Before I get started on your critique, I figured I'd answer your question, just to avoid forgetting about it.

When it comes to design of any sort - which of course includes character design - the course that had the biggest impact on me was one I took at Concept Design Academy, called Intro to Form Language, taught by John Park. It was an in-person class, but since then John had started his own school, Brainstorm School - while I can't speak to any of the specific courses they offer, since I haven't taken them myself, I can at least vouch for the fact that John Park is an excellent educator, and I'd expect anything he's involved in to be of fairly solid quality. They have a specific Intro to Character Design course as well. Note that these kinds of courses are not cheap - they're intended for students studying to pursue jobs in the entertainment industry, and so like most other schools, they are priced as an investment.

If you're on a much more constrained budget, the Foundation Patreon's Gumroad (they had a patreon where they'd release videos regularly, but you could purchase them one-off as well through their gumroad) has a lot of content talking about all sorts of design topics. For example, this one explores character creation/ideation, and it seems they're giving away the first part for free.

The thing to keep in mind with these cheaper options is that you're not getting a complete structured course - you're getting bite sized pieces of information. If your foundations are well developed, these can still be very useful, but without feedback and without the larger context of a full course stringing everything together, you're going to be doing a lot of the heavy lifting yourself.

Going back to that Intro to Form Language course I took, what it explored were the core elements of a design, the visual aspects of which largely come down to shape/form language, proportion, and balance. I'm not going to get into it in any significant depth but shape/form language speaks to the simple pieces that are combined into creating your design, which generally are most prominent and visible when we fill the entirety of the character in to view only their silhouette. Proportion speaks to how those different components relate to one another in terms of their size (both in 2D space when we're talking about shape language, and in 3D space when we're talking about form language). Lastly, balance speaks to how they're arranged, and often comes down to achieving a sense of balance (where the elements on one side are balanced out by elements on the other) without necessarily resorting to full on symmetry, which is often quite boring.

We did exercises such as taking existing character designs and analyzing them to break down their shape language and how they relate to the kinds of characters they were, creating a ton of silhouettes focusing on specific combinations of shape language (there's traditionally 3 major shapes, circle/rectangle/triangle, and so most of the time we'd pick two of the three and do our best to generate some kind of character silhouette - I remember this being especially difficult as I was getting used to it), taking an existing object and doing pages and pages of studies of it before finally doing a few more pages where the proportions of those objects would be changed to explore how those changes affected the impact/impression of the object's design, and so forth - I went with a 1967 Shelby Mustang GT500, and it was really interesting to see how increasing the height of the cab relative to the rest of the structure took a very serious, intimidating looking muscle car and turned it into something more of a clown car instead.

Ultimately a lot of this came back to the idea that design is problem solving - I come from a programming background, so problem solving is my bread and butter, but it wasn't until I took that course that I understood the tools that were at my disposal for solving those design problems. Prior to that it all felt random and unstructured, which is why the course had such a significant impact.

Touching briefly on the idea of problem solving, what's left is identifying the nature of the problem you're trying to solve. I actually talk about my main strategy for that (which I refer to as the "What if" strategy) in this video. The title is entirely clickbaity, it's not about art block, it's about thinking through what it is you're actually creating in order to reveal and explore the "problems" you're trying to solve through your design.

Lastly, when you say "visual structure" I'm assuming you mean the layout of your comic pages. Back when I first started drawing my own web comic (which is currently on pause to focus my efforts on Drawabox), I had "Making Comics" by Scott McCloud recommended to me. I know he's very well respected when it comes to understanding the basics of comic making, and he's got other books on the topic like "Understanding Comics", and I did buy what had been recommended to me, but... to this day I've never cracked it open. So, as yet I don't really know if it's everything people say it is, but I can pass on the second hand recommendation to you.

What I do know however is that a lot of planning out your pages comes down to thinking about how the viewer is going to engage with it. In western comics, which are traditionally read left to right, the reader generally will go first to the speech bubble in the top left, and will slowly move down and to the right. Recognizing this hierarchy, and how one thing being higher or further to the left may cause it to be seen first by the viewer is pretty much the basis for structuring pages. My recommendation regardless of what resources you ultimately use to learn from, is that you start drawing comics sooner than later, so you can actually learn from the experience of it.

