Mouse pointer is left visible. Included is all the download links ( in the description )

https://www.youtube.com/watch?v=-1WfNwqBQDc

Grand Theft Auto IV requires pre-defined rendering types in order to draw polygons in its shader-based RAGE engine. All of these shaders can be found in the directory:

Grand Theft Auto IV/common/shaders/win32_30

However, you will only find shader names in this location, and no details about them. This tutorial will help you understand how each material should be set up with its corresponding shader

All materials in Zmodeler must have one of these shaders specified:

• badges
  
• basic
  
• chrome
  
• disc
  
• generic
  
• interior
  
• interior2
  
• lightsemissive
  
• mesh
  
• paint1
  
• paint2
  
• paint3
  
• rims1
  
• rims2
  
• rubber
  
• shuts
  
• tire
  
• vehglass
  

I have searched online, and I haven't figured out how to set up all of these materials as of yet; however, the ones in bold are commonly used on GTA vehicles and should suffice in any car mod. The rest that aren't in bold can be ignored for now.

------

Naming materials:

Every material must be set up with a name and a shader. For example:

vehicle_generic_tyrewallblack tire

Note that the section in blue is the name of the material and the section in red is the shader name.

The name of the material can be anything you want, but it is commonly easier to name it the same as the main texture that the material is using. In this case, the vehicle_generic_tirewallblack.png texture is being used for this material.

In addition, many of the GTA shaders can render polygons differently, resulting in varied appearances. All of these settings can be found in the directory:

Grand Theft Auto IV/common/shaders/db

If you look at some IV vehicles, such as the Merit or the Lokus, you will notice that polygons of the headlight glass have this mixed shader type applied to them. In this case, the mixed type is lights(vehglass)

lights represents the shader, and vehglass represents the setting for that shader. This type of material is commonly used for when you have glass in front of headlights that are mapped to a lightsemissive material, which both the Merit and Lokus have.

------

Paint colors:

Many shaders in IV also allow them to be painted different colors.

Examples:

paint1 [CLR:2]

This means that this material will use the second color number value in the carcols data line (#1,#2,#3,#4,) and it will use the paint1 shader.

paint2 [CLR:1]

This means that this material will use the first color number value in the carcols data line (#1,#2,#3,#4,) and it will use paint2's metallic paint shader.

[CLR:3] and [CLR:4] use the 3rd and 4th carcols color number values, respectively.

------

Material Setup:

Each material requires that they are properly set up in order to export and appear properly in GTA IV. This requires that you apply the corresponding textures to the corresponding texture layers for each material.

In order to specify the settings for each texture layer, press the "Edit" button and set up the layer accordingly.

Note that you do not want to UV map any object(s) until you have properly set up the material. This could lead mapping issues afterwards.

badges

Layer 1: Badge texture. Apply as GTA: Texture, UV#1

Layer 2: Badge normal-map texture. Apply as GTA: DontProduct3, UV#1

Layer 3: Specular texture, typically vehicle_generic_smallspecmap. Apply as GTA: Specular, UV#1

interior/interior2

Layer 1: Main interior texture. Apply as GTA: Texture, UV#1

Layer 2: Specular texture for interior, typically vehicle_generic_smallspecmap. Apply as GTA: Specular, UV#1

lightsemissive

Layer 1: Base texture, typically vehicle_generic_glasswindows2. Apply as GTA: Texture, UV#1

Layer 2: Dirt texture, typically vehicle_genericmud_car. Apply as GTA: Dirt/Mud/Reflection, UV#2

Layer 3: Specular texture, typically vehicle_generic_smallspecmap. Apply as GTA: Specular, UV#1

mesh

Layer 1: Main vehicle texture. Apply as GTA: Texture, UV#1

Layer 2: Dirt texture. Apply as GTA: Dirt/Mud/Reflection, UV#2

Layer 3: Normal-map texture. Apply as GTA: DontProduct3, UV#1

Layer 4: Specular texture for main texture. Apply as GTA: Specular, UV#1

paint1

Layer 1: Base texture, typically vehicle_generic_smallspecmap. Apply as GTA: Texture, UV#1

Layer 2: Dirt texture, typically vehicle_genericmud_car. Apply as GTA: Dirt/Mud/Reflection, UV#2

Layer 3: Specular texture, typically vehicle_generic_alloy_silver_spec. Apply as GTA: Specular, UV#1

paint2

Layer 1: Base texture, typically vehicle_generic_smallspecmap. Apply as GTA: Texture, UV#1

Layer 2: Dirt texture, typically vehicle_genericmud_car. Apply as GTA: Dirt/Mud/Reflection, UV#1

Layer 3: Specular texture, typically vehicle_generic_alloy_silver_spec. Apply as GTA: Specular, UV#1

paint3*

The paint3 shader involves the use of livery textures.

