BetterSpace

Adding Detail to Spaceships with Texture

So far we have used our the detail panel image that we created in the last lesson to apply a fairly generic texture to some of our object. While this technique is fine for smaller areas, larger and more prominent areas need to have texture maps specially painted for them, in order that the texture accentuates the detail that is already there.

In this lesson we will paint a map for the curved forward surface of our object, and also for some of the panels that protrude from this surface. I will not be explaining exactly how to paint the map, as the painting is your own work, to be done as you see fit. What I will be doing is explaining how to prepare yourself to paint a map... read on...

Open Modeler and load your object.

Open the Polygon Statistics window and use it to select all polygons with the following surfaces:

You should find that the polygons shown below are selected.

polygons selected

Copy (do not Cut) these polygons and Paste them into a new layer.

In this new layer, select one of the polygons that make up the base surface (called Main Forward Upper) then press ] (close square-bracket) to select all attached polygons. Now Cut and Paste these polygons into another new layer.

In the side view, carefully select those polygons that are not a part of the main curved surface (ie: they are part of bevelled areas) and delete them. You should be left with something like this seen in the side view:

only main curved surface remaining

Press d to access the Display Options and set the following options:

display options

(Lightwave 5.6 users: set the display options to hide everything except faces.)

Set the Modeler view layout to one single viewport showing the top view, then maximise the Modeler window, so that it occupies as much screen space as possible. It is very important that you can return the window to exactly this size again in a minute, so I recommend using Maximise rather than manually resizing the window. Now press a to zoom in to the object so that it fills the window.

maximise button

Lightwave [6]

Lightwave 5.6

Press q to assign these polygons a new surface, and apply a suface called <whatever you like> to them. Now open the Surface Editor and change the surface to be R.200, G.200, B.200, with 100% luminosity and 0% diffuse. Exit the surface editor and change the top view to OpenGL mode. Now press Alt+PrintScreen (usually, this is the key above 'Insert' on your keyboard). This will take a screengrab of your Modeler window and store it in the Windows clipboard.

Press Alt+PrintScreen (usually, this is the key above 'Insert' on your keyboard). This will take a screengrab of your Modeler window and store it in the Windows clipboard.

NB: The instructions about maximising the window and performing the screengrab were designed for Windows users... no offence to you Mac and SGI (and all the other Lightwave platforms) users, but hey, you know how to maximise your window and do a screengrab don't you? :-)

Now that we have a screengrab of our object, we need to edit it, so open Photoshop.

When Photoshop has finished loading, set the background colour to black, then press Ctrl+n to create a new image, and you will notice that it automatically sets the dimensions for the new image to be equal to the dimensions of the screengrab we just made... convenient! Make sure that we are creating an RGB image and the Contents are set to Background Color , then press OK to confirm these settings.

After the new image appears, press Ctrl+v to paste the screengrab into the image.

You should have an image that looks like this (of course, the colour of the Modeler backdrop is defined by your preferences in Modeler, and is not important, while Lightwave 5.6 users will have a wireframe view, not OpenGL)

first screengrab

Now leave this image as it is in Photoshop and return to Modeler.

Return the Modeler to exactly the size it was before we opened Photoshop. DO NOT alter the zoom of the display, or use alt+drag to move the viewports around.

Now swap the layers so that the layer with only the four panels is showing.

Press q to assign a new surface, and give this surface a colour of R.180, G.180, B.180.

Now press Alt+printscreen to screengrab the window.

Leave Modeler as it is. We will be returning to these altered layers later...

Return to Photoshop and press Ctrl+v to paste the new screengrab into the same canvas as we were working with a moment ago, so that the first screengrab is the Layer 1 and the second screengrab is Layer 2.

