I recently began a project to design and build a custom mechanical keyboard, with a custom ergonomic design. From what I know so far I am using CherryMX Blue mechanical switches, a Teensy++ 2.0 for the processing and programmability. I am currently researching lighting for the keyboard, and case design, while designing the rest of the circuitry.
I have, for the past 6 years, been working on techniques of processing photos and other digital artwork using GIMP. As a result of that study I have developed a number of advanced techniques, one of which is using individual layers tuned for specific areas of the image to correct colors.
The first step in the process is to decide what needs to be changed about the image. For this generally I try for color correction on each color or texture of the image, along with any specific areas that need work. Figure 1 shows an example image with different generalized areas highlighted for correction. This step is essential to the success of the rest of the processing, because it allows you to decide what each new layer you work on should do for the image.
Figure 1: Choosing the regions of an image.
After you decide on the sections you then need to make a general outline of what you are going to be doing to each section. Starting with the Red outlined region, I decided to make significant adjustments to the saturation and lightness of this section, primarily to make it look the way it did to me at the time, but also because it looked rather dull as it was. The Yellow region I decided needed a careful attention to make sure that it kept all of the detail, while meshing with the neighboring regions. The Green section needed the tree removed, because of the poor stitching job on it, and other than that I wanted to process it the same as the Magenta sectors. The Cyan section needed contrast and saturation enhancements. Blue needed similar saturation, but no contrast adjustments. Finally the Magenta sections needed to be processed for optimal white balance and contrast.
You can start hard or easy while processing in this fashion, and I tend to start hard, but in this case I started with the easy part. The cloning for the Green section was what I decided needed taking care of right away, because it would make all of the processing layer based, instead of having to finalize a version of it and then do the cloning. I won't go into detail about cloning in this article, but the clone was particularly easy, despite the large area of image because of the almost solid background color. From there I decided to take care of regions Red, Cyan, Blue, Magenta, further adjustments to Red, and some overall tweaks, the whole time balancing Yellow.
In the Part 2 article I will discuss the particular layer modes I used to do the adjustments. Another important skill that I will be covering soon that I used extensively is color correction via the levels and curves tools. For now I will leave you with a screen-shot of the open GIMP file used for this image.
Finished edit file for the pano. All layers are labeled descriptively.
For several projects I am currently working on I am I have begun to learn the Git Version Control System. I thought I would share some of my initial ideas on the system, and give a little overview of VCS software in general.
VCS software such as Git is used primarily for keeping track of software development, allowing development by multiple people while avoiding or resolving many conflicts between work done simultaneously. Though coding is the primary use of such software, it can be used for many other things as well. Some of the other uses include keeping track of personal writing or artwork projects, or working with others on such a project.
The usability of Git is not quite as easy as I would like to see out of such a system initially, but it is superbly functional. As I have begun to use the software more I am becoming more comfortable with it, though I still have much to learn. The features of the software is easily enough to overcome the difficulty of learning initially.
As I use Git more I will try to post more of my thoughts about it, and possibly even a short guide to using it.
I was just asked by a friend to give an overview of the tools available in GIMP, so I decided to share that with all of you readers.
Figure 1
I will start off with showing the basic groupings of tools, which is shown in illustration 1. The red (first) group is the selection tools, which may be combined into a single tool in the future. Each of these tools is particularly useful in certain situations. The next tool isn't highlighted at all because it doesn't fit into any of the categories. It is the path tool, used to create vector shapes, sometimes useful for marking out shapes, or storing selections. The magenta (second) group is what I like to call support tools, used for various tasks throughout some design projects. The green (third) group consists of transform tools, used to move rotate, align, and skew layers within the image. Technically this group should contain the two tools highlighted in yellow, but I find that those two are usually best used in special situations. In the middle of that group there is a knife icon, which is the crop tool, and directly following it is the text tool. The cyan (fourth) group is the fill tools, the first for solid color, the second for gradients. The blue (last) group is the paint tools, used for repairing the image, and painting.
Within the selection tools, in order: box, circle, free, fuzzy, color, intelligent scissors, and foreground. The box and circle tools are used for basic shape selection and can be useful for image cropping. The free select tool is also shape based, but the shape is user defined, so it becomes really useful for selecting objects or regions within the image that contain varied color or texture. Fuzzy and color select are quite similar, both working off of similar color, differentiated by the fact that fuzzy will only select connected areas. Intelligent scissors follow the edges of brightness or color, and can make selection of complex objects that are differentiated from their background. The foreground select tool is used when there is a single color background, and uses the background color of the pallet to determine what not to select.
I won't elaborate on the path tool for the present.
The 'support' tools are, in order, eye dropper, zoom, and measure. The eyedropper is used to pick colors from the image for use during editing. A quick tip relating to colors is to use the X key to swap the foreground and background colors in the pallet. The zoom tool is becoming less useful, but can serve as a reminder of the Ctrl+scroll shortcut to zoom in/out on the picture. If you have a mouse which allows clicking with the wheel you can scroll around the canvas that way. The measure tool (Figure 2) is pretty self explanatory, it is just used to measure distances and angles.
Figure 2
Figure 3
Transform tools (including special ones) in order are, move, align, rotate, scale, sheer, perspective, flip, and cage transform. I will skip further explanation of move, rotate, scale, and flip, as they are quite intuitive. Align is used to align layers to different layers or guides. Sheer is rarely useful, but shifts the lines within the image one way or another on either the horizontal or vertical axes. Perspective (Figure 3) is sometimes quite useful, such as when correcting lens distortion, or you simply want a somewhat 3D effect for your image. Cage transform allows you to select an area defined by a series of points, and then stretch/modify the area based on movement of those points.
The text tool is used for adding text to an image, useful for posters etc.
The fill and gradient tools are fairly standard also, so I will leave further explanation until I am ready to post an advanced tutorial on their use.
The last grouping of tools, in order: pencil, paintbrush, eraser, airbrush, calligraphy (only useful with a graphics tablet), clone, heal, perspective clone, blur/sharpen, smear, and dodge/burn. The pencil, paintbrush and airbrush tools are quites similar, differentiated by the fact that the pencil doesn't support partial transparency for brushes, and the airbrush doesn't lay all the color down at the same time (how fast is governed by a 'rate' setting in the tool preferences). The eraser either puts down the background color, or turns and area transparent. clone and heal are very similar also both using a point of the image (set by Ctrl-clicking) and paints it onto another part of the image. That becomes very useful for removing parts of a picture or removing dust spots. The smear brush is also has an obvious functionality.