We will start by selecting the Rectangle Select Tool, which is conveniently the first icon in the toolbox.

Most options must be set before selecting a rectangle.

First, we will feather the edges.

The smaller the radius, the sharper the edges will look.  For this example, I am using a radius of 10 pixels.

We will also be applying rounded corners to aid in the look and feel.

Once again, the smaller the radius, the more the corners will look like, well, corners.  A radius of zero will give sharp corners.  This can be adjusted after making the selection.

It is now time to create the selection.  Prior to making the selection, I’ve created a new, layer with a transparent background.

Create the selection by starting at one corner and dragging to the corner diagonal from the start point.

At this point, you will see the rounded corners, and if you’ve selected a huge feathering radius, you will also see that.

Now that we have the selection, grab the Blend Tool.

Be sure to select both a foreground and a background color, as both will be used in the blend.  Also, while not needed, I’ve applied a bit of transparency to the blend.  This can be useful if you plan on overlaying the final result on top of something else.  Everything else is set to the default values.

This next part I wasn’t able to get a screenshot of, but to make the blend, click in the upper region of the selection, and drag to the lower region, it may take a time or two to get it just how you like it.  In order to keep the blend perfectly straight, hold the Crtl key down while dragging.  This is what it will look like just after the blend is applied:

To see the final product of your hard work, deselect (Ctrl+Shift+a).

If you remembered to put it on it’s own transparent layer, you can turn off the background layer (which is currently why the background is white).

With a little bit of tweaking to this technique, you can do quite a bit:

Each of the many modes provide a different way to display the layer, and the opacity determines how transparent/visable the layer is.  Below is a sampling of just a handful of the different modes and how they effect a layer.

Normal

Normal is the default mode for all layers.  The above shows the yellow oval with a 50% opacity.

Dissolve

Again showed at 50% opacity, dissolve essentially dissolves the layer pixel by pixel with nearly all of the layer showing at 100% opacity, and nearly nothing showing as the opacity approaches 0%.

Multiply

Multiply makes the layer feel almost like a lens filter and gives the effect of mixing colors.

Difference

Difference shows the color remaining after removing the layer’s color from the background.

A) The New Layer button does just that, it allows you to create a new layer.  New layers will appear above the currently selected layer.

B) The Raise and Lower buttons allow you to move the layer up and down the list. Personally, I tend not to use these buttons because layers can also be dragged up and down in the list, which I find to be much faster and convenient.

Below: The layer ‘Green Circle’ has been moved from below ‘Orange Rectangle’ to above it.

C) If you just need to make a small change to a layer, but still want the original to fall back on, the Duplicate Layer button is for you.  Once again, another button that does exactly what it sounds like it does — it copies the currently selected layer and inserts a copy of it above.

D) Don’t want a layer? Just click the Delete button.  You can also drag the layer onto the button.

E) Most of the time, the active layer will be listed here. Occasionally other information will be shown here, generally relating to the active tool.

F) Layer Visibility.  When the eye is shown, so is the layer.  A single click toggles the visiblity.

G) The active layer will always be highlighted.  Any changes performed will affect this layer.

H) Layer Preview.  Use this to quickly get an idea of what the layer holds.

I) Layer Name.  When working with just a few layers, I generally don’t name them, either because I have so few, or because I know the preview will give me a good idea of what the layer hold.  With that said, it is still a good idea to name all your layers to prevent confusion.  The easiest way to do this is to double click on the current name.

As mentioned at the start, this is only a brief introduction to the interface.  Look for the next part later this week.  (Ok, who am I kidding, it will probably be next week after SELF.)

Sometimes you would like to make one part of an image stand out from the rest. In this case, I’ve turned the image into a black and white photo, with the exception of the bottom of the ski that is being held up in the air.

Here is the original photo that I am using for this:

We will begin by duplicating the layer. Since we just opened the image, there is only one layer. Select the ‘Duplicate Later’ button as pointed to in this screenshot. Alternatively, you can right click the layer you wish to duplicate and select it from that menu.