Learn how to set up and UV map paint3 materials in corleone's tutorial here.

shuts

Layer 1: Inner door shuts texture, typically vehicle_generic_doorshut. Apply as GTA: Texture, UV#1

Layer 2: Normal-map texture. typically vehicle_generic_detail2_normal, Apply as GTA: DontProduct3, UV#1

Layer 3: Specular texture, typically vehicle_generic_smallspecmap. Apply as GTA: Specular, UV#1

tire

Layer 1: Main vehicle texture, typically vehicle_generic_tyrewallblack. Apply as GTA: Texture, UV#1

Layer 2: Dirt texture, typically vehicle_generic_tyrewall_dirt. Apply as GTA: Dirt/Mud/Reflection, UV#2

Layer 3: Normal-map texture. typically vehicle_generic_tyrewall_normal, Apply as GTA: DontProduct3, UV#1

Layer 4: Specular texture for main texture, typically vehicle_generic_tyrewall_spec. Apply as GTA: Specular, UV#1

vehglass [lights(vehglass), etc]

Layer 1: Base texture, typically vehicle_generic_glasswindows2. Apply as GTA: Texture, UV#1

Layer 2: Dirt texture, typically vehicle_glassdirt. Apply as GTA: Dirt/Mud/Reflection, UV#2

Layer 3: Specular texture, typically vehicle_generic_smallspecmap. Apply as GTA: Specular, UV#1

------

Reference resources for this tutorial:

forum.zmodeler3.com

generic

mesh

vehglass

lightsemissive

tire

interior

interior2

[CRL:1]

[CRL:2]

badges

chrome

http://www.youtube.com/watch?v=0TSDzSeuJ8M

If this helped please subscribe or give it a like. I hope this helped some people at least.

Thank you,

Bossnix

Find and extract the emerglights.png located in the car's .wtd file.

It should look something like this..

Now. There are 2 areas you need to edit here. The bit with the little mini images of the lightbars and the big colorful things.

Those..

Now with the lightbar image bit, you need to find your car's lightbar type. i.e. if its useing the Vector then its obviously the one that looks like the Vector.. so you would edit this one..

But in this case it's not that one but this one...

..as I am editing FORD CROWN VICTORIA / CVPI-V4.6A ELS-H

Now to editing it. It is very important to note that the color is slightly transparent! So you will need a program like GIMP or Photoshop to do this. Paint (to my knowledge) will NOT WORK!

This one's easy as it is just a rectangle..

First make a new transparent layer. Next take your box/select tool and select the entire bar. Make sure you are currently on the new layer.

Then with the fill/bucket tool select 'fill whole selection' and set the opacity level to about 50% and fill that area with the desired color. Then with the pencil or paintbrush tool go over the black separator lines, but with no opacity! These have to be solid. Oh, and be sure the area is still box selected. Should look like this..

Now, switch to the original layer, and clear the boxed contents. Should look..

Do the same thing to the smaller one next to it.

Alright you should be here by now..

Good. Onto this section...

Now these are the textures that the individual 'bulbs' in the lightbar will have when they are off.

In this case this part of editing is very easy because we only have one color on the lightbar. So most are going to be the same color. If we had different colors we would need to figure out what sections of the bar used what. For example I already know where certain thing go for this car. This should be pretty obvious when you 1st open up the original texture. For this car it like this...

5 is for the rear secondary's and i'm not sure what 6 is for.

Anyhow this not so hard for this car. You are just going to make 1,2,3,and 4 orange.

Get yo box select tool and select one of em.

Then will the fill tool, set the opacity to around 50% (this can varies per car and color) and fill that box.

Now for the trick.. with that select tool box still there, right click and go to 'copy visible'.

Then just copy and paste a few time over the others and boom.

Now say you wanted that back secondary light to be green. Use a similar process.

Use the circle select tool (cause its a circle!) and do the same. Bucket fill with a green at about 75% opacity (only cause we are covering red this time so it takes a little more).

Should get something like this...

And that is it for the emerglights file!

Save as a .png with any name.

Next is the emerglights_emis.png

Extract the emerglights_emis.png from the .wtd

Open it up!

Should look something like this..

What this file does is give texture to the individual 'bulbs' when they are on.

Now again this will be easy as our lightbar is all one color. If it were not we'd have to figure out which ones go to which parts of the lightbar. This can be really annoying sometimes. Just remember these were designed with some sense so its not going to be toooo complicated. Oh and those labels are there for a reason.

But we don't have to worry about that!

Because our bar all has the same color we are just going to make all the ones to the right orange.

These..

With your box select tool, select any of those and fill with the bucket tool with an opacity of about 65% (again this will vary depending on the colors, just experiment)..

Now, with the box still selected, right click and go to 'copy visible'.

Then it is just a simple process of copying and pasting that over all of those that i showed we had to make orange.

You can even use a trick to make this go faster. Cover the top row then box select it, copy visible and paste onto the lower two.

Should end up like this...

Now since we concluded that we wanted the back secondary's to be green we have to change the top static, generally the secondary's will be one of the statics. So do the same as before but this time only the one and making it green. I suggest a bucket fill opacity of about 75%.

Oh whoops i missed one.. this one needs to be orange too for some reason.

Generally you don't need to touch that one.

Thats it!

