Jumping in with your form intersections, you're doing a great job in demonstrating a strong understanding of the relationships between these forms. My only concern here is that your linework - at least on the first page - is decidedly rougher, which suggests that insofar as the application of the ghosting methods three distinct stages go, you're probably allowing it to get a little mixed together, falling back to trying to draw in a more instinctual fashion. Remember - Drawabox is designed in such a way that every mark we put down in our work here is planned out and considered. It is this which helps us improve the way in which we draw when doing so in an instinctual, fluid manner - and so all the work you do in this course needs to implement that kind of specificity and care, in order to be leveraged to its best effect. It is easy to lose sight of that.

Continuing onto your cylinders in boxes, your approach here appears to line up well with the instructions, so you are indeed still applying this exercise correctly. That's really all we're double checking here, to ensure that as you continue to practice this in the future on your own, you'll still be doing so with the line extensions being applied correctly. I should mention that in the upper left you had an issue in that the ellipses only touched two of the four edges of their enclosing planes, but as this was not present in the other cases, I expect this was more of a one-off issue, and you did roll with it well, in terms of how you went about applying the line extensions based on that.

Moving onto your vehicle constructions, you continue to handle this well, and the care and specificity of your linework has definitely returned here. You're demonstrating an inordinate amount of patience in approaching your constructions, with your approach yielding a great deal of precision in many cases.

That said, there are a few points I want to mention, and share my thoughts upon.

• In the drawing of the Trabant P601, I noticed that there were some areas in which your approach did not offer as much precision as it could have, due to certain elements being overlooked. I can see that you're leaning into working with a grid with set, consistent divisions of space, but this predisposes you to overlooking certain things. For example, take a look at what I've marked out here. For now, ignore the door handle and focus on the part defining the structural element of the windshield. In both the construction and the orthographic plan, we can see that no decision was established for where this would go - you decided where it would connect to the main body of the car through approximation/eyeballing, rather than making a decision ahead of time, leading to a drop in precision. Ideally we'd decide in the orthographic plan, where it looks like the element would span from the 3 and 1/3 mark to the 3 and 2/3 mark (something we can establish by subdividing that column into thirds). That can then also be applied in your 3D construction, to pin down the specific positions. While that's not exactly what you ended up drawing in the construction (looks closer to 3 and 1/2 to 3 and 3/4), it's in the same rough ballpark. What matters is that the decision is made first, separately. Looking at the door handle however, it's clear that you made a decision when tackling your 3D construction (for the door handle to span between the 5 and 6 marks), but it's very clear that this decision differed from what you were leaning towards in your orthographic plan. That is still better than having no decision made at all, but again - you'd ideally try to make as many of the decisions as you can in the planning phase, prior to transferring that information into three dimensions. Or, if you find a decision that hasn't been made yet, you can always go back to your plan and figure it out there.

• I'm breaking this off into a separate point, but I did want to note the fact that leaning so firmly into consistent grids of set units can make us less inclined to add the in-between subdivisions. You may feel yourself split between deciding to subdivide everything to a significant degree (as you did with this covered wagon), but that is not necessary. You can apply entirely different levels of subdivision where they're needed. The only focus is on making the decisions ahead of time, and being able to transfer them from your orthographic plan to three dimensions. I suspect this may have played a role in your decision to leave the covered part of the wagon out - it was much harder to define in terms of your specific grid. Of course, I believe part of it was also the fact that you didn't fall back to the core principles of breaking things down into simple forms. No need to worry about the wavy nature of the canopy - build it as a solid simple form first, then add the rest of those concerns towards the end.

• This is just a minor point, but remember that filled areas of solid black are reserved only for cast shadows, and most cases of filling in an existing shape or surface that is already present in your construction is likely to be form shading, where you're darkening the surface based on its orientation. Cast shadows will generally have their own separate shape that you yourself must design - it's that shape which establishes the relationship between the form casting the shadow, and the surface receiving it. The construction of the trabant p601 is a good example for this point as well, because filling the interior of the cab with solid black is entirely fine. It seems like an exception, but in essence we're saying that the entire interior is covered in shadow, which is cast by the structure of the cab itself. Obviously this isn't 100% correct (since light would penetrate through the windows) but for our purposes, it's fine. Filling in the side of the wheel, or the wheelwell however, is definitely form shading, and should be left out when doing these kinds of exercises (just to avoid confusion between what is a cast shadow and what isn't).

As to your question about the whole stage you constructed, I think there's a lot of value to what you're doing here (and it's very similar to how James Gurney talks about working with maquettes in his book, Imaginative Realism: How to Paint What Doesn't Exist, but there's an important distinction to make. What you're doing here is effectively taking the kinds of things we tackle in Drawabox, and turning it into an exercise that is more observational. Or at least, it very much has the potential to be that way.

You'll definitely find this kind of exercise in more traditionally perspective-focused courses, but it is distinct from what we're doing in this course. What we're doing here does not concern itself with accuracy to the reference - our proportions can be off, but as long as those proportions were decided upon ahead of time, that's fine. Nor does what we're doing here concern itself with perfect perspective, and thus we don't worry about concrete vanishing points, station points, laying out a cone of vision, and all that. What we're doing here speaks only to developing your brain's internal model of 3D space, and the flexibility with which you can employ it to create whatever it is you want.

Now, I don't know what your goals are. There are definitely lots of courses that'll applaud you for devising and employing this kind of exercise, and I should mention that those instructors are far more qualified than I. Still, I think the way in which we work here, without concrete vanishing points, without everything being just-so, is far more effective for the demands of my own profession. That is, concept art specifically, although illustration as well.

Setting aside the value of what you're doing here as an exercise, I think there's a lot to be said about staging things like this for the purposes of making a specific illustration and ensuring that everything is as correct as you can manage - but that is where we're our goal is all towards the production of this "final image". That is, as opposed to what we're doing here, which has no concern with the final image, but rather focuses on the process as part of an exercise. That's a very important distinction to make. The drawings we produce in this course are inconsequential. It's the production of them, and the specific approaches we employ, that matters to us.

Ultimately I would definitely try and get other opinions on this - I've clearly tried to voice the value that lies here for others, for those whose focus is different from mine - simply because I wouldn't go to these lengths myself. I've explained why I wouldn't do it for an exercise intended to develop one's spatial reasoning skills, but I also don't really do it for my own illustrations, both because the return on investment isn't that high (I imagine you spent a ton of time setting this all up), but also because I feel it would lock me into one specific trajectory and make me reticent to change it even if I felt that would improve the piece. But then, that's just me.

Anyway! Back to the lesson, you've done quite well, and while I outlined ways in which your exercises could be yet improved, you've demonstrated a solid grasp of the material, and a clear capacity to leverage these exercises as you continue forward on your own work. Just remember - everything serves a purpose, and changing the nature of a given exercise can change the purpose it achieves.

So! I'll go ahead and mark this lesson, and the course as a whole, as complete. Congratulations on this culmination of nearly eight months of hard work.