Lesson 7: Applying Construction to Vehicles
1967 Shelby Mustang Demo
While there is another car demo in this set, this one is by far the most detailed and complex of the lot. It is worth mentioning that the Camaro in the next demo was done in October 2016, whereas this one is from December 2018.
It is also at this point that I truly accept that cars - especially classic muscle and sports cars - are like the boss fight of the RPG that is Drawabox. Here you will be applying all of the spatial awareness you've been developing over the last while as you've worked through the rest of the lessons.
To put it simply, cars are extremely difficult. They are composed of very specific curves in all three dimensions, combined to convey such a striking and precise sense of character. Similar to the concept of the "uncanny valley" when depicting or reproducing humans, there's something in us that can easily recognize that a drawing of a car is wrong - even if we can't put our finger on it.
Anyway, lets get started. The demo video for this is an hour or so in length, and involves full discussion as we move through the various stages of construction. Below you'll find the major steps laid out with text description.
Here's the reference I'm using. As you can see, there are a number of different shots of the same vehicle. I was pretty lucky and found exactly the car I wanted (a 1967 Shelby Mustang GT500 Fastback, the one named "Eleanor" from Gone in 60 Seconds) with high resolution images in a blog post by West Coast Shipping.
My actual drawing will be based on the large image in the bottom left.
To start with, we're going to take a look at the proportions of our vehicle.
While here I've done the study from a single image at a 3/4 view to save myself some time and keep this long demonstration as short as I can, you should really do separate studies of the front and side, using the fact that the height of the car in both images represents the same physical measurement to provide a consistent comparison between the measurements in the different photographs.
I find using the wheel to be an extremely useful measuring tool. Being a circle, the plane that encompasses it will be a square, and therefore an effective scale that can be applied to all three dimensions of our vehicle.
What you see in this proportion study image is a combination of a few techniques to build a grid off the square that encompasses each wheel. Keep in mind, I do take certain liberties with approximation and guesswork, but I try to keep them limited.
I'm able to find the length of the car (approximately 7 wheels, mind the fact that I've actually started that count at 0 instead of 1) by simply repeating that measurement in perspective, a technique covered in lesson 6.
The front is more difficult - I can carry the height over from the side, but in order to project that same distance in this other dimension in order to create a square, I need to apply the circle-in-plane techniques discussed as part of the cylinder challenge. Once the square is established, I repeat that measurement over.
The height is found by repeating the unit square up once, and then adding a little extra for good measure. I assumed it to be about a quarter - this is the measurement I'll be using, though upon later reflection it was more like one eighth.
Establishing our box
To start with, I've drawn two lines on my page. The horizontal one is my horizon - while I'm not drawing explicit vanishing points, knowing where the horizon is will help me estimate their position as I'm converging lines towards them.
The other line represents a vertical edge of the box that will encompass the entire car - specifically the edge closest to the viewer.
The two new lines along the bottom each establish where our encompassing box will touch the ground. Each line goes off to its own vanishing point, implying their position (based on where each line intersects with the horizon line.
Be careful of where you "place" these vanishing points (or more specifically, where you imply them to be). If they are too far apart, you're going to end up distorting your drawing considerably.
Try and think about how they relate to this central vertical line. For example, if we were to straighten the line going off to the left to be more parallel to the horizon line (pushing that left vanishing point further out), then the front of our car is going to end up feeling stretched.
If you want to learn more about the mechanics of how to place your vanishing points, this is covered in lesson 1.
The line I've added here represents what I want to be the top of my wheel. If you look closely near the vertical line, you can see where I'd added a little dot to mark that height before drawing the full line.
The height I've chosen for my wheel is arbitrary, but it does determine the rest of the scale for my car, since our "unit" measurement is built around the height and width of our wheels.
Next, I've drawn a line going to the left to match the height established in the previous step, and have drawn ellipses on both sides of this corner.
These ellipses are meant to represent circles in 3D space, and therefore they must fit the criteria discussed in the cylinder challenge:
Minor axis must go to the appropriate vanishing point (right ellipse's goes to the left VP, left ellipse's goes to the right VP - DON'T GET THIS BACKWARDS!)
Edges touching the top and bottom go to the opposite VP of the minor axis, setting them perpendicular to the minor axis in 3D space
The points at which the ellipse contacts the top and bottom edges should align towards the vertical vanishing point (in this case we're working in 2 point perspective, so the vertical VP is at infinity and all our verticals run straight up and down, perpendicular to the horizon line). The red lines in this image represent the line passing through these "contact" points.
Using basic repetition of measurements in perspective, we construct a grid of 4 unit squares towards the left vanishing point (the front of the car) and 7 unit squares towards the right vanishing point (the side of the car).
Note: The diagonals we use to repeat the wheel-plane back in space must pass through the far end of the plane at the vertical middle. This technique is explained on page 1 of this lesson.
Using the same technique, we can transfer the unit squares up once to establish a height of 2, and then add on an estimated amount for the extra quarter. You can actually be more exact with this without too much difficulty (increasing the height to 3 and then subdividing the last twice) but I didn't bother.
Next I draw a cylinder in that box, applying what we've learned from the 250 cylinder challenge. I did include a bunch of inset ellipses within the one facing us, along with a few that are recessed to create the illusion of a wheel.
You don't need to do this just now - I mainly did it because with the digital tool I was using to draw my ellipses, it's not particularly easy to line it up again. To this end, I would recommend noting down the degrees of ellipses you're using (if you're using an ellipse guide), so you can find them again quickly later on.