Save it as a .png with any name.

Finally import these over the ones you extracted from the car's .wtd

Lastly to get the els reflections right edit your els.ini. Depending on the car's name go to its section in the els.ini file. For example say it was just the regular police. Find the section for police...

[ POLICE ]



Active = 1

Halogen_Only = 0

Color_Light1L = 0

Color_Light2L = 0

Color_Light3R = 2

Color_Light4R = 0

Color_Light7L = 3

Color_Light8C = 3

Color_Light9R = 3

Offst_SLights = 0

The primary light are Lights 1 to 4. The secondary lights are 7 to 9.



Since we made the primary's orange we'll set the primarys to amber, the closed to orange. Only red, blue, white, amber, and green are supported. They have corresponding numbers as seen in the reference section of the els.ini (located at the bottom).

// REFERENCE

//

// Active Determines whether the ELS can be enabled on that vehicle / 0 for OFF, 1 for ON

// Halogen_Only Determines whether the model is intended for halogen only lighting patterns

// Offst_SLights Y axis offset of secondary environment lights relative to vehicle

//

// Red 0

// Blue 1

// White 2

// Amber 3

// Green 4

So to set the primarys to yellow we change 1-4 to 3 and to make the secondarys green(since we make em green) set 7-9 to 4. Oh and don't forget that this car is a Halogen.



End result show look like this..

[ POLICE ]



Active = 1

Halogen_Only = 1

Color_Light1L = 3

Color_Light2L = 3

Color_Light3R = 3

Color_Light4R = 3

Color_Light7L = 4

Color_Light8C = 4

Color_Light9R = 4

Offst_SLights = 0

DONE. Lets go test it out.

In-game:

Questions? Comments? PM or email me (simplestarwarsfan@gmail.com)

I'll concentrate on vehicles for GTA IV, but most of these tips work for other game textures such as pedestrians or signs, as well as other games like GTA SA, or vehicles in flight or truck simulators.

While I strive to be as accurate and informative as I can, I am by no means a professional artist. I'm a hobbyist, learning his way as he goes along, just like many of you. Feel free to add your suggestions and/or corrections.

Most images here have been scaled down and can be clicked to view in larger sizes. I'd also like to thank all the photographers and artists whose images appear in this tutorial. Without such visual aids, it would be much more difficult to teach.

I've covered many different aspects, each in its own lesson:

Lesson 1: Basics

Lesson 2: Layers

Lesson 3: Positioning & Reference Points

Lesson 4: Text & Fonts

Lesson 5: Normal Maps

Lesson 6: Optimizing a Simple Logo a: Colour Correction, b: Adding Realistic Detail

Lesson 7: Creating a Complex Logo From a Car Photo

Lesson 8: Creating Simulated Reflective Vinyl

Lesson 1: Basics

If you're completely new to modding GTA, I suggest you acquaint yourself with installing mods before you create your own skin.

Editing Software:

Any Windows PC comes pre-installed with MS Paint, but I'm afraid it'll be too limited for what we need. Adobe Photoshop (PS) or Corel Paint Shop Pro are great, if you can afford them. But there are free alternatives that have nearly all of PS's functionality. paint.net http://www.getpaint.net/ is popular, but I haven't used it much. Since I'm familiar with The GIMP http://www.gimp.org/, I'll be referring to it throughout this tutorial. Regardless of which program you use, remember there is a wealth of great tutorials on the web, specific to each program.

If you haven't used any of these programs before, I'd try editing an existing texture first, before you make your own from scratch. Just be aware that if you change someone else's work and release it on the web, that most artists require their permission for you to do so, or at least that you credit them for their original work. Check the readme if one's included, or contact the artist if you're not sure. The last thing anyone wants to hear is "Mr. X stole a logo from Mr. Y's car. Mr. X is a thief! Mr. Y's a selfish jerk!". We don't need that drama, we're here to have fun.

File Types:

GTA IV imports & exports textures in .png format so our final product will need to be saved as a .png. However, while we're editing, .png has its limitations. You'll likely want to save yourt work as you go along, before it's finished. I suggest you save your work in one the following file types:

-.psd for Photoshop,

-.xcf for GIMP,

-paint.net has an available plugin to read .psd files

This allows you to save your work while keeping information such as seperate layers. What are layers? I'll explain in our second lesson.

You've got a car downloaded, it came with skins already installed, and it likely came with a template. Since it's such a popular car, I'll use Bxbugs123's CVPI as an example:

Note that the above template is an example only, it's been scaled-down for easy viewing here. For best results, use the full-size one included with your downloaded car.

The template shows you reference to where the skin will be applied, but it is not intended to be included as a part of the skin. If you fill in the template itself with your desired colurs, text, etc., it will look poor. I'll asume that you want a quality skin, or you wouldn't be here. So, how do we utilize the template, 'painting' on top of it, and then getting rid of the template later? By using layers.

Lesson 2: Layers

"Layers? Sounds kinda technical, what about text, or stripes or something?" Text, stripes, logos, etc. are all best used in seperate layers. "GIMPtrick"'s video on YouTube does an excellent job of describing the "whats" and "hows" of layers, I have the "whys" for us below.

