Bones Drawing Guide
Welcome to the Bones Drawing Guide, ready as promised. This entry is designed to rapidly teach the important basics about how to draw. The first rule is, don't assume you aren't capable of drawing!
The art for these lessons was drawn on paper and scanned; it also applies for drawing vector art in Powerpoint along the lines of my previous guides (1, 2). The text was originally handwritten but is also typed below each pagescan in case you can't read my handwriting, plus expanded on a little.
It is partly inspired by Bruce Mcintyre's "How to Draw", which is the text I learned to draw from, and I highly recommend. (Here's what the book looks like.) That book has tons of lessons and a system of important elements of drawing. So as not to plagiarize, and because I think his system is a bit incomplete, I've formed my own system that I think is more logical. But if you want a more complete collection of practice lessons, try to get your hands on that book or a similar such book.
A very important point he made is that everybody learns how to write, but contradictorily, we aren't properly taught how to draw. If you're one of those people who always utters the cliche "I can't draw/I have no talent/etc.", ask yourself, can you write? You might not have the talent to become a "writer", but you can put a pencil to paper and draw letters that form words. Why? Because you learned how -- and because that helps you communicate. Or same with typing.
Likewise, you might not have the talent to become a famous artist or the like, but you ARE capable of learning how to draw, so that you can communicate visual ideas. Or maybe you DO have the talent. You might not even know it yet.
To treat this entry as lessons, I recommend spending a day drawing your own stuff based on each rule, then spending a day on each column in the examples category. Both draw the same things I showed and draw your own, with emphasis on your own. Or you can just read it or treat it as reference; I've designed it to be easily referenced.
The main thing that you'll find lacking here is huge amounts of lessons -- this IS freeware after all.
To those that would say "I was already taught how to draw" -- well, there are many bad methods being taught out there in schools (for example, the wrongway cube method; see below), and there's a lot of different things that I've compressed all into one guide. This also has a Bionicle focus you might find helpful. Even if nothing in this is new to you, it may still be interesting.
Rules - the logically organized rules for drawing
A: 3D Shape
1: Outer Shape
3: Avoid Flat
8: Vanishing Points
4: Surface Shine
5: Cast Shadows
8: Metal Shading
Examples - five pages of various examples
1: Wrongway/Rightway cubes and cylinders, foreshortening, and vanishing points.
2: Metallic shading, doughtnut foreshortening, etc.
3: Facial expressions/emotions on a simple Bionicle-esque face, and Human/Toa proportions.
4: More about vanishing points; parabolic fully accurate system, and skyscraper system.
5: Various stuff to close out the guide, especially 3D letters.
This section presumes that we're not dealing in abstract / heavily stylized art, but in realism. When you get into abstraction these rules can be bent or even ignored at will -- but it helps even there to know to draw in realism first so you know what to bend/ignore.
Section also presumes we're drawing the shapes with black lines, either on paper or on the computer, rather than a more photorealistic no-line edge system (as with paint or the computer equivalent). It's the easiest way to draw, though no-line systems are more realistic. For the most part the rules apply regardless of this, just wanted to be clear.
On paper, I recommend using something erasable (I prefer Papermate erasable black pens since you don't have to sharpen them unlike pencils). If you use pencil, try to get a real black pencil, not just the silvery gray school pencils, and sharpen it often, preferably with an electric pencil sharpener.
On the computer, obviously I would recommend using Powerpoint (see the previous guides as linked above) or another vector program, instead of bitmap/raster programs, because the lines can be edited much more easily with vectors.
Finally, I'm not dealing with color at all here. Just one bit of advice for coloring on paper -- I strongly recommending scanning your lineart WITHOUT color, then printing it, and coloring the printout. That way if your particular chosen method of coloring messes it up in your opinion, you can always try again. (And I'd recommend Prismacolor pencils for that.)
Aaand without further ado...
A: 3D Shape
1: Outer Shape
See in your mind the 3D outer shape, or a basic idea of it, before you begin drawing. Be able to rotate it in your mind. Think of drawing as a way to photograph your imagination -- you have a "camera" that lets you draw a picture of the 3D shape inside your head. NOTE that once you begin on paper sometimes you have adjust the 3D shape in your mind as you go as the pen does something different from what you planned, but you should still remember to form a(n updated) 3D shape to guide where your pen goes next.
Not all shapes have an inner structure, but keep the concept in mind. Organic (and biomechanical) creatures especially have a skeleton & muscles underneath that shows through. NOTE the bone structure of the "Tuskhop mouse" (ignoring that I ran out of room to make him as normalish as I wanted ) -- those are the same basic bone shapes, connection points, and angles that most mammals share.
