Lesson 7: Applying Construction to Vehicles
1967 Chevy Camaro Demo
This demo is a little older, having been published in October 2016. As such, while I have decided that there is still something of value here, any techniques or approaches outlined in demonstrations not flagged with this message should be considered to take precedence over what is covered here. This is a natural part of Drawabox being an evolving, growing resource.
I love cars, but I hate drawing them. The thing about cars is that they all have very specific proportions to them, very specific curves, very specific form language. It's very easy to stray ever so slightly, and people will notice. That said, the technique applied to the computer mouse in the last lesson is pretty effective for this sort of thing.
This demo recording has no accompanying audio commentary.
Step by step
Whenever dealing with cars, ALWAYS think in terms of wheels as measurement units. I mentioned this in the truck demo, but it's even more important here.
You'll notice that I've got a set of somewhat lighter wheels - six of them - running along the length of the car, not actually aligned to the car's actual wheels. Looking at the reference image, I noted that the car is roughly six wheels long and two wheels tall. I did extend my car a little further in each direction, largely to compensate for little discrepancies with this method of estimation.
In order to properly establish the proportions of my car, I want to use the perfect-circle approach. In order to do so, I define a vertical and two horizontal lines - the vertical to act as the bounds for the ellipse I'll be drawing, and the two horizontals to establish a vanishing point off in the distance, and define my perspective system.
Remember the criteria you're trying to hit. You want your ellipse to fit snugly between these three bounds. You also want the points of contact it has with the top and bottom lines to sit perfectly above/below each other. Use an ellipse guide if you have one. If you don't, accept that freehanding it won't be perfect, but it'll be good enough. Probably. Once you've got your ellipse down, close off the plane to establish what is now a square in 3D space.
Okay, this image definitely looks confusing at first, because there's two things happening here. First, I took the measurement of that square in 3D space and carried it down four more times to establish five equal squares. Remember that the car is roughly 6 wheels long? Well, in observing my reference, I also noted that there are three wheel-lengths between the front and back wheels. This leaves roughly one wheel to be split up between the section in front of the front wheel, and behind the back wheel. To my eye, the front looks to be about 1/3 of a wheel, and the back appears to be 2/3. So what do we do with this information?
First, I don't want to clutter up my drawing by first measuring out six wheels, then placing the actual wheels in a way that doesn't actually align with any of the wheels I've put down in order to measure things out. So, I drew five wheel lengths, with the first and last ones actually corresponding with the car's actual wheels. All that's left is to tack on the front and back sections. But how?
That's where the additional marks come in - first I eyeballed a point that is about 1/3 of the way into the front wheel. Then I mirrored this measurement up and across the left side of that wheel's plane, giving me that same measurement directly to the left of the wheel. Did the same thing for the back - eyeballed a point 2/3 of the way into the back wheel, then reflected it towards the right.
These transferring techniques are extremely useful when trying to figure out little finicky things like this, while maintaining a degree of accuracy.
Now we're using the same technique to transfer the height of the wheel up. Since this drawing is in 2 point perspective, our verticals run straight up and down and do not converge towards a vanishing point. This means we could technically just take a measurement of the left-most wheel's left side and transfer it up, but I generally avoid that kind of approach wherever possible. It's just a lot better to get used to thinking about these techniques as tools at your disposal that can be used in any situation, and getting used to reaching for them rather than a measuring tool.
I will admit something - I skipped a pretty important part of all of this. I did a proportion study for the side of the car, but the front is also very important. Make sure you do proportion studies for every major side of the vehicles you attempt, as it will come in very handy. Since I have no proportional information to go off of, I've pretty much decided to wing it and extend the box out to a point that seems roughly correct to me.
Now that we have our box fleshed out, we can finally use all of our subdivisions to transfer the forms we identified during the proportion study and build them up by first stamping them onto the side plane of the box, and extruding them across. Notice that I'm not using any curves here - straight lines are in most situations going to maintain a sense of solidity that you want to hold onto as long as possible. Once you've got all of your forms fleshed out as solid blocks, then you can start smoothing them out. Don't do it any earlier than that.
If you think about it, all of the parts you need have already been drawn - all that's left is to refine your forms and add detail. With all of the subdivision, you can construct the remaining forms in a sort of connect-the-dots fashion, and then smooth over the resulting hard edges wherever necessary. That's the thing about construction - 99% of it is just preparation, but the last 1% of it is where the drawing suddenly jumps out at you and comes alive.