There's a few reasons to use layers. You can have the template on the lowermost layer (like the 'tabletop" in the video), and 'paint' the solid background colour on a middle layer, your stripes on an upper layer, and text on another layer. I go a step further, and make each stripe colour on its own layer. This way, there's no gaps or bleed-through with two or more adjoining stripes. just like the dots in the video. If we need to correct the position or colour of one stripe, we can do so without affecting the other layers.

I'm going to make a replica of the Ontario Provincial Police Crown Vic:

Russell County (Rockland) O.P.P. by smitty93, on Flickr

So, let's open our CVPI template with GIMP, and right away, save it as an .xcf file. Our main background colour is black, so let's make a new black layer.

Right-click on the background layer, create a new layer, and for simplicity, call it "black". GIMP will automatically select the newest layer for us to paint on, but if we needed to work on the another layer, we can click on it in the layers dialog, and edit on that layer.

"Whoa! Where'd the template go?"

Don't worry, we can hide it. Either completely, by clicking the little 'eyeball'... :

...or partially, by lowering its opacity (making it partly transparent), both found in the layers dialog:

We're now done with the black, so let's hide it completely, so we can see what we're doing. If you hide the template as well, you'll see the canvas is grey checkers, representing transparency. We'll need the template though, so let's leave it visible.

--Note regarding white backgrounds--

Some GTA IV vehicles support 'transparent' textures. They already have a white base, so no 'white' layer is needed. If you leave the skin's background transparent, you'll achieve a smaller file size, and a texture with slightly more detail that "pops" from the car, as if it were really decals. Most of Bxbug's cars require soild backgrounds, while LUXART's support transparency. If you're not sure if your car uses transparent textures, just paint the background solid white. You can tell whether or not by viewing the original textures in SparkIV. This example of LUXART's texture is transparent, notice the grey checkered background:

Let's make some more layers. The white paint, the stipes, logos, etc. will all be in seperate layers. Let's do the white paint next, by first creating a new "white" layer. We can verify that we're actually working on the "white" layer, and not the "black" or "template" because "white" is selected in the layers dialog. The roof's simple, it's all-white, so we'll draw a box and fill it with the bucket tool.

The front doors need a little more precision, so that we don't paint the rear doors as well. We'll lower the opacity of the while layer, to around 50%, so we can see both the white and the template. We'll zoom in and paint with the brush or the pencil, erase some if we need to, all the while not affecting the black layer, template layer, or any other layer. Here. I just need to finish painting along the left gap between the doors:

Repeat the process for each new colour, stripe, logo, etc. Creating text will automatically create a new layer as well. I'll give more tips on text & fonts in lesson 4.

Remember, you can keep saving your work as an .xcf or .psd as you work. When you've finished, you'll need to save it as a .png., which doesn't allow multiple layers. GIMP will ask you if you'd like to merge the layers, or you can do so yourself before hand. Right-click on each layer in the layers dialog, and select "merge down". You can now delete the template layer also. Don't worry, If you need to change something later, you still have the .xcf copy in all its layers.

Lesson 3: Positioning & Reference Points

How do you know where to put details like logos, text, & stripes, or what size they should be? We can use a photo as a layer, and use it like we would use tracing paper. It should be noted that we are NOT using the photo as part of our final texture. We are using it for reference only. Think of it as using a projector on a white wall, projecting a photo on the wall, and painting what we see in the image. When we're done, we'd turn off the projector, or in this case, toggle the layer off or delete it.

Since the above reference photo is a nice side-shot, let's save it, and drag it into our texture in GIMP.

The photo is now its own new layer. We'll get rid of the excess background by placing a box around the edges of the car and selecting "crop to selection" from the Layer menu. (selecting "crop to selection" from the Image menu will crop the entire image, we don't want to do that). We'll resize it so that it's roughly the same size as the car in the template.

Lower the opacity of the photo layer, and position it over the car in the template.

If the photo still doesn't fit, you can rotate it, or even change its perspective. This is an extreme example, to give you an idea.

Resizing, rotating, and such can degrade the image quality a little bit each time, so we'd normally be concerned about overdoing it. But we aren't concerned here, since we won't be using any part of the photo itself in our texture. When we're ready to save the final texture as a .png, we'll delete the photo layer completely.

I'm sure you can see the benefit of putting the real car over the template, especially if this car had a more complex design.

Now you can perfectly place all of your details, in their own seperate layers. Here, I've added details such as the crown, "O.P.P.", "POLICE", black and white stripes. None of these details are actually sourced from the reference photo. The text is made with the text tool, stripes with the brush tool, and the crown is taken from a different, high-res image (I'll describe how in a future lesson). When we delete the photo layer, the details will remain, in perfect proportions and positions:

Lesson 4: Text & Fonts

Since lettering makes up such a big part of our textures, I've tried to put a lot of info in this lesson. Many of the methods here are also applicable to logos, stripes, or whatever.

