The first of what I hope to be a series of tutorials explaining from first principles how I go about making my vector traces. This is basically an introductory chapter in which I go into some detail about what tracing is, why I (and others) bother with it, and also go over the structure of the upcoming tutorials. I guess if you’re looking for anything incredibly profound or insightful, you might be a little disappointed.
What is a Vector?
In digital graphics, we divide images up into two distinct classes: raster graphics and vector graphics. Most people are familiar with raster graphics, which are by far the most common form of digital image type. In essence, a raster graphic is a graphic that is constructed from a two-dimensional array of points, called pixels. This image class is found throughout the Internet – GIF, JPEG, PNG and BMP being the most common formats. It is also the basis for the technology in your monitor.
Conversely a vector image represents the image’s content in terms of mathematically defined primitives, such as lines, circles, curves and the like. Generally speaking, for our purposes, a vector graphic breaks higher-level primitives like circles and squares down to their components - curves. Or more correctly, Cubic Bézier Curves. Don’t worry about the math, though. We don’t need to see it at all.
he main disadvantage of a raster-based image is that it does not scale. If we blow up a raster image to 4 times its original size, then we will encounter blocky degradation of the image – pixellation. Making something bigger means adding extra pixels, and this implies that we have to invent new data to stick into them. Vectors, however, can be scaled up and down with no loss in image quality, which is their greatest strength.
Why bother?
Obsessive-compulsive disorder combined with anal-retentitiveness. 
If you have a decent resolution monitor, then undoubtedly you will have run into times when you find a wallpaper which is just perfect for you, but you can’t use it because it’s far too small. Or maybe it’s covered in text or gone through the photoshop filter gauntlet. Maybe you’ve been watching anime and seen a frame that you really want as your desktop. Generally compressed video frames are very badly degraded, due to artifacts introduced by the compression agorithms and the fact that the image has often been compressed and recompressed several times. Or maybe you have a great image that has had large chunks obscured by text or something else. This is where vectoring comes in. Using the original image as a base, we trace over the image using vectors, creating a scalable, high-resolution version of the original picture. Here’s a before / after comparison:
Vectors are uniquely suited for anime-related work, because of the 2D nature of the medium. There is generally not a the same level of detail in a capture from an anime series, for example, as there is in a photograph. Taking an original image and then re-drawing vector graphics over the top of it is generally referred to as vector tracing and it’s actually quite fun. Also surprisingly relaxing, at least for me.
You don’t necessarily have to have a dedicated vector graphics application - Photoshop, a Raster application, has limited vector support in the form of paths - however you’ll find that vector tracing is easier, faster, and a lot more appealing in a proper vector graphics application. I use Adobe Illustrator, and will be writing the tutorials with this in mind, however the same principles should be applicable to any other vector graphics program, from Macromedia Flash to Corel Draw and Inkscape.
Tracing Steps & Tutorial Structure
There are several general steps to going from nothing to completed vector trace:
- Select a good starting image and prepare it for tracing
- Trace the line work
- Add the colour layers
- Convert the vector back into a high-resolution raster image
I’ll be basing the installments of this tutorial around these basic steps, with each major step getting some detail. I’ll then go over some ‘advanced’ techniques, including pathed line art, the subtractive layers method, and possibly the use of clipping masks and gradient meshes.
So that’s the introduction done. Join me for part two. Eventually.
Unrelated note: I’ve been considering putting my traces into some kind of more friendly gallery or something, but I’d like to try and keep it within the blog. Anyone know of any good options, or should I just write something myself?
I was waiting for that tutorial! Great work and thanks alot!
You could try something like WP-Gallery2 which integrates into wordpress or a standalone like zenphoto (which is great btw).
WP-Gallery2: http://wpg2.ozgreg.com/
zenphoto: http://www.zenphoto.org/
Goofus and Gallant make wallpapers is awesome. This is a great start by the way.
Oh and seconding zenphoto, I think there is actually a way to integrate it with wordpress but you’d have to check the site forums for that info as I’m using that silly korean blog script. Zenphoto is pretty slick and I’m moving my own vector trace stuff to it once the next version releases with subfolder support.
A very good introduction to vector-tracing. I like it.
Still, I’m not so fervent a Jack Sparrow, and stealing Illustrator is stealing, so do you know of any open-source/free vector graphics applications that I can use? I can only think of Inscape at the moment.
On a somewhat n00bish note, we can still use our traditional mouse to trace the character outlines to form a vector, right? I tried that for some time, and it was kinda hard to trace the character, let alone trying to trace all the facial features. (I might figured that you would cover those things in future tutorials, so I’ll let you decide the content of your tutorials.)
Sweet!
I was actually quite curious on how you did your traces.
I’ll be sure to follow your tutorials.
TP,
Inkscape is open-source and works quite nicely once you spend about 5 minutes getting used to the interface. I’ve found it can do just about everything Illustrator can besides gradient meshes and a few other tools. And yes, a mouse works fine.
As for the tutorial, I’m looking forward to the gradient meshes part if you decide to put it in. My utter inability to figure them out is the only thing keeping me from migrating to Illustrator.
Gradient meshes are pretty straight forward concept (gradients are created between all node points with different colors) but execution of use is nightmarish, which is why I think not many tutorials exist on them.
Personally, I think gradient meshes need to be tried out in a trail-and-error fashion to get used to them. (Once some basics are thrown down like picking the correct base shapes to apply the mesh to, and then applying the mesh in an efficient density.)
One thing I love about Illustrator is that once you use both gradient meshes and gradient masks, you can do almost anything… try googling Yukio Miyamoto for some interesting and beautiful examples.
Mind you I don’t know anything about masks, and I still have trouble with meshes so I’m really looking forward to those tutorials!
TP: Piracy isn’t stealing, it’s copying. There’s a big difference - don’t believe what the recording industry tells you.
Regardless, Inkscape is pretty much your only option. There’s also Sodipodi, but it sucks.
Woo nice! I have no skills as an artist, but it’s nice to know the process behind the work.
I could use a Inkscape tutorial…
By the way, if you want to set up a WP Gallery: http://codex.wordpress.org/Plugins/Images
AniGA Gallery is amazing (I tested it myself, anyone can leave comments on your images - or not, it’s configurable; you can decide how to show the image; and you can upload the images without an FTP). Don’t try PhotoPress because it’s broken.
The rest of the plugins are based on FTP uploads (like, searching a folder for images like /images/) or integration of 3rd party services like Flickr or Gallery2.
Please stop asking for Inkscape tutorials. I can hardly give people a tutorial for a program that I’ve used for 15 minutes a year ago and currently cannot get to run.
Need other parts tbh :