8:49 PM, Friday August 6th 2021
Looking at your new drawings, I don't really get the impression that you fully understood some of the major points I raised in my feedback. One of these that stands out, is as follows:
One of the main reasons it went off the rails, so to speak, was because when drawing the various sets of parallel lines that the construction would follow, you did not make any attempt to have them converge. You effectively placed those vanishing points at infinity, even though none of those sets of lines were actually running perpendicular to the viewer's angle of sight. Given their orientation, they definitely should have been converging towards their own vanishing points.
I also addressed this issue again at the end of the critique, when attempting to summarize things so you'd be more likely to keep it in mind:
Make sure you're considering how the sets of lines that are parallel in 3D space converge towards a vanishing point when drawn on a flat page. The only situation where the vanishing point goes to infinity is when that set of lines runs perpendicular to the viewer's angle of view.
To put it simply, you're not drawing in perspective. What you're doing is more similar to axonometric or isometric projection, rather than perspective projection. When drawing those bounding boxes, you're keeping all of its sets of parallel edges (in 3D space) parallel in 2D space as well, as you draw them on the flat page. You're not considering how they ought to be converging towards their shared vanishing points.
When objects are smaller - like if these were toys rather than real objects - you'd get fairly shallow foreshortening. The lines would converge a little, but it'd be fairly subtle. With objects of greater size however, that convergence becomes more noticeable. If you look at this step of the shelby mustang demo, you can see that there is a lot of noticeable convergence as those lines go back towards their vanishing points. Compare this to how you're approaching drawing these bounding boxes.
I've made some notes directly on your last drawing to demonstrate this point - see how your lines run parallel on the page?
The other point I raise on those notes pertains to the issue you raised yourself - "I think I am struggling with getting curved lines to converge towards the vanishing points."
That is certainly true, and for good reason. When we talk about how lines converge towards a vanishing point, we're generally talking about lines that follow a single, consistent path - in other words, straight lines. Curving lines themselves are, at face value, much more complicated - their trajectory changes throughout their length. But there is a much easier way to consider them. We can simplify them down into a series of straight lines, as discussed here in Lesson 6.
Each of those straight sub-segments will be subject to a different vanishing point, because they're all oriented differently in space, but they are far easier for us to build up in our construction, rather than worrying about trying to eyeball a single complex curved line.
I talked about this in my critique a little, though perhaps not enough, in the context of adhering to your box structures. In your vehicles, you have a tendency to create your general bounding box, and you'll break it down somewhat further, but you always hit a point where you decide to draw the vehicle itself, as a series of curving lines, by eyeballing and approximating things. You're not treating the straight lines, flat planes and boxy forms as a solid framework - you seem to treat it like more of a loose suggestion or a vague plan to keep in mind.
That's not the case. Think of it more like you've started with a giant block of wood, and early on you're cutting away big pieces of it to fashion it into a vaguely car-like shape. Then, as you get further along, you start cutting away smaller pieces, and by the end, you're rounding out those corners. You're not drawing a whole new car - you're just refining the structure that you've been building up the whole time.
The point I linked from Lesson 6 (I'll link it again here) is integral - you're laying down the structure that becomes your curve, first by breaking it down into a series of straight lines, then rounding out those corners at the end. This way you can establish exactly how those structures occupy space in simpler terms (boxes are far simpler than curving forms), and then increase their complexity once everything is in its place.
If you have questions, feel free to ask, but otherwise I do want you to try the same revisions again - that is, 3 pages of car constructions. It's worth mentioning that only 3 and a half days have passed since I sent out my critique (ignoring the time that may have passed between me posting it and you seeing it). Taking into account that you were spending mornings/lunchbreaks/etc. on them, as well as the time required to go through my feedback and review the lesson material, that's not a lot of time given to each of these three constructions. Though I can see that you've tried to spread it out and to give each drawing plenty of time, I still think you are underestimating just how long each of these drawings requires from you.
Next Steps:
Try the 3 car constructions again.