Finding the right font:

If you have a sample image, and don't recognise what font is used, you can try asking in the forums, or searching yourself. I have samples of some common fonts below. Identifont lets you identify a typeface by answering a series of questions. WhatTheFont lets you submit an image, and then automatically searches its database. Just remember to crop your example image so that only the lettering is visible.

Note that not all lettering is necessarily a "font", and therefore may not be type-able. An example would be the scripted "Ford" in the blue oval logo. Such lettering would need to be "hand-drawn" (literally or digitally) in order to duplicate.

Font Samples:

These are just some of the many fonts that I use for vehicles, in not-much of a particular order. When making licence plates, "Roadway", "Mandatory", and obviously "Licence Plate" are useful.

Editing the text:

By this, I don't mean editing what the text says, but how it appears. When we input text with the text tool, GIMP creates a new text layer. We can tell it's a text layer by the 'T' icon:

While it's a text layer, we can change things such as the typeface (font), the font size, colour, whether each line is aligned left, center, or right, and the spacing between each line or between each letter. We can also change the word itself, in case we need to correct a typo. If you've closed the text editor box shown in the centre here, you can reopen it by right-clicking the text layer in the layers dialog and selecting "Text Tool".

Some editing techniques will alter the layer in such a way that it's treated like any other layer. From then on, we can't edit it with the tools in the text editor, such as changing font, etc. These methods may be used with things other than text as well, such as logos and stripes, but are especially useful here.

For example, if we need "slanted" text, and there's no italic version of a typeface, we can use GIMP's shear tool to angle it right or left:

After. Note how the layer called "POLICE" no longer has a 'T' icon. This means that GIMP no longer considers it to be a text layer:

Often, the text is either too tall or too short, relative to its width. We can scale the text either vertically or horizontally to achieve the proportions we want. Here, I've reduced the vertical size, in order to 'squish' the text:

After:

Let's make a red outline to the text. I've selected red as my foreground colour, and created a duplicate of the text layer. To do this, right-click the layer and select "Duplicate Layer". The new layer is called "POLICE #1". Here' I've selected the "POLICE" layer to work on, you can tell by the blue in the layer dialog. Right-click the "POLICE" layer again, and select "Alpha to Selection". This puts dotted lines chasing each other around the text:

Now from the "Select" drop-down menu, select "Grow". Note that only the dotted lines grow, not the image. Here I've grown the selection by about 4 or 5 pixels:

We can now fill that selection with the bucket fill tool. To fill the entire selection at once, hold down the shift key, instead of filling each letter seperately. The black layer is on top of the red layer, but if they're mixed up, just move them around with the little green arrows in the layers dialog.

To make a fade between two colours, use the Gradient tool (just to the right of the bucket tool). Here, I've selected the outline again by using "Alpha to Selection". I have grey as the foreground colour, and red as the background colour. I've clicked the cursor at the bottom of the text, and dragged it to the top.

Here's a basic square to illustrate how placement of the cursor affects the gradient fade:

You can fade at any angle you wish. By going diagonally, it makes the fade go diagonally:

Like I mentioned before, these methods are also very useful for stripes, logos, or whatever you need.

Lesson 5: Normal Maps

You've likely noticed in SparkIV some strange textures, usually licence plates, that look like this:

It's called a Normal Map, and creating it is actually quite easy once you know how.

"Wait, what is it, and why do I need one"

It's a type of bump map, used for 'faking' the lighting of bumps and dents, without using more polygons. Here's a video:

If you aren't using GIMP or Photoshop, you can create them with programs such as ShaderMap (free) and ShaderMap Pro (payware) or CrazyBump (payware).

If you are using GIMP or PS, there's free plugins available for each, that let you create Normal Maps right in the program. For Photoshop, nVIDIA makes one here: http://developer.nvi...adobe-photoshop, and for GIMP, you can find one here: http://code.google.c...imp-normalmap/. Apparently, there was an nVIDIA plugin for Paint Shop Pro, but it doesn't work with the latest version.

As always, I'll be using GIMP for this lesson, so go ahead, install the pluin, and open GIMP.

Each normal map texture is paired with a regular texture, the same dimensions, and for the most part, identical other than the obvious strange blue/purple colour.

Some recent vehicle mods have the licence plates as a part of the main texture, but many still have the plates as seperate textures, usually 512x256 pixels. They're usually called "backplate" and "frontplate", and their normal maps are called "frontplate_n" and "backplate_n" Here's a 512x256 "frontplate" that I made:

We'll make a copy of "frontplate" and call it "frontplate_n", which will end up as our normal texture.