3: Avoid Flat
Real objects are 3D. Keep that in mind as you draw "2D" art -- your paper is 2D, but it is merely a flat window into a 3D world. NOTE: avoid the "wrongway cube" technique that is sadly taught in many public schools (I speak from personal experience >_<) that has you draw a flat square and then magically have the rest of the cube appearing at a ridiculous angle behind it -- such a shape is NOT a cube. You CAN draw a house from the "flat" angle shown above, but the chances of your art "camera" actually aiming at a house from exactly that angle are slim, and it doesn't look realistic.
Your "camera" points at the 3D shape you're drawing from an angle. Choose the angle you want.
Generally make sure that the angle is the same on all things in a single image. Exceptions are when you're close enough to things (or they're large enough) that the vanishing point system and the like is needed; see later rules on that.
Also Ojh and I ran into the situation when making our RPG of a false angle system built into it -- the chipset "ground" tiles are forced into a pure top view, yet you usually see the sides of charset people/objects. So the angle system we used there was to always go with the angle that best showed the object, so in the same area you could have somethings topview, some angled sideview, and some pure sideview.
Regardless, a big chunk of the artistry of an image is the angle you choose, so choose wisely.
Closer objects look "lower" on a floor.
Farther = smaller (for two objects of the same size, and also for different parts of a close/big object).
Usually, farther objects/parts have less detail or texture density, and are drawn more lightly. Especially distant mountains due to atmospheric blueing.
Exceptions to the above Focus rule; curved textures look denser on the farther edges, where you're looking at the texture at a greater angle, like the tree bark, leaves, and wireframes above.
Multiple objects/parts overlapping can be tricky. Try imagining the lines of the shape behind (re: dotted lines). ALSO when drawing on paper, always start with the nearest overlapping objects first, then draw back (on computer generally do the opposite).
At angles, circles (as in at the top of cylinders) become ovals, not circles with malformed thicknesses magically appearing. Slightly "football" shaped, too. And foreshortening applies to all shapes.
NOTE with the cube that ALL parts of it must be foreshortened. The only way you could realistically draw a perfect un-foreshortened square would be if your "camera angle" was directly perpendicular to the cubeface, and in that case, the thickness would NOT be visible, so all you would draw is a square.
Can't emphasize the cube thing enough, as you've probably had gradeschool art teachers that didn't know better. Try looking at an actual physical cube -- you'll see I'm right.
8: Vanishing Points
Most accurate way to foreshorten. Use especially for: 1) large objects, 2) closeup objects, 3) room interiors, 4) simple geometric shapes, 5) roads. NOTE especially that this applies to curved and complex objects, NOT just to rectangular ones, though it's usually harder to tell, hence the foreshortened circle inside a vanishing point grid.
You may want to skip down to the larger illustration on vanishing points in the Examples section, then come back up here, as it makes this much clearer.
Easiest way to create depth to objects, and applies to many real things, not just imaginary wireframes.
Examples to take note of: pipes/tubes (note the use of this for a Kanohi tube in the Examples section), rocklayers (note in the above example the thicker and thinner varying layers as in real sedimentarily laid-down rock and in volcanic rocks often), and the wood/bark. With the first two of these examples, the lines are perpendicular to the cylindrical direction of the shape, thus curve around it, but with the bark, the lines are parallel with the branches. Also try crisscrossing diagonals and more.
More examples: fur; notice how instead of using solid outlines on the two creatures above I made the outlines out of "//////" shapes for the rodentlike creature and interlocking curves for the horselike creature's mane. For things with wiry hair (like a mostly bald guy's forehead :-P) you would go and draw a solid outline and draw the wiry hairs coming off of it, but these things are rare.
Also stripes, wrinkles, and on biological creatures especially mammals, underlying structure that shows through the skin and fur such as ribs, other bones, and muscles.
Note the infected Hau to show you that spots don't necessarily have to be cheesy polkadots. :-P Be sure to foreshorten. Also note the two dots on the shirt on the right shoulder -- not only must they be foreshortened, but they must also be curved to wrap with the curve of the cloth there.
A specific line texture technique that adds age to things, especially metal, stone, wood, plastic etc., while also adding depth. Usually you want to make sure each nick is at a very different angle from the others, so it's not confused with parallel or perpendicular line texture. Also for curved objects (like the biomechanical leg shown above) make the nicks more common on the edges (wrap).
In some cases like reptiles or rooves. With scales like the reptile's above (or like fish scales), draw the "frontmost" scales first; the ones that overlap the others behind it.