Both layers after duplication:

From now on, we will only be working with the new layer, ‘Background copy’.

With this layer selected, choose ‘Desaturate’ from the ‘Colors’ menu. We will use this to turn the layer from color to black and white.

Select the default options (or play around with them to see what works best for you).

We will now add a Layer Mask to the layer.  The mask will allow us to select what parts of the layer we want to show, and which parts we want to be transparent.

Since we want most of the new layer to show, we are going to choose to initialize the layer mask to white.  When the mask is white, the layer shows; when the mask is black, the layer is transparent. Shades of gray result in partial transparency.

This is what they layers will look like when we’ve added the mask.  The white box shows a thumbnail of the mask.

We are now ready to start carving into the mask. There are many ways to do this, from using the paintbrush or pencil tool to creating a selection (using the rectangle, circle, freehand, or path tools) and filling them with the paint bucket.

Because of the shape and size of the ski, I have elected to use the path tool.

Once you create the path around the object, create a selection from the path. For more help on this, visit the tutorial on using the path tool.

Selected ski:

With it selected, grab the paint bucket and dump black into that area, this will turn that portion of the layer mask to black, and allow the untouched layer to show.

Deselect and you are ready to share your creation.

While there are many ways to approach fixing this, I had the best luck with GIMP’s Colorize tool. This tool give you the ability to adjust the image’s Hue, Saturation, and Lightness.

Adjusting the hue up will add more red to the image, while down will add more yellow.  Less saturation will essentially remove color from the image, and more saturation will make the colors more profound. For my case, even though I wanted the snow and lights to be nearly white, I didn’t want it to be completely white.  The adjustment to the hue managed to change most of the turquoise to a blue and once the saturation was turned down, the blue was left just as an accent.

Here is the final product after those adjustments:

When preparing a picture for the web, often you don’t want to include the background. In the last tutorial, the background of my desk was quite obvious. Lets use the path tool to remove it.

We are going to start with the same base image from last time.

In the last tutorial, I showed one way to zoom in on an image. A faster, and much more convenient way is to hold “Ctrl” and scroll up to zoom in, or down to zoom out. In order to put a path around the mouse, we are going to zoom in on the mouse.

After zooming in on the object that we are about to put the path around, we will select the path tool.

The path tool is a bit tricky to figure out, but once you understand it, it is very easy to use. The following screen capture shows me quickly putting a path around the mouse. Following the video, I explain exactly what I’m doing.

The first thing to do is select the major corners if there are any and the parts where curves change. To close the path, hold “Ctrl” and click on the first point. Once the rough path is complete, extra points are added in the curve by holding “Ctrl”, clicking on the line, and dragging out to the edge of the mouse. When these points are added, we can use the adjusting handles to adjust the curve. If “Shift” is being held, both handles will move together.

A curve can also be put in just by dragging the line without holding “Ctrl”. This generally produces a steeper curve, but once again, can be adjusted using the same adjustment handles, but from either size point.

Additionally, holding “Ctrl” and dragging from any point will add the handles for further adjustments.

After the path is finalized, it is time to turn the path into a selection. This is easily done by selecting the “Selection from Path” button.

With just the object selected, we will crop the image down to just the selected object. Just like in the last tutorial, select “Image” then “Crop to Selection”. The outlying image is cropped down to just a rectangle around the selected object.

Next, it is time to remove the rest of the area. The first thing to do is to select the area we want to delete. Currently, the object is selected. To do this, we want to invert the selection. Select this from the menu as seen below, or “Ctrl+I”.

Once the outlying area is selected, hit Delete. You will be left with a white background. At this point, you can save for web as shown in the last tutorial.

A few side notes: You may notice the brown tint around the completed image above. This was caused by reflection of light off the desk. Most of the time when I do this I use a white background instead of a desk or other color for this reason. It can be adjusted, but that is beyond the scope of this tutorial.