We want the numbers to appear as though they're higher than the background. This particular style of plate has "CA EXEMPT" painted on, but it isn't stamped into the plate like the numbers are:

1044243 by Carbon Arc, on Flickr

For realism's sake, we don't want the "CA EXEMPT" to appear any higher than the background, so we'll just fill it in:

We're now ready to create our normal map, and actually, we're almost finished. Select Filters - Map - Normalmap:

And this window will appear:

There's a lot of options here, but you don't need to mess around with much. Two things you will need to adjust are:

a: Filter type. There's 4sample, Sobel 3x3, Sobel 5x5, Prewitt 3x3, etc. These will all have different effects on the intensity of the normal map. You can play around with them, and it'll show you how it'll look in the preview. I find Sobel 3x3 works best for my uses.

b: Notice that I have "Invert X" and "Invert Y" checked. This is because the filter creates a normal map based on lightness and darkness. Normally, it will make dark areas low, and light areas high. Since this is the opposite of our licence plate, we need to invert them with the checkboxes. An alternative would be to make a plate texture with a dark background and light letters, but this is simpler.

Press OK, save your work, and that's it!

Lesson 6a: Optimizing a Simple Logo: Colour Correction

Sometimes, a Google image search will provide us with a perfect replica of a logo such as a door crest. Often, we aren't so lucky, and we'll cover those instances in Lesson 7. For now, let's do some minor editing on otherwise good logos.

If you click & view the largest size of this photo, you can see the rear side window decal is for Nevada Crime Stoppers:

IMG_9156 "Nevada Highway Patrol" by saxxon57, on Flickr

Google image search only comes up with some red or black logos, not yellow:

We're going to turn the above solid image into a transparent yellow one.

In the Layers dialog, mouse over the layer and right-click. Select "Add Alpha Channel". This allows us to add transparency. With the "Select By Colour" tool, click the white area, then right-clic, select "Edit", and select "Cut". There might be some stray white pixels around the edges, but that's fine for now:

If we just went ahead and bucket-filled everything, it would either leave rough edges, or not fill in everything, or both. Instead, we'll use the "Colourize" function.

The next step is only needed since there's more than one colour, in this case black and red. From the colours menu, select "Brightess-Contrast" and slide the contrast all the way to the left. This will make everyting the same shade of grey. Otherwise, in the next step, the portions that are currently red will always look lighter than the black ones.

Now, to make it yellow, from the "Colors" menu again, select "Colorize". By using the "Colour Picker" tool on the priginal photo, I know the yellow color should be around Hue: 33 and Saturation: 51. Adjust lightness as needed:

To match the photo, we can delete the "OF NEVADA" bit, and resize the numbers, but I think you get the idea.

Lesson 6b: Optimizing a Simple Logo: Adding realistic detail

It's easy to find a great quality logo for a force as large as the RCMP, such as this one from Wikipedia:

Heck, it's even a transparent .svg file available in sizes from 200px to 2000px, and if you weren't as picky as me, you could scale it to fit and just use it as-is. But, I think what seperates a good skin from a great one are the details. On this close-up of an RCMP car door, we can see the decal has a white background that is a little larger than the logo:

RCMP Door Logo by mile105, on Flickr

It's greyish-white because it's made from reflective vinyl:

DSC_0043 by C-GCLL, on Flickr

We're going to add the greyish-white surround to the RCMP logo from Wikipedia, in just few simple steps.

I've opened the 500px version of it in GIMP. We can see the canvas is no larger than the edges of the logo:

We need a bigger canvas to put our white surround onto, so from the "Image" menu, select "Canvas size", and change the width from 500 to 550 pixels or so. Move the logo to the centre of the canvas, and you'll see the layer is still 500px wide, as shown by the yellow dotted lines:

From the "Layer" menu, select "Layer to image size". We're going to make a copy of the layer to make our white with. Mouse over the background layer in the Layers dialog, right-click, and select "Duplicate Layer". The top layer we'll leave unchanged. Working on the bottom layer now, mouse over the bottom layer in the layers dialog, righ-click, and select "Alpha To Selection". This will put dotted lines around the border of the logo. From the "Select" menu, select "Grow". In this case, I chose to grow by 6 pixels. How much you need to grow may take some trial and error. You can see the selection's dotted lines are bigger than the logo now:

Finally, with a suitable greyish-white as your foreground colour, just bucket fill the selection:

Holding shift while you bucket fill will fill the entire selection, leaving no gaps between the white and the logo. That's why I chose to leave the original logo in an upper layer. Merge, and you're done!

Round, oval, square or rectangular decals are even simpler. Just make your lower layer with the "Rectangle Select" or "Ellipse Select" tool. For instance, my Los Angeles Sheriff and Police Departments skins both have simulated clear backing to the "Emergency 9-1-1" decals on the fenders, such as here in the lower left of the photo:

LOS ANGELES POLICE DEPARTMENT (LAPD) by Navymailman, on Flickr

Here, I'd just make a rectangle, a little lighter than the black paint, and place my white "Emergency 9-1-1" lettering above the rectangle. Here's the final result (click for full size):

Lesson 7: Creating a Complex Logo From a Car Photo

As I stated before, sometimes finding a suitable logo isn' so easy. Search for "Michigan State Patrol logo", and you'll find many like this:

On its own, it looks good, but it's not quite the same as what is actually on the cars:

Michigan State Police cars -- Door Seal on 1992 Ford Mustang 5.0 by Corvair Owner, on Flickr

Using the photo above, I'm going to show you how to create a perfect logo nearly from scratch, and end up with something like this:

If you'd like to work on the same image yourself, you can download the original size herehere A big thanks to Corvair Owner on Flickr for the photo.

