corner of the document then center it with the Align dialog (Mask | Align; figure 14).

Click on the Art Brush ('F5') and update the following properties : Paint Mode = Multiply, Nib Size = 160, Transparency = 90, Soft Edge = 100. Set the Paint well to Black. In the Art Brush properties, click on 'Stroke a Mask with a Brush' then choose 'Inside of Mask' from the options. Click OK, remove the mask and disable 'Lock Transparency'.

One final step for the 'grid_dots' : Set the Merge Mode to 'Add' and the Opacity to 50%. Add mode will liven up the color that is partially lost via the level of Opacity and, at the same time, allow the texture of the 'torus' object to penetrate the dots. Note also how the black that surrounds the dots from the Color Mask result is a non-issue. Save your work.

It's back to Draw to create the 'Snake'; the bold array of boxes curving around in sundry directions.

Add a new page then select the Freehand tool and draw a line similar to figure 15. If it serves you, I've provided the point placement as part of the figure whereby you can map the nodes, then work with each to get your shape.

The real issue with this initial curve is this : Avoid straight lines and create sharp curves. This will ensure the boxes will get their own 'space' and not blend into each other. That's the actual reason for my curve looking as it does.

When you're finished, go to the page that has the large cube from the beginning or our exercise. We'll tweak it up a bit then use it.

Make a duplicate of the cube and ungroup it; then Combine it ('Ctrl Q'). The fill will shift and that's fine. We'll fix it in a moment. Reduce the size of the cube diagonally until the Width reads 100. Then drag the top-middle handle upward until the height reads 125. That will square it up a bit.

Now, zoom in and select the Interactive Fill tool. The old fill handles are already there and we want to modify them. Select 'Conical' for the Fill Type. If 'Snap to Objects' is still on, snap the center anchor to the top-right node of the cube and rotate the darker anchor chip to the center node (figure 16). Set the 'Fountain Fill Midpoint' property to 30.

Cut this cube, go back to the page of the curve and paste it. Make a duplicate of it and set it apart. Take this duplicate and shrink it diagonally until the height reads 35. One will act as the 'head' and the other as the 'tail' (figure 17).

Before we blend, it is important to ensure that the 'head' is in front of the 'tail'. To do this quickly, select the 'head' box then press 'Sh PgUp'.

Select the Interactive Blend tool and blend the 'head' and 'tail'; change the Number of Steps to 85. Go to the Path Properties and choose New Path, then click on the curve (figure 18).

We need to change the speed of the array. The two accelerators are 'Object' and 'Color' and are in the form of triangles at the base of our snake; the lower is the 'Object' accelerator and is what we want to move to the right. So, double-click on the lower triangle then drag it right as in figure 19. This will shift the boxes to help even the flow a bit.

Of course, this last part is determined by how closely you follows my own path. If yours is different, then realize that the idea is to create an even distribution across the path.

Make a copy of the blend and move over to PhotoPaint the paste it directly into your original; rename it 'snake'. We will deepen the brown to a beautiful red then apply the Plastic effect to give it a little depth.

Resize and position the 'snake' according to figure 20.

I have you resize here instead of Draw because I noted that resampling the object re-identified the lines of the cubes which adds a touch more detail to this object than otherwise.

Likewise please note the placement...it's offset according to the red rows/columns of dots. You won't want to place it 'inside' the red dots which can lead to it looking confined.

Once it is in place, make a duplicate. Rename the top snake 'snake' and the lower one, 'snake shadow'. Select 'snake'.

Open Contrast Enhancement (Image | Adjust), press 'Reset' then place 90 / 130 for the Input Value Clipping range. And there is your red.

Next, go to Effects | Texture | Plastic... click 'Reset' then bump up the 'Smoothness' to 50. And that's it! Isn't it wonderful how the Plastic makes it look as though there were smaller boxes inside the originals? Nifty!

For the 'snake_shadow', nudge it down and right 20 pixels. Change its Merge Mode to 'Multiply' and Opacity to 80%. Apply a Gaussian Blur (Effects | Blur) of 9.0. Save your work.

The larger offset of the drop shadow on the sphere shape of the dots give almost a curved look to it. More importantly, because the offset allows for additional light to penetrate between 'snake' and 'snake_shadow', there is the added sensation of depth that would not occur without it.

