Archive | February 2013

Using Reference Images

Referring to reference images when sketching and painting can make a world of difference. If you’re like me, your mind isn’t photographic. You spend a reasonable amount of time looking at a baboon and think you have a pretty good idea what it looks like. However, once a little time has passed, your memory of it becomes more of a caricature than an accurate representation. In fact, even as you stared dreamily into the baboon’s limpid gaze you probably only really noticed features that your mind found interesting.

This can work to your advantage, helping you to capture the essence of baboon rather than a baboon-shaped husk devoid of baboon attitude. However, to go off on a different tangent, mediæval art and architecture are littered with examples of strange looking lions. Like a big dog with a curly mane, they are clearly the work of someone who hasn’t actually seen a lion. Working with a description and whatever four-legged carnivores they could find, they can be excused any discrepancies between their image and a real lion. However, in this day and age, with an internet full of images (many of them not of cats) at our fingertips, we have no excuse for drawing from memory that which we have no memory of.

Anyway, the gist of all this waffle is I decided to make more of an effort to seek out reference images before beginning a sketch, or rather I decided to make it a lot less effort for me to collect reference material and display it on the monitor screen I’m not using for painting. To this end, I have written a small program which fetches images, either from a folder on the computer, or by searching the internet with Google Image Search. It then allows you to arrange as many images as you like within the viewport, rotating, scaling, moving and cropping them to get as much reference material as possible visible at once.

RefImage

Imaginative name, eh? I decided that getting on with writing it was more important than thinking up a clever name, and now I can’t be bothered to change it (I haven’t thought of a better name yet anyway).

Screenshot showing RefImage in use (with "Oxygen" Qt widget style; It should match your native widgets and colour scheme).

Screenshot showing RefImage in use (with “Oxygen” Qt widget style; It should match your native widgets and colour scheme).

This screenshot shows RefImage in use. Search results are shown as thumbnails in the carousel at the bottom. Clicking a thumbnail loads/downloads the image in the main viewport. These images have then been scaled and cropped to show only the subject we’re interested in. (Composition is of course very important in a photo, unless it’s being used as a reference image, in which case the subject’s surroundings are just extra junk that takes up screen space and gets in the way.) While this image shows the toolbar and thumbnails, they can be hidden once you have the images you want, in order to maximise the space available for the images themselves.

The user interface is designed to make it possible to use with a tablet instead of a mouse if desired (no double clicking required!), although I usually have the tablet mapped just to the monitor I’m painting on, and run RefImage on the other monitor.

Another screenshot. Less practical, but shows a few more features, including the fact you can basically chuck images wherever you like. You can even zoom, pan and rotate the entire viewport!

Another screenshot. Less practical, but shows a few more features, including the fact you can basically chuck images wherever you like. You can even zoom, pan and rotate the entire viewport!

Development has been somewhat infrequent since it became more-or-less functional. If I discover issues or have ideas for new features while I’m using it, I try to note them down and then work on fixing/implementing many at one time. After all, there’s no point having a tool to save time if you then spend ages tweaking and improving it when you should be using it.

The source code is available from github. It is licensed under the GNU GPL, a copy of which is included with the source. In theory it should work on any platform with Qt. In practice, I have only tested it on Linux, so it may need some minor modifications before it will work on other platforms. Please report any bugs here.

Dependencies:

  • Qt: The master branch requires Qt5. There is a also a qt4 branch, which works with Qt 4. It has been tested with Qt version 4.8.4. However, it should work with any reasonably recent 4.x version.
  • qjson (only required for Qt4 branch.)

Building

Assuming a suitable development environment is set up, it should be quite simple to build. Instructions can be found in the file INSTALL.md.

Running

A .desktop file is included. Otherwise, run it from the command line using the command refimage. Information on how to use it can be found in the “Help” menu under “Introduction” and “Tips”.

Advertisements

Swimming Seal Screen-Print Design

A Design I did for a corporate swimming team.

A Design I did for a corporate swimming team.

I enjoy designing for screen printing. Because the ink is either there or not there, the design is limited to only two colours[1], usually contrasting, with no intermediate tones[2]. However, rather than feeling restricted, I like the fact that it forces you to be bold. You can revel in the stark contrast and aggressive outlines, safe in the knowledge that this is a medium which favours bold, striking designs over subtlety.

[1] Unless different stencils are used to build up multiple layers.

[2] With the exception of what can be achieved with stippling and/or hatching.

Creative Commons Licence CC-BY-NC-SA 3.0

This image is licensed under a
Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.

Pancakes

Image of pancake ingredients

Recently I’ve been experimenting with using textures in digital painting.

Using natural media, it is near impossible to get a perfectly flat, even colour. Chalk collects on the peaks as it is ground across the rough paper landscape, and the battle between surface tension and gravity dumps molecules of watercolour pigment in the canvas’ hollows or draws them out over vast expanses until the boundary between wet and dry is reached. The results can be incredible; A single brush stroke can be beautiful, its simple elegant form enhanced by the intricate level of detail that the media provide.

By contrast, digital brush strokes will do only what they are programmed to do; any interest or variation in the stroke is entirely artificial. At its most basic, a simple, flat, round brush is sufficient for most digital sketching and painting, but zoom in too far and the edges of said brush are all too obvious and none too attractive. To my mind, there are three main solutions to consider, depending on the nature of the painting:

  1. Refine the painting to the extent where the brush strokes are no longer visible. This is perhaps the most obvious solution, and it can yield impressive results. However, there are two drawbacks. Firstly, it takes time, especially for large, high-resolution paintings. Secondly, it can ruin what was a lovely loose, free, and lively drawing by meticulously dragging every last pixel into line. To spend hours refining a painting, only to realise that you’ve been slowly but surely draining it of all the life and energy it once had, is disappointing to say the least.
  2. Ignore them. They’re part of the painting. In fact, they are the painting.
    “I can see your brush strokes”
    “How careless of me. Next time I’ll make sure to cover the painting up with a cloth”
  3. Try to make them look nice. This is what I’ve been experimenting with lately, using textures to add variation to areas of flat colour and break up excessively uniform stroke edges. I find it’s easy to go too far, ending up with a result that looks somewhat like it was created by smearing paint across corrugated cardboard, so subtlety is key if textures are to be an improvement rather than a distracting novelty. Ideally, I don’t want to try to completely imitate real world painting so much as to add a little interest to a brush which, by reason of its conception in the clinical world of numbers and logic, has no soul beyond what we devise.

I’m still experimenting, so hopefully in the future I’ll have some examples of digital paintings with a freer, looser style and subtler texturing.

Creative Commons Licence CC-BY-NC-SA 3.0

This image is licensed under a
Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.