Step 1a: Finding a suitable photo.

Large, high-resolution photos, taken straight on to the subject work best. The larger, the better. Even though the final logo we place on the template will be shrunk down to around 125 to 200 pixels high, the process we will use to create it will work best if it is much larger.

Shadows, reflections, or body creases can be dealt with, but are best avoided if possible. Smaller photos, or ones shot at an angle aren't impossible to work with, but will make our job more diffucult.

The photo I've chosen isn't perfect. If we used the logo as-is, the body lines and reflections would show. Our goal is to remove these imperfections. Fortunately, it is high-res, in focus, and shot straight-on.

Step 1b: Optimizing smaller photos, in case you can't find a large one.

The CVPI's door is 250 pixels high from the window to the rocker panel. The logo we place on it won't be any larger than that, but a larger image will be easier to work with. If the largest image you can find is only 250 or 300 pixels high, it's best to enlarge it before further editing. In GIMP, from the Image menu, select "Scale Image". A good rule of thumb is to double its width and height. It won't appear any more detailed at first, but it gives us four times the pixels to manipulate.

For reference, in our case, the logo will be 194 pixels high. Using my method from Lesson 3, I placed a photo over the template to determine placement and size of the logo. I measured the height using the "Measure" tool:

Step 2: Seperate each colour into its own layer.

Stay with me, it's easier than it sounds. With our Michigan State Police logo, we have only four colours: gold, navy blue, black, and a little bit of red. At least, that's how the vinyl was printed. In the photo, none of the colours are uniform, due to the reflections. The gold and blue are lighter at the top, and the black is nearly grey. We'll correct each one seperatly, in its own layer, and keep the original photo in its own layer as well.

I cropped the image to just the logo area, just to make life simpler. Now's a good time to save the work as an .xcf, so we can save as we go and keep our layers intact. Let's start with the gold. With the "Select By Colour" tool, left-click on the gold area. The dotted lines show what has been selected.

As you can see, it only picked some of the darker area below. To add the lighter areas, hold "Shift" and click the missing areas. To un-select an a shade, hold "Ctrl" and click it. You'll likely need to zoom in to see what you're doing:

You might not need to go too crazy picking every little stray pixel. In this screencap, sone of the edges near the black aren't selected, and some of the black is selected, such as the sun:

We'll do our best to preserve these details, but in reality, many of them will be lost when we scale down to fit the template. On the other hand, had we started with a small photo, these details might not even be picked up at all.

Once you're satisfied that you have enough gold selected, mouse over anywhere in the main window, right-click, select "Edit", select "Copy". Right-click again, select "Edit", select "Paste". We now have our gold on its own, so let's make it its own layer. Mouse over "Floating Selection (Pasted Layer)", right-click, select "New Layer":

I've hidden the lower layer to show the gold by itself. Now, it's even more apparent that it's all sorts of different shades:

We're ready to make it all one colour. With the "Colour Picker" tool, select the pedominant colour, or whatever you feel is the correct shade. In this case, I picked the upper part of the elk. In the "Layers, Channels, Paths, Undo" dialogue, mouse over the new layer, right-click, and select "Alpha To Selection". This will select all the gold area with dotted lines again. Now, we just use the "Bucket Fill" tool to paint everything the same colour. Hold "Shift" as you click it, and it will fill the entire selection in one go. If you need to un-select the gold again, just click the main window with the "Rectange Select" tool.

Repeat the process for the black layer. As long as you dont select and drag them around, each layer will stay aligned. Sometimes, there will be some portions that were selected, but aren't part of the logo (black horizontal lines, in this case). Just erase them from the black layer, as I've started to do here:

Another issue with my black layer is that some areas weren't picked up at all, for instance where the black and gold meet. To fix it, I paint it by hand, purposely "painting outside the lines" and going into the gold. Here, the black layer is above the gold:

But, if we move the black layer below the gold, it's easier to work with. Using this method ensures that there's no gap between black and gold. Since the gold layer's edges are nearly perfect, it looks best on top:

**Tip: you can quickly and easily change the paintbrush and eraser size with the "[" and "]" keys.**

Correcting these errors is the most tedious part of the process. You don't have to be perfect, but major errors will likely show up after we scale it down.

Here's my finished black layer. Looks awful, right?:

But with the gold layer above it:

For the blue, we don't even need to select/copy/paste it. Create a new layer, below the gold and black ones. Just paint it all in by hand, like I did when I corrected the black. This will ensure there's no gaps between the blue and the other colours.

Lastly, we'll tackle the red. This is a very colourful and detailed area, notice that the gold layer didn't pick up the lettering:

Since it has no errors on its own, We'll just copy it as-is from the photo layer:

Copy, paste, and create new layer as before. Using the "Eraser" tool, we'll erase the surrounding blue. Since we already created a nice black layer (which we'll place above), we can erase some of the black edges so that we're sure there's no blue left:

That should just about do it. Saving a copy your work as an .xcf file will keep each layer seperate, in case we need to go back and darken the blue layer, for instance. But for the final product, we need to merge all the layers. From top to bottom: gold, black, red, blue.

