Lesson 1: Lines, Ellipses and Boxes

Now that we've gotten used to using vanishing points in a direct, explicit fashion, let's look at removing those training wheels and trying to estimate our convergences a little. You'll find this pattern to my approach - I'll introduce some rigid, technical tools, and then gradually remove them one by one to push those mechanisms into your subconscious.

The goal of this entire course is to make it so you don't have to consciously think about how things need to be drawn, so you can focus on what it is you're drawing, thus maximizing the time you spend in the "flow" state where you're most creative.

In order to achieve that goal however, we need to be hyper attentive to everything we do throughout this course, in order to rewire how our brains approach drawing in general.

We start off similarly to the previous exercise, by breaking our sheet of paper into 3 rectangular frames with a ruler.

Using a ruler again, add a horizon line to the middle of the frame, and place a single vanishing point towards its center.

Because we're doing more estimation in this exercise than the last one, it's very common for students to... well, panic. And when they panic, they don't think. So if you catch yourself panicking, take a step back and go back through these instructions.

As we're using one concrete vanishing point for all the boxes in the scene, it means that all of these boxes are aligned in a very specific fashion. Their faces are not slanted at all relative to us or the ground plane.

What this ultimately means is that for this step, all we need to do is draw a rectangle. Do this using the ghosting method - meaning, plot all 4 points, focusing primarily on keeping them aligned with one another. Keep in mind we're going for a rectangle here, not a square, so your proportions don't matter.

With your points plotted out, go through the preparation and execution phases of the ghosting method to connect them with confident edges.

Remember that the ghosting method is all about breaking the process into steps so your final execution can focus on being confident, rather than worrying about accuracy. This may of course lead to your lines going a little off track, but that is of course entirely okay, and simply something that you'll get better at with practice.

As you can see here, I'm making plenty of little mistakes, as I'm admittedly rusty from spending most of my time teaching, rather than drawing.

Now things get a little tricky. The main limitation of this exercise is that you're not allowed to draw lines all the way back to the vanishing point. That means this is going to take a great deal of estimation - which is why it's called the rough perspective exercise, not exact perspective.

Start with just one corner. You can ghost with your pen between the vanishing point and the front corner of your box to get a sense of the path between them - but when you feel comfortable, place a point somewhere along that path.

What makes it so tricky is that the moment we stop ghosting through that path, our recollection of its orientation immediately starts to fade - but if we rush to place the point, we're also prone to being sloppy and making a mistake. So, it's a tug of war between not taking so long that you lose track of the path, and not being so quick that you get sloppy. As with all things, you'll get the rhythm of it with practice.

Repeat the same process for the other 3 back corners. Obviously this gets time consuming when done correctly. Get used to taking breaks when you need them, and don't be afraid to spread this exercise - or any exercise throughout this course - across multiple sessions if you need to.

Working tired is a great way to work sloppy, and falling short of your personal responsibility to tackle the work to the best of your current ability.

Again, connect the remaining dots using the ghosting method to complete your box.

You've drawn one box! Congratulations. Now draw another 4 or 5 to fill out the rest of the frame.

Once you're done a whole page, grab a pen of a different colour, or a pencil (as long as it's sharp), or something you can visibly separate from the rest of your work. Using your ruler, take all of your depth lines (those that are meant to converge towards the vanishing point) and extend them to where they intersect with the horizon.

In doing this, we're analyzing how far off from the vanishing point our lines' trajectories actually hit the horizon. Of course, if everything were perfect each of these red lines would hit the vanishing point, but this is unlikely to be the case.

By looking at the points where these lines do hit the vanishing point, we can see just how far off we are - and more importantly, we can identify patterns of behaviour where these margins of error tend to be greater.

A common one, which you can see in my example, is that the further out to the side a box is, the further off from the vanishing point its edges are. This means that for these boxes, we need to consciously push ourselves to exaggerate their orientation, pushing those edges to turn more towards the vanishing point than we think we actually need.

In a perfect world, our line extensions would follow along the edge we've drawn, and simultaneously pass through both of the corners on either end of the edge. But this is not a perfect world, and we make mistakes resulting in the line we draw not necessarily connecting those two points perfectly.

This begs the question, in such a case do we draw our line extensions so they pass through the corners (in effect testing our intent), or do we extend the edge that was drawn (to test our execution)?

The answer is that both are acceptable, and both have their own advantages. What matters most is that you don't mix-and-match within the same frame. You can try different ones between different frames of the exercise, or you can just stick to whichever one you feel most comfortable with through the whole exercise.

Every now and then I'll have a student submit work for this exercise that has more than one vanishing point. I want you to use only one because it allows you to focus more on the challenge of aligning your lines back to that one VP. Giving yourself multiple things to worry about isn't going to help you learn any quicker, it'll instead slow you down. As such, each exercise is designed with a strategically chosen set of goals.

When adhering to a strict perspective system (one point, two point, three point), you immediately give up a fair bit of flexibility in terms of how the objects you're drawing will be oriented. For example, if everything is drawn according to the same two vanishing points, it will be as though everything was placed on a set grid.

There is a vast advantage to this however - because the orientation of objects becomes restricted, we limit the possible behaviour of every line to only a few options. Many students ignore this fact, however - they lay out their vanishing point(s), then assume they'll know by instinct how every line should behave. Instead of knowing, they guess.

If you ever catch yourself guessing or uncertain about how to draw a line, stop and step back. Take a look at the system you're working with, and think about the short list of possible behaviours. In one point perspective (which we are using for this exercise), you have the following behaviours:

1. All lines that go off into the distance converge at the vanishing point

2. All horizontal lines run perfectly parallel to the horizon

3. All vertical lines run perfectly perpendicular to the horizon

There are simply no other options. Every single line will adhere to one of these three behaviours. So, find which one matches the line you're trying to draw, and then apply it.

It is fair to say that in one point perspective, you can have lines that don't run parallel or perpendicular to the horizon, but - these rules are going to hold fast here however because for the purposes of this exercise, all of our boxes are going to run parallel to the ground plane, rather than being slanted or angled.

So the point of this exercise is to get used to estimating how your lines converge back to the VP based on the section that is actually drawn, rather than actually drawing them all the way back on the page. So, instead of drawing those lines, you can work on your accuracy by ghosting the lines all the way back to get a good angle, then drawing only the segment required for the box you're drawing.

It's extremely important that you apply your line extensions correctly, ensuring that you are extending along the trajectory of the edges that exist in the depth dimension of space.

One issue we see sometimes from students is that they'll start their line extension at the corner closest to them, then extend in some arbitrary direction. While logically there's no reason to do this, it comes down to confusion, which results in panicking.

Just as noted in the "Guessing" mistake, it's critical that you not take actions for which you cannot explain their reasoning. It is entirely okay to make a mistake because you misunderstood something, as that misunderstanding can be identified and corrected - but if you act on random impulses, that is much more difficult to address.

So, if you find yourself confused and panicking, always go back to the instructions. If they don't clarify things, feel free to join our discord chat server and ask for help.

The lower example in this diagram shows a similar mistake, but one that follows a certain logic (even if that logic is incorrect). Here the student is drawing their line extensions from the front corner to the vanishing point. While this is not what the instructions have us doing (specifically because it's much harder to measure how far off we are, as it requires us to judge the angles between the edges and the extensions, whereas the correct method allows us to judge distances instead), it does hold some merit, and more importantly the reasoning behind the mistake is something that can be discussed and corrected.