Lightwave 5.6

You Lightwave 5.6 users will have noticed that Lightwave [6] users are able to view their objects in shaded OpenGL mode in all viewports, which aids us in making texture maps. Lightwave 5.6 can only display OpenGL in the perspective window, so your screengrabs will show a wireframe mesh, rather than a block of solid colour. We're going to need to fix those grabbed images so that we can work with them. The following instructions are for Lightwave 5.6 users ONLY, Lightwave [6] users should skip to the next step. With the eyedropper tool carefully sample the colour of the wireframe lines. Hint: if you haven't altered the default colours of the Lightwave interface using the LWCOM.CFG file, the wireframe lines will be R.176, G.176, B.176. Now proceed to use the bucket fill tool to fill in the areas of the image that are occupied by polygons. Do this for both screengrabs (Layers 1 and 2). When you have done this, the geometry in Layers 1 and 2 should look like this... (although your versions will still have remnants of the Modeler screen showing). Layer 1 Layer 2 Next, set your foreground colour to R.200, G.200, B.200 and use the bucket tool to fill the whole of the light grey area in Layer 1 with the new colour. When this is done, change to Layer 2 and set your foreground colour to 180, 180, 180, then use the bucket fill tool to fill all four polygons with the new colour. Your layers should now be more or less identical to those grabbed by Lightwave [6] users.

Now both LW5.6 and LW[6] users have their screengrabs in a common format, with the polygonal areas represented by blocks of contiguous colour, we can continue to prepare the screengrabs to become image maps. We will begin by cropping the image to get rid of the Modeler controls and the uneccesary border around the objects.

Hide the top layer of the image by pressing the eye icon next to Layer 2 in the layers palette (shown under the pointer in the image below). This will reveal Layer 1 once again.

Photoshop layers palette

Begin by using the Crop tool to crop away the Modeler controls from your image, so that you are left with your objects in the center of the image and a single-coloured border around a border around the edge.

first image crop

Ensure that your crop border is within the Modeler controls, and does not overlap them, as shown above.

USEFUL HINT: You may find that it is very difficult to see the Photoshop cursors when working with your Modeler screengrabs, because they are a very similar colour to that which Modeler uses as a background colour. To overcome this, create a new layer in your image and use the bucket fill tool to fill the entire layer with black, then set the layer opacity to 50% and move it so that it is the top layer, and hey presto, you can see the cursors perfectly. Just remember to remove the temporary black layer before you save the image!

creating a temporary layer in order to better see the Photoshop cursors

Now that we have cropped away the Modeler controls, we will further crop the image to remove the border from around the objects. You could do this by eye, but there is an easier way to do it, which ensures precision.

We will begin by clearing away all of the border and excess background colour from Layers 1 and 2.

Open the Select Colour Range tool from the Select \ Color Range menu and set the options as shown below.

select colour range dialogue

Now click somewhere on the background colour of your image, and you will find that the area representing the background colour in the dialogue will turn white. When you click OK, all the white areas will be selected, and all the black areas will not. Make sure your object appears black in the Color Range dialogue, and click OK.

After you click OK, you will find that hopefully, all of the background colour (in my case, green) has been selected. Press Delete on your keyboard to clear these areas.

Next, turn Layer 2 back on by clicking on it in the Layers Palette, then repeat the Select Color Range operation to clear the background colour from this layer too. When this is done, you will see the darker grey shapes of Layer 2 on top of the lighter grey shape of Layer 1, with the back background showing through the holes.

layers after background deletion

Select the Photoshop Magic Wand tool and set the Magic Wand Options to those shown below.

Photoshop magic wand options

Next, click on the black border area of the image, and the magic wand will automatically select the whole border area. Invert the selection by pressing Ctrl+Shift+i. Now we have the whole of the useful area of our image selected.

Press Crtl+r to turn on Photoshop's image rulers, if you don't already have them turned on. Also, check that Snap to Guides is enabled, in the View menu.

Next, drag out guides from the rulers at the sides of the image to the edges of the object in the middle of the image. The guides should snap into place when they come into the proximity of the edge of the selection.

guides in place

FInally, use the Crop tool to crop out the central image. Note: your crop selection must start within the area defined by the guides, not outside it.

image cropped

Next, we will instruct Photoshop to only allow us to paint on areas of the layers where there is already image data. To do this, check the Preserve Transparency button in the Layers Palette for both Layer 1 and Layer 2.

preserve transparency option

As a final adjustment to make painting easier, rotate the entire canvas by 180 degrees, so that it appears as though we were looking from the front/top of the object.