The MSP logos are different from the right side of the car to the left, so we need to reverse ours for the other side. After merging, make a copy of the image. In the new image, make a copy of the layer by right-clicking it and selecting "Duplicate Layer". Flip the new layer with the "Flip" tool. Here, I've lowered its opacity so I can tell them apart. You might need to move it side-to-side a little to line up with the lower layer:

Hide the upper layer, and erase the "lightning bolt" part from the lower layer:

Now, working on the upper layer, erase the edges of the crest:

Move the upper layer down below the lower layer, merge the two, and fill in any gold near the edge of the crest and lightning bolt, and it's done:

From the Image menu, select "Scale Image", and make it 194 pixels high. Copy & paste it onto the template, using the reference photo for near-perfect fit:

Since it's much smaller now, many details have been lost. Some were good, and some were bad. In scaling it down, the image was smoothed a bit. We lost some sharp details that looked nice, and lost some rough edges and imperfections as well. That's why we don't have to be super-picky with errors in the high-res version. If, after scaling down, we still spot errors, we can easliy go back to our full-size .xcf file and correct them. If I choose to make the blue darker, or the gold yellower, etc., I can easily do so, since I still have the original copy, seperated into layers.

I hope this helped you create your own high-quality logo.

Lesson 8: Creating Simulated Reflective Vinyl

Ridgerunner had asked for this lesson long ago, but until now, I hadn't posted it to the public.?

Note that I said "simulated". This lesson is intended to teach how to make you graphics appear as though they're made from reflective vinyl, but they will not react to the light as, say, a specular map can do.

The basic steps are

1: Create our graphics such as stripes, lettering, logos normally, as if they weren't reflective.

2: Create a seamless pattern that looks like our desired reflective material.

3: Apply the pattern to the desired areas, but in a new layer.

Regardless of the type of material, the process of applying it to the texture is the same.

Step 1: Creating the base graphics.

We likely want the reflective pattern to appear on the stripes only, and not the background "paint" of the car. When creating the graphics, leave them in their own layer, don't merge them to a soild background before applying the pattern. If you want the pattern to be applied only to certain sections, it's even better to leave those sections in their own layer, although it's not essential.

Step 2a: Creating a seamless pattern from scratch.

We can make a pattern from scratch, or from an image. First, we'll make one from scratch, and later, after I've shown step 3, I'll show you how to make one from an image. Making one from an image is a little more complicated, and it might make more sense if we've already learned what we're trying to achieve.

Create a new image in GIMP, only 2x2 pixels large, with a transparent background.

Make one pixel 100% black. Create a new layer, and make two more pixels, in opposite corners. Make them also 100% black, but lower the layer's transparency to around 50%.

Merge the layers and save the image. Your scratch-made pattern is done, but leave its window open for now. (There's a way of saving it as a pattern instead of an image, but this method works fine for me). This is just a basic pattern, and might not be best for every type of reflective decal, so feel free to experiment with different sizes and arrangements.

Step 3: Applying the pattern.

Open your car texture. In this example of my RCMP Taurus, all of the graphics are in a single transparent layer, with the template in a separate layer. Here's a closeup of the driver's door:

Select the layer you want the pattern applied to, right-click the layer, and select "Alpha to Selection". This will apply "crawling ants" dotted lines around the edges of the graphics:

Create a new layer, since we won't be applying the pattern directly to the graphics yet. Leave the window open, and go back to the pattern you made. Right-click it and select "copy", then go back to the car texture again. Select the "Bucket Fill" tool from the toolbox. In the lower part of the toolbox, there's some options. Under "Fill Type", select "Pattern Fill". The default pattern will likely be "Pine", but we'll change it regardless. Click the little "Pine" box, and a window with a bunch of patterns will show up. the one in the top left is called "Clipboard". It's whatever the last item we copied was, in this case, or scratch-made pattern. Select it. The "crawling ants" should still be there around our graphics. In the new layer, use the "Bucket Fill" tool to fill the "crawling ants" with our pattern:

If there's areas of the graphics that you don't want the pattern applied to, just remove the pattern with the eraser tool, as I'm doing here on the crest:

Of course it's way too dark, that's why we kept the pattern in its own layer. Depending on the pattern, you may need to experiment with the layer to get it to look right. In this case, I just reduced the pattern layer's opacity to 14%:

But, you may want to try different modes of merging. In the layers dialog, it says "Mode: Normal". Beside it is a drop-down menu, with dissolve, dodge, burn, etc. If you haven't used them before, try each one to see how it effects the texture, and remember you can still adjust the opacity as well. Once satisfied, merge the layers, and you're done! Here's the same pattern, applied to a different livery, and what it looks like in real life. Click for larger sizes in order to see the reflective detail:

City of London Police, Octavia Police Car by BillKatyGemma, on Flickr