Also worth noticing - my 'cylinder' isn't actually perfectly cylindrical. The edges along its length flare out, a bit like a bubble. If you're not sure why, then you probably haven't completed the 25 wheel challenge - one of the prerequisites for this lesson.
I also extend the bottom lines of the enclosing box I constructed in the previous step to the opposite side so I'd at least know where the other wheel was going to sit, even if I wasn't necessarily going to draw it.
The hard part
You may feel that up til now things have been difficult - after all, cylinders are anything but easy, and there's been a lot of perspective to handle. In truth however, that's been the tedious part. Now things are really getting hard.
I'm going to try and keep my construction here as broken down as I can, but you will find that I'm not going to be constructing everything with forms as simple as I usually do. Instead, I'm leveraging the enclosing box and perspective lines I've established as well as my general spatial awareness (which you'll hopefully have developed while working through the rest of the lessons - which is why this topic is the last lesson of the curriculum) to build forms that are solid, but somewhat more complex and nuanced than we're used to.
What I've added here is the chassis. It's still pretty simple, but I'm showing some awareness for how the form swoops up towards the back and slants downwards towards the front. Studying your reference carefully is a must. Cars are full of all kinds of subtle shifts in form that you're simply not going to pick up on at first, but the only way you'll learn to do so is to really look.
I'm using some contour lines along the hood to reinforce how it slants down towards the front. I'm also subdividing the space where the headlights and grille will eventually go.
I'm not going to jump too deep just yet however - I don't want to get too caught up in any one part just yet. I'm still trying to tackle it big-picture style, hammering out all of the car, little by little, only gradually getting into more specific details.
Pushing back, I've captured the hood scoop as well as the cabin.
You may notice that I haven't put much into the hood scoop in the middle, despite it being a somewhat significant form as far as the whole construction goes. This is because I would generally advise you not to work strictly from front to back, or back to front. Don't push in one direction - instead, establish everything on one end, then jump to the opposite end. Once you've got that nailed down, it'll be easier to place whatever falls in between in relation to two points of reference, rather than letting your positioning errors accumulate as you go.
It's a lot like when you try and write out a title nice and neat, but because you're working from left to right, you end up underestimating the amount of space you're going to need and run out of room. The better way to approach it is to place your first and last letters, split up the space in between and allot them accordingly to the remaining characters.
I haven't forgotten about the opposite side of the car - you'll see that I've blocked in the far corner of the interior/cabin to maintain my awareness of what's going on over there.
I've also added my wheel well detail, and have paid special attention to the contour/cut lines coming down the side of the driver's door. Again - observe your reference carefully. These surfaces are designed to a very particular specification, and you need to sharpen your senses to pick up on all the little curves and bends and twists and turns.
Now we're back towards the front. Using the subdivision lines I put down before, I'm starting to flesh out more of the complex curves and structure. Try to identify major landmarks and separate things out. I've immediately separated the grille from where the headlights sit, for instance.
In addition to this, don't ignore the silhouette. Any feature that breaks up the silhouette is going to be critical, as the silhouette of an object is the most important element that conveys its character. Here I've made a particular point to flare out the bumper.
At this point I'm only including the rest of these images for the sake of showing a steady progression, rather than dumping "draw the rest of the owl" on you. It's all just elements I've identified from my reference images that I'm dropping in, bit by bit.
I'm spending the majority of my time studying my reference, and whenever I pick out a specific element that I want to carry over, I'll transfer it a few lines at a time, always going back to refresh my memory before drawing any more.
I'm starting to get a little bolder as I reach my finishing point. I'll drop in some stronger black areas (which currently look a bit strange because they're highlighted in red). I'm more willing to commit now - not that I wasn't committed to the strokes I was putting down before, but I was ready to face the accumulation of mistakes I'd been making til now, and having to quickly pivot or adjust accordingly.
The last thing here is to put down a nice, solid cast shadow. I'll always draw these as an outline first, before deciding whether or not I want to fill it in. In this case, I decided to go ahead and fill it with a solid black to make the rest of the car pop (when working in ink, you may want to use a brush pen for this - just be careful not to let it go outside your set outline).
I've also put down a little bit of detail - nothing rushed, but nothing particularly distracting - to the ground to help make the drawing feel more natural. Lastly, I leveraged a puddle near the front wheel as an opportunity to put a few more solid blacks to balance out my overall composition.
As I said when we started, if your construction went catastrophically wrong, that's not surprising. This is a complicated task, with many layers of construction and a significant amount of dependence on the skills you've developed through these lessons. It's no longer all about putting each and every simple form down on the page - due to its complexity you're forced to rely more and more on your mental model of 3D space, your ability to visualize or at least understand how those forms interact with one another, and your ability to pivot in response to problems and mistakes that may come up.
I'd recommend that before you attempt a car like this, start with something more like an old vehicle from the 1920s, like the Ford Model T, or even a boxy Volvo from the 70s or 80s. It's not a simple one by any stretch, but it is going to lend itself a little better to bridging the gap between what we've tackled so far and more modern cars.
The Science of Deciding What You Should Draw
Right from when students hit the 50% rule early on in Lesson 0, they ask the same question - "What am I supposed to draw?"
It's not magic. We're made to think that when someone just whips off interesting things to draw, that they're gifted in a way that we are not. The problem isn't that we don't have ideas - it's that the ideas we have are so vague, they feel like nothing at all. In this course, we're going to look at how we can explore, pursue, and develop those fuzzy notions into something more concrete.