Convex lines point out from "center" of surface/angle. I put center in quotes because on the bumpy bone-forehead of the reptilian creature above, there isn't an exact circular center. So more properly you could say angling outward perpendicular from the surface. Note also the atom-like bumpy sphere, where the bumps are angling away from a true center.
Yes, I know the numbering is wrong starting with this rule in the image above. I don't have time now to fix. Basically this is the opposite of bumps, using concave lines pointing in towards the "center." Note the sandy ground texture in the lowermost image above, and that the footprints use indentations inside bumps, basically.
They must foreshorten and curve around textures. This may very well be the most difficult part of realistic art (on paper), as your muscle memory constantly tries to lure you back into normal 2D writing mode. See the Examples section for much more illustrations of this.
Weight distorts texture, depending on what it is. This rule corresponds to the following things in the above image: the candlewax drips, the footprint impressions and pushed-up-and-out sand around them, and the greater amount of grass at the base of the tree, and the slightly higher ground there. The ways this rule manifests abound.
Remember that light is radiating energy. It comes from a source, and whether that source is "onscreen" or off in the image your "camera" happens to "take", all lighting and shading and cast shadows and the like must take this into account.
The paper version above is confusing on this one, so lemme be clear:
With lamps and most similar artificial "terrestrial" light sources, the shadows and shading radiate away at many different angles, all radiating from the source. Even the two different sides of the shadow are at least a little angled differently from each other. Note especially the shadow of the lower-left cylinder in the image with the lamp. Also shadows of farther objects are often longer than closer objects (depending on height and yadda).
With sunlight, all the light is mostly from one direction at a time, since the sun is so far away (and is bigger than Earth by a ton).
The three balls above cannot possibly be in the same image because the light direction differs randomly (barring weird atmospheric bending effects :-P). Choose your lighting directions carefully. By default I usually go with light coming from the upper left corner if I don't wanna bother figuring out where a source is (for when I'm just drawing an object and not its background just to illustrate the object itself etc.).
Refers to darkened surfaces of objects, opposite the light sources. Note the gradient fading effect of the shading on curved objects; realistically there is no definite linebreak between light and dark on them, unlike cornered objects. Exceptions can include nighttime shading and especially outer space shading (and simple cartoon styles).
(For shading in .ppt I generally use layers of trans-black.)
4: Surface Shine
The opposite of shading; refers to lightened surfaces of objects, facing the light sources. On paper, 'tis the abscence of shading. On computer art such as .ppt, I generally use either layers of trans-white, or radial "shine" ovals (or both).
5: Cast Shadows
Refers to darkened area cast BY the object onto other objects, the floor, etc. Again, shading refers to the darkened areas of the object itself, which is not the same thing as shadows.
In direct sunlight, the edges of shadows are sharp edges; not blurred at all. Shadows cast by lamps generally begin with sharp edges but gradually fade into blurry edges the farther away from the object and lamp they get. Shadows in partial cloudy sunlight are extremely blurry.
In cases of multiple sources of light, there are multiple shadows, and the shading/lighting on the objects themselves are blended between the multiple systems too. When two cast shadows overlap, they are twice as dark (or to put it another way, where they don't overlap, they are only half as dark). I recommend doing all the shading for a single light source first lightly, then doing all of it for the other, then adding extra darkness to the overlapped shadow areas.
NOTE that shadow edge blending is technically a case of tons of multisources -- one the original bright source's center, two all diffuse light reflecting off other nearby objects, walls, floor, through clouds, etc., and three the fact that close/large light sources have light coming from all over them, not just their point-centers.
Direct sunlight can be treated as a point-center-only light source, as can distant streetlamps or sports stadium floodlights, etc. But a household lamp that is near the object in question has light coming from a wide area, from the top of the lampshade to the bottom, and from the left side to the right. So essentially it's thousands of multisources, times a bazillion when you factor in reflected diffuse light.
Reflected images get complex fast. Basically imagine a dotted line perpendicular to the mirroring surface -- the incoming angles of light with respect to this line equal the outgoing angles. Do this for every point on the mirror inside your imagination's 3D shape, and you can come to a habitual understanding of reflections. Also just try holding various shapes in front of a real mirror and observe the real world.
Explanations of the other text in the above cardscan:
The image at the top shows the reflection of the tall rectangular box shape in a mirror that's at an angle. If you go to a shoestore, they often have mirrors against benches in this exact position. Study those mirrors and note how wacky the reflection is compared to the real objects.