The rest of the exercise will be done predominantly in PhotoPaint with just a few small items left in Draw. So, we'll continue in Paint for now.

Figure 21 has the elements we're about to create : four placeholders and three pieces of text. The placeholders are derived from the Rectangle tool ('F6') and that's where we'll start.

Click on the Rectangle tool, choose Texture Fill then Edit Fill. When the Texture Fill dialog appears, find 'Threads Embossed' in 'Styles'. There are two color properties at right; choose Black for 'Shadow' and 30% Black for 'Light'. Click on the 'Preview' button once (figure 22) then OK. Finally, set the 'Shape Width' to 5.

Best to use the Info Docker for the following : Drag a rectangle to 320 x 65 pixels (position doesn't matter yet).

From the Rectangle tool properties, change the Fill to 'Solid', the 'Shape Width' to 0, then right click Black in the RGB color palette. Create a rectangle, 220 x 25 pixels then make two duplicates of it.

With all four in play you can now distribute and align according to figure 21. Note here that these objects are behind the 'snake/snake_shadow' objects in the Docker.

We'll add a little noise to the black shapes. Select one of them, go to Add Noise... (Effects | Noise), set the 'Level' to 75, the 'Density' to 50, Gaussian type, Intensity mode. Then, selecting the remaining two one at a time, press 'Ctrl F' to repeat the Noise.

The text elements are 24pt BankGothic Md BT, upper case. Their color is : R255 G163 B5. The order - cubic, triangular, spherical - is in accordance to the sequence of the shapes that will be placed in the texture-filled placeholder. Align your text carefully and save your work.

Great. With those finished, we have just one more element to do in Draw.

In Draw, add a new page. We've done a sufficient amount of cubes to know how to do that. The end result for you, though, should be a cube that measures approximately 45 pixels in height.

The triangle is painless as well. Start with a perfect triangle, 100 px. Rotate it clockwise 60 degrees then skew - using the left-middle skew handle - down 30 degrees. Duplicate and mirror vertically then snap the vertical edges together.

As you can see from viewing figure 22, the triangle's size is a bit larger but only in width. The height is a deception. I'll generalize here for flexibility : I created the height of the triangle by placing it inside the cube then matched it...point-for-point, top to bottom. I then resized the width to exceed the cube 3 pixels on each side (figure 23). I set the triangle away from the cube then gave each shape a linear fill with the colors we've been using.

As a tip : without annoying yourself about which colors were used for the original cube in order to attribute them identically to the triangles, simply right-click and drag from one of the cube faces to a triangle until a bull's-eye appears replacing the cursor. An options menu will appear upon release allowing you to 'Copy Fill Here', which is what you want to do. Select the Interactive Fill tool and adjust the handles accordingly.

The circle is a perfect 43 pixels round. According to figure 24, make a duplicate of the circle, resize it up halfway (a), then shift-click and drag diagonally (b). Open the Trim Docker (Arrange | Shaping), check only 'Target Object(s)' then click on the perfect circle. Your trimmer will disappear, leaving you with the circle and a partial. Fill these two shapes with the same colors but use 'Radial' instead of 'Linear'. Save your work.

If you haven't already, group each of your shapes individually.

While still in Draw, align your shapes 'Center - Horizontally' and distribute them about 1/3 the width of the cube (figure 22). Select all your shapes and right-click 80% black (for the outline of all shapes). Copy all three shapes and let's get back into PhotoPaint.

(Figure 25 gives the overall result of the following)

Paste your objects into your original, place them over the texture-filled placeholder, then make a duplicate. Rename the duplicate 'sml_shapes' and the one below it, 'sml_shapes_shadow'.

Nudge 'sml_shapes_shadow' down 10 pixels. Fill it with Black then attribute a Gaussian Blur of 6.0. Change its Merge Mode to 'Multiply' and reduce the Opacity to 80%.

Next, select 'sml_shapes'. We want to brighten these guys up so open the Contrast Enhancement dialog, Reset it, then put 40 / 150 for the 'Input Value Clipping'. This creates a nice, golden tone to the brown and really identifies them even though they are smaller than the larger (but subdued) objects in our composition.

Save your work and have a look...