With the image correctly cropped and oriented conveniently, we are now ready to start painting details onto the map.

To begin with, I will paint some basic weathering onto the edges of the panels in Layer 2. I will do this using the Paintbrush tool with a dark-grey colour and 25% opacity, gradually building up the darkness of the colour around sharper edges, especially on the leading edges of the panels and around where the smaller details protrude.

weathering added to Layer 2

As you paint, you will notice that is is impossible to paint outside the existing painted areas of the layer, as we are using the Preserve Transparency option.

When you are happy with your progress on Layer 2, switch to Layer 1 and paint some weathering there.

weathering added to Layer 1

The details that you paint following this are entirely up to you. As you can see from the image below, I have added a little more weathering to various surfaces, and I have also added some various panel images, which I butchered from the Lightwave GENERICPANELS.IFF image map. I attached these by pasting the generic panels image onto my own map, then changing the compositing mode to overlay and altering the opacity and scale of the panels image.

painting completed

I also added a little text and a a warning symbol, just because...

All that remains in the painting of this map is a little clean-up and preparation for mapping.

As you may remember, we rotated this image through 180 degrees earlier so that we could view it more easily for painting. This must now be reversed, so rotate the image through 180 degrees again.

Next, you will no doubt be aware that the background colour of this image is black. This must now be changed, in case our map doesn't fit the object exactly, so that we don't have ugly black edges to our image.

Select the background layer and fill it with 200,200,200 grey.

image map finished

Now you may flatten the layers of your image (if you want to) and scale it down to a more manageable size if necessary, then save it.

Back to Modeler now, and you need to select the two Modeler layers of which we took screengrabs earlier.

Copy the polygons from these two layers into a fresh new layer, and assign all the polygons a new surface (it doesn't matter what you call it, but I will call mine 'texturePlacement').

new surface for these polygons

Lightwave [6] users can play with the new map here in Modeler, but 5.6 users must follow the instructions below:

Lightwave 5.6
Save the layer containing the polygons shown in the image above and load it into Layout, then open the surface editor.

In the surface editor, select the newly created surface and apply the following settings to it:

Now add a new colour texture as follows: Set the image used as our newly painted texture, that we just saved from Photoshop. Project this image through the Y axis, with no texture anti-aliasing or pixel blending, and press the Automatic Sizing button.

Copy this surface into the clipboard, and then paste the surface settings to the following surfaces:

This will apply the map to each of these surfaces, correctly positioned and scaled as we want it.

textures applied

Using the techniqes described above, I then proceeded to paint six more texture maps for the object, some which could be applied to multiple panels.

These are the images I painted:

texture map 1

texture map 2

texture map 4

texture map 5

texture map 6

texture map 3

texture map 7

The image below is colour-coded according to the map applied to that particular surface, so you can see what map has been applied to each surface.

maps used

The next thing I will do is apply the GENERICPANELS.IFF image map as a specular, diffuse and bump map to some of the surfaces. I will be using exactly the same techniques to apply this map as we did in the previous lesson where we added our detail-panel map.

This time, I will use a size of X=50m, Y=50m, z=50m for all the textures, as the realism of the surface is quickly reduced by using this image at too small or too large a size. I will also be using a Cubic projection type, so that the image is applied with the correct mapping axis according to each surface's orientation. You can clearly see the influence of the generic panels image if you look at the large curved panel on the left of the front of the object.

After applying the map to several surfaces, this is how my object is looking:

image mapping complete

image mapping complete

I have finished applying maps to my object, but for you, the only limit to the amount and complexity of your maps is your computer's memory, and your imagination.

In the next lesson we will complete the texturing of the object when we add some grime and scratches to some of its surfaces.

Do you have any comments of questions regarding this lesson?