Or, to replicate that scene without going through the dotted line process for every point, you can follow three steps: 1) draw the exact flip of the original object in a mirror at a 45 degree angle to both the object and your "camera", 2) angle the mirror down, and angle the reflection down by the same distance. So the reflection is twice as angled as the mirror, and 3) Move your camera up, and angle the reflection so now you're more up than sideways. The ceiling will generally be partly visible.
There's another even easier way to do that image, which I'll get back to. First lemme explain it for a simple puddle.
With the cylinder above the puddle, notice two key things. 1) The upside-down base of the reflection is the base of the object, NOT the top of the puddle. Since the puddle begins a ways away, the reflection looks cut off. (Also I showed another cuttoff part of the reflection farther out where the ground rises out of the puddle again; so treat the reflection as if it's "behind" the outline of the ground; as if the ground is overlapping it.) 2) If you flip the image upside down, you're now looking "up" at the pillar in the water, instead of down at the original pillar. The top of the original pillar shows a foreshortened circle, but the top of the reflection does NOT.
So back to the angled mirror, another way to draw it (and the way I used incidentally) is to mentally or actually angle the paper so the mirror looks "flat" to you, like the puddle, and then just draw the upside-down-flipped version of everything. If you get good enough you will be able to do this totally mentally without tilting your paper at all.
Finally with the mountain lake scene, notice again the bases and the upside-down-flipped aspects. There are three different (basic) bases; the shore earth's base which is right on the edge of the lake, the bases of the trees, and the bases of the mountains.
The shore base actually curves and bends, and where the shore juts out to the left, you see what could be considered a fourth base for that portion of the shore. Also, since the shore curves/angles away from the water instead of being perpendicular to it, the reflection is thinner.
The tree base is the most complex. The tree trunks go into the earth higher up than the lake. So the reflection base is actually somewhere inside the dirt under the trees, not at the visible base of the trees themselves. So imagine that each tree is actually on top of a pillar of dirt inside the ground, and the base of the pillar is at water level (and that might very well be where the ground water level is). In addition to that, each tree's base is farther away, thus "higher" on the puddle's plane, so less likely to peek over the shore, much less the other trees. So the only trees you see in the reflection are the closest ones.
The mountains are the easiest -- they're so far away the horizon line acts like their reflection base. The deviance from that line is too slight to worry about. NOTE that the middle mountain's reflection is actually not done quite right; there should be far less of it visible since it's the farthest of the three mountains judging by overlap.
Now, look at that image upside-down (if you can manage that onscreen :-P) and look at the reflection. Doesn't it look as if you're looking up at the mountain scene through a hole in the ground? That's basically what you're doing.
8: Metal Shading
Metallic curved surfaces almost always have a bright edge on the farthest side from the light source (reflected light from the ground on their mirror-like surfaces, intensity depending on how reflective the metal is), more extreme contrast (gradient goes to pure white towards the light source and to pure black away from it before the bright-edge), and a leading edge darker area.
For flat metal, use many randomly spaced diagonal gradient lines. All on one surface are parallel, but lines on the next surface align differently. This isn't pure realism here, but pure realism often makes it harder to understand as metal at a glance.
Another technique for flat metal is to ripple and patchy-ify the lighting based on the slight ripples and bumps in the metal's surface, which I didn't happen to draw, but yeah.
Even trickier than reflection. Usually water is all you need to worry about. It's another case where it's best to imagine the actual light rays. Also taking a college physics course would help. :-P
Note most extreme angle does not refect at all but reflects down off the air back into the water. That's basically what fiber optics takes advantage of incidentally.
There is a slight foreshortening of the whole shape, which gets more extreme the more extreme the angle of the "camera" is.
This section shows some of the most commonly important concepts in art (with a Bionicle focus). Each of these pages basically goes down columns then up to the next column etc.
To start this off, look at the wrongway cube that is sadly often taught in art schools. The two images below that show what that mangled shape would actually look like if you angled it different ways. It's NOT a cube -- it's a parallelogram prism.
Note that the first example under Rightway cube started out too short so I gave it a base. The three examples below that show actual cubes, and below that you see the only situation in which you should actually draw a perfect square that isn't foreshortened -- here you're looking directly at one of the cube's six faces, and thus you do NOT see the other faces at all.
Then there are some wireframe stacked cubes and such.
Back up to the topo' the next column, note the wrongway cylinder, which is actually a... diagonally squished pop can? :-P Then note the rightway, and the top view.
Third column shows some examples of foreshortening, under the vanishing point system.
First column starts with metallic lighting. There's another example of multisource lighting with a metallic cylinder -- notice the multiple bright lines in the cylinder's gradient. Then some random stuff.
Second column starts with a rounded cube. The shading is solid on the flat parts but a gradient on the curved. The doughnut and snakes show what happens when a tubelike shape is foreshortened. And the rattle cobra shows various snake features in one creature for the heck of it.
Third column shows various stuff, especially letters.
First column is devoted to facial expressions, on a simple Bionicle-esque face, and with VeggieTales faces.
At the top of the second column is a geometric basic idea of a Toa Metru ish face, and then all the parts shown seperately.
Below that is basic human/Toa proportions. Note that for humans the arms should be a little shorter; Toa arms tend to be longer. Basic idea is to use the head heighth as a measuring unit for the rest of the body.
Also some general tips: Try drawing real plastic Kanohi and Bionicle/LEGO pieces, as your eyes actually see them. Try sketching lightly first then adding detail. Also try molding from clay or Sculpy or the like, then drawing that shape.
Exact text (mostly), starting with the emotion column:
Simple Bionicle-esque example face to show emotion.
"that's weird", yawn, annoyed
VeggieTales style emotes (satisfied/halfsmile, annoyed glare, "you're crazy" glare, super-scared, coolguy grin, insane, very annoyed, and singing uplook.
Example Toa-style head, using simple geometric shapes.
NOTE: Average 0nly. Contrary to popular myth, Bionicle characters do not need to be in human proportions, as they're not related to humans and are biomechanical. However, Toa tend to be close; same with similar taller beings. Matoran (and human children, by the way) tend to be more like 4 to 6 times the head, etc. And even adult humans vary.
Total height of a proportionate humanoid is 8 times the head.
Legs in proportionate humanoid equal half the body height.
Arm length may vary in biomechanical/etc. beings.
These arms are longer than a human's by about half or 3/4 of a head; about average for a Toa.
(Human belly button at 3 times the head (down from top of head).)
This page shows the Parabolic 3 Vanishing Point System, and the Skyscraper 2 Vanishing Point System.
Note the lower left diagram, which shows that a fully accurate vanishing point system for all objects actually has 5 vanishing points; one for the distant horizon you're facing, two for left and right parts of horizon, and two for up and down.
Example: if you look down at a cube or up at a skyscraper, the verticle lines should actually be slightly bent towards the corresponding up/down point. The skyscraper's top here bends in towards the up vanishing point because you are looking up at it.
This applies for close/large objects. Small objects you're looking at from a distance don't necessarily need this.
The above typed text says what I handwrote on the pagescan, but says it better, so I won't bother typing the exact wording of the handwriting here, except that the random building is shaped like a rocket from top view thus I've dubbed it the "Rocket Hotel". Reasoning being to show off the "knife-edge" style that many modern skyscrapers now have, instead of all 90 degree corners.
And finally some various things to finish out this guide.
Topleft is a reminder to add "thicknesses" to things like Kanohi masks or windowsills, etc. Only paper actually has no (noteworthy) thickness. That's a randomly coolified Miru there.
Below that are four simple steps for drawing a plastic/metal/etc. 8-point star:
1) Draw a foreshortened "paper" square.
2) Divvy it up into four parts by cutting it in half with a line lengthwise and widthwise, and draw a verticle up from the center.
3) Draw a circle within the foreshortened square and halve its sections with lines (giving you eight radiating lines from the center of the circle, all foreshortened automatically), plus another smaller circle inside.
4) Draw the final lines between the points in question -- the center at the top of the verticle line, the outer tips of the star points where the eight radiating lines hit the outer circle, and the eight points on the inner circle exactly between the radiating eight lines. Voila.
Then there are various things on writing 3-dimensionally. Here's the text of all that:
ART. Practice this way of writing by making all your letters do these kinds of things for a while. Then add the boundaries.
Or try writing normally, then tilting the page and drawing what you just wrote.
Note the wrongway to tilt your letters; you can't just write at an angle, you must mutate the lines of the letters themselves.
Note that in the wrongway, the verticle parts of letters are parallel to the verticle edges of the foreshortened square, and the horizontal lines are angled so they're parallel to the angled "horizontal" edges of the foreshortened square.
Notice how even the parts of the same letter bend down, not just each letter being lower than the other. Notice the oval as the foreshortened letter "o."
Avoiding wrong angles (in foreshortened writing) is tricky; may take a lot of practice.
Next to last is a bending serrated tube. Note the (|) shape at the point most parallel to the "camera", and the scrunching of the foreshortened bands at the parts most perpendicular to the "camera."
Finally, four common LEGO/Bionicle pieces.
A'ight, there's the guide. Comments/questions/yadda?