So for example, where earlier on in the comic I would sometimes make mistakes where the speech bubbles would compete with one another (so here on page 58, the "Fine, but you're going to have to find your way out of this audit" is further to the right but also higher than the "Can you even handle ghosts?" which, though obvious when you think about it, requires a moment of thought from the reader to figure out what they should be reading first - especially since the former seems to follow the path of "Can we take this one? Puh-leaaaaaase?" potentially causing the reader to totally miss the bit about ghosts.

Compare that to page 205, which is similarly busy, but follows a much more predictable arrangement of dialogue that helps the reader flow through the page with more ease. Learning the rules to how this stuff can work is one thing, but at the end of the day you're going to need to develop the mileage and experience with both following and bending those rules in order to properly get comfortable with them.

Anyway, I went way longer into that stuff than I ever intended - so I hope it helps in some fashion. Let's get onto your actual critique.

Jumping in with your cylinders around arbitrary minor axes, your work here is a bit mixed, although overall I'm seeing a decent understanding of how to go about constructing these forms and how they sit in 3D space, but there are a number of issues that do come up:

• In the instructions' reminders section, I mention that it's very important that we avoid forcing our vanishing points to infinity (which would result in the side edges being drawn as parallel on the page, regardless of how the cylinder itself is meant to be oriented). While you didn't do this all over the place, which is fortunate, but I did see it in a number of cases throughout the set. It's possible that these were cases where you weren't intending to keep the side edges parallel, but I still wanted to highlight them as it does suggest you might want to push yourself to make the farther ellipse a little smaller, in case you have a tendency to execute them a bit larger than you planned for. Just to name a few cases of where this is happening, so you've got some examples of what I'm talking about: 8, 31, 33, 69, 79, 83, 150.

• While you have lots of cases where you're demonstrating a solid understanding of how the shift in degree works (where the end farther away from us is to be drawn with a wider degree than the end closer to us, as we see prominently in cases like 132), there are a number of cases where you seem to very intentionally reverse this relationship - for example, 76, 134 and 135 are very prominent cases of this, although we can also see this in cases like 63, 99, 120, and so forth - those are ones where the degree of both ellipses is pretty much the same.

• This one's not really an overt issue, more because it's not something I explain directly, as I want to give students the opportunity to pick up on this on their own, even if only subconsciously. Basically, our cylinders feature two "shifts". There's the degree shift we were just talking about, and the shift in overall scale that comes from the side edges converging according to the rules of perspective, squishing that far end to be smaller overall. Both of these shifts convey the same thing - the rate of foreshortening, which itself is what tells the viewer just how much of the cylinder's length can be measured directly on the flat, two dimensional page, and how much of that length exists in the "unseen" dimension of depth. Because they represent the same thing, they work in concert - as the far end ellipse gets pinched down with perspective and made smaller, it will also get wider at a similar rate. So, ideally we'd always increase the degree in tandem with the shrinking of the overall scale to avoid things looking "off". So for example, 149 is a great example of these two shifts working together to create a consistent impression, whereas 100's degree shift is fairly minimal, while the scale shift is more significant, resulting in the cylinder feeling a bit weird - even if the viewer isn't sure why.

Continuing onto your cylinders in boxes, while you're mostly applying the instructions for this exercise correctly, there is one key area where you seem to be a bit confused in regards to which lines we're meant to be extending. This exercise is really all about helping develop students' understanding of how to construct boxes which feature two opposite faces which are proportionally square, regardless of how the form is oriented in space. We do this not by memorizing every possible configuration, but rather by continuing to develop your subconscious understanding of space through repetition, and through analysis (by way of the line extensions).

Where the box challenge's line extensions helped to develop a stronger sense of how to achieve more consistent convergences in our lines, here we add three more lines for each ellipse: the minor axis, and the two contact point lines. In checking how far off these are from converging towards the box's own vanishing points, we can see how far off we were from having the ellipse represent a circle in 3D space, and in turn how far off we were from having the plane that encloses it from representing a square.

As I see it, it doesn't appear you are actually extending the minor axes for either of your ellipses. Instead, you appear to be extending the boxes' 4 edges for that dimension, and then what appears to be a side edges for your cylinder, rather than the minor axes of each ellipse individually. Here I've marked out the lines you're extending incorrectly, and those you should be extending instead, directly on your work.

Aside from that, your work is otherwise done well, but this is a pretty significant issue that I do want to ensure you understand - so I'll be assigning some more limited revisions